MECM90002

0）先把“考试结构/评分逻辑”搞清：你要赢的是分数结构，不是背百科

0.1 三个任务（你 final 前最该做的是把它们当成“一条流水线”）

• A1 Essay（50%｜2500w）：从三道大题里选一道，写一篇论证型文章，预期大约 ~20 个来源，必须“立场明确 + 证据支撑 + 可反驳”。[3]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.[12]Source: asksia-bible-mecm90002-bilingual.pdfWeight Length Due / timing A1 . Argumentative essay 50% 2500w 19 Apr A2 . Oral presentation 25% 3 min / 3 slides Wk 10-11 A3 . Written case study 25% 1500w 7 Jun A1: pick one of three set questions (national policy rationale · UN- style transnational brief . platform responsibility); ~ 20 references expected, peer-reviewed + policy/report/news. A2: step into a policymaker's shoes and pitch a policy intervention - it directly feeds A3. A3: develops that same intervention into an in-depth, single-case study (a company/practice, a country's policy or its absence, or one international instrument). A1: 从三道既定题目中选一道(国家政策依据 · UN 风格的 跨国简报 · 平台责任);预期约 20 篇参考文献,须包含同 行评审文献+政策 /报告/新闻。A2:设身处地成为一名决 策者,推介一项政策干预 -- 它直接为 A3 提供素材。A3: 将同一项干预深化为一份深入的单一案例研究(一家公司/ 一种做法、某国的政策或政策的缺位,或一项国际工具)。 The twohurdles&the AIrule 两道门槛与 AI 规则 ! Hurdles - clear both or you cannot pass 门槛 -- 两道都过,否则无法通过 Attendance: a minimum 80% of tutorials/seminars, monitored; falling below without documented evidence means you cannot pass. Submission: every piece must be submitted as a genuine attempt. Late work loses 5% of available marks per day without an approved extension. 出勤:导修课/研讨课出勤率至少为 80%,并受监 控;若无书面证明而低于此线,将无法通过。提交: 每一项作业都必须作为真实尝试提交。未获批准延期 的迟交,每天扣除可得分数的5%。 MECM90002 . Global Data Policy & Governance i AI policy in brief (MPF1310) AI 政策速览 (MPF1310) All submitted work must be your own. AI is permitted to define hard terms, locate or summarise sources for comprehension, and check spelling/grammar - with disclosure. It is not permitted to write or rephrase your sentences, to be cited as a source, or to translate your assignment. Turnitin auto-scans for AI; undeclared use triggers an integrity investigation. 所有提交的作业都必须是你本人完成的。AI 可以用来 解释难懂的术语、为帮助理解而查找或概括来源、检 查拼写与语法 -- 但必须声明。Al 不可以用来撰写或 改写你的句子、作为来源被引用,或翻译你的作业。 Turnitin 会自动扫描 AI 痕迹;未申报的使用会触发学 术诚信调查。 The one strategic insight behind every task ★ 贯穿每项任务背后的那一个战略洞见 The whole subject is a tour of a huge, global debate - but you are never asked to cover it. You are asked to narrow it: take one borderless problem and compress it into a single, sharp, evidence-backed policy argument. Essay = argue one position; presentation = pitch one intervention; case study = analyse one bounded case. Markers reward a tight scope and a defended claim, and punish broad description. Every chapter that follows gives you the concepts, regions and cases to make that one argument land.[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。
• A2 Pitch（25%｜3分钟｜3页 slides）：你要“扮演政策制定者”，点名一个具体监管问题 + 推介一个干预（intervention）；而且它会直接“喂给”A3。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构[7]Source: asksia-bible-mecm90002-bilingual.pdfSlide Job What goes on it Time The case, broadly Introduce the case / context - one platform, country or instrument - and why it matters. One image, ≤5 lines. ~45s 2 The specific issue Zoom in to the one regulatory problem you will fix. Name it precisely; ground it in a real incident. ~60s 3 The policy angle Pitch the intervention + a first-pass solution. Who acts (state / platform / IGO / civil society)? ~75s What lever? First-cut, not final. ! Reading your slides verbatim = mark downgrade 逐字照念你的幻灯片=降分 The brief explicitly penalises reading slides word-for- word. Slides carry keywords and one visual; you carry the argument. Rehearse to the clock - three minutes is brutally short, so each slide is one idea, spoken in plain sentences while the slide shows the anchor. 作业说明明确惩罚逐字念幻灯片。幻灯片承载关键词 和一张图;你承载论点。对着钟表去排练 -- 三分钟 极其短暂,所以每张幻灯片只讲一个想法,在幻灯片 展示锚点的同时,用平实的句子把它说出来。 i The policymaker frame 决策者视角框架 Speak as the regulator, not about regulation: "As an adviser to the ACCC, I propose . . . ". Choosing the actor (national government, platform, an IGO like the OECD/ITU, or civil society) forces clarity about who has the lever - the exact judgement A3 will test in depth. 要作为监管者去说,而不是谈论监管:“作为 ACCC 的一名顾问,我提议, . . . . . ”。选择行动者(国家政府、 平台、像 OECD/ITU 这样的 IGO,或公民社会)会迫 使你想清楚谁握有杠杆 -- 这正是A3 将深入考查的那 个判断。 MECM90002 . Global Data Policy & Governance ✓ Pace it: 3 minutes is ~390 words 把控节奏:3 分钟约 390 字 At a calm speaking rate you have room for roughly 390-420 spoken words total. Write a short script, time it twice, and cut until it fits with a few seconds' buffer. Practise the transition lines ("so the specific problem is . . . ", "my intervention is . . . ") - that is where nervous over-runs happen. 以平稳的语速,你总共大约有390-420 个口述词的余 地。写一个短脚本,计时两遍,删到能在留出几秒缓 冲的情况下讲完为止。练习那几句过渡语(“所以具体 的问题是 ………. ”‘ ”“我的干预是 . . . . . ") 一紧张超时往往 就发生在那里。 ★ Non-presenters: peer feedback is part of the task 非展示者:同伴反馈也是任务的一部分 When you are not presenting you are expected to give structured peer feedback. Listen for the move this chapter teaches: did they narrow to one problem and one fix, or drift into a survey? That feedback is gold for your own A3. 当你不是在做报告时,会被要求给出有结构的同伴反 馈。注意听本章所教的那个动作:他们是否收窄到了 一个问题与一个对策,还是飘成了一篇概述?这份反 馈对你自己的A3而言是金子。[12]Source: asksia-bible-mecm90002-bilingual.pdfWeight Length Due / timing A1 . Argumentative essay 50% 2500w 19 Apr A2 . Oral presentation 25% 3 min / 3 slides Wk 10-11 A3 . Written case study 25% 1500w 7 Jun A1: pick one of three set questions (national policy rationale · UN- style transnational brief . platform responsibility); ~ 20 references expected, peer-reviewed + policy/report/news. A2: step into a policymaker's shoes and pitch a policy intervention - it directly feeds A3. A3: develops that same intervention into an in-depth, single-case study (a company/practice, a country's policy or its absence, or one international instrument). A1: 从三道既定题目中选一道(国家政策依据 · UN 风格的 跨国简报 · 平台责任);预期约 20 篇参考文献,须包含同 行评审文献+政策 /报告/新闻。A2:设身处地成为一名决 策者,推介一项政策干预 -- 它直接为 A3 提供素材。A3: 将同一项干预深化为一份深入的单一案例研究(一家公司/ 一种做法、某国的政策或政策的缺位,或一项国际工具)。 The twohurdles&the AIrule 两道门槛与 AI 规则 ! Hurdles - clear both or you cannot pass 门槛 -- 两道都过,否则无法通过 Attendance: a minimum 80% of tutorials/seminars, monitored; falling below without documented evidence means you cannot pass. Submission: every piece must be submitted as a genuine attempt. Late work loses 5% of available marks per day without an approved extension. 出勤:导修课/研讨课出勤率至少为 80%,并受监 控;若无书面证明而低于此线,将无法通过。提交: 每一项作业都必须作为真实尝试提交。未获批准延期 的迟交,每天扣除可得分数的5%。 MECM90002 . Global Data Policy & Governance i AI policy in brief (MPF1310) AI 政策速览 (MPF1310) All submitted work must be your own. AI is permitted to define hard terms, locate or summarise sources for comprehension, and check spelling/grammar - with disclosure. It is not permitted to write or rephrase your sentences, to be cited as a source, or to translate your assignment. Turnitin auto-scans for AI; undeclared use triggers an integrity investigation. 所有提交的作业都必须是你本人完成的。AI 可以用来 解释难懂的术语、为帮助理解而查找或概括来源、检 查拼写与语法 -- 但必须声明。Al 不可以用来撰写或 改写你的句子、作为来源被引用,或翻译你的作业。 Turnitin 会自动扫描 AI 痕迹;未申报的使用会触发学 术诚信调查。 The one strategic insight behind every task ★ 贯穿每项任务背后的那一个战略洞见 The whole subject is a tour of a huge, global debate - but you are never asked to cover it. You are asked to narrow it: take one borderless problem and compress it into a single, sharp, evidence-backed policy argument. Essay = argue one position; presentation = pitch one intervention; case study = analyse one bounded case. Markers reward a tight scope and a defended claim, and punish broad description. Every chapter that follows gives you the concepts, regions and cases to make that one argument land.[15]Source: asksia-cheatsheet-mecm90002.pdfConcl. - restate . synthesise . "why it matters" Narrow the huge question to one sharp, defensible scope (one region, one law, one platform) - breadth is the #1 trap. 16 . A2 . Presentation 25% . ~ 3 MIN . 3 SLIDES Be a policymaker: name a specific regulatory problem & pitch an intervention. Slide 1 the case (broad) . Slide 2 the specific issue . Slide 3 the policy angle + first solutions. Don't read slides verbatim (mark penalty). Feeds directly into A3. 15b . Essay Traps AVOID THESE · Too broad - "regulate the internet" can't be argued in 2500w · Theory-dropping - naming Zuboff without using her · One region when the question invites comparison 16b . Pick Your Question ALIGN A1+A2+A3 Choose the essay question whose case you can reuse across all three tasks - one well-chosen case (a platform, a country, an IGO) can power the essay, the pitch and the case study. Pick where you have real evidence & a clear tension, not the "easiest" topic. 16c . Slide Design A2 TIPS 3 slides, ~1 idea each - keep them visual, not wordy (reading text aloud is penalised). One clear problem statement, one concrete intervention, one "why it's feasible. " Rehearse to land inside 3 minutes; the pitch is a dry-run of your A3 thesis. Practice once out loud - 3 minutes is shorter than it sounds. Compiled by AskSia . mapped to the MECM90002 curriculum . asksia. ai/cheatsheet/unimelb-mecm90002 17 . A3 · Case Study 25% . 1500W Develop the A2 idea into a rigorous study of one tightly-bounded case. Pick a lane: · Topic 1 - a company / practice / tech field x a policy dimension (e. g. a platform x misinfo moderation) · Topic 3 - a specific international policy/body (e. g. the UN Global Digital Compact) 3-PART STRUCTURE
• A3 Case study（25%｜1500w）：把 A2 的同一项干预，打磨成一个边界非常清晰的单一案例研究（更深，不更广）。[3]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.[12]Source: asksia-bible-mecm90002-bilingual.pdfWeight Length Due / timing A1 . Argumentative essay 50% 2500w 19 Apr A2 . Oral presentation 25% 3 min / 3 slides Wk 10-11 A3 . Written case study 25% 1500w 7 Jun A1: pick one of three set questions (national policy rationale · UN- style transnational brief . platform responsibility); ~ 20 references expected, peer-reviewed + policy/report/news. A2: step into a policymaker's shoes and pitch a policy intervention - it directly feeds A3. A3: develops that same intervention into an in-depth, single-case study (a company/practice, a country's policy or its absence, or one international instrument). A1: 从三道既定题目中选一道(国家政策依据 · UN 风格的 跨国简报 · 平台责任);预期约 20 篇参考文献,须包含同 行评审文献+政策 /报告/新闻。A2:设身处地成为一名决 策者,推介一项政策干预 -- 它直接为 A3 提供素材。A3: 将同一项干预深化为一份深入的单一案例研究(一家公司/ 一种做法、某国的政策或政策的缺位,或一项国际工具)。 The twohurdles&the AIrule 两道门槛与 AI 规则 ! Hurdles - clear both or you cannot pass 门槛 -- 两道都过,否则无法通过 Attendance: a minimum 80% of tutorials/seminars, monitored; falling below without documented evidence means you cannot pass. Submission: every piece must be submitted as a genuine attempt. Late work loses 5% of available marks per day without an approved extension. 出勤:导修课/研讨课出勤率至少为 80%,并受监 控;若无书面证明而低于此线,将无法通过。提交: 每一项作业都必须作为真实尝试提交。未获批准延期 的迟交,每天扣除可得分数的5%。 MECM90002 . Global Data Policy & Governance i AI policy in brief (MPF1310) AI 政策速览 (MPF1310) All submitted work must be your own. AI is permitted to define hard terms, locate or summarise sources for comprehension, and check spelling/grammar - with disclosure. It is not permitted to write or rephrase your sentences, to be cited as a source, or to translate your assignment. Turnitin auto-scans for AI; undeclared use triggers an integrity investigation. 所有提交的作业都必须是你本人完成的。AI 可以用来 解释难懂的术语、为帮助理解而查找或概括来源、检 查拼写与语法 -- 但必须声明。Al 不可以用来撰写或 改写你的句子、作为来源被引用,或翻译你的作业。 Turnitin 会自动扫描 AI 痕迹;未申报的使用会触发学 术诚信调查。 The one strategic insight behind every task ★ 贯穿每项任务背后的那一个战略洞见 The whole subject is a tour of a huge, global debate - but you are never asked to cover it. You are asked to narrow it: take one borderless problem and compress it into a single, sharp, evidence-backed policy argument. Essay = argue one position; presentation = pitch one intervention; case study = analyse one bounded case. Markers reward a tight scope and a defended claim, and punish broad description. Every chapter that follows gives you the concepts, regions and cases to make that one argument land.[19]Source: asksia-cheatsheet-mecm90002.pdf· Too broad - "regulate the internet" can't be argued in 2500w · Theory-dropping - naming Zuboff without using her · One region when the question invites comparison 16b . Pick Your Question ALIGN A1+A2+A3 Choose the essay question whose case you can reuse across all three tasks - one well-chosen case (a platform, a country, an IGO) can power the essay, the pitch and the case study. Pick where you have real evidence & a clear tension, not the "easiest" topic. 16c . Slide Design A2 TIPS 3 slides, ~1 idea each - keep them visual, not wordy (reading text aloud is penalised). One clear problem statement, one concrete intervention, one "why it's feasible. " Rehearse to land inside 3 minutes; the pitch is a dry-run of your A3 thesis. Practice once out loud - 3 minutes is shorter than it sounds. Compiled by AskSia . mapped to the MECM90002 curriculum . asksia. ai/cheatsheet/unimelb-mecm90002 17 . A3 · Case Study 25% . 1500W Develop the A2 idea into a rigorous study of one tightly-bounded case. Pick a lane: · Topic 1 - a company / practice / tech field x a policy dimension (e. g. a platform x misinfo moderation) · Topic 3 - a specific international policy/body (e. g. the UN Global Digital Compact) 3-PART STRUCTURE 1 describe case + why it must change 2 define the problem w/ real incidents 3 propose a concrete, targeted intervention Graded on: problem & context · intervention design . critical analysis > description . writing & citing. Case study # essay: the essay argues a position; A3 diagnoses one case & designs a fix. Go deeper, not broader - figures/tables don't count to the word limit. 18c . Evidence Checklist A3[26]Source: asksia-cheatsheet-mecm90002.pdf1 describe case + why it must change 2 define the problem w/ real incidents 3 propose a concrete, targeted intervention Graded on: problem & context · intervention design . critical analysis > description . writing & citing. Case study # essay: the essay argues a position; A3 diagnoses one case & designs a fix. Go deeper, not broader - figures/tables don't count to the word limit. 18c . Evidence Checklist A3 · The law/policy text itself (primary source) · Real incidents - a fine, a breach, a court case · 2-3 scholars framing the debate · A counter-view you then answer 18 . Choosing a CaseSET YOURSELF UP Narrow - one figure/law/event, never "social media" Has a visible tension (sovereignty vs free flow; rights vs market) · 2-3 set scholars genuinely apply · Concrete evidence you can cite (law text, reports, coverage) 18b . Worked . Mini Case THE SHAPE Case: Australia's News Media Bargaining Code. Problem: platform market power starves journalism funding. Evidence: Google paid up; Meta "went dark," signed, then exited news (2024). Intervention: mandatory bargaining + final-offer arbitration. So what: a state forcing platform accountability - but is it durable when Meta can just drop news? 18d . "Targeted" Means A3 KEY WORD Grading rewards a narrow, feasible, specific intervention: "regulate Big Tech" fails; "extend the bargaining code's arbitration to AI-training licensing" passes. Name the actor, the mechanism, and why it's enforceable. Then pre-empt the obvious objection - feasibility & unintended effects - in a sentence.
• 复习策略：选一个“你愿意写三遍”的同一案例/同一干预，让 A1 的阅读为 A2/A3 铺路（这就是这门课的“省命打法”）。[3]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.[15]Source: asksia-cheatsheet-mecm90002.pdfConcl. - restate . synthesise . "why it matters" Narrow the huge question to one sharp, defensible scope (one region, one law, one platform) - breadth is the #1 trap. 16 . A2 . Presentation 25% . ~ 3 MIN . 3 SLIDES Be a policymaker: name a specific regulatory problem & pitch an intervention. Slide 1 the case (broad) . Slide 2 the specific issue . Slide 3 the policy angle + first solutions. Don't read slides verbatim (mark penalty). Feeds directly into A3. 15b . Essay Traps AVOID THESE · Too broad - "regulate the internet" can't be argued in 2500w · Theory-dropping - naming Zuboff without using her · One region when the question invites comparison 16b . Pick Your Question ALIGN A1+A2+A3 Choose the essay question whose case you can reuse across all three tasks - one well-chosen case (a platform, a country, an IGO) can power the essay, the pitch and the case study. Pick where you have real evidence & a clear tension, not the "easiest" topic. 16c . Slide Design A2 TIPS 3 slides, ~1 idea each - keep them visual, not wordy (reading text aloud is penalised). One clear problem statement, one concrete intervention, one "why it's feasible. " Rehearse to land inside 3 minutes; the pitch is a dry-run of your A3 thesis. Practice once out loud - 3 minutes is shorter than it sounds. Compiled by AskSia . mapped to the MECM90002 curriculum . asksia. ai/cheatsheet/unimelb-mecm90002 17 . A3 · Case Study 25% . 1500W Develop the A2 idea into a rigorous study of one tightly-bounded case. Pick a lane: · Topic 1 - a company / practice / tech field x a policy dimension (e. g. a platform x misinfo moderation) · Topic 3 - a specific international policy/body (e. g. the UN Global Digital Compact) 3-PART STRUCTURE[19]Source: asksia-cheatsheet-mecm90002.pdf· Too broad - "regulate the internet" can't be argued in 2500w · Theory-dropping - naming Zuboff without using her · One region when the question invites comparison 16b . Pick Your Question ALIGN A1+A2+A3 Choose the essay question whose case you can reuse across all three tasks - one well-chosen case (a platform, a country, an IGO) can power the essay, the pitch and the case study. Pick where you have real evidence & a clear tension, not the "easiest" topic. 16c . Slide Design A2 TIPS 3 slides, ~1 idea each - keep them visual, not wordy (reading text aloud is penalised). One clear problem statement, one concrete intervention, one "why it's feasible. " Rehearse to land inside 3 minutes; the pitch is a dry-run of your A3 thesis. Practice once out loud - 3 minutes is shorter than it sounds. Compiled by AskSia . mapped to the MECM90002 curriculum . asksia. ai/cheatsheet/unimelb-mecm90002 17 . A3 · Case Study 25% . 1500W Develop the A2 idea into a rigorous study of one tightly-bounded case. Pick a lane: · Topic 1 - a company / practice / tech field x a policy dimension (e. g. a platform x misinfo moderation) · Topic 3 - a specific international policy/body (e. g. the UN Global Digital Compact) 3-PART STRUCTURE 1 describe case + why it must change 2 define the problem w/ real incidents 3 propose a concrete, targeted intervention Graded on: problem & context · intervention design . critical analysis > description . writing & citing. Case study # essay: the essay argues a position; A3 diagnoses one case & designs a fix. Go deeper, not broader - figures/tables don't count to the word limit. 18c . Evidence Checklist A3

0.2 两道硬门槛（别在 final 前翻车）

• 出勤门槛：导修/研讨课出勤 ≥80%，否则可能无法通过。[11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.[18]Source: asksia-cheatsheet-mecm90002.pdfAskSia STUDY SHEET SERIES 100% assignment-assessed . check your current subject outline . @ 2026 good luck. write sharp. KNOW THE TRADE SIDE 2/2 APPLY & WRITE . IGOS & interventions . The future . Scholar bank . Case bank . The essay . Presentation . Case study . Argument craft . Integrity 100% ASSIGNMENT . NO EXAM MECM90002 Global Data Policy & Governance UNIVERSITY OF MELBOURNE . MEDIA & COMMUNICATIONS STUDY & ESSAY TOOLKIT Sem 1 2026 . SIDE 1 OF 2 100% assignment-assessed · no exam SIDE 1/2 · Latin America 0 · How to Use This READ FIRST * This subject is graded entirely by writing & a pitch - no exam : an argumentative essay (50%, 2500w), an oral presentation (25%, ~3 min . 3 slides), and a written case study (25%, 1500w). Hurdles: ≥80% tutorial attendance + submit every piece. The marked skill is narrowing a huge global debate into one sharp, evidence-backed policy argument. So this sheet is a thinking + citing bank: the toolkit + the regional map (Side 1), the scholars, cases & writing engine (Side 2). SIA + The whole subject is one move: pick a region/case, name a specific problem, argue a targeted intervention. Description loses marks; analysis ("so what for power, rights, the public interest?") wins them. 1 . The Big Picture W1-2 Old media (TV, radio, print) was bounded & territorial - easy for a nation to regulate. Today's data is the fluid data ecology: it sits on servers anywhere, constantly crosses borders, and it's unclear where it's stored or who can access it (cloud, IoT, 5G). That fluidity breaks nation-bound regulation - the core problem of the whole subject. Governance # government. Government = the state's law over its territory. Governance = multi-actor steering by states + IGOs (UN/ITU/OECD) + blocs (EU) + platforms + civil society. The course's spine = a spectrum: EU rights-based + US/UK/AU market-based . Global South (Africa/Asia/LatAm) navigating development, sovereignty & platform power. 1b . How the Subject Maps 3 PARTS
• 提交门槛：每个作业都必须作为“真实尝试”提交；迟交会每天扣分。[1]Source: asksia-bible-mecm90002-bilingual.pdf核对 AI 给出的想法是否经得起可靠来源的检验 · 用 Al 来解释或定义难懂的术语 · 帮助查找来源 · 为你自 己的理解而对学术来源进行概括或翻译 · 非 AI 的拼写 / 语法工具(关闭 AI 功能的 Grammarly / Hemingway)。所有这些都必须在你的提交中声明。 ! Not permitted - ever 绝不允许 Al writing whole sentences / paragraphs / papers . presenting AI output as your own research or analysis · citing Al output as a source . Al rewriting / rephrasing / 'improving' your text . Al translating your assignment into English. A chatbot is a next- word predictor that hallucinates - treated like paying a ghost-writer. Al 撰写整句/ 整段/整篇论文 · 把 AI 的输出当作你 自己的研究或分析呈现 · 把 AI 的输出作为来源引用 · 让 Al 改写/重述/“润色”你的文字 · 让 AI 把你的作 业翻译成英文。聊天机器人是一个会胡编的下一词预 测器 -- 会被等同于花钱雇了一个代笔人。 MECM90002 . Global Data Policy & Governance ! Late penalty is real and compounding 迟交罚分真实存在且会累积 5% of available marks per day (or part-day) without an approved extension; not accepted at all after 20 University business days. Extensions ≤10 working days are requested by email to the coordinator/tutor; longer needs Special Consideration. Submit early enough to survive an upload glitch. 未获批准延期者,每天(或不足一天按一天计)扣除 可得分数的5%;超过20个大学工作日后将完全不予 接受。≤10个工作日的延期,通过电子邮件向课程协 调员/导师申请;更长的则需 Special Consideration。要提早提交,以便挺过一次上传故 障。 ★ Declare it - every time, even for permitted use 声明它 -- 每一次都要,即便是被允许的使用 Any AI use at any stage must be declared in the submission (the University links disclosure templates). Undeclared or out-of-scope use triggers an academic-integrity investigation. The safe habit: keep a one-line log as you go ("used ChatGPT to define 'data localisation', verified against Kuner") and paste it into the declaration. Disclosure of permitted use costs you nothing; the alternative can cost you the subject. 在任何阶段对 AI的任何使用,都必须在提交中声明(大学提供了披露模板的链接)。未声明或超出范围的使用会触发学 术诚信调查。稳妥的习惯:边做边记一行日志(“用 ChatGPT 定义‘data localisation’,并对照 Kuner 核实”),再粘贴 进声明里。披露被许可的使用对你毫无代价;另一种做法则可能让你赔上这门课。 All submitted work must be your own. Permitted Al helps you understand and find - never write, rephrase, translate, or be cited. Declare every use. 所有提交的作业都必须是你本人完成的。被许可的 AI 帮你理解与查找 -- 绝不替你撰写、改写、翻译或被引用。每一次 使用都要声明。 MPF1310 . STUDENT ACADEMIC INTEGRITY MECM90002 . Global Data Policy & Governance CH 11 . NARROWING DRILLS - CHAPTER 11 . PRACTICE BANK NO EXAM . DRILL THE CRAFT Drill the writing, with model answers 操练写作,附范文 Ten AskSia-authored items in the subject's style - narrowing, models, case planning, concepts 十道由 AskSia 按本课程风格撰写的题目 -- 收窄、模式、案例规划、概念 There is no exam to cram, so these drills target the moves the three assessments reward: narrowing a huge question, naming the right model and where it breaks, planning a case study, and the concept checks that keep your analysis honest. Work each one cold, then read the model answer and the marker's-eye note. 这里没有需要临时抱佛脚的考试,所以这些练习针对的是三项评估所奖励的招式:收窄一个宏大问题、点出正确的模型及其失 效之处、规划一份案例研究,以及那些让你的分析保持诚实的概念检查。每一道都先冷启动独立做完,再去读范例答案与评卷 人视角的批注。 ★ How to use this bank 如何使用这个题库 Cover the answer. These are skill drills, not facts to memorise. Write or speak your response first, then compare. The narrowing drills are the highest-value - if you can reliably cut a planet-sized question to one defensible claim, you have the whole subject. These items are AskSia's own, built in the subject's style; they paraphrase and transform the real prompts - never reproduce them. Always answer the official brief on Canvas. 把答案盖住。这些是技能演练,不是要背的事实。先 写出或说出你的回答,再对照。收窄类演练是价值最 高的 -- 如果你能可靠地把一个庞大无边的问题切到 一个可辩护的主张,你就掌握了整门课。这些题目是 AskSia 自创的,按本课风格构建;它们对真实题目做 了转述与改造 -- 绝不照搬。请始终回答 Canvas 上 的官方说明。[11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.

1）本课“唯一核心技能”（也是你复习时最该练的）：收窄 + 论证（别写成综述）

1.1 评分者最怕你做什么：只描述、不论证

• 你会丢分的典型错误：写成“讲了很多事实”的描述/综述，而不是“提出主张并证明”的论证。[2]Source: asksia-bible-mecm90002-bilingual.pdf3 · 论证 An assessment is due. Go to Ch 10 (writing the essay, presentation & case study) and the glossary. The single skill rewarded everywhere is narrowing a huge question into one specific, evidence-backed claim. This bible is built to help you do exactly that. 临近交作业的时候。请直接看第 10 章(写论文、口头报告与案例 研究)以及术语表。这门课处处 奖励的唯一一项技能,就是把一 个庞大的问题收窄成一个具体、 有证据支撑的主张。这本 bible 的全部设计,就是帮你做到这一 点。 MECM90002 . Global Data Policy & Governance ! The mistake that sinks marks: describing instead of arguing 拖垮分数的错误:只描述而不论证 Every task here is argumentative or analytical, never a summary. The essay must take a position and defend it; the case study must analyse one tightly-bounded case, not narrate it; the presentation must propose an intervention. Students lose marks by surveying "everything about data governance. " The fix is on the next page and in Ch 10: pick one country / platform / instrument, name one problem, defend one intervention. 这里的每一项任务都是论辩型或分析型的,绝不是简单的概述。论文必须提出一个立场并为之辩护;案例研究必须分析 一个边界清晰的案例,而不是叙述它;口头报告必须提出一项干预措施。学生失分往往是因为去通盘罗列“关于数据治理 的一切”。解决办法在下一页和第10 章里:挑一个国家/平台/法律工具,点明一个问题,为一项干预辩护。 i How this bible was built - and the two-layer rule 这本 bible 是如何构建的 -- 以及双层规则 Grounded entirely in this subject's own materials: the subject outline, all twelve week pages, the lecture and seminar slide decks, and the assessment briefs. Standard facts - public laws (GDPR, DSA, DMA, the Al Act), named theories, real cases - are stated plainly; the lecturer's specific framings and examples are paraphrased, never copied. Always verify dates and weights against your own Canvas before submitting. 本书完全立足于这门课自己的材料:课程大纲、全部十二周的周页面、讲座与研讨课的幻灯片,以及各项作业说明。标 准事实 -- 公开的法律(GDPR、DSA、DMA、Al Act)、有名的理论、真实的案例一直接平实地陈述;讲师特有的框 架与例子则一律转述,绝不照搬。提交前请务必对照你自己的 Canvas 核实日期与权重。 MECM90002 . Global Data Policy & Governance THE BLUEPRINT - THE ASSESSMENT BLUEPRINT 100% WRITTEN & ORAL No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, two hurdles, and one strategic move that decides your grade 三项任务、两道门槛,以及决定你成绩的一个战略动作 MECM90002 has no sit-down exam. Your entire mark comes from an argumentative essay, a short policymaker presentation and a written case study - with two hurdles you must clear regardless of those marks. Read this page first; it is the map for everything else. MECM90002 没有闭卷考试。你的全部成绩来自一篇论证性论文、一段简短的决策者展示,以及一份书面案例研究 -- 此外 还有两道无论分数如何都必须通过的及格门槛(hurdle)。请先读这一页;它是其余一切内容的地图。 50% ESSAY (2500W) 论文(2500 字) 25% PRESENTATION 口头报告 25% CASE STUDY (1500W) 案例研究 (1500 字) EXAM - NONE 考试 -- 无 The three assessed tasks 三项受评任务 Task[3]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。
• 材料里把这句话讲得很直白：描述告诉读者发生了什么；论点告诉读者该怎么想、以及为什么。[3]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。
• 复习时的自检问题（很关键）：
  • “我这一段有没有在推进一个可争辩的主张？”
  • “有没有一句话在做 This matters because...（这很重要因为……）？”
  • 如果没有，那大概率你在“写背景材料”，不是在得分。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构

1.2 这门课的“战略洞见”：永远不是覆盖全球，而是把它做小

• 你被要求做的是：把一个“无国界的大问题”压缩成一个具体、可防守、证据支撑的政策主张。[10]Source: asksia-bible-mecm90002-bilingual.pdf整门课是对一场宏大而全球性的辩论的巡览 -- 但从来没有要求你去覆盖它的全部。要求你做的是把它收窄:取一个无 国界的问题,把它压缩成一个单一、犀利、有证据支撑的政策论点。论文=论证一个立场;口头报告=推介一项干预; 案例研究=分析一个边界清晰的案例。评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的描述。后面每一章都会为你 提供概念、地区与案例,让你把那一个论点落到实处。 MECM90002 . Global Data Policy & Governance CONTENTS CONTENTS The whole field, in one ordered companion 整个领域,一本循序渐进的伴读手册 Toolkit - debates - world tour - the writing that is graded 工具箱→辩论→ 全球巡览→ 被评分的写作 Ch Chapter Core idea Part 1 . The toolkit & the debates 1 The conceptual toolkit datafication · fluid data ecology · sovereignty → 2 Regulation models & key debates self- vs state- vs co-regulation → Part 2 . A region-by-region tour 3 Europe / GDPR the 'data subject' . DSA . DMA . AI Act → 4 USA, UK & Australia neoliberalism · antitrust · selective intervention → 5 Africa development comms · mobile spheres → 6 Asia government control vs commercialisation → 7 Latin America post-colonial media . conglomerates → 8 IGOs, interventions & the future[12]Source: asksia-bible-mecm90002-bilingual.pdfWeight Length Due / timing A1 . Argumentative essay 50% 2500w 19 Apr A2 . Oral presentation 25% 3 min / 3 slides Wk 10-11 A3 . Written case study 25% 1500w 7 Jun A1: pick one of three set questions (national policy rationale · UN- style transnational brief . platform responsibility); ~ 20 references expected, peer-reviewed + policy/report/news. A2: step into a policymaker's shoes and pitch a policy intervention - it directly feeds A3. A3: develops that same intervention into an in-depth, single-case study (a company/practice, a country's policy or its absence, or one international instrument). A1: 从三道既定题目中选一道(国家政策依据 · UN 风格的 跨国简报 · 平台责任);预期约 20 篇参考文献,须包含同 行评审文献+政策 /报告/新闻。A2:设身处地成为一名决 策者,推介一项政策干预 -- 它直接为 A3 提供素材。A3: 将同一项干预深化为一份深入的单一案例研究(一家公司/ 一种做法、某国的政策或政策的缺位,或一项国际工具)。 The twohurdles&the AIrule 两道门槛与 AI 规则 ! Hurdles - clear both or you cannot pass 门槛 -- 两道都过,否则无法通过 Attendance: a minimum 80% of tutorials/seminars, monitored; falling below without documented evidence means you cannot pass. Submission: every piece must be submitted as a genuine attempt. Late work loses 5% of available marks per day without an approved extension. 出勤:导修课/研讨课出勤率至少为 80%,并受监 控;若无书面证明而低于此线,将无法通过。提交: 每一项作业都必须作为真实尝试提交。未获批准延期 的迟交,每天扣除可得分数的5%。 MECM90002 . Global Data Policy & Governance i AI policy in brief (MPF1310) AI 政策速览 (MPF1310) All submitted work must be your own. AI is permitted to define hard terms, locate or summarise sources for comprehension, and check spelling/grammar - with disclosure. It is not permitted to write or rephrase your sentences, to be cited as a source, or to translate your assignment. Turnitin auto-scans for AI; undeclared use triggers an integrity investigation. 所有提交的作业都必须是你本人完成的。AI 可以用来 解释难懂的术语、为帮助理解而查找或概括来源、检 查拼写与语法 -- 但必须声明。Al 不可以用来撰写或 改写你的句子、作为来源被引用,或翻译你的作业。 Turnitin 会自动扫描 AI 痕迹;未申报的使用会触发学 术诚信调查。 The one strategic insight behind every task ★ 贯穿每项任务背后的那一个战略洞见 The whole subject is a tour of a huge, global debate - but you are never asked to cover it. You are asked to narrow it: take one borderless problem and compress it into a single, sharp, evidence-backed policy argument. Essay = argue one position; presentation = pitch one intervention; case study = analyse one bounded case. Markers reward a tight scope and a defended claim, and punish broad description. Every chapter that follows gives you the concepts, regions and cases to make that one argument land.
• 官方写作引擎（你要背下来的复习口诀）：
  • 挑一个行动者（国家/平台/IGO 等）
  • 点名一个工具或事件（某法律、某事故、某判例、某政策文本）
  • 只选一个维度（隐私/竞争/内容/安全……）
  • 然后形成一条可争辩的 thesis。[14]Source: asksia-bible-mecm90002-bilingual.pdfi The two recurring moves 两个反复出现的操作 Move 1 - narrow: add one actor + one instrument/incident + one dimension until the scope fits 2500 (essay) or 1500 (case) words. Move 2 - analyse: name a model from Ch 1-2 (sovereignty/jurisdiction; self/state/co-regulation; rights/market/state-control) and say where it breaks on your case. Every drill below trains one of these. 第 1招 -- 收窄:不断加上一个行动者+一项工具 / 事件+一个维度,直到范围适配 2500 字(论文)或 1500 字(案例)。第2招 -- 分析:点出第1-2 章的 一个模型(主权/管辖权;自我/国家/共同监管; 权利/市场/国家管控),并说出它在你的案例上从何 处失效。下面每一道演练都训练其中之一。 D1 NARROWING DRILL essay scope A classmate's draft essay opens: "This essay examines how data sovereignty is understood and applied by governments around the world, with reference to many countries and laws. " Rewrite this as a single narrowed thesis suitable for a 2500-word argument, and name the one model you would use as the lens. 一位同学的论文初稿这样开头:“本文考察世界各国政府如何理解和运用数据主权,并参照众多国家与法律。”把它重写为 一个适用于 2500 字论证的、收窄过的单一论点,并点出你将用作透镜的那一个模型。 MODEL ANSWER Model answer. "Russia's 2019 'Sovereign Internet' data-localisation regime is best read not as data protection but as a tool of state control - and that conflation is why national digital sovereignty, however legitimate elsewhere, fails as a justification here. " Lens: Couture & Toupin's plural sovereignty, contrasting national vs state-control framings. MECM90002 . Global Data Policy & Governance - 范例答案。“俄罗斯 2019年的‘主权互联网’数据本地化体制,最好不被解读为数据保护,而被解读为一种国家管控工具 -- 而 正是这种混淆,使得国家数字主权(无论在别处多么正当)在此处无法成立为正当理由。”透镜:Couture 与 Toupin 的多元 主权,对照国家(national) 与国家管控(state-control)两种框架表述。 Marker's eye. The fix is one country, one instrument, one dimension, and a contestable claim. "Many countries and laws" guarantees a survey; markers cap survey essays well below a defended argument no matter how much is 'covered. ' 评卷人视角。修正之道是一个国家、一项工具、一个维度,外加一个可争辩的主张。“众多国家与法律”注定写成综述;无论 “覆盖”了多少,评卷人都会把综述类论文的分数压得远低于一篇有立场可辩护的论证。 D2 NARROWING DRILL case-study scope Too broad for a 1500-word case study: "Platforms should do more to stop misinformation. " Narrow it to a targeted intervention with a named actor and lever, grounded in a real incident. 对一篇1500 字的案例研究而言过于宽泛:“平台应当采取更多措施来阻止错误信息。”把它收窄为一项有的放矢的干预, 带有指名道姓的行为主体与杠杆,并扎根于一桩真实事件。 MODEL ANSWER Model answer. "Following the failures exposed by Rohingya v Facebook, Meta should be designated a 'gatekeeper' under a DMA-style threshold and subjected to independent, audited reporting of its misinformation-moderation outcomes in at-risk markets - a co-regulatory duty, not a vague call for 'responsibility' . " 范例答案。“鉴于 Rohingya v Facebook 所暴露的种种失败,Meta 应被依据一道 DMA 式的门槛指定为‘守门人’,并就其在 高风险市场中错误信息审核结果,接受独立的、经审计的报告 -- 这是一项共同监管义务,而非对‘责任’的含糊呼吁。” Marker's eye. "Targeted" is the graded word. Look for a named actor (Meta), a named lever (audited reporting under a designation threshold) and a real anchor (Rohingya v Facebook). "Do more" scores nothing. 评卷人视角。“有的放矢(targeted)”是那个被计分的词。要看到指名道姓的行为主体(Meta)、指名道姓的杠杆(在指定门 槛下经审计的报告)以及一个真实的锚点(Rohingya v Facebook)。“采取更多措施”一分不得。 MECM90002 . Global Data Policy & Governance CH 11 . ESSAY VARIANTS - CHAPTER 11 . ESSAY - QUESTION VARIANTS A1 REHEARSAL . ARGUE Practise the three lanes on fresh prompts 用新题目练习三条赛道 AskSia-written variants that transform the set questions - pick one, draft a thesis 由 AskSia 撰写、改造指定题目的变体 -- 选一道,起草一个论点 These are our own question variants, built to rehearse the same three lanes as the real essay (national rationale · transnational brief · platform responsibility) without reproducing the official prompts. For each, write the one-sentence thesis and the model you would deploy. Treat them as timed planning sprints (10 minutes each).[16]Source: asksia-bible-mecm90002-bilingual.pdf2 Test the scope against the word count. 2500 words is roughly 4-5 body paragraphs. If your scope needs ten paragraphs to cover, it is still too broad - cut again. 用字数来检验范围。2500 字大约是4-5个主体段落。如 果你的范围要用十个段落才覆盖得了,那它仍然太宽 -- 再 砍。 3 Convert the narrowed scope into a thesis - a single sentence that takes a contestable position. If a reasonable classmate couldn't disagree with it, it is description, not a thesis. 把收窄后的范围转化为论点 -- 一个采取可争议立场的单 句。如果一个讲道理的同学无法对它表示异议,那它就是描 述,而不是论点。 4 Deploy a named model from Ch 1-2 as your analytical lens (sovereignty vs jurisdiction; self- vs state- vs co-regulation; rights- vs market- vs state-control). The lens is what turns your case into analysis. 部署一个第1-2 章的具名模型作为你的分析视角(主权 vs 管辖权;自我 vs 国家 vs 共同监管;权利 vs 市场 vs 国家 管控)。正是这个视角把你的案例变成分析。 ✓ Thesis templates (fill the blanks) 论点模板(填空使用) Stance + because + scope: "[Country/Platform] should [do X], because [reason grounded in a model], as the case of [instrument/incident] shows. " Tension form: "Although [the borderless counter- argument], [your actor] is justified in [its policy] because [model + evidence] - though only where [the limit you concede]. " The conceded limit signals critical engagement and stops the essay reading as one-sided. 立场+因为+范围:“[国家/平台]应当[做 X],因为 [基于某个模型的理由],正如[工具/事件]这一案例所 示。” 张力句式:“尽管[那条无国界的反方论点],[你的行动 者]仍有理由[采取其政策],因为[模型+证据] -- 尽管 只在[你所让步的限度]之内。” 所让步的限度表明你具有批判性的投入,并能阻止论 文读起来一边倒。 ! Scope creep is the No. 1 essay killer 范围蔓延是头号论文杀手 "Should the internet be regulated?" is not an essay - it is a degree. Narrow until the question is answerable in 2500 words with ~20 sources: one country, one instrument, one dimension, one defended claim. “互联网应当被监管吗?”不是一篇论文 -- 它是一个 学位。一路收窄,直到这个问题能在 2500 字、约 20 条来源里被回答:一个国家、一项工具、一个维度、 一个有辩护的主张。 MECM90002 . Global Data Policy & Governance ✓ Do this 要这样做 "Australia's News Media Bargaining Code is a legitimate exercise of national digital sovereignty, because it corrects a specific market failure - platform gatekeeper power over news distribution - that no global body was positioned to fix. " One actor, one instrument, one defended claim. “澳大利亚的 News Media Bargaining Code 是国家 数字主权的一次正当行使,因为它纠正了一个具体的 市场失灵 -- 平台对新闻分发的守门人权力 -- 而没 有任何全球机构能够补救这一失灵。”一个行动者、一 项工具、一个有辩护的主张。 ! Not this 不要这样做 "This essay will discuss data sovereignty around the world, looking at China, the EU, the US and Australia, and many platforms and laws. " No actor, no claim, no chance - a survey, not an argument. “本文将讨论世界各地的数据主权,考察中国、欧盟、 美国与澳大利亚,以及众多平台与法律。”没有行动 者,没有主张,没有胜算 -- 这是一篇概述,不是一 个论点。 MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . STRUCTURE CHAPTER 10 . THE ESSAY (A1) - STRUCTURE[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。
• 反面例子（一定要避开）：
  • “讨论世界各地数据主权，涉及很多国家与法律”——这注定写成综述。[16]Source: asksia-bible-mecm90002-bilingual.pdf2 Test the scope against the word count. 2500 words is roughly 4-5 body paragraphs. If your scope needs ten paragraphs to cover, it is still too broad - cut again. 用字数来检验范围。2500 字大约是4-5个主体段落。如 果你的范围要用十个段落才覆盖得了,那它仍然太宽 -- 再 砍。 3 Convert the narrowed scope into a thesis - a single sentence that takes a contestable position. If a reasonable classmate couldn't disagree with it, it is description, not a thesis. 把收窄后的范围转化为论点 -- 一个采取可争议立场的单 句。如果一个讲道理的同学无法对它表示异议,那它就是描 述,而不是论点。 4 Deploy a named model from Ch 1-2 as your analytical lens (sovereignty vs jurisdiction; self- vs state- vs co-regulation; rights- vs market- vs state-control). The lens is what turns your case into analysis. 部署一个第1-2 章的具名模型作为你的分析视角(主权 vs 管辖权;自我 vs 国家 vs 共同监管;权利 vs 市场 vs 国家 管控)。正是这个视角把你的案例变成分析。 ✓ Thesis templates (fill the blanks) 论点模板(填空使用) Stance + because + scope: "[Country/Platform] should [do X], because [reason grounded in a model], as the case of [instrument/incident] shows. " Tension form: "Although [the borderless counter- argument], [your actor] is justified in [its policy] because [model + evidence] - though only where [the limit you concede]. " The conceded limit signals critical engagement and stops the essay reading as one-sided. 立场+因为+范围:“[国家/平台]应当[做 X],因为 [基于某个模型的理由],正如[工具/事件]这一案例所 示。” 张力句式:“尽管[那条无国界的反方论点],[你的行动 者]仍有理由[采取其政策],因为[模型+证据] -- 尽管 只在[你所让步的限度]之内。” 所让步的限度表明你具有批判性的投入,并能阻止论 文读起来一边倒。 ! Scope creep is the No. 1 essay killer 范围蔓延是头号论文杀手 "Should the internet be regulated?" is not an essay - it is a degree. Narrow until the question is answerable in 2500 words with ~20 sources: one country, one instrument, one dimension, one defended claim. “互联网应当被监管吗?”不是一篇论文 -- 它是一个 学位。一路收窄,直到这个问题能在 2500 字、约 20 条来源里被回答:一个国家、一项工具、一个维度、 一个有辩护的主张。 MECM90002 . Global Data Policy & Governance ✓ Do this 要这样做 "Australia's News Media Bargaining Code is a legitimate exercise of national digital sovereignty, because it corrects a specific market failure - platform gatekeeper power over news distribution - that no global body was positioned to fix. " One actor, one instrument, one defended claim. “澳大利亚的 News Media Bargaining Code 是国家 数字主权的一次正当行使,因为它纠正了一个具体的 市场失灵 -- 平台对新闻分发的守门人权力 -- 而没 有任何全球机构能够补救这一失灵。”一个行动者、一 项工具、一个有辩护的主张。 ! Not this 不要这样做 "This essay will discuss data sovereignty around the world, looking at China, the EU, the US and Australia, and many platforms and laws. " No actor, no claim, no chance - a survey, not an argument. “本文将讨论世界各地的数据主权,考察中国、欧盟、 美国与澳大利亚,以及众多平台与法律。”没有行动 者,没有主张,没有胜算 -- 这是一篇概述,不是一 个论点。 MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . STRUCTURE CHAPTER 10 . THE ESSAY (A1) - STRUCTURE

2）本课的“总主线/总矛盾”（你所有论证的指南针）

2.1 核心张力（你写任何题都可以从这里起笔）

• 这门课反复强调的总矛盾是：
  • 数据在全球流动、无视国界（borderless / fluid）
  • 但 权力与法律仍然以国家领土组织（territorial）
• 你论文/案例的强开头，往往就是把这张力说清楚，然后立刻收窄到一个国家/平台/工具。[4]Source: asksia-bible-mecm90002-bilingual.pdfITU . OECD . UN Compact → Part 3 . Turn it into marks 9 Bilingual glossary key terms EN + +X gloss → 10 Writing the essay, presentation & case study narrowing . structure . citing → 11 Practice set questions · intervention drills → i Why this order 为何采用这一顺序 Build the toolkit first (Ch 1) so the vocabulary - datafication, the fluid data ecology, sovereignty, jurisdiction - is yours; then the models and debates (Ch 2) that every region argues over. Part 2 is the subject's signature world tour (Ch 3-8): Europe's rights-based model, the Anglo-American market model, then Africa, Asia and Latin America, closing on the global bodies and the future. Part 3 turns all of it into the essay, pitch and case study - with a bilingual glossary and a graded practice bank. 先搭好概念工具箱(第1章),让 datafication、流动的数据生态、主权、管辖权这些词汇真正成为你的;再掌握每个地 区都在争论的模型与辩论(第2章)。第二部分是本课标志性的世界巡览(第3-8章):欧洲以权利为本的模型、英美 的市场模型,接着是非洲、亚洲与拉丁美洲,最后落在全球机构与未来议题上。第三部分把这一切转化为论文、推介与 案例研究 -- 并配有双语术语表和一套已评分的练习题库。 MECM90002 . Global Data Policy & Governance CH 1 . TOOLKIT THE TOOLKIT CHAPTER 1 The conceptual toolkit 概念工具箱 Eighteen ideas that turn 'the internet is hard to regulate' into an argument 把“互联网难以监管”变成论点的十八个概念 Every later chapter - Europe, the US, Africa, Asia, Latin America, the global bodies - is one of these concepts applied to a place or a problem. Learn them now and the whole subject becomes a set of variations on a single tension: data flows globally, but power and law are still organised nationally. Master that tension and you can build the sharp, narrow argument the essay, presentation and case study all reward. 后面每一章 -- 欧洲、美国、非洲、亚洲、拉丁美洲、各全球机构 -- 都是把这些概念中的某一个应用到一个地方或一个问题 上。现在就把它们学透,整个科目便成为围绕同一张力的一组变奏:数据在全球流动,但权力与法律仍按国家来组织。掌握这 一张力,你就能构建出论文、展示与案例研究都看重的那种锋利而聚焦的论证。 ★ Why this chapter is the highest-leverage one in the bible 为何本章是整本 bible 中杠杆率最高的一章 You are never marked on reciting definitions - you are marked on using them to frame a specific case. A strong essay opens by naming the tension precisely ("the fluid data ecology strains jurisdiction, so states reach for data sovereignty . . . ") and then narrows to one country or platform. Each concept below ends with a "use it in your writing" angle - that is the part that earns marks. 评分从来不看你能否背诵定义 -- 而是看你能否运用它们去框定一个具体案例。强论文的开头会精确点明张力(“流动的 数据生态给管辖权带来压力,于是各国转向数据主权, . . ”),然后收窄到一个国家或平台。下面每个概念都以一个“在写 作中如何用它”的角度收尾––那才是真正得分的部分。 1. 1 Datafication - the raw material 1. 1 数据化––原材料 Start at the bottom. Datafication is the relentless conversion of everyday life, behaviour and personal activity into machine-readable data points that platforms can capture, store and monetise. Mejias & Couldry (cited in Flew & Martin 2022) frame platforms as "colonising" the internet and datafying personal information - the supply side of the surveillance-capitalism economy Zuboff describes. Everything else in this subject is a response to the sheer volume and mobility of this data. 从最底层开始。数据化是把日常生活、行为与个人活动不停地转化为机器可读数据点的过程,平台得以捕获、存储并将其货币 化。Mejias 与 Couldry(转引自 Flew & Martin 2022) 将平台描述为“殖民”互联网并对个人信息进行数据化 -- 这正是 Zuboff 所描述的监控资本主义经济的供给侧。本科目其余的一切,都是对这种数据巨大的体量与流动性的回应。[24]Source: asksia-cheatsheet-mecm90002.pdfThe semester runs in three movements - strong essays keep all three live: 1. Concepts (W1-4) - fluid data ecology, sovereignty, jurisdiction, monopolies 2. Regions (W5-10) - the comparative tour: EU -> US/UK/AU -> Africa -> Asia -> LatAm 3. Global & future (W11-12) - IGOs, intervention types, AI & data divides The through-question: in a borderless data world, who should govern data - the nation-state, the market, or a global order - and how? Every assignment is a focused answer to this. 1c . The Core Tension YOUR COMPASS Every topic is a clash of two truths : data is borderless (flows ignore the map) but law is territorial (states rule patches of ground). Whoever you back to resolve it - nation, market, or global body - becomes your thesis. Reading any case, ask: which actor asserts control, over what, and at whose expense? 2 . Core Concepts I W2-4 Datafication . converting everyday life & behaviour into machine-readable, monetisable data points (Mejias & Couldry; platforms "colonise" the net). Sovereignty (classical) . supreme authority of a state over its land & people, free of outside interference - since Westphalia (1648). Grimm: it has no fixed meaning & is eroding as states cede power upward. Data / digital sovereignty · the claim to control data, infrastructure & standards. Plural & contested (Couture & Toupin): national (localisation, sovereign cloud), individual/subjective (Estonia e-ID), Indigenous (Māori data), even platform sovereignty. Jurisdiction . a court's right to apply & enforce law over subject matter in its territory (procedural . substantive · enforcement). Online tension: the net is global, but jurisdiction presumes a world cut into territories - hard to "anchor. " Extraterritoriality . a state reaching beyond its borders - e. g. Australia's Assistance & Access Act 2018 catches anyone with one Australian end-user. 2b . The Framing Theory W1 . CITE- READY Network society . Castells - power & production run through global "flows" that transcend time & space, weakening the nation-state's grip. Van Dijck: a "culture of connectivity"; Volkmer: an emerging global public sphere. Deterritorialisation · new data "territories" where "code is law" (Lessig) - software logic shapes the communicative space & shrinks sovereign power (Volkmer). Surveillance capitalism · Zuboff - human behaviour captured as raw material for prediction products; the economic engine beneath datafication. 2c . Sovereignty . 4 Types USE PRECISELY TYPE CLAIM / EXAMPLE National[20]Source: asksia-cheatsheet-mecm90002.pdfMECM90002 Global Data Policy & Governance UNIVERSITY OF MELBOURNE . MEDIA & COMMUNICATIONS STUDY & ESSAY TOOLKIT Sem 1 2026 . SIDE 1 OF 2 100% assignment-assessed · no exam SIDE 1/2 · Latin America 0 · How to Use This READ FIRST * This subject is graded entirely by writing & a pitch - no exam : an argumentative essay (50%, 2500w), an oral presentation (25%, ~3 min . 3 slides), and a written case study (25%, 1500w). Hurdles: ≥80% tutorial attendance + submit every piece. The marked skill is narrowing a huge global debate into one sharp, evidence-backed policy argument. So this sheet is a thinking + citing bank: the toolkit + the regional map (Side 1), the scholars, cases & writing engine (Side 2). SIA + The whole subject is one move: pick a region/case, name a specific problem, argue a targeted intervention. Description loses marks; analysis ("so what for power, rights, the public interest?") wins them. 1 . The Big Picture W1-2 Old media (TV, radio, print) was bounded & territorial - easy for a nation to regulate. Today's data is the fluid data ecology: it sits on servers anywhere, constantly crosses borders, and it's unclear where it's stored or who can access it (cloud, IoT, 5G). That fluidity breaks nation-bound regulation - the core problem of the whole subject. Governance # government. Government = the state's law over its territory. Governance = multi-actor steering by states + IGOs (UN/ITU/OECD) + blocs (EU) + platforms + civil society. The course's spine = a spectrum: EU rights-based + US/UK/AU market-based . Global South (Africa/Asia/LatAm) navigating development, sovereignty & platform power. 1b . How the Subject Maps 3 PARTS The semester runs in three movements - strong essays keep all three live: 1. Concepts (W1-4) - fluid data ecology, sovereignty, jurisdiction, monopolies 2. Regions (W5-10) - the comparative tour: EU -> US/UK/AU -> Africa -> Asia -> LatAm 3. Global & future (W11-12) - IGOs, intervention types, AI & data divides The through-question: in a borderless data world, who should govern data - the nation-state, the market, or a global order - and how? Every assignment is a focused answer to this. 1c . The Core Tension YOUR COMPASS Every topic is a clash of two truths : data is borderless (flows ignore the map) but law is territorial (states rule patches of ground). Whoever you back to resolve it - nation, market, or global body - becomes your thesis.

2.2 你可以把所有题都理解成同一个问法

• “在无国界的数据世界里，到底该谁来治理数据：国家？市场？全球秩序？平台？以及用什么工具？”
• 你选择“谁该出手”，其实就是你的 thesis 方向。[20]Source: asksia-cheatsheet-mecm90002.pdfMECM90002 Global Data Policy & Governance UNIVERSITY OF MELBOURNE . MEDIA & COMMUNICATIONS STUDY & ESSAY TOOLKIT Sem 1 2026 . SIDE 1 OF 2 100% assignment-assessed · no exam SIDE 1/2 · Latin America 0 · How to Use This READ FIRST * This subject is graded entirely by writing & a pitch - no exam : an argumentative essay (50%, 2500w), an oral presentation (25%, ~3 min . 3 slides), and a written case study (25%, 1500w). Hurdles: ≥80% tutorial attendance + submit every piece. The marked skill is narrowing a huge global debate into one sharp, evidence-backed policy argument. So this sheet is a thinking + citing bank: the toolkit + the regional map (Side 1), the scholars, cases & writing engine (Side 2). SIA + The whole subject is one move: pick a region/case, name a specific problem, argue a targeted intervention. Description loses marks; analysis ("so what for power, rights, the public interest?") wins them. 1 . The Big Picture W1-2 Old media (TV, radio, print) was bounded & territorial - easy for a nation to regulate. Today's data is the fluid data ecology: it sits on servers anywhere, constantly crosses borders, and it's unclear where it's stored or who can access it (cloud, IoT, 5G). That fluidity breaks nation-bound regulation - the core problem of the whole subject. Governance # government. Government = the state's law over its territory. Governance = multi-actor steering by states + IGOs (UN/ITU/OECD) + blocs (EU) + platforms + civil society. The course's spine = a spectrum: EU rights-based + US/UK/AU market-based . Global South (Africa/Asia/LatAm) navigating development, sovereignty & platform power. 1b . How the Subject Maps 3 PARTS The semester runs in three movements - strong essays keep all three live: 1. Concepts (W1-4) - fluid data ecology, sovereignty, jurisdiction, monopolies 2. Regions (W5-10) - the comparative tour: EU -> US/UK/AU -> Africa -> Asia -> LatAm 3. Global & future (W11-12) - IGOs, intervention types, AI & data divides The through-question: in a borderless data world, who should govern data - the nation-state, the market, or a global order - and how? Every assignment is a focused answer to this. 1c . The Core Tension YOUR COMPASS Every topic is a clash of two truths : data is borderless (flows ignore the map) but law is territorial (states rule patches of ground). Whoever you back to resolve it - nation, market, or global body - becomes your thesis.[24]Source: asksia-cheatsheet-mecm90002.pdfThe semester runs in three movements - strong essays keep all three live: 1. Concepts (W1-4) - fluid data ecology, sovereignty, jurisdiction, monopolies 2. Regions (W5-10) - the comparative tour: EU -> US/UK/AU -> Africa -> Asia -> LatAm 3. Global & future (W11-12) - IGOs, intervention types, AI & data divides The through-question: in a borderless data world, who should govern data - the nation-state, the market, or a global order - and how? Every assignment is a focused answer to this. 1c . The Core Tension YOUR COMPASS Every topic is a clash of two truths : data is borderless (flows ignore the map) but law is territorial (states rule patches of ground). Whoever you back to resolve it - nation, market, or global body - becomes your thesis. Reading any case, ask: which actor asserts control, over what, and at whose expense? 2 . Core Concepts I W2-4 Datafication . converting everyday life & behaviour into machine-readable, monetisable data points (Mejias & Couldry; platforms "colonise" the net). Sovereignty (classical) . supreme authority of a state over its land & people, free of outside interference - since Westphalia (1648). Grimm: it has no fixed meaning & is eroding as states cede power upward. Data / digital sovereignty · the claim to control data, infrastructure & standards. Plural & contested (Couture & Toupin): national (localisation, sovereign cloud), individual/subjective (Estonia e-ID), Indigenous (Māori data), even platform sovereignty. Jurisdiction . a court's right to apply & enforce law over subject matter in its territory (procedural . substantive · enforcement). Online tension: the net is global, but jurisdiction presumes a world cut into territories - hard to "anchor. " Extraterritoriality . a state reaching beyond its borders - e. g. Australia's Assistance & Access Act 2018 catches anyone with one Australian end-user. 2b . The Framing Theory W1 . CITE- READY Network society . Castells - power & production run through global "flows" that transcend time & space, weakening the nation-state's grip. Van Dijck: a "culture of connectivity"; Volkmer: an emerging global public sphere. Deterritorialisation · new data "territories" where "code is law" (Lessig) - software logic shapes the communicative space & shrinks sovereign power (Volkmer). Surveillance capitalism · Zuboff - human behaviour captured as raw material for prediction products; the economic engine beneath datafication. 2c . Sovereignty . 4 Types USE PRECISELY TYPE CLAIM / EXAMPLE National

3）必须会的“概念/定义”（你写作里要用它们做分析，而不是背诵）

3.1 Datafication（数据化）

• 定义要点：把日常生活/行为/个人活动不断转成机器可读的数据点，平台得以捕获、存储并货币化；整个课程后面的争论都在回应这种数据的体量与流动。[4]Source: asksia-bible-mecm90002-bilingual.pdfITU . OECD . UN Compact → Part 3 . Turn it into marks 9 Bilingual glossary key terms EN + +X gloss → 10 Writing the essay, presentation & case study narrowing . structure . citing → 11 Practice set questions · intervention drills → i Why this order 为何采用这一顺序 Build the toolkit first (Ch 1) so the vocabulary - datafication, the fluid data ecology, sovereignty, jurisdiction - is yours; then the models and debates (Ch 2) that every region argues over. Part 2 is the subject's signature world tour (Ch 3-8): Europe's rights-based model, the Anglo-American market model, then Africa, Asia and Latin America, closing on the global bodies and the future. Part 3 turns all of it into the essay, pitch and case study - with a bilingual glossary and a graded practice bank. 先搭好概念工具箱(第1章),让 datafication、流动的数据生态、主权、管辖权这些词汇真正成为你的;再掌握每个地 区都在争论的模型与辩论(第2章)。第二部分是本课标志性的世界巡览(第3-8章):欧洲以权利为本的模型、英美 的市场模型,接着是非洲、亚洲与拉丁美洲,最后落在全球机构与未来议题上。第三部分把这一切转化为论文、推介与 案例研究 -- 并配有双语术语表和一套已评分的练习题库。 MECM90002 . Global Data Policy & Governance CH 1 . TOOLKIT THE TOOLKIT CHAPTER 1 The conceptual toolkit 概念工具箱 Eighteen ideas that turn 'the internet is hard to regulate' into an argument 把“互联网难以监管”变成论点的十八个概念 Every later chapter - Europe, the US, Africa, Asia, Latin America, the global bodies - is one of these concepts applied to a place or a problem. Learn them now and the whole subject becomes a set of variations on a single tension: data flows globally, but power and law are still organised nationally. Master that tension and you can build the sharp, narrow argument the essay, presentation and case study all reward. 后面每一章 -- 欧洲、美国、非洲、亚洲、拉丁美洲、各全球机构 -- 都是把这些概念中的某一个应用到一个地方或一个问题 上。现在就把它们学透,整个科目便成为围绕同一张力的一组变奏:数据在全球流动,但权力与法律仍按国家来组织。掌握这 一张力,你就能构建出论文、展示与案例研究都看重的那种锋利而聚焦的论证。 ★ Why this chapter is the highest-leverage one in the bible 为何本章是整本 bible 中杠杆率最高的一章 You are never marked on reciting definitions - you are marked on using them to frame a specific case. A strong essay opens by naming the tension precisely ("the fluid data ecology strains jurisdiction, so states reach for data sovereignty . . . ") and then narrows to one country or platform. Each concept below ends with a "use it in your writing" angle - that is the part that earns marks. 评分从来不看你能否背诵定义 -- 而是看你能否运用它们去框定一个具体案例。强论文的开头会精确点明张力(“流动的 数据生态给管辖权带来压力,于是各国转向数据主权, . . ”),然后收窄到一个国家或平台。下面每个概念都以一个“在写 作中如何用它”的角度收尾––那才是真正得分的部分。 1. 1 Datafication - the raw material 1. 1 数据化––原材料 Start at the bottom. Datafication is the relentless conversion of everyday life, behaviour and personal activity into machine-readable data points that platforms can capture, store and monetise. Mejias & Couldry (cited in Flew & Martin 2022) frame platforms as "colonising" the internet and datafying personal information - the supply side of the surveillance-capitalism economy Zuboff describes. Everything else in this subject is a response to the sheer volume and mobility of this data. 从最底层开始。数据化是把日常生活、行为与个人活动不停地转化为机器可读数据点的过程,平台得以捕获、存储并将其货币 化。Mejias 与 Couldry(转引自 Flew & Martin 2022) 将平台描述为“殖民”互联网并对个人信息进行数据化 -- 这正是 Zuboff 所描述的监控资本主义经济的供给侧。本科目其余的一切,都是对这种数据巨大的体量与流动性的回应。
• 写作怎么用：用它解释“为什么平台拥有结构性权力/为什么治理问题变得急迫”，然后落到一个具体案例。

3.2 Fluid data ecology（“流动的数据生态”）

• 核心意思（按材料表述）：数据可以在任何服务器上、跨境流动，导致“以国家领土为边界的监管”被不断拉扯甚至失效。[4]Source: asksia-bible-mecm90002-bilingual.pdfITU . OECD . UN Compact → Part 3 . Turn it into marks 9 Bilingual glossary key terms EN + +X gloss → 10 Writing the essay, presentation & case study narrowing . structure . citing → 11 Practice set questions · intervention drills → i Why this order 为何采用这一顺序 Build the toolkit first (Ch 1) so the vocabulary - datafication, the fluid data ecology, sovereignty, jurisdiction - is yours; then the models and debates (Ch 2) that every region argues over. Part 2 is the subject's signature world tour (Ch 3-8): Europe's rights-based model, the Anglo-American market model, then Africa, Asia and Latin America, closing on the global bodies and the future. Part 3 turns all of it into the essay, pitch and case study - with a bilingual glossary and a graded practice bank. 先搭好概念工具箱(第1章),让 datafication、流动的数据生态、主权、管辖权这些词汇真正成为你的;再掌握每个地 区都在争论的模型与辩论(第2章)。第二部分是本课标志性的世界巡览(第3-8章):欧洲以权利为本的模型、英美 的市场模型,接着是非洲、亚洲与拉丁美洲,最后落在全球机构与未来议题上。第三部分把这一切转化为论文、推介与 案例研究 -- 并配有双语术语表和一套已评分的练习题库。 MECM90002 . Global Data Policy & Governance CH 1 . TOOLKIT THE TOOLKIT CHAPTER 1 The conceptual toolkit 概念工具箱 Eighteen ideas that turn 'the internet is hard to regulate' into an argument 把“互联网难以监管”变成论点的十八个概念 Every later chapter - Europe, the US, Africa, Asia, Latin America, the global bodies - is one of these concepts applied to a place or a problem. Learn them now and the whole subject becomes a set of variations on a single tension: data flows globally, but power and law are still organised nationally. Master that tension and you can build the sharp, narrow argument the essay, presentation and case study all reward. 后面每一章 -- 欧洲、美国、非洲、亚洲、拉丁美洲、各全球机构 -- 都是把这些概念中的某一个应用到一个地方或一个问题 上。现在就把它们学透,整个科目便成为围绕同一张力的一组变奏:数据在全球流动,但权力与法律仍按国家来组织。掌握这 一张力,你就能构建出论文、展示与案例研究都看重的那种锋利而聚焦的论证。 ★ Why this chapter is the highest-leverage one in the bible 为何本章是整本 bible 中杠杆率最高的一章 You are never marked on reciting definitions - you are marked on using them to frame a specific case. A strong essay opens by naming the tension precisely ("the fluid data ecology strains jurisdiction, so states reach for data sovereignty . . . ") and then narrows to one country or platform. Each concept below ends with a "use it in your writing" angle - that is the part that earns marks. 评分从来不看你能否背诵定义 -- 而是看你能否运用它们去框定一个具体案例。强论文的开头会精确点明张力(“流动的 数据生态给管辖权带来压力,于是各国转向数据主权, . . ”),然后收窄到一个国家或平台。下面每个概念都以一个“在写 作中如何用它”的角度收尾––那才是真正得分的部分。 1. 1 Datafication - the raw material 1. 1 数据化––原材料 Start at the bottom. Datafication is the relentless conversion of everyday life, behaviour and personal activity into machine-readable data points that platforms can capture, store and monetise. Mejias & Couldry (cited in Flew & Martin 2022) frame platforms as "colonising" the internet and datafying personal information - the supply side of the surveillance-capitalism economy Zuboff describes. Everything else in this subject is a response to the sheer volume and mobility of this data. 从最底层开始。数据化是把日常生活、行为与个人活动不停地转化为机器可读数据点的过程,平台得以捕获、存储并将其货币 化。Mejias 与 Couldry(转引自 Flew & Martin 2022) 将平台描述为“殖民”互联网并对个人信息进行数据化 -- 这正是 Zuboff 所描述的监控资本主义经济的供给侧。本科目其余的一切,都是对这种数据巨大的体量与流动性的回应。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构[20]Source: asksia-cheatsheet-mecm90002.pdfMECM90002 Global Data Policy & Governance UNIVERSITY OF MELBOURNE . MEDIA & COMMUNICATIONS STUDY & ESSAY TOOLKIT Sem 1 2026 . SIDE 1 OF 2 100% assignment-assessed · no exam SIDE 1/2 · Latin America 0 · How to Use This READ FIRST * This subject is graded entirely by writing & a pitch - no exam : an argumentative essay (50%, 2500w), an oral presentation (25%, ~3 min . 3 slides), and a written case study (25%, 1500w). Hurdles: ≥80% tutorial attendance + submit every piece. The marked skill is narrowing a huge global debate into one sharp, evidence-backed policy argument. So this sheet is a thinking + citing bank: the toolkit + the regional map (Side 1), the scholars, cases & writing engine (Side 2). SIA + The whole subject is one move: pick a region/case, name a specific problem, argue a targeted intervention. Description loses marks; analysis ("so what for power, rights, the public interest?") wins them. 1 . The Big Picture W1-2 Old media (TV, radio, print) was bounded & territorial - easy for a nation to regulate. Today's data is the fluid data ecology: it sits on servers anywhere, constantly crosses borders, and it's unclear where it's stored or who can access it (cloud, IoT, 5G). That fluidity breaks nation-bound regulation - the core problem of the whole subject. Governance # government. Government = the state's law over its territory. Governance = multi-actor steering by states + IGOs (UN/ITU/OECD) + blocs (EU) + platforms + civil society. The course's spine = a spectrum: EU rights-based + US/UK/AU market-based . Global South (Africa/Asia/LatAm) navigating development, sovereignty & platform power. 1b . How the Subject Maps 3 PARTS The semester runs in three movements - strong essays keep all three live: 1. Concepts (W1-4) - fluid data ecology, sovereignty, jurisdiction, monopolies 2. Regions (W5-10) - the comparative tour: EU -> US/UK/AU -> Africa -> Asia -> LatAm 3. Global & future (W11-12) - IGOs, intervention types, AI & data divides The through-question: in a borderless data world, who should govern data - the nation-state, the market, or a global order - and how? Every assignment is a focused answer to this. 1c . The Core Tension YOUR COMPASS Every topic is a clash of two truths : data is borderless (flows ignore the map) but law is territorial (states rule patches of ground). Whoever you back to resolve it - nation, market, or global body - becomes your thesis.
• 写作怎么用：常用来引出 jurisdiction（管辖权）断点 或国家转向 data sovereignty（数据主权） 的动机。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构

3.3 Sovereignty vs. Jurisdiction（主权 vs 管辖权）

• 这是一对你非常常用的分析透镜：
  • 为什么国家觉得“控制在流失”（主权叙事）
  • 以及法律执行为什么困难（管辖权叙事）
• 材料给了一个“得分级”的段落示范：把“数据本地化”作为对主权流失的回应，并把关键分析落在“因为流动生态让数据漂向国外服务器，法院触及不到”这个因果链上。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构
• 复习方式：你不用把它背成哲学史，你要能在段落里写出一句“This matters because...（这之所以重要，是因为……）”把证据通过这个概念连回主张。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构

3.4 Governance ≠ Government（治理 ≠ 政府）

• government：国家对其领土内的法律统治
• governance：多方参与的“引导/规制”，包括国家 + 平台 + IGOs（UN/ITU/OECD）+ 公民社会等。[18]Source: asksia-cheatsheet-mecm90002.pdfAskSia STUDY SHEET SERIES 100% assignment-assessed . check your current subject outline . @ 2026 good luck. write sharp. KNOW THE TRADE SIDE 2/2 APPLY & WRITE . IGOS & interventions . The future . Scholar bank . Case bank . The essay . Presentation . Case study . Argument craft . Integrity 100% ASSIGNMENT . NO EXAM MECM90002 Global Data Policy & Governance UNIVERSITY OF MELBOURNE . MEDIA & COMMUNICATIONS STUDY & ESSAY TOOLKIT Sem 1 2026 . SIDE 1 OF 2 100% assignment-assessed · no exam SIDE 1/2 · Latin America 0 · How to Use This READ FIRST * This subject is graded entirely by writing & a pitch - no exam : an argumentative essay (50%, 2500w), an oral presentation (25%, ~3 min . 3 slides), and a written case study (25%, 1500w). Hurdles: ≥80% tutorial attendance + submit every piece. The marked skill is narrowing a huge global debate into one sharp, evidence-backed policy argument. So this sheet is a thinking + citing bank: the toolkit + the regional map (Side 1), the scholars, cases & writing engine (Side 2). SIA + The whole subject is one move: pick a region/case, name a specific problem, argue a targeted intervention. Description loses marks; analysis ("so what for power, rights, the public interest?") wins them. 1 . The Big Picture W1-2 Old media (TV, radio, print) was bounded & territorial - easy for a nation to regulate. Today's data is the fluid data ecology: it sits on servers anywhere, constantly crosses borders, and it's unclear where it's stored or who can access it (cloud, IoT, 5G). That fluidity breaks nation-bound regulation - the core problem of the whole subject. Governance # government. Government = the state's law over its territory. Governance = multi-actor steering by states + IGOs (UN/ITU/OECD) + blocs (EU) + platforms + civil society. The course's spine = a spectrum: EU rights-based + US/UK/AU market-based . Global South (Africa/Asia/LatAm) navigating development, sovereignty & platform power. 1b . How the Subject Maps 3 PARTS[20]Source: asksia-cheatsheet-mecm90002.pdfMECM90002 Global Data Policy & Governance UNIVERSITY OF MELBOURNE . MEDIA & COMMUNICATIONS STUDY & ESSAY TOOLKIT Sem 1 2026 . SIDE 1 OF 2 100% assignment-assessed · no exam SIDE 1/2 · Latin America 0 · How to Use This READ FIRST * This subject is graded entirely by writing & a pitch - no exam : an argumentative essay (50%, 2500w), an oral presentation (25%, ~3 min . 3 slides), and a written case study (25%, 1500w). Hurdles: ≥80% tutorial attendance + submit every piece. The marked skill is narrowing a huge global debate into one sharp, evidence-backed policy argument. So this sheet is a thinking + citing bank: the toolkit + the regional map (Side 1), the scholars, cases & writing engine (Side 2). SIA + The whole subject is one move: pick a region/case, name a specific problem, argue a targeted intervention. Description loses marks; analysis ("so what for power, rights, the public interest?") wins them. 1 . The Big Picture W1-2 Old media (TV, radio, print) was bounded & territorial - easy for a nation to regulate. Today's data is the fluid data ecology: it sits on servers anywhere, constantly crosses borders, and it's unclear where it's stored or who can access it (cloud, IoT, 5G). That fluidity breaks nation-bound regulation - the core problem of the whole subject. Governance # government. Government = the state's law over its territory. Governance = multi-actor steering by states + IGOs (UN/ITU/OECD) + blocs (EU) + platforms + civil society. The course's spine = a spectrum: EU rights-based + US/UK/AU market-based . Global South (Africa/Asia/LatAm) navigating development, sovereignty & platform power. 1b . How the Subject Maps 3 PARTS The semester runs in three movements - strong essays keep all three live: 1. Concepts (W1-4) - fluid data ecology, sovereignty, jurisdiction, monopolies 2. Regions (W5-10) - the comparative tour: EU -> US/UK/AU -> Africa -> Asia -> LatAm 3. Global & future (W11-12) - IGOs, intervention types, AI & data divides The through-question: in a borderless data world, who should govern data - the nation-state, the market, or a global order - and how? Every assignment is a focused answer to this. 1c . The Core Tension YOUR COMPASS Every topic is a clash of two truths : data is borderless (flows ignore the map) but law is territorial (states rule patches of ground). Whoever you back to resolve it - nation, market, or global body - becomes your thesis.

4）本课最像“公式/套路”的东西：写作结构引擎（你 final 最需要的是这个）

4.1 A1 Essay 的结构骨架（你可以当成“模板公式”）

• **引言（约 250–300w）**四件事：
  • hook（用一个具体事件/事实开场）
  • context（把事件放回大辩论：流动生态/平台守门人等）
  • thesis（一句可争辩主张，别藏）
  • roadmap（按顺序说你接下来怎么论证）[13]Source: asksia-bible-mecm90002-bilingual.pdfSTRUCTURE . ~ 4-5 BODY PARAS The scaffold that carries 2500 words 撑起 2500 字的骨架 Intro that lands a thesis . body paragraphs that argue . conclusion that earns its weight 落定论点的引言 · 进行论证的正文段 · 撑得起分量的结论 1 The introduction (~250-300w) 1引言(约 250-300 字) 1 Hook - one concrete sentence that opens the problem (a real incident or a striking fact: "In 2021 a $150bn suit accused Facebook of fuelling violence in Myanmar. "). 钩子 -- 一个把问题打开的具体句子(一桩真实事件或一个 惊人事实:“2021年,一桩 $150bn的诉讼指控 Facebook 助长了缅甸的暴力。”)。 2 Context - two or three sentences situating it in the broader debate (the fluid data ecology breaks nation-bounded regulation; platforms act as de facto gatekeepers). 背景 -- 两三句话把它置于更广的辩论中(流动的数据生态 瓦解了以国家为界的监管;平台充当事实上的守门人)。 3 Thesis - your one-sentence contestable claim, stated outright. Do not bury it. 论点 -- 你那一句可争议的主张,直截了当地说出来。不要 把它埋起来。 4 Roadmap - one sentence naming the steps your argument will take, in order. 路线图 -一句话,按顺序点明你的论证将走的几个步骤。 3 The conclusion (~200-250w) 3 结论(约 200-250 字) 1 Restate the thesis in fresh words - not a copy of the intro. 用新的措辞重述论点 -- 而不是照抄引言。 2 Synthesise, don't summarise: show how the body paragraphs together prove the claim, not what each said in turn. 综合,而非概括:展示各主体段落是如何合在一起证明了主 张,而不是逐一复述各段说了什么。 3 So-what - one sentence of implication: what should the policymaker / body / platform do next, or what does your case reveal about the broader debate? 那又如何 -- 一句话说明其意涵:政策制定者 /机构/平台 接下来该做什么,或者你的案例对更广的辩论揭示了什么? 4 No new evidence. A fresh citation in the conclusion is a structural error. 不引入新证据。在结论里出现一处全新的引用,是一个结构 性错误。 2 Body paragraphs - the CETAL move, every time 2 正文段 -- 每一段都用 CETAL 动作 Each body paragraph (aim for 4-5 at ~400-450 words each) is a self-contained argument unit. Run the same five-beat move so the marker can see the reasoning, not just the facts: 每一个正文段落(目标为4-5段,每段约400-450字)都是一个自成体系的论证单元。运用同一套五拍式操作,让评卷人 看见推理,而不只是事实: CETAL . THE BODY - PARAGRAPH SPINE C Claim - the paragraph's sub-argument in one sentence (topic sentence). E Evidence case - a real instrument or incident . T Theory - the named model that frames it. A Analysis - why the evidence, read through the theory, supports the claim. L Link -
• 正文段落（4–5 段）用 CETAL 动作反复跑：[13]Source: asksia-bible-mecm90002-bilingual.pdfSTRUCTURE . ~ 4-5 BODY PARAS The scaffold that carries 2500 words 撑起 2500 字的骨架 Intro that lands a thesis . body paragraphs that argue . conclusion that earns its weight 落定论点的引言 · 进行论证的正文段 · 撑得起分量的结论 1 The introduction (~250-300w) 1引言(约 250-300 字) 1 Hook - one concrete sentence that opens the problem (a real incident or a striking fact: "In 2021 a $150bn suit accused Facebook of fuelling violence in Myanmar. "). 钩子 -- 一个把问题打开的具体句子(一桩真实事件或一个 惊人事实:“2021年,一桩 $150bn的诉讼指控 Facebook 助长了缅甸的暴力。”)。 2 Context - two or three sentences situating it in the broader debate (the fluid data ecology breaks nation-bounded regulation; platforms act as de facto gatekeepers). 背景 -- 两三句话把它置于更广的辩论中(流动的数据生态 瓦解了以国家为界的监管;平台充当事实上的守门人)。 3 Thesis - your one-sentence contestable claim, stated outright. Do not bury it. 论点 -- 你那一句可争议的主张,直截了当地说出来。不要 把它埋起来。 4 Roadmap - one sentence naming the steps your argument will take, in order. 路线图 -一句话,按顺序点明你的论证将走的几个步骤。 3 The conclusion (~200-250w) 3 结论(约 200-250 字) 1 Restate the thesis in fresh words - not a copy of the intro. 用新的措辞重述论点 -- 而不是照抄引言。 2 Synthesise, don't summarise: show how the body paragraphs together prove the claim, not what each said in turn. 综合,而非概括:展示各主体段落是如何合在一起证明了主 张,而不是逐一复述各段说了什么。 3 So-what - one sentence of implication: what should the policymaker / body / platform do next, or what does your case reveal about the broader debate? 那又如何 -- 一句话说明其意涵:政策制定者 /机构/平台 接下来该做什么,或者你的案例对更广的辩论揭示了什么? 4 No new evidence. A fresh citation in the conclusion is a structural error. 不引入新证据。在结论里出现一处全新的引用,是一个结构 性错误。 2 Body paragraphs - the CETAL move, every time 2 正文段 -- 每一段都用 CETAL 动作 Each body paragraph (aim for 4-5 at ~400-450 words each) is a self-contained argument unit. Run the same five-beat move so the marker can see the reasoning, not just the facts: 每一个正文段落(目标为4-5段,每段约400-450字)都是一个自成体系的论证单元。运用同一套五拍式操作,让评卷人 看见推理,而不只是事实: CETAL . THE BODY - PARAGRAPH SPINE C Claim - the paragraph's sub-argument in one sentence (topic sentence). E Evidence case - a real instrument or incident . T Theory - the named model that frames it. A Analysis - why the evidence, read through the theory, supports the claim. L Link -
  • C = Claim（本段小主张/主题句）
  • E = Evidence（真实案例/法律工具/事件）
  • T = Theory（点名一个模型当透镜）
  • A = Analysis（最值钱：写清“因为……所以证据支持主张”）
  • L = Link（拉回 thesis 并过渡到下一段）
• 结论（约 200–250w）：
  • 重述 thesis（换词，不复制）
  • 综合（说明正文合在一起如何证明主张）
  • so-what（一句政策含义/下一步）
  • 不要引入新证据/新引用。[13]Source: asksia-bible-mecm90002-bilingual.pdfSTRUCTURE . ~ 4-5 BODY PARAS The scaffold that carries 2500 words 撑起 2500 字的骨架 Intro that lands a thesis . body paragraphs that argue . conclusion that earns its weight 落定论点的引言 · 进行论证的正文段 · 撑得起分量的结论 1 The introduction (~250-300w) 1引言(约 250-300 字) 1 Hook - one concrete sentence that opens the problem (a real incident or a striking fact: "In 2021 a $150bn suit accused Facebook of fuelling violence in Myanmar. "). 钩子 -- 一个把问题打开的具体句子(一桩真实事件或一个 惊人事实:“2021年,一桩 $150bn的诉讼指控 Facebook 助长了缅甸的暴力。”)。 2 Context - two or three sentences situating it in the broader debate (the fluid data ecology breaks nation-bounded regulation; platforms act as de facto gatekeepers). 背景 -- 两三句话把它置于更广的辩论中(流动的数据生态 瓦解了以国家为界的监管;平台充当事实上的守门人)。 3 Thesis - your one-sentence contestable claim, stated outright. Do not bury it. 论点 -- 你那一句可争议的主张,直截了当地说出来。不要 把它埋起来。 4 Roadmap - one sentence naming the steps your argument will take, in order. 路线图 -一句话,按顺序点明你的论证将走的几个步骤。 3 The conclusion (~200-250w) 3 结论(约 200-250 字) 1 Restate the thesis in fresh words - not a copy of the intro. 用新的措辞重述论点 -- 而不是照抄引言。 2 Synthesise, don't summarise: show how the body paragraphs together prove the claim, not what each said in turn. 综合,而非概括:展示各主体段落是如何合在一起证明了主 张,而不是逐一复述各段说了什么。 3 So-what - one sentence of implication: what should the policymaker / body / platform do next, or what does your case reveal about the broader debate? 那又如何 -- 一句话说明其意涵:政策制定者 /机构/平台 接下来该做什么,或者你的案例对更广的辩论揭示了什么? 4 No new evidence. A fresh citation in the conclusion is a structural error. 不引入新证据。在结论里出现一处全新的引用,是一个结构 性错误。 2 Body paragraphs - the CETAL move, every time 2 正文段 -- 每一段都用 CETAL 动作 Each body paragraph (aim for 4-5 at ~400-450 words each) is a self-contained argument unit. Run the same five-beat move so the marker can see the reasoning, not just the facts: 每一个正文段落(目标为4-5段,每段约400-450字)都是一个自成体系的论证单元。运用同一套五拍式操作,让评卷人 看见推理,而不只是事实: CETAL . THE BODY - PARAGRAPH SPINE C Claim - the paragraph's sub-argument in one sentence (topic sentence). E Evidence case - a real instrument or incident . T Theory - the named model that frames it. A Analysis - why the evidence, read through the theory, supports the claim. L Link -

4.2 CETAL 里“最重要的一拍”：A（分析）

• 材料直接说：很多人 C/E/T 做得不错，但没有“显式因为句”就会变成描述。
• 自检：如果一段里没有“这很重要因为…”“这表明…”这种分析句，你就很危险。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构

4.3 A2 Pitch（三页 slide 结构=你的“3步模型”）

• Slide 1：介绍案例/背景（一个平台/国家/工具 + 为什么重要），文字 ≤5 行 + 一张图。[7]Source: asksia-bible-mecm90002-bilingual.pdfSlide Job What goes on it Time The case, broadly Introduce the case / context - one platform, country or instrument - and why it matters. One image, ≤5 lines. ~45s 2 The specific issue Zoom in to the one regulatory problem you will fix. Name it precisely; ground it in a real incident. ~60s 3 The policy angle Pitch the intervention + a first-pass solution. Who acts (state / platform / IGO / civil society)? ~75s What lever? First-cut, not final. ! Reading your slides verbatim = mark downgrade 逐字照念你的幻灯片=降分 The brief explicitly penalises reading slides word-for- word. Slides carry keywords and one visual; you carry the argument. Rehearse to the clock - three minutes is brutally short, so each slide is one idea, spoken in plain sentences while the slide shows the anchor. 作业说明明确惩罚逐字念幻灯片。幻灯片承载关键词 和一张图;你承载论点。对着钟表去排练 -- 三分钟 极其短暂,所以每张幻灯片只讲一个想法,在幻灯片 展示锚点的同时,用平实的句子把它说出来。 i The policymaker frame 决策者视角框架 Speak as the regulator, not about regulation: "As an adviser to the ACCC, I propose . . . ". Choosing the actor (national government, platform, an IGO like the OECD/ITU, or civil society) forces clarity about who has the lever - the exact judgement A3 will test in depth. 要作为监管者去说,而不是谈论监管:“作为 ACCC 的一名顾问,我提议, . . . . . ”。选择行动者(国家政府、 平台、像 OECD/ITU 这样的 IGO,或公民社会)会迫 使你想清楚谁握有杠杆 -- 这正是A3 将深入考查的那 个判断。 MECM90002 . Global Data Policy & Governance ✓ Pace it: 3 minutes is ~390 words 把控节奏:3 分钟约 390 字 At a calm speaking rate you have room for roughly 390-420 spoken words total. Write a short script, time it twice, and cut until it fits with a few seconds' buffer. Practise the transition lines ("so the specific problem is . . . ", "my intervention is . . . ") - that is where nervous over-runs happen. 以平稳的语速,你总共大约有390-420 个口述词的余 地。写一个短脚本,计时两遍,删到能在留出几秒缓 冲的情况下讲完为止。练习那几句过渡语(“所以具体 的问题是 ………. ”‘ ”“我的干预是 . . . . . ") 一紧张超时往往 就发生在那里。 ★ Non-presenters: peer feedback is part of the task 非展示者:同伴反馈也是任务的一部分 When you are not presenting you are expected to give structured peer feedback. Listen for the move this chapter teaches: did they narrow to one problem and one fix, or drift into a survey? That feedback is gold for your own A3. 当你不是在做报告时,会被要求给出有结构的同伴反 馈。注意听本章所教的那个动作:他们是否收窄到了 一个问题与一个对策,还是飘成了一篇概述?这份反 馈对你自己的A3而言是金子。
• Slide 2：收窄到一个具体监管问题（用真实事件落地）。[7]Source: asksia-bible-mecm90002-bilingual.pdfSlide Job What goes on it Time The case, broadly Introduce the case / context - one platform, country or instrument - and why it matters. One image, ≤5 lines. ~45s 2 The specific issue Zoom in to the one regulatory problem you will fix. Name it precisely; ground it in a real incident. ~60s 3 The policy angle Pitch the intervention + a first-pass solution. Who acts (state / platform / IGO / civil society)? ~75s What lever? First-cut, not final. ! Reading your slides verbatim = mark downgrade 逐字照念你的幻灯片=降分 The brief explicitly penalises reading slides word-for- word. Slides carry keywords and one visual; you carry the argument. Rehearse to the clock - three minutes is brutally short, so each slide is one idea, spoken in plain sentences while the slide shows the anchor. 作业说明明确惩罚逐字念幻灯片。幻灯片承载关键词 和一张图;你承载论点。对着钟表去排练 -- 三分钟 极其短暂,所以每张幻灯片只讲一个想法,在幻灯片 展示锚点的同时,用平实的句子把它说出来。 i The policymaker frame 决策者视角框架 Speak as the regulator, not about regulation: "As an adviser to the ACCC, I propose . . . ". Choosing the actor (national government, platform, an IGO like the OECD/ITU, or civil society) forces clarity about who has the lever - the exact judgement A3 will test in depth. 要作为监管者去说,而不是谈论监管:“作为 ACCC 的一名顾问,我提议, . . . . . ”。选择行动者(国家政府、 平台、像 OECD/ITU 这样的 IGO,或公民社会)会迫 使你想清楚谁握有杠杆 -- 这正是A3 将深入考查的那 个判断。 MECM90002 . Global Data Policy & Governance ✓ Pace it: 3 minutes is ~390 words 把控节奏:3 分钟约 390 字 At a calm speaking rate you have room for roughly 390-420 spoken words total. Write a short script, time it twice, and cut until it fits with a few seconds' buffer. Practise the transition lines ("so the specific problem is . . . ", "my intervention is . . . ") - that is where nervous over-runs happen. 以平稳的语速,你总共大约有390-420 个口述词的余 地。写一个短脚本,计时两遍,删到能在留出几秒缓 冲的情况下讲完为止。练习那几句过渡语(“所以具体 的问题是 ………. ”‘ ”“我的干预是 . . . . . ") 一紧张超时往往 就发生在那里。 ★ Non-presenters: peer feedback is part of the task 非展示者:同伴反馈也是任务的一部分 When you are not presenting you are expected to give structured peer feedback. Listen for the move this chapter teaches: did they narrow to one problem and one fix, or drift into a survey? That feedback is gold for your own A3. 当你不是在做报告时,会被要求给出有结构的同伴反 馈。注意听本章所教的那个动作:他们是否收窄到了 一个问题与一个对策,还是飘成了一篇概述?这份反 馈对你自己的A3而言是金子。
• Slide 3：提出干预（谁行动：国家/平台/IGO/公民社会？用什么杠杆？先给一版可行的初解）。[7]Source: asksia-bible-mecm90002-bilingual.pdfSlide Job What goes on it Time The case, broadly Introduce the case / context - one platform, country or instrument - and why it matters. One image, ≤5 lines. ~45s 2 The specific issue Zoom in to the one regulatory problem you will fix. Name it precisely; ground it in a real incident. ~60s 3 The policy angle Pitch the intervention + a first-pass solution. Who acts (state / platform / IGO / civil society)? ~75s What lever? First-cut, not final. ! Reading your slides verbatim = mark downgrade 逐字照念你的幻灯片=降分 The brief explicitly penalises reading slides word-for- word. Slides carry keywords and one visual; you carry the argument. Rehearse to the clock - three minutes is brutally short, so each slide is one idea, spoken in plain sentences while the slide shows the anchor. 作业说明明确惩罚逐字念幻灯片。幻灯片承载关键词 和一张图;你承载论点。对着钟表去排练 -- 三分钟 极其短暂,所以每张幻灯片只讲一个想法,在幻灯片 展示锚点的同时,用平实的句子把它说出来。 i The policymaker frame 决策者视角框架 Speak as the regulator, not about regulation: "As an adviser to the ACCC, I propose . . . ". Choosing the actor (national government, platform, an IGO like the OECD/ITU, or civil society) forces clarity about who has the lever - the exact judgement A3 will test in depth. 要作为监管者去说,而不是谈论监管:“作为 ACCC 的一名顾问,我提议, . . . . . ”。选择行动者(国家政府、 平台、像 OECD/ITU 这样的 IGO,或公民社会)会迫 使你想清楚谁握有杠杆 -- 这正是A3 将深入考查的那 个判断。 MECM90002 . Global Data Policy & Governance ✓ Pace it: 3 minutes is ~390 words 把控节奏:3 分钟约 390 字 At a calm speaking rate you have room for roughly 390-420 spoken words total. Write a short script, time it twice, and cut until it fits with a few seconds' buffer. Practise the transition lines ("so the specific problem is . . . ", "my intervention is . . . ") - that is where nervous over-runs happen. 以平稳的语速,你总共大约有390-420 个口述词的余 地。写一个短脚本,计时两遍,删到能在留出几秒缓 冲的情况下讲完为止。练习那几句过渡语(“所以具体 的问题是 ………. ”‘ ”“我的干预是 . . . . . ") 一紧张超时往往 就发生在那里。 ★ Non-presenters: peer feedback is part of the task 非展示者:同伴反馈也是任务的一部分 When you are not presenting you are expected to give structured peer feedback. Listen for the move this chapter teaches: did they narrow to one problem and one fix, or drift into a survey? That feedback is gold for your own A3. 当你不是在做报告时,会被要求给出有结构的同伴反 馈。注意听本章所教的那个动作:他们是否收窄到了 一个问题与一个对策,还是飘成了一篇概述?这份反 馈对你自己的A3而言是金子。[15]Source: asksia-cheatsheet-mecm90002.pdfConcl. - restate . synthesise . "why it matters" Narrow the huge question to one sharp, defensible scope (one region, one law, one platform) - breadth is the #1 trap. 16 . A2 . Presentation 25% . ~ 3 MIN . 3 SLIDES Be a policymaker: name a specific regulatory problem & pitch an intervention. Slide 1 the case (broad) . Slide 2 the specific issue . Slide 3 the policy angle + first solutions. Don't read slides verbatim (mark penalty). Feeds directly into A3. 15b . Essay Traps AVOID THESE · Too broad - "regulate the internet" can't be argued in 2500w · Theory-dropping - naming Zuboff without using her · One region when the question invites comparison 16b . Pick Your Question ALIGN A1+A2+A3 Choose the essay question whose case you can reuse across all three tasks - one well-chosen case (a platform, a country, an IGO) can power the essay, the pitch and the case study. Pick where you have real evidence & a clear tension, not the "easiest" topic. 16c . Slide Design A2 TIPS 3 slides, ~1 idea each - keep them visual, not wordy (reading text aloud is penalised). One clear problem statement, one concrete intervention, one "why it's feasible. " Rehearse to land inside 3 minutes; the pitch is a dry-run of your A3 thesis. Practice once out loud - 3 minutes is shorter than it sounds. Compiled by AskSia . mapped to the MECM90002 curriculum . asksia. ai/cheatsheet/unimelb-mecm90002 17 . A3 · Case Study 25% . 1500W Develop the A2 idea into a rigorous study of one tightly-bounded case. Pick a lane: · Topic 1 - a company / practice / tech field x a policy dimension (e. g. a platform x misinfo moderation) · Topic 3 - a specific international policy/body (e. g. the UN Global Digital Compact) 3-PART STRUCTURE
• 关键扣分点：逐字念 slides 会降分；三分钟大约只能讲 ~390–420 words，一定要排练计时。[7]Source: asksia-bible-mecm90002-bilingual.pdfSlide Job What goes on it Time The case, broadly Introduce the case / context - one platform, country or instrument - and why it matters. One image, ≤5 lines. ~45s 2 The specific issue Zoom in to the one regulatory problem you will fix. Name it precisely; ground it in a real incident. ~60s 3 The policy angle Pitch the intervention + a first-pass solution. Who acts (state / platform / IGO / civil society)? ~75s What lever? First-cut, not final. ! Reading your slides verbatim = mark downgrade 逐字照念你的幻灯片=降分 The brief explicitly penalises reading slides word-for- word. Slides carry keywords and one visual; you carry the argument. Rehearse to the clock - three minutes is brutally short, so each slide is one idea, spoken in plain sentences while the slide shows the anchor. 作业说明明确惩罚逐字念幻灯片。幻灯片承载关键词 和一张图;你承载论点。对着钟表去排练 -- 三分钟 极其短暂,所以每张幻灯片只讲一个想法,在幻灯片 展示锚点的同时,用平实的句子把它说出来。 i The policymaker frame 决策者视角框架 Speak as the regulator, not about regulation: "As an adviser to the ACCC, I propose . . . ". Choosing the actor (national government, platform, an IGO like the OECD/ITU, or civil society) forces clarity about who has the lever - the exact judgement A3 will test in depth. 要作为监管者去说,而不是谈论监管:“作为 ACCC 的一名顾问,我提议, . . . . . ”。选择行动者(国家政府、 平台、像 OECD/ITU 这样的 IGO,或公民社会)会迫 使你想清楚谁握有杠杆 -- 这正是A3 将深入考查的那 个判断。 MECM90002 . Global Data Policy & Governance ✓ Pace it: 3 minutes is ~390 words 把控节奏:3 分钟约 390 字 At a calm speaking rate you have room for roughly 390-420 spoken words total. Write a short script, time it twice, and cut until it fits with a few seconds' buffer. Practise the transition lines ("so the specific problem is . . . ", "my intervention is . . . ") - that is where nervous over-runs happen. 以平稳的语速,你总共大约有390-420 个口述词的余 地。写一个短脚本,计时两遍,删到能在留出几秒缓 冲的情况下讲完为止。练习那几句过渡语(“所以具体 的问题是 ………. ”‘ ”“我的干预是 . . . . . ") 一紧张超时往往 就发生在那里。 ★ Non-presenters: peer feedback is part of the task 非展示者:同伴反馈也是任务的一部分 When you are not presenting you are expected to give structured peer feedback. Listen for the move this chapter teaches: did they narrow to one problem and one fix, or drift into a survey? That feedback is gold for your own A3. 当你不是在做报告时,会被要求给出有结构的同伴反 馈。注意听本章所教的那个动作:他们是否收窄到了 一个问题与一个对策,还是飘成了一篇概述?这份反 馈对你自己的A3而言是金子。

4.4 A3 Case Study（1500w 的“3段式公式”）

• Part 1：描述案例 + 为什么必须改变（唯一允许“描述”的部分，也要写紧）[11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.[19]Source: asksia-cheatsheet-mecm90002.pdf· Too broad - "regulate the internet" can't be argued in 2500w · Theory-dropping - naming Zuboff without using her · One region when the question invites comparison 16b . Pick Your Question ALIGN A1+A2+A3 Choose the essay question whose case you can reuse across all three tasks - one well-chosen case (a platform, a country, an IGO) can power the essay, the pitch and the case study. Pick where you have real evidence & a clear tension, not the "easiest" topic. 16c . Slide Design A2 TIPS 3 slides, ~1 idea each - keep them visual, not wordy (reading text aloud is penalised). One clear problem statement, one concrete intervention, one "why it's feasible. " Rehearse to land inside 3 minutes; the pitch is a dry-run of your A3 thesis. Practice once out loud - 3 minutes is shorter than it sounds. Compiled by AskSia . mapped to the MECM90002 curriculum . asksia. ai/cheatsheet/unimelb-mecm90002 17 . A3 · Case Study 25% . 1500W Develop the A2 idea into a rigorous study of one tightly-bounded case. Pick a lane: · Topic 1 - a company / practice / tech field x a policy dimension (e. g. a platform x misinfo moderation) · Topic 3 - a specific international policy/body (e. g. the UN Global Digital Compact) 3-PART STRUCTURE 1 describe case + why it must change 2 define the problem w/ real incidents 3 propose a concrete, targeted intervention Graded on: problem & context · intervention design . critical analysis > description . writing & citing. Case study # essay: the essay argues a position; A3 diagnoses one case & designs a fix. Go deeper, not broader - figures/tables don't count to the word limit. 18c . Evidence Checklist A3
• Part 2：界定具体问题 + 用真实事件证明伤害真实存在 [11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.
• Part 3：提出具体、可行、**有的放矢（targeted）**的干预，并回应明显反对意见（可行性/副作用）。[11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.[26]Source: asksia-cheatsheet-mecm90002.pdf1 describe case + why it must change 2 define the problem w/ real incidents 3 propose a concrete, targeted intervention Graded on: problem & context · intervention design . critical analysis > description . writing & citing. Case study # essay: the essay argues a position; A3 diagnoses one case & designs a fix. Go deeper, not broader - figures/tables don't count to the word limit. 18c . Evidence Checklist A3 · The law/policy text itself (primary source) · Real incidents - a fine, a breach, a court case · 2-3 scholars framing the debate · A counter-view you then answer 18 . Choosing a CaseSET YOURSELF UP Narrow - one figure/law/event, never "social media" Has a visible tension (sovereignty vs free flow; rights vs market) · 2-3 set scholars genuinely apply · Concrete evidence you can cite (law text, reports, coverage) 18b . Worked . Mini Case THE SHAPE Case: Australia's News Media Bargaining Code. Problem: platform market power starves journalism funding. Evidence: Google paid up; Meta "went dark," signed, then exited news (2024). Intervention: mandatory bargaining + final-offer arbitration. So what: a state forcing platform accountability - but is it durable when Meta can just drop news? 18d . "Targeted" Means A3 KEY WORD Grading rewards a narrow, feasible, specific intervention: "regulate Big Tech" fails; "extend the bargaining code's arbitration to AI-training licensing" passes. Name the actor, the mechanism, and why it's enforceable. Then pre-empt the obvious objection - feasibility & unintended effects - in a sentence.
• 核心评分词：targeted（有的放矢）决定分数。[11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.[14]Source: asksia-bible-mecm90002-bilingual.pdfi The two recurring moves 两个反复出现的操作 Move 1 - narrow: add one actor + one instrument/incident + one dimension until the scope fits 2500 (essay) or 1500 (case) words. Move 2 - analyse: name a model from Ch 1-2 (sovereignty/jurisdiction; self/state/co-regulation; rights/market/state-control) and say where it breaks on your case. Every drill below trains one of these. 第 1招 -- 收窄:不断加上一个行动者+一项工具 / 事件+一个维度,直到范围适配 2500 字(论文)或 1500 字(案例)。第2招 -- 分析:点出第1-2 章的 一个模型(主权/管辖权;自我/国家/共同监管; 权利/市场/国家管控),并说出它在你的案例上从何 处失效。下面每一道演练都训练其中之一。 D1 NARROWING DRILL essay scope A classmate's draft essay opens: "This essay examines how data sovereignty is understood and applied by governments around the world, with reference to many countries and laws. " Rewrite this as a single narrowed thesis suitable for a 2500-word argument, and name the one model you would use as the lens. 一位同学的论文初稿这样开头:“本文考察世界各国政府如何理解和运用数据主权,并参照众多国家与法律。”把它重写为 一个适用于 2500 字论证的、收窄过的单一论点,并点出你将用作透镜的那一个模型。 MODEL ANSWER Model answer. "Russia's 2019 'Sovereign Internet' data-localisation regime is best read not as data protection but as a tool of state control - and that conflation is why national digital sovereignty, however legitimate elsewhere, fails as a justification here. " Lens: Couture & Toupin's plural sovereignty, contrasting national vs state-control framings. MECM90002 . Global Data Policy & Governance - 范例答案。“俄罗斯 2019年的‘主权互联网’数据本地化体制,最好不被解读为数据保护,而被解读为一种国家管控工具 -- 而 正是这种混淆,使得国家数字主权(无论在别处多么正当)在此处无法成立为正当理由。”透镜:Couture 与 Toupin 的多元 主权,对照国家(national) 与国家管控(state-control)两种框架表述。 Marker's eye. The fix is one country, one instrument, one dimension, and a contestable claim. "Many countries and laws" guarantees a survey; markers cap survey essays well below a defended argument no matter how much is 'covered. ' 评卷人视角。修正之道是一个国家、一项工具、一个维度,外加一个可争辩的主张。“众多国家与法律”注定写成综述;无论 “覆盖”了多少,评卷人都会把综述类论文的分数压得远低于一篇有立场可辩护的论证。 D2 NARROWING DRILL case-study scope Too broad for a 1500-word case study: "Platforms should do more to stop misinformation. " Narrow it to a targeted intervention with a named actor and lever, grounded in a real incident. 对一篇1500 字的案例研究而言过于宽泛:“平台应当采取更多措施来阻止错误信息。”把它收窄为一项有的放矢的干预, 带有指名道姓的行为主体与杠杆,并扎根于一桩真实事件。 MODEL ANSWER Model answer. "Following the failures exposed by Rohingya v Facebook, Meta should be designated a 'gatekeeper' under a DMA-style threshold and subjected to independent, audited reporting of its misinformation-moderation outcomes in at-risk markets - a co-regulatory duty, not a vague call for 'responsibility' . " 范例答案。“鉴于 Rohingya v Facebook 所暴露的种种失败,Meta 应被依据一道 DMA 式的门槛指定为‘守门人’,并就其在 高风险市场中错误信息审核结果,接受独立的、经审计的报告 -- 这是一项共同监管义务,而非对‘责任’的含糊呼吁。” Marker's eye. "Targeted" is the graded word. Look for a named actor (Meta), a named lever (audited reporting under a designation threshold) and a real anchor (Rohingya v Facebook). "Do more" scores nothing. 评卷人视角。“有的放矢(targeted)”是那个被计分的词。要看到指名道姓的行为主体(Meta)、指名道姓的杠杆(在指定门 槛下经审计的报告)以及一个真实的锚点(Rohingya v Facebook)。“采取更多措施”一分不得。 MECM90002 . Global Data Policy & Governance CH 11 . ESSAY VARIANTS - CHAPTER 11 . ESSAY - QUESTION VARIANTS A1 REHEARSAL . ARGUE Practise the three lanes on fresh prompts 用新题目练习三条赛道 AskSia-written variants that transform the set questions - pick one, draft a thesis 由 AskSia 撰写、改造指定题目的变体 -- 选一道,起草一个论点 These are our own question variants, built to rehearse the same three lanes as the real essay (national rationale · transnational brief · platform responsibility) without reproducing the official prompts. For each, write the one-sentence thesis and the model you would deploy. Treat them as timed planning sprints (10 minutes each).[26]Source: asksia-cheatsheet-mecm90002.pdf1 describe case + why it must change 2 define the problem w/ real incidents 3 propose a concrete, targeted intervention Graded on: problem & context · intervention design . critical analysis > description . writing & citing. Case study # essay: the essay argues a position; A3 diagnoses one case & designs a fix. Go deeper, not broader - figures/tables don't count to the word limit. 18c . Evidence Checklist A3 · The law/policy text itself (primary source) · Real incidents - a fine, a breach, a court case · 2-3 scholars framing the debate · A counter-view you then answer 18 . Choosing a CaseSET YOURSELF UP Narrow - one figure/law/event, never "social media" Has a visible tension (sovereignty vs free flow; rights vs market) · 2-3 set scholars genuinely apply · Concrete evidence you can cite (law text, reports, coverage) 18b . Worked . Mini Case THE SHAPE Case: Australia's News Media Bargaining Code. Problem: platform market power starves journalism funding. Evidence: Google paid up; Meta "went dark," signed, then exited news (2024). Intervention: mandatory bargaining + final-offer arbitration. So what: a state forcing platform accountability - but is it durable when Meta can just drop news? 18d . "Targeted" Means A3 KEY WORD Grading rewards a narrow, feasible, specific intervention: "regulate Big Tech" fails; "extend the bargaining code's arbitration to AI-training licensing" passes. Name the actor, the mechanism, and why it's enforceable. Then pre-empt the obvious objection - feasibility & unintended effects - in a sentence.

5）三条 Essay 赛道（你选题时就要知道“各自怎么收窄”）

Q1：国家为什么仍要有自己的数字政策（national rationale）

• 关键不是为“主权”做抽象辩护，而是：选一个国家 + 一个理由（文化/安全/产业等）+ 一个工具来论证。[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。
• 材料示范里提到可用的工具例子：例如中国 2017 数据本地化/网络安全法这类“工具-论证”写法。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。

Q2：如果你坐在 UN 这类全球机构里，你优先推进什么（transnational brief）

• 反面陷阱：“什么都重要”。
• 正确做法：只选一个优先事项 + 一个机制/工具，论证为什么它该优先。[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。

Q3：平台责任（platform responsibility）

• 做法同样是：一个平台 + 一个义务/责任维度 + 一个锚点事件，用模型分析并回应反对意见。[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。

6）证据与引用（final 前最容易“丢容易分/踩雷”的部分）

6.1 来源数量与混合

• 论文大致预期 ~20 条引用，要混合：
  • 同行评审学术（课程具名学者/理论）
  • 一手政策/法律文本（例如法规正文）
  • 监管机构报告（ACCC、Ofcom、OECD 等）
  • 权威新闻用于“活案例”。[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。[21]Source: asksia-cheatsheet-mecm90002.pdfDescription = what the policy is. Analysis = what it means for power, rights, who profits, who's excluded. End each ‘| on a "so what. " · Critical voice - "this suggests", "however", "more convincingly" · Use a model (§4) + a case (§14) + a scholar (§13) in every body ‘| CRITICAL PHRASES . STEAL THESE Contrast: however . yet . by contrast. Analyse: this suggests . this reveals . which implies . more convincingly. Engage: as X argues . extending X . X overstates · against X. 20 . Referencing & Integrity DON'T LOSE EASY MARKS ~20 sources: peer-reviewed + policy docs/reports/news. Cite consistently; build a full list. AI policy (MPF1310): work auto-scanned by Turnitin AI detection . OK (disclose): verify ideas, define terms, locate/summarise sources, non-AI grammar tools. Not OK: AI writing sentences/paragraphs, citing AI as a source, AI rephrasing/"improving" or translating your work. Declare any AI use. 20b . Source Mix ~20 REFS Blend peer-reviewed (the set readings + beyond), policy/law (the GDPR text, an ACCC report), and quality news (Guardian, ABC). Avoid blogs & AI output as sources. Cite as you write - never reconstruct references at 2am. 21 . HD Checklist BEFORE YOU SUBMIT · Arguable thesis, answered throughout Model + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2
• 重要原则：聊天机器人不是来源。[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。

6.2 学术诚信与 AI 使用（final 前务必照做）

• 允许（但要声明）：解释术语、找来源/概括来源帮助理解、核对想法是否经得起可靠来源检验、拼写语法工具（非 AI 功能）等。[1]Source: asksia-bible-mecm90002-bilingual.pdf核对 AI 给出的想法是否经得起可靠来源的检验 · 用 Al 来解释或定义难懂的术语 · 帮助查找来源 · 为你自 己的理解而对学术来源进行概括或翻译 · 非 AI 的拼写 / 语法工具(关闭 AI 功能的 Grammarly / Hemingway)。所有这些都必须在你的提交中声明。 ! Not permitted - ever 绝不允许 Al writing whole sentences / paragraphs / papers . presenting AI output as your own research or analysis · citing Al output as a source . Al rewriting / rephrasing / 'improving' your text . Al translating your assignment into English. A chatbot is a next- word predictor that hallucinates - treated like paying a ghost-writer. Al 撰写整句/ 整段/整篇论文 · 把 AI 的输出当作你 自己的研究或分析呈现 · 把 AI 的输出作为来源引用 · 让 Al 改写/重述/“润色”你的文字 · 让 AI 把你的作 业翻译成英文。聊天机器人是一个会胡编的下一词预 测器 -- 会被等同于花钱雇了一个代笔人。 MECM90002 . Global Data Policy & Governance ! Late penalty is real and compounding 迟交罚分真实存在且会累积 5% of available marks per day (or part-day) without an approved extension; not accepted at all after 20 University business days. Extensions ≤10 working days are requested by email to the coordinator/tutor; longer needs Special Consideration. Submit early enough to survive an upload glitch. 未获批准延期者,每天(或不足一天按一天计)扣除 可得分数的5%;超过20个大学工作日后将完全不予 接受。≤10个工作日的延期,通过电子邮件向课程协 调员/导师申请;更长的则需 Special Consideration。要提早提交,以便挺过一次上传故 障。 ★ Declare it - every time, even for permitted use 声明它 -- 每一次都要,即便是被允许的使用 Any AI use at any stage must be declared in the submission (the University links disclosure templates). Undeclared or out-of-scope use triggers an academic-integrity investigation. The safe habit: keep a one-line log as you go ("used ChatGPT to define 'data localisation', verified against Kuner") and paste it into the declaration. Disclosure of permitted use costs you nothing; the alternative can cost you the subject. 在任何阶段对 AI的任何使用,都必须在提交中声明(大学提供了披露模板的链接)。未声明或超出范围的使用会触发学 术诚信调查。稳妥的习惯:边做边记一行日志(“用 ChatGPT 定义‘data localisation’,并对照 Kuner 核实”),再粘贴 进声明里。披露被许可的使用对你毫无代价;另一种做法则可能让你赔上这门课。 All submitted work must be your own. Permitted Al helps you understand and find - never write, rephrase, translate, or be cited. Declare every use. 所有提交的作业都必须是你本人完成的。被许可的 AI 帮你理解与查找 -- 绝不替你撰写、改写、翻译或被引用。每一次 使用都要声明。 MPF1310 . STUDENT ACADEMIC INTEGRITY MECM90002 . Global Data Policy & Governance CH 11 . NARROWING DRILLS - CHAPTER 11 . PRACTICE BANK NO EXAM . DRILL THE CRAFT Drill the writing, with model answers 操练写作,附范文 Ten AskSia-authored items in the subject's style - narrowing, models, case planning, concepts 十道由 AskSia 按本课程风格撰写的题目 -- 收窄、模式、案例规划、概念 There is no exam to cram, so these drills target the moves the three assessments reward: narrowing a huge question, naming the right model and where it breaks, planning a case study, and the concept checks that keep your analysis honest. Work each one cold, then read the model answer and the marker's-eye note. 这里没有需要临时抱佛脚的考试,所以这些练习针对的是三项评估所奖励的招式:收窄一个宏大问题、点出正确的模型及其失 效之处、规划一份案例研究,以及那些让你的分析保持诚实的概念检查。每一道都先冷启动独立做完,再去读范例答案与评卷 人视角的批注。 ★ How to use this bank 如何使用这个题库 Cover the answer. These are skill drills, not facts to memorise. Write or speak your response first, then compare. The narrowing drills are the highest-value - if you can reliably cut a planet-sized question to one defensible claim, you have the whole subject. These items are AskSia's own, built in the subject's style; they paraphrase and transform the real prompts - never reproduce them. Always answer the official brief on Canvas. 把答案盖住。这些是技能演练,不是要背的事实。先 写出或说出你的回答,再对照。收窄类演练是价值最 高的 -- 如果你能可靠地把一个庞大无边的问题切到 一个可辩护的主张,你就掌握了整门课。这些题目是 AskSia 自创的,按本课风格构建;它们对真实题目做 了转述与改造 -- 绝不照搬。请始终回答 Canvas 上 的官方说明。[11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.[21]Source: asksia-cheatsheet-mecm90002.pdfDescription = what the policy is. Analysis = what it means for power, rights, who profits, who's excluded. End each ‘| on a "so what. " · Critical voice - "this suggests", "however", "more convincingly" · Use a model (§4) + a case (§14) + a scholar (§13) in every body ‘| CRITICAL PHRASES . STEAL THESE Contrast: however . yet . by contrast. Analyse: this suggests . this reveals . which implies . more convincingly. Engage: as X argues . extending X . X overstates · against X. 20 . Referencing & Integrity DON'T LOSE EASY MARKS ~20 sources: peer-reviewed + policy docs/reports/news. Cite consistently; build a full list. AI policy (MPF1310): work auto-scanned by Turnitin AI detection . OK (disclose): verify ideas, define terms, locate/summarise sources, non-AI grammar tools. Not OK: AI writing sentences/paragraphs, citing AI as a source, AI rephrasing/"improving" or translating your work. Declare any AI use. 20b . Source Mix ~20 REFS Blend peer-reviewed (the set readings + beyond), policy/law (the GDPR text, an ACCC report), and quality news (Guardian, ABC). Avoid blogs & AI output as sources. Cite as you write - never reconstruct references at 2am. 21 . HD Checklist BEFORE YOU SUBMIT · Arguable thesis, answered throughout Model + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2
• 绝不允许：让 AI 写整句整段/改写润色你的文字/把 AI 输出当成你自己的分析/引用 AI 当来源/让 AI 翻译你的作业成英文。[1]Source: asksia-bible-mecm90002-bilingual.pdf核对 AI 给出的想法是否经得起可靠来源的检验 · 用 Al 来解释或定义难懂的术语 · 帮助查找来源 · 为你自 己的理解而对学术来源进行概括或翻译 · 非 AI 的拼写 / 语法工具(关闭 AI 功能的 Grammarly / Hemingway)。所有这些都必须在你的提交中声明。 ! Not permitted - ever 绝不允许 Al writing whole sentences / paragraphs / papers . presenting AI output as your own research or analysis · citing Al output as a source . Al rewriting / rephrasing / 'improving' your text . Al translating your assignment into English. A chatbot is a next- word predictor that hallucinates - treated like paying a ghost-writer. Al 撰写整句/ 整段/整篇论文 · 把 AI 的输出当作你 自己的研究或分析呈现 · 把 AI 的输出作为来源引用 · 让 Al 改写/重述/“润色”你的文字 · 让 AI 把你的作 业翻译成英文。聊天机器人是一个会胡编的下一词预 测器 -- 会被等同于花钱雇了一个代笔人。 MECM90002 . Global Data Policy & Governance ! Late penalty is real and compounding 迟交罚分真实存在且会累积 5% of available marks per day (or part-day) without an approved extension; not accepted at all after 20 University business days. Extensions ≤10 working days are requested by email to the coordinator/tutor; longer needs Special Consideration. Submit early enough to survive an upload glitch. 未获批准延期者,每天(或不足一天按一天计)扣除 可得分数的5%;超过20个大学工作日后将完全不予 接受。≤10个工作日的延期,通过电子邮件向课程协 调员/导师申请;更长的则需 Special Consideration。要提早提交,以便挺过一次上传故 障。 ★ Declare it - every time, even for permitted use 声明它 -- 每一次都要,即便是被允许的使用 Any AI use at any stage must be declared in the submission (the University links disclosure templates). Undeclared or out-of-scope use triggers an academic-integrity investigation. The safe habit: keep a one-line log as you go ("used ChatGPT to define 'data localisation', verified against Kuner") and paste it into the declaration. Disclosure of permitted use costs you nothing; the alternative can cost you the subject. 在任何阶段对 AI的任何使用,都必须在提交中声明(大学提供了披露模板的链接)。未声明或超出范围的使用会触发学 术诚信调查。稳妥的习惯:边做边记一行日志(“用 ChatGPT 定义‘data localisation’,并对照 Kuner 核实”),再粘贴 进声明里。披露被许可的使用对你毫无代价;另一种做法则可能让你赔上这门课。 All submitted work must be your own. Permitted Al helps you understand and find - never write, rephrase, translate, or be cited. Declare every use. 所有提交的作业都必须是你本人完成的。被许可的 AI 帮你理解与查找 -- 绝不替你撰写、改写、翻译或被引用。每一次 使用都要声明。 MPF1310 . STUDENT ACADEMIC INTEGRITY MECM90002 . Global Data Policy & Governance CH 11 . NARROWING DRILLS - CHAPTER 11 . PRACTICE BANK NO EXAM . DRILL THE CRAFT Drill the writing, with model answers 操练写作,附范文 Ten AskSia-authored items in the subject's style - narrowing, models, case planning, concepts 十道由 AskSia 按本课程风格撰写的题目 -- 收窄、模式、案例规划、概念 There is no exam to cram, so these drills target the moves the three assessments reward: narrowing a huge question, naming the right model and where it breaks, planning a case study, and the concept checks that keep your analysis honest. Work each one cold, then read the model answer and the marker's-eye note. 这里没有需要临时抱佛脚的考试,所以这些练习针对的是三项评估所奖励的招式:收窄一个宏大问题、点出正确的模型及其失 效之处、规划一份案例研究,以及那些让你的分析保持诚实的概念检查。每一道都先冷启动独立做完,再去读范例答案与评卷 人视角的批注。 ★ How to use this bank 如何使用这个题库 Cover the answer. These are skill drills, not facts to memorise. Write or speak your response first, then compare. The narrowing drills are the highest-value - if you can reliably cut a planet-sized question to one defensible claim, you have the whole subject. These items are AskSia's own, built in the subject's style; they paraphrase and transform the real prompts - never reproduce them. Always answer the official brief on Canvas. 把答案盖住。这些是技能演练,不是要背的事实。先 写出或说出你的回答,再对照。收窄类演练是价值最 高的 -- 如果你能可靠地把一个庞大无边的问题切到 一个可辩护的主张,你就掌握了整门课。这些题目是 AskSia 自创的,按本课风格构建;它们对真实题目做 了转述与改造 -- 绝不照搬。请始终回答 Canvas 上 的官方说明。[11]Source: asksia-bible-mecm90002-bilingual.pdfThe three-part structure 三段式结构 1 1 . Describe + why change 1 · 描述+为何需要改变 Set up the case concisely and state why it needs to change. This is the only descriptive part - keep it tight (~300-350w) so the analysis has room. 简洁地设定案例背景,并说明为 何它需要改变。这是唯一的描述 性部分 -- 要把它写紧(约 300-350 字),好给分析留出空 间。 2 2 . Define the problem 2 · 界定问题 Pin down the specific regulatory problem and ground it in real incidents / examples for context. This is where you show the harm is real, not hypothetical (~500-550w). 锁定那个具体的监管问题,并把 它落地到真实事件/例子上以提 供脉络。这是你证明危害是真 实、而非假设的地方(约 500- 550 字)。 3 3 . Propose the fix 3 · 提出解决方案 Propose a concrete, feasible, targeted intervention - not a broad "there ought to be a law". Say who acts, what lever, and answer the obvious objection (~550-600w). 提出一项具体、可行、有针对性 的干预 -- 而不是泛泛的“应该 有一部法律”。说清谁来行动、 用什么杠杆,并回应那个显而易 见的反对意见(约 550-600 字)。 MECM90002 . Global Data Policy & Governance ! "Targeted" is the word that decides the mark “有的放矢(targeted)”这个词决定了分数 The brief rewards a targeted intervention and punishes a broad one. "Platforms should be more responsible" is broad and worthless. "Designate Meta a gatekeeper under a DMA-style threshold and mandate independent audit of its misinformation- moderation outcomes" is targeted: a named actor, a named lever, a measurable duty. 作业说明奖励有针对性的干预,惩罚宽泛的干预。“平 台应当更负责任”是宽泛而无价值的。“依照类似 DMA 的门槛把 Meta 指定为守门人,并强制对其虚假信息 审核结果进行独立审计”才是有针对性的:一个具名的 行动者、一个具名的杠杆、一项可衡量的义务。 ✓ Use a model as the analytical engine 把某个模型当作分析引擎 Frame the problem with a Ch 2 model and use it to justify the fix: is this a job for self-, state- or co- regulation? A rights-, market- or state-control approach? Naming the model - and the case where it breaks - is what makes part 3 analysis rather than opinion. 用第2 章的一个模型来框定问题,并用它来论证对 策:这是该由自我监管、国家监管,还是共同监管来 做的活?是该用以权利为本、以市场为本,还是以国 家管控为本的路径?点明那个模型 -- 以及它失效的 那个案例 -- 正是让第三部分成为分析、而非意见的 关键。 What the marker scores 评阅者评分的依据 Criterion What earns it Problem outline & context A precise problem, grounded in real incidents. Intervention design Concrete, feasible, targeted; named actor + lever; justified, with the objection answered.
• 最安全习惯：边写边记一行“我用了 AI 做什么 + 我用什么来源核对”，然后贴到声明里。[1]Source: asksia-bible-mecm90002-bilingual.pdf核对 AI 给出的想法是否经得起可靠来源的检验 · 用 Al 来解释或定义难懂的术语 · 帮助查找来源 · 为你自 己的理解而对学术来源进行概括或翻译 · 非 AI 的拼写 / 语法工具(关闭 AI 功能的 Grammarly / Hemingway)。所有这些都必须在你的提交中声明。 ! Not permitted - ever 绝不允许 Al writing whole sentences / paragraphs / papers . presenting AI output as your own research or analysis · citing Al output as a source . Al rewriting / rephrasing / 'improving' your text . Al translating your assignment into English. A chatbot is a next- word predictor that hallucinates - treated like paying a ghost-writer. Al 撰写整句/ 整段/整篇论文 · 把 AI 的输出当作你 自己的研究或分析呈现 · 把 AI 的输出作为来源引用 · 让 Al 改写/重述/“润色”你的文字 · 让 AI 把你的作 业翻译成英文。聊天机器人是一个会胡编的下一词预 测器 -- 会被等同于花钱雇了一个代笔人。 MECM90002 . Global Data Policy & Governance ! Late penalty is real and compounding 迟交罚分真实存在且会累积 5% of available marks per day (or part-day) without an approved extension; not accepted at all after 20 University business days. Extensions ≤10 working days are requested by email to the coordinator/tutor; longer needs Special Consideration. Submit early enough to survive an upload glitch. 未获批准延期者,每天(或不足一天按一天计)扣除 可得分数的5%;超过20个大学工作日后将完全不予 接受。≤10个工作日的延期,通过电子邮件向课程协 调员/导师申请;更长的则需 Special Consideration。要提早提交,以便挺过一次上传故 障。 ★ Declare it - every time, even for permitted use 声明它 -- 每一次都要,即便是被允许的使用 Any AI use at any stage must be declared in the submission (the University links disclosure templates). Undeclared or out-of-scope use triggers an academic-integrity investigation. The safe habit: keep a one-line log as you go ("used ChatGPT to define 'data localisation', verified against Kuner") and paste it into the declaration. Disclosure of permitted use costs you nothing; the alternative can cost you the subject. 在任何阶段对 AI的任何使用,都必须在提交中声明(大学提供了披露模板的链接)。未声明或超出范围的使用会触发学 术诚信调查。稳妥的习惯:边做边记一行日志(“用 ChatGPT 定义‘data localisation’,并对照 Kuner 核实”),再粘贴 进声明里。披露被许可的使用对你毫无代价;另一种做法则可能让你赔上这门课。 All submitted work must be your own. Permitted Al helps you understand and find - never write, rephrase, translate, or be cited. Declare every use. 所有提交的作业都必须是你本人完成的。被许可的 AI 帮你理解与查找 -- 绝不替你撰写、改写、翻译或被引用。每一次 使用都要声明。 MPF1310 . STUDENT ACADEMIC INTEGRITY MECM90002 . Global Data Policy & Governance CH 11 . NARROWING DRILLS - CHAPTER 11 . PRACTICE BANK NO EXAM . DRILL THE CRAFT Drill the writing, with model answers 操练写作,附范文 Ten AskSia-authored items in the subject's style - narrowing, models, case planning, concepts 十道由 AskSia 按本课程风格撰写的题目 -- 收窄、模式、案例规划、概念 There is no exam to cram, so these drills target the moves the three assessments reward: narrowing a huge question, naming the right model and where it breaks, planning a case study, and the concept checks that keep your analysis honest. Work each one cold, then read the model answer and the marker's-eye note. 这里没有需要临时抱佛脚的考试,所以这些练习针对的是三项评估所奖励的招式:收窄一个宏大问题、点出正确的模型及其失 效之处、规划一份案例研究,以及那些让你的分析保持诚实的概念检查。每一道都先冷启动独立做完,再去读范例答案与评卷 人视角的批注。 ★ How to use this bank 如何使用这个题库 Cover the answer. These are skill drills, not facts to memorise. Write or speak your response first, then compare. The narrowing drills are the highest-value - if you can reliably cut a planet-sized question to one defensible claim, you have the whole subject. These items are AskSia's own, built in the subject's style; they paraphrase and transform the real prompts - never reproduce them. Always answer the official brief on Canvas. 把答案盖住。这些是技能演练,不是要背的事实。先 写出或说出你的回答,再对照。收窄类演练是价值最 高的 -- 如果你能可靠地把一个庞大无边的问题切到 一个可辩护的主张,你就掌握了整门课。这些题目是 AskSia 自创的,按本课风格构建;它们对真实题目做 了转述与改造 -- 绝不照搬。请始终回答 Canvas 上 的官方说明。

7）给你一份“考前最后冲刺清单”（按优先级，从救命到加分）

7.1 必做（最能立刻提分）

• 把你的 topic 用“漏斗”收窄到：一个行动者 + 一个工具/事件 + 一个维度，写成一句可争辩 thesis。[16]Source: asksia-bible-mecm90002-bilingual.pdf2 Test the scope against the word count. 2500 words is roughly 4-5 body paragraphs. If your scope needs ten paragraphs to cover, it is still too broad - cut again. 用字数来检验范围。2500 字大约是4-5个主体段落。如 果你的范围要用十个段落才覆盖得了,那它仍然太宽 -- 再 砍。 3 Convert the narrowed scope into a thesis - a single sentence that takes a contestable position. If a reasonable classmate couldn't disagree with it, it is description, not a thesis. 把收窄后的范围转化为论点 -- 一个采取可争议立场的单 句。如果一个讲道理的同学无法对它表示异议,那它就是描 述,而不是论点。 4 Deploy a named model from Ch 1-2 as your analytical lens (sovereignty vs jurisdiction; self- vs state- vs co-regulation; rights- vs market- vs state-control). The lens is what turns your case into analysis. 部署一个第1-2 章的具名模型作为你的分析视角(主权 vs 管辖权;自我 vs 国家 vs 共同监管;权利 vs 市场 vs 国家 管控)。正是这个视角把你的案例变成分析。 ✓ Thesis templates (fill the blanks) 论点模板(填空使用) Stance + because + scope: "[Country/Platform] should [do X], because [reason grounded in a model], as the case of [instrument/incident] shows. " Tension form: "Although [the borderless counter- argument], [your actor] is justified in [its policy] because [model + evidence] - though only where [the limit you concede]. " The conceded limit signals critical engagement and stops the essay reading as one-sided. 立场+因为+范围:“[国家/平台]应当[做 X],因为 [基于某个模型的理由],正如[工具/事件]这一案例所 示。” 张力句式:“尽管[那条无国界的反方论点],[你的行动 者]仍有理由[采取其政策],因为[模型+证据] -- 尽管 只在[你所让步的限度]之内。” 所让步的限度表明你具有批判性的投入,并能阻止论 文读起来一边倒。 ! Scope creep is the No. 1 essay killer 范围蔓延是头号论文杀手 "Should the internet be regulated?" is not an essay - it is a degree. Narrow until the question is answerable in 2500 words with ~20 sources: one country, one instrument, one dimension, one defended claim. “互联网应当被监管吗?”不是一篇论文 -- 它是一个 学位。一路收窄,直到这个问题能在 2500 字、约 20 条来源里被回答:一个国家、一项工具、一个维度、 一个有辩护的主张。 MECM90002 . Global Data Policy & Governance ✓ Do this 要这样做 "Australia's News Media Bargaining Code is a legitimate exercise of national digital sovereignty, because it corrects a specific market failure - platform gatekeeper power over news distribution - that no global body was positioned to fix. " One actor, one instrument, one defended claim. “澳大利亚的 News Media Bargaining Code 是国家 数字主权的一次正当行使,因为它纠正了一个具体的 市场失灵 -- 平台对新闻分发的守门人权力 -- 而没 有任何全球机构能够补救这一失灵。”一个行动者、一 项工具、一个有辩护的主张。 ! Not this 不要这样做 "This essay will discuss data sovereignty around the world, looking at China, the EU, the US and Australia, and many platforms and laws. " No actor, no claim, no chance - a survey, not an argument. “本文将讨论世界各地的数据主权,考察中国、欧盟、 美国与澳大利亚,以及众多平台与法律。”没有行动 者,没有主张,没有胜算 -- 这是一篇概述,不是一 个论点。 MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . STRUCTURE CHAPTER 10 . THE ESSAY (A1) - STRUCTURE[17]Source: asksia-bible-mecm90002-bilingual.pdf描述告诉读者发生了什么(GDPR 于 2018 年通过; 它创设了数据主体)。论点告诉读者该如何想、以及为 什么(正是 GDPR 的域外触达使它成为事实上的全球 标准 -- 而这恰恰就是为何不应由单一集团去设定全 世界的隐私底线)。每一段都必须靠推进一个主张来挣 得它的位置,而不是靠报告一个事实。 i Source expectations & what "credible" means here 对来源的预期,以及此处“可信”意味着什么 The essay expects roughly 20 references. Mix peer-reviewed scholarship (the named theorists in Ch 1-2 - Castells, Volkmer, Flew & Martin, Pickard, Grimm, Couture & Toupin, Kuner) with credible non-academic evidence: primary policy documents (the GDPR text, the DMA, the AU bargaining code, the UN Global Digital Compact), regulator reports (ACCC, Ofcom, OECD), and reputable news for live incidents. A chatbot is not a source - see the Al policy on the last page of this chapter. 论文预期大约20条参考文献。把同行评审学术(第1-2 章中具名的理论家 -- Castells、Volkmer、Flew & Martin、 Pickard、Grimm、Couture & Toupin、Kuner)与可信的非学术证据混合使用:一手政策文件(GDPR 文本、DMA、 AU 议价准则、UN Global Digital Compact)、监管机构报告(ACCC、Ofcom、OECD),以及报道鲜活事件的权威新 闻。聊天机器人不是来源 -- 见本章最后一页的 AI 政策。 Markers reward a tight scope and a defended claim, and punish a broad survey. Pick one country, platform or instrument; name one problem; defend one fix. 评分者奖励紧凑的范围与有辩护的主张,惩罚宽泛的概述。挑一个国家、平台或工具;点明一个问题;为一项对策辩 护。 THE WHOLE CHAPTER IN ONE LINE MECM90002 . Global Data Policy & Governance CH 10 . ESSAY . NARROWING - CHAPTER 10 . THE ESSAY (A1, 50%) ARGUMENTATIVE . 2500 WORDS Pick one of three, then make it small 三选一,然后把题目做小 The three set questions, and how to cut a planet-sized prompt down to a thesis 三道指定题目,以及如何把行星级别的题目削减为一个论点 You choose one of three broad questions. Each is deliberately vast; the marks are in how sharply you narrow it. Below are the three lanes (paraphrased), each with the kind of tightly-scoped argument that wins - not the broad survey that loses. 你从三道宽泛的题目中选一道。每道都被刻意设计得极为宏大;分数在于你把它收窄得有多锋利。下面是三条赛道(已意 译),每条都附有那种能赢的、范围收得很紧的论证 -- 而不是那种会输的宽泛综述。 Q The broad prompt (paraphrased) How to narrow it Q1 National-policy rationale. Why does a country still need its own digital policy when data flows ignore borders? Don't defend 'sovereignty' in the abstract. Pick one country and one rationale (culture / cybersecurity / local-economy protection) and argue it via one instrument - e. g. China's 2017 data-localisation mandate, or the AU News Media Bargaining Code. Q2 UN-style transnational brief. If you sat inside a global rule-making body, what would you prioritise to govern the worldwide digital space? Resist 'everything matters. ' Choose one priority (e. g. equitable access vs protection) and one mechanism (the UN Global Digital Compact, OECD AI Principles) and argue why that should lead. Q3 Platform responsibility. What duties does a globe-spanning platform owe its users, societies and democratic processes? Pick one platform and one duty - e. g. Meta and content moderation of disinformation (Rohingya v Facebook), or the 'cooperative responsibility' of Helberger et al. against the 'too big to be governed' critique. MECM90002 . Global Data Policy & Governance The narrowing funnel 收窄漏斗 1 Start broad, end narrow. Take the chosen question and add three constraints: one actor (a country / platform / body), one instrument or incident, and one dimension (privacy, competition, content, security). 从宽开始,到窄收尾。取来所选的问题,加上三道约束:一 个行动者(一个国家/平台/机构)、一项工具或事件,以 及一个维度(隐私、竞争、内容、安全)。
• 用 CETAL 写 4–5 个正文段的“主题句”（只读主题句就能串起整篇论证）。[13]Source: asksia-bible-mecm90002-bilingual.pdfSTRUCTURE . ~ 4-5 BODY PARAS The scaffold that carries 2500 words 撑起 2500 字的骨架 Intro that lands a thesis . body paragraphs that argue . conclusion that earns its weight 落定论点的引言 · 进行论证的正文段 · 撑得起分量的结论 1 The introduction (~250-300w) 1引言(约 250-300 字) 1 Hook - one concrete sentence that opens the problem (a real incident or a striking fact: "In 2021 a $150bn suit accused Facebook of fuelling violence in Myanmar. "). 钩子 -- 一个把问题打开的具体句子(一桩真实事件或一个 惊人事实:“2021年,一桩 $150bn的诉讼指控 Facebook 助长了缅甸的暴力。”)。 2 Context - two or three sentences situating it in the broader debate (the fluid data ecology breaks nation-bounded regulation; platforms act as de facto gatekeepers). 背景 -- 两三句话把它置于更广的辩论中(流动的数据生态 瓦解了以国家为界的监管;平台充当事实上的守门人)。 3 Thesis - your one-sentence contestable claim, stated outright. Do not bury it. 论点 -- 你那一句可争议的主张,直截了当地说出来。不要 把它埋起来。 4 Roadmap - one sentence naming the steps your argument will take, in order. 路线图 -一句话,按顺序点明你的论证将走的几个步骤。 3 The conclusion (~200-250w) 3 结论(约 200-250 字) 1 Restate the thesis in fresh words - not a copy of the intro. 用新的措辞重述论点 -- 而不是照抄引言。 2 Synthesise, don't summarise: show how the body paragraphs together prove the claim, not what each said in turn. 综合,而非概括:展示各主体段落是如何合在一起证明了主 张,而不是逐一复述各段说了什么。 3 So-what - one sentence of implication: what should the policymaker / body / platform do next, or what does your case reveal about the broader debate? 那又如何 -- 一句话说明其意涵:政策制定者 /机构/平台 接下来该做什么,或者你的案例对更广的辩论揭示了什么? 4 No new evidence. A fresh citation in the conclusion is a structural error. 不引入新证据。在结论里出现一处全新的引用,是一个结构 性错误。 2 Body paragraphs - the CETAL move, every time 2 正文段 -- 每一段都用 CETAL 动作 Each body paragraph (aim for 4-5 at ~400-450 words each) is a self-contained argument unit. Run the same five-beat move so the marker can see the reasoning, not just the facts: 每一个正文段落(目标为4-5段,每段约400-450字)都是一个自成体系的论证单元。运用同一套五拍式操作,让评卷人 看见推理,而不只是事实: CETAL . THE BODY - PARAGRAPH SPINE C Claim - the paragraph's sub-argument in one sentence (topic sentence). E Evidence case - a real instrument or incident . T Theory - the named model that frames it. A Analysis - why the evidence, read through the theory, supports the claim. L Link -[8]Source: asksia-cheatsheet-mecm90002.pdfModel + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2 AskSia STUDY SHEET SERIES 100% assignment-assessed . check your current subject outline . @ 2026 good luck. write sharp. KNOW THE TRADE SIDE 2/2 APPLY & WRITE . IGOS & interventions . The future . Scholar bank . Case bank . The essay . Presentation . Case study . Argument craft . Integrity 100% ASSIGNMENT . NO EXAM
• 加一段反方/反对意见 + 你的回应（这会直接提升“批判性参与”的评分信号）。[6]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构[21]Source: asksia-cheatsheet-mecm90002.pdfDescription = what the policy is. Analysis = what it means for power, rights, who profits, who's excluded. End each ‘| on a "so what. " · Critical voice - "this suggests", "however", "more convincingly" · Use a model (§4) + a case (§14) + a scholar (§13) in every body ‘| CRITICAL PHRASES . STEAL THESE Contrast: however . yet . by contrast. Analyse: this suggests . this reveals . which implies . more convincingly. Engage: as X argues . extending X . X overstates · against X. 20 . Referencing & Integrity DON'T LOSE EASY MARKS ~20 sources: peer-reviewed + policy docs/reports/news. Cite consistently; build a full list. AI policy (MPF1310): work auto-scanned by Turnitin AI detection . OK (disclose): verify ideas, define terms, locate/summarise sources, non-AI grammar tools. Not OK: AI writing sentences/paragraphs, citing AI as a source, AI rephrasing/"improving" or translating your work. Declare any AI use. 20b . Source Mix ~20 REFS Blend peer-reviewed (the set readings + beyond), policy/law (the GDPR text, an ACCC report), and quality news (Guardian, ABC). Avoid blogs & AI output as sources. Cite as you write - never reconstruct references at 2am. 21 . HD Checklist BEFORE YOU SUBMIT · Arguable thesis, answered throughout Model + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2
• 每段都补上一个 “so what”（权力/权利/公共利益/谁获益谁受损），把描述拉回分析。[21]Source: asksia-cheatsheet-mecm90002.pdfDescription = what the policy is. Analysis = what it means for power, rights, who profits, who's excluded. End each ‘| on a "so what. " · Critical voice - "this suggests", "however", "more convincingly" · Use a model (§4) + a case (§14) + a scholar (§13) in every body ‘| CRITICAL PHRASES . STEAL THESE Contrast: however . yet . by contrast. Analyse: this suggests . this reveals . which implies . more convincingly. Engage: as X argues . extending X . X overstates · against X. 20 . Referencing & Integrity DON'T LOSE EASY MARKS ~20 sources: peer-reviewed + policy docs/reports/news. Cite consistently; build a full list. AI policy (MPF1310): work auto-scanned by Turnitin AI detection . OK (disclose): verify ideas, define terms, locate/summarise sources, non-AI grammar tools. Not OK: AI writing sentences/paragraphs, citing AI as a source, AI rephrasing/"improving" or translating your work. Declare any AI use. 20b . Source Mix ~20 REFS Blend peer-reviewed (the set readings + beyond), policy/law (the GDPR text, an ACCC report), and quality news (Guardian, ABC). Avoid blogs & AI output as sources. Cite as you write - never reconstruct references at 2am. 21 . HD Checklist BEFORE YOU SUBMIT · Arguable thesis, answered throughout Model + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2[28]Source: asksia-cheatsheet-mecm90002.pdf19 · Argument Craft DESCRIPTION - ANALYSIS THESIS TEMPLATE "While [region/platform] appears to , this essay argues it actually -- , revealing about [sovereignty / rights / the public interest]. " Description = what the policy is. Analysis = what it means for power, rights, who profits, who's excluded. End each ‘| on a "so what. " · Critical voice - "this suggests", "however", "more convincingly" · Use a model (§4) + a case (§14) + a scholar (§13) in every body ‘| CRITICAL PHRASES . STEAL THESE Contrast: however . yet . by contrast. Analyse: this suggests . this reveals . which implies . more convincingly. Engage: as X argues . extending X . X overstates · against X. 20 . Referencing & Integrity DON'T LOSE EASY MARKS ~20 sources: peer-reviewed + policy docs/reports/news. Cite consistently; build a full list. AI policy (MPF1310): work auto-scanned by Turnitin AI detection . OK (disclose): verify ideas, define terms, locate/summarise sources, non-AI grammar tools. Not OK: AI writing sentences/paragraphs, citing AI as a source, AI rephrasing/"improving" or translating your work. Declare any AI use. 20b . Source Mix ~20 REFS Blend peer-reviewed (the set readings + beyond), policy/law (the GDPR text, an ACCC report), and quality news (Guardian, ABC). Avoid blogs & AI output as sources. Cite as you write - never reconstruct references at 2am. 21 . HD Checklist BEFORE YOU SUBMIT · Arguable thesis, answered throughout

7.2 加分（冲高分用）

• 每个正文段都做到：model + case + scholar（模型+案例+学者）齐全，且都服务同一条 thesis。[21]Source: asksia-cheatsheet-mecm90002.pdfDescription = what the policy is. Analysis = what it means for power, rights, who profits, who's excluded. End each ‘| on a "so what. " · Critical voice - "this suggests", "however", "more convincingly" · Use a model (§4) + a case (§14) + a scholar (§13) in every body ‘| CRITICAL PHRASES . STEAL THESE Contrast: however . yet . by contrast. Analyse: this suggests . this reveals . which implies . more convincingly. Engage: as X argues . extending X . X overstates · against X. 20 . Referencing & Integrity DON'T LOSE EASY MARKS ~20 sources: peer-reviewed + policy docs/reports/news. Cite consistently; build a full list. AI policy (MPF1310): work auto-scanned by Turnitin AI detection . OK (disclose): verify ideas, define terms, locate/summarise sources, non-AI grammar tools. Not OK: AI writing sentences/paragraphs, citing AI as a source, AI rephrasing/"improving" or translating your work. Declare any AI use. 20b . Source Mix ~20 REFS Blend peer-reviewed (the set readings + beyond), policy/law (the GDPR text, an ACCC report), and quality news (Guardian, ABC). Avoid blogs & AI output as sources. Cite as you write - never reconstruct references at 2am. 21 . HD Checklist BEFORE YOU SUBMIT · Arguable thesis, answered throughout Model + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2[8]Source: asksia-cheatsheet-mecm90002.pdfModel + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2 AskSia STUDY SHEET SERIES 100% assignment-assessed . check your current subject outline . @ 2026 good luck. write sharp. KNOW THE TRADE SIDE 2/2 APPLY & WRITE . IGOS & interventions . The future . Scholar bank . Case bank . The essay . Presentation . Case study . Argument craft . Integrity 100% ASSIGNMENT . NO EXAM
• 选一个“未来议题”（偏见/鸿沟/抓取/数据主权等），用“rights/market/state”三哲学 + 行动层级（政府/平台/IGO/公民社会）跑一遍分析引擎，形成有政策落点的结论。[5]Source: asksia-bible-mecm90002-bilingual.pdf对数据痕迹的抓取 -- 既包括可信来源,也包括用户在不知 情下产生的活动数据 -- 是一场具有定义意义的 AI 争论。 Meta 把 Instagram/Facebook 数据用于 Al: EU 用户依据 GDPR “可以选择退出(opt out)”,但 Meta 不保证在其他 地方也如此 -- 那么其他国家的公民受到保护了吗? Society mustremain in control:the choices AI poses are collective choices. 社会必须保持掌控:AI 所抛出的那些选择,是集体的选择。 MARTIN TISNE, PARIS AI ACTION SUMMIT (COURSE FRAMING) § The closing synthesis - tie the regional spine back together 收束式综合 -- 把区域主线重新串回一起 The wrap-up reiterates the analytical spine of the whole subject: the EU (data subject, Brussels effect), the US/UK/Australia (neoliberalism, free speech, antitrust, 'duty of care', selective interventionism), and the shared Global-South issues across Asia and Latin America (digital divide, no uniform data protection, mis/disinformation, rising digital authoritarianism). The future issues cut across all three - which is exactly why no single regulatory philosophy resolves them. 总结部分重申了整门课的分析脊柱:欧盟(数据主体、布鲁塞尔效应)、美国/英国/澳大利亚(新自由主义、言论自 由、反垄断、“注意义务”、选择性干预主义),以及横跨亚洲与拉丁美洲的共同全球南方议题(数字鸿沟、无统一数据保 护、错误/虚假信息、上升中的数字威权主义)。未来的议题横切这三者 -- 而这正是为何没有任何单一的监管哲学能解 决它们。 ✓ Essay move - pick a future issue, run the engine 论文操作 -- 选定一个未来议题,开动这台引擎 Take one future issue (bias, divides, scraping, data sovereignty, infrastructure colonisation), then run it through the analytical engine: which philosophy (rights/market/state) frames it best? Which level (government/platform/IGO/civil society) should act? Soft or hard law? That combination - a live problem analysed with the subject's full toolkit - is the shape of a first-class answer. 取一个未来议题(偏见、鸿沟、抓取、数据主权、基础设施殖民),然后让它过一遍分析引擎:哪种哲学(权利/市场/ 国家)最能框定它?哪个层级(政府/平台/IGO/公民社会)应当行动?软法还是硬法?这种组合 -- 用本课全套工具 去分析一个鲜活的问题 -- 正是一流答案的样子。 MECM90002 . Global Data Policy & Governance GLOSSARY . 1 OF 3 STUDY COMPANION - REFERENCE Bilingual glossary 双语术语表 Key terms of global data policy - English . #X . one-line definition 全球数据政策核心术语 -- English · 中文 · 一句话定义 This glossary collects the recurring vocabulary of MECM90002 in one place: each term carries an English label, a > translation, and a single-line working definition. It is a study companion, not a substitute for the lectures and readings - use it to fix the words before you argue with them. Terms are grouped by theme across three pages. 本词汇表把 MECM90002 中反复出现的词汇集中于一处:每个术语都带有一个英文标签、一个中文译名,以及一行实用定 义。它是一份学习伴读,而非课堂讲座与阅读材料的替代品 -- 用它在你与这些词争辩之前先把它们固定下来。术语按主题分 组,分布于三页之中。 i How to read each row 如何逐行解读 Column one is the English term; column two is the +> translation (natural, not literal); column three is a one-line definition in English. Where a concept is tied to a scholar or instrument, the name is given in brackets so you can trace it back to the readings. 第一列是英文术语;第二列是中文译法(自然,而非字面直译);第三列是英文的一行定义。凡概念与某位学者或某项工 具相关,名字都在括号中给出,以便你追溯回阅读材料。 Term (EN) 中文 One-line definition Core concepts & political economy of data Data governance 数据治理[23]Source: asksia-cheatsheet-mecm90002.pdf10c . Reading a Region A METHOD For any region, pin four things: its philosophy (rights / market / state-control), its signature law, a vivid case, and a scholar. That quartet = a ready-built body paragraph. Then contrast two regions to show you grasp the spectrum, not one dot on it. 100% assignment-assessed . check your current subject outline . @ 2026 flip - for side 2 . scholars, cases & the writing engine asksia. ai/cheatsheet/ unimelb-mecm90002 . side 1/2 AskSia STUDY SHEET SERIES 100% ASSIGNMENT . NO EXAM THE TOOLKIT & THE MAP . Fluid data ecology . Sovereignty . Jurisdiction . Regulation models . EU/GDPR . US/UK/AU . Africa . Asia MECM90002 Global Data Policy & Governance UNIVERSITY OF MELBOURNE . MEDIA & COMMUNICATIONS STUDY & ESSAY TOOLKIT Sem 1 2026 . SIDE 2 OF 2 Apply . cite . write 11 . IGOs & Interventions W11 Inter-governmental bodies: UN agencies - ITU (1865, spectrum/telecoms), WIPO, UNESCO; thematic - OECD (AI Principles 2019, first intergov. Al standard); WTO/GATS (trade lens). Soft vs hard law: soft = non-binding, fast, flexible (OECD AI Principles, guidelines); hard = binding treaty (Budapest Cybercrime Convention). 4 intervention levels (Gorwa, platform governance) - a ready essay frame: 1. Government - laws, bans, taxes, infrastructure, co-regulation 2. Platform - moderation, ToS, oversight boards, ad libraries 3. IGO - soft/hard law, summit diplomacy 4. Civil society - petitions, submissions, campaigns, protest 11b . Interventions . Examples

8）我需要你补充 2 个信息，我就能给你“按你题目定制”的超详细复习方案

• A1 论文 / A2 pitch / A3 case study（选一个）

• Q1 国家政策 / Q2 全球机构 / Q3 平台责任（选一个）

你好，我是 Sia（Asksia 的学科教授）。欢迎来到你的学习会话：MECM90002《Global Data Policy & Governance》考前复习地图。

从你这次的资料来看，你手上有两份核心工具：

• bible（双语精读）：帮你把“全球数据治理”这种超大议题，收窄成一个具体、可辩护、有证据的论点；课程没有考试，成绩来自论文/展示/案例研究，最怕“只描述不论证”[1]Source: asksia-bible-mecm90002-bilingual.pdf3 · 论证 An assessment is due. Go to Ch 10 (writing the essay, presentation & case study) and the glossary. The single skill rewarded everywhere is narrowing a huge question into one specific, evidence-backed claim. This bible is built to help you do exactly that. 临近交作业的时候。请直接看第 10 章(写论文、口头报告与案例 研究)以及术语表。这门课处处 奖励的唯一一项技能,就是把一 个庞大的问题收窄成一个具体、 有证据支撑的主张。这本 bible 的全部设计,就是帮你做到这一 点。 MECM90002 . Global Data Policy & Governance ! The mistake that sinks marks: describing instead of arguing 拖垮分数的错误:只描述而不论证 Every task here is argumentative or analytical, never a summary. The essay must take a position and defend it; the case study must analyse one tightly-bounded case, not narrate it; the presentation must propose an intervention. Students lose marks by surveying "everything about data governance. " The fix is on the next page and in Ch 10: pick one country / platform / instrument, name one problem, defend one intervention. 这里的每一项任务都是论辩型或分析型的,绝不是简单的概述。论文必须提出一个立场并为之辩护;案例研究必须分析 一个边界清晰的案例,而不是叙述它;口头报告必须提出一项干预措施。学生失分往往是因为去通盘罗列“关于数据治理 的一切”。解决办法在下一页和第10 章里:挑一个国家/平台/法律工具,点明一个问题,为一项干预辩护。 i How this bible was built - and the two-layer rule 这本 bible 是如何构建的 -- 以及双层规则 Grounded entirely in this subject's own materials: the subject outline, all twelve week pages, the lecture and seminar slide decks, and the assessment briefs. Standard facts - public laws (GDPR, DSA, DMA, the Al Act), named theories, real cases - are stated plainly; the lecturer's specific framings and examples are paraphrased, never copied. Always verify dates and weights against your own Canvas before submitting. 本书完全立足于这门课自己的材料:课程大纲、全部十二周的周页面、讲座与研讨课的幻灯片,以及各项作业说明。标 准事实 -- 公开的法律(GDPR、DSA、DMA、Al Act)、有名的理论、真实的案例一直接平实地陈述;讲师特有的框 架与例子则一律转述,绝不照搬。提交前请务必对照你自己的 Canvas 核实日期与权重。 MECM90002 . Global Data Policy & Governance THE BLUEPRINT - THE ASSESSMENT BLUEPRINT 100% WRITTEN & ORAL No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, two hurdles, and one strategic move that decides your grade 三项任务、两道门槛,以及决定你成绩的一个战略动作 MECM90002 has no sit-down exam. Your entire mark comes from an argumentative essay, a short policymaker presentation and a written case study - with two hurdles you must clear regardless of those marks. Read this page first; it is the map for everything else. MECM90002 没有闭卷考试。你的全部成绩来自一篇论证性论文、一段简短的决策者展示,以及一份书面案例研究 -- 此外 还有两道无论分数如何都必须通过的及格门槛(hurdle)。请先读这一页;它是其余一切内容的地图。 50% ESSAY (2500W) 论文(2500 字) 25% PRESENTATION 口头报告 25% CASE STUDY (1500W) 案例研究 (1500 字) EXAM - NONE 考试 -- 无 The three assessed tasks 三项受评任务 Task[11]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.。
• cheatsheet（速查表）：把概念工具箱、区域对比、写作结构、常见陷阱、引用与学术诚信要点压缩成“随手查”的清单，强调“分析大于描述”“范围不要太大”[2]Source: asksia-cheatsheet-mecm90002.pdfModel + case + scholar in each · Analysis > description; every | has a "so what" · Counter-argument raised & answered · Each | links back to the thesis SIA > Read only your topic sentences end-to- end - they should tell the whole argument alone. If they don't, the structure isn't there yet. asksia. ai/cheatsheet/ unimelb-mecm90002 · side 2/2 AskSia STUDY SHEET SERIES 100% assignment-assessed . check your current subject outline . @ 2026 good luck. write sharp. KNOW THE TRADE SIDE 2/2 APPLY & WRITE . IGOS & interventions . The future . Scholar bank . Case bank . The essay . Presentation . Case study . Argument craft . Integrity 100% ASSIGNMENT . NO EXAM[3]Source: asksia-cheatsheet-mecm90002.pdfConcl. - restate . synthesise . "why it matters" Narrow the huge question to one sharp, defensible scope (one region, one law, one platform) - breadth is the #1 trap. 16 . A2 . Presentation 25% . ~ 3 MIN . 3 SLIDES Be a policymaker: name a specific regulatory problem & pitch an intervention. Slide 1 the case (broad) . Slide 2 the specific issue . Slide 3 the policy angle + first solutions. Don't read slides verbatim (mark penalty). Feeds directly into A3. 15b . Essay Traps AVOID THESE · Too broad - "regulate the internet" can't be argued in 2500w · Theory-dropping - naming Zuboff without using her · One region when the question invites comparison 16b . Pick Your Question ALIGN A1+A2+A3 Choose the essay question whose case you can reuse across all three tasks - one well-chosen case (a platform, a country, an IGO) can power the essay, the pitch and the case study. Pick where you have real evidence & a clear tension, not the "easiest" topic. 16c . Slide Design A2 TIPS 3 slides, ~1 idea each - keep them visual, not wordy (reading text aloud is penalised). One clear problem statement, one concrete intervention, one "why it's feasible. " Rehearse to land inside 3 minutes; the pitch is a dry-run of your A3 thesis. Practice once out loud - 3 minutes is shorter than it sounds. Compiled by AskSia . mapped to the MECM90002 curriculum . asksia. ai/cheatsheet/unimelb-mecm90002 17 . A3 · Case Study 25% . 1500W Develop the A2 idea into a rigorous study of one tightly-bounded case. Pick a lane: · Topic 1 - a company / practice / tech field x a policy dimension (e. g. a platform x misinfo moderation) · Topic 3 - a specific international policy/body (e. g. the UN Global Digital Compact) 3-PART STRUCTURE。

这门课的复习主线可以理解成一条“写作引擎”：

• 先用概念工具箱（如：流动的数据生态、主权/管辖权、不同监管模型）建立分析框架，再把它应用到不同地区与机构上；最后回到写作任务，用“收窄→论证→提出干预→回应反对意见”的链条拿分[6]Source: asksia-bible-mecm90002-bilingual.pdfUganda (lower-middle-income) embodies "a gap between strategy documents and real-world implementation. " It has Uganda Vision 2040 alongside a stack of restrictive communications laws (Regulation of Interception of Communications Act 2010, Computer Misuse Act 2011, Uganda Communications Act 2013) and repeated election- time shutdowns. Matsiko & Kersting (2023) study bottom-up digital participation under that state restriction. 乌干达(Uganda)(中低收入)体现了“战略文件与现实世 界落实之间的鸿沟”。它既有《乌干达2040 愿景》 (Uganda Vision 2040),又有一摞限制性通信法律 (《2010 年通信拦截监管法》、《2011年计算机滥用法》、 《2013 年乌干达通信法》)以及反复出现的选举时段断网。 Matsiko 与 Kersting (2023) 研究了那种国家限制之下自 下而上的数字参与。 MECM90002 . Global Data Policy & Governance AskSia Library SUBJECT BIBLE FACULTY OF ARTS . MEDIA & COMMUNICATIONS SEMESTER 1 . 2026 - A COMPLETE STUDY COMPANION Global Data Policy & Governance 全球数据政策与治理 HOW THE WORLD ACTUALLY GOVERNS THE BORDERLESS INTERNET - AND HOW TO ARGUE ONE SHARP PIECE OF IT. 墨尔本大学 MECM90002 · 双语学科精读 · 论文 / 报告/ 案例三大写作全攻略 MECM90002 . UNIVERSITY OF MELBOURNE 中英双语版 · BILINGUAL EDITION 英文主讲,中文随行 一 考试要点与术语保留英文原词 A region-by-region tour of how governments, blocs, platforms and global bodies try to regulate a digital sphere that ignores borders - built to serve the three written tasks that are this subject: a 2500-word argumentative essay, a 3-minute policymaker pitch, and a 1500-word case study. There is no exam. Success turns on narrowing a vast debate into one defensible argument. Independent study companion Not affiliated with or endorsed by the University of Melbourne Corrections: takedowns@asksia. ai PREFACE - HOW TO USE THIS BIBLE One field, three written tasks 一个领域,三项写作任务 Read it to understand the debate - then to win the marks 先读懂这场辩论 -- 再去赢得分数 This is not a transcript of the lectures. It is a self-contained companion to the whole field of global data governance - every key concept defined, every regional model mapped, every live debate laid out with both sides - written so the same pages serve you across the semester and feed directly into the writing that is assessed. There is no exam in this subject; everything here points at the essay, the presentation and the case study. 这并非课堂讲座的文字实录。它是一份自成体系、覆盖全球数据治理整个领域的伴学读本 -- 每个关键概念都有定义,每种区 域模式都被梳理映射,每场进行中的争论都摆出正反双方 -- 其写作方式让同样的几页内容能在整个学期为你服务,并直接对 接受评估的写作任务。本科目没有考试;这里的一切都指向论文、展示与案例研究。 A 1 . ORIENT 1 · 定位 You're new to the field. Read Ch 1-2 to build the toolkit: datafication, the fluid data ecology, sovereignty vs jurisdiction, and the regulation models. Every later chapter is these concepts applied to a region or a body. 你是这一领域的新手。读第 1-2 章来搭建工具箱:数据化、流动 的数据生态、主权对阵管辖权, 以及各监管模型。后面每一章都 是把这些概念应用到某个区域或 某个机构上。 B 2 . COMPARE 2 · 对比 You're mid-semester. Use the regional chapters (Europe, USA/UK/AU, Africa, Asia, Latin America) as a comparison bank. The whole subject is a tour - your edge is being able to set one region's approach against another with real cases. 你正处于学期中段。把各区域章 节(欧洲、美国/英国/澳大利 亚、非洲、亚洲、拉丁美洲)当 作一个比较题库来用。整个科目 是一场巡览 -- 你的优势在于能 用真实案例把一个区域的路径与 另一个区域对照起来。 C 3 · ARGUE[16]Source: asksia-bible-mecm90002-bilingual.pdfi The two recurring moves 两个反复出现的操作 Move 1 - narrow: add one actor + one instrument/incident + one dimension until the scope fits 2500 (essay) or 1500 (case) words. Move 2 - analyse: name a model from Ch 1-2 (sovereignty/jurisdiction; self/state/co-regulation; rights/market/state-control) and say where it breaks on your case. Every drill below trains one of these. 第 1招 -- 收窄:不断加上一个行动者+一项工具 / 事件+一个维度,直到范围适配 2500 字(论文)或 1500 字(案例)。第2招 -- 分析:点出第1-2 章的 一个模型(主权/管辖权;自我/国家/共同监管; 权利/市场/国家管控),并说出它在你的案例上从何 处失效。下面每一道演练都训练其中之一。 D1 NARROWING DRILL essay scope A classmate's draft essay opens: "This essay examines how data sovereignty is understood and applied by governments around the world, with reference to many countries and laws. " Rewrite this as a single narrowed thesis suitable for a 2500-word argument, and name the one model you would use as the lens. 一位同学的论文初稿这样开头:“本文考察世界各国政府如何理解和运用数据主权,并参照众多国家与法律。”把它重写为 一个适用于 2500 字论证的、收窄过的单一论点,并点出你将用作透镜的那一个模型。 MODEL ANSWER Model answer. "Russia's 2019 'Sovereign Internet' data-localisation regime is best read not as data protection but as a tool of state control - and that conflation is why national digital sovereignty, however legitimate elsewhere, fails as a justification here. " Lens: Couture & Toupin's plural sovereignty, contrasting national vs state-control framings. MECM90002 . Global Data Policy & Governance - 范例答案。“俄罗斯 2019年的‘主权互联网’数据本地化体制,最好不被解读为数据保护,而被解读为一种国家管控工具 -- 而 正是这种混淆,使得国家数字主权(无论在别处多么正当)在此处无法成立为正当理由。”透镜:Couture 与 Toupin 的多元 主权,对照国家(national) 与国家管控(state-control)两种框架表述。 Marker's eye. The fix is one country, one instrument, one dimension, and a contestable claim. "Many countries and laws" guarantees a survey; markers cap survey essays well below a defended argument no matter how much is 'covered. ' 评卷人视角。修正之道是一个国家、一项工具、一个维度,外加一个可争辩的主张。“众多国家与法律”注定写成综述;无论 “覆盖”了多少,评卷人都会把综述类论文的分数压得远低于一篇有立场可辩护的论证。 D2 NARROWING DRILL case-study scope Too broad for a 1500-word case study: "Platforms should do more to stop misinformation. " Narrow it to a targeted intervention with a named actor and lever, grounded in a real incident. 对一篇1500 字的案例研究而言过于宽泛:“平台应当采取更多措施来阻止错误信息。”把它收窄为一项有的放矢的干预, 带有指名道姓的行为主体与杠杆,并扎根于一桩真实事件。 MODEL ANSWER Model answer. "Following the failures exposed by Rohingya v Facebook, Meta should be designated a 'gatekeeper' under a DMA-style threshold and subjected to independent, audited reporting of its misinformation-moderation outcomes in at-risk markets - a co-regulatory duty, not a vague call for 'responsibility' . " 范例答案。“鉴于 Rohingya v Facebook 所暴露的种种失败,Meta 应被依据一道 DMA 式的门槛指定为‘守门人’,并就其在 高风险市场中错误信息审核结果,接受独立的、经审计的报告 -- 这是一项共同监管义务,而非对‘责任’的含糊呼吁。” Marker's eye. "Targeted" is the graded word. Look for a named actor (Meta), a named lever (audited reporting under a designation threshold) and a real anchor (Rohingya v Facebook). "Do more" scores nothing. 评卷人视角。“有的放矢(targeted)”是那个被计分的词。要看到指名道姓的行为主体(Meta)、指名道姓的杠杆(在指定门 槛下经审计的报告)以及一个真实的锚点(Rohingya v Facebook)。“采取更多措施”一分不得。 MECM90002 . Global Data Policy & Governance CH 11 . ESSAY VARIANTS - CHAPTER 11 . ESSAY - QUESTION VARIANTS A1 REHEARSAL . ARGUE Practise the three lanes on fresh prompts 用新题目练习三条赛道 AskSia-written variants that transform the set questions - pick one, draft a thesis 由 AskSia 撰写、改造指定题目的变体 -- 选一道,起草一个论点 These are our own question variants, built to rehearse the same three lanes as the real essay (national rationale · transnational brief · platform responsibility) without reproducing the official prompts. For each, write the one-sentence thesis and the model you would deploy. Treat them as timed planning sprints (10 minutes each).[24]Source: asksia-bible-mecm90002-bilingual.pdf★ Pulling the toolkit together into one argument 把整个工具箱凝聚成一个论点 A high-scoring move chains three or four of these concepts into a single sentence of analysis, then narrows hard: the fluid data ecology erodes jurisdiction, so states assert data sovereignty through localisation - risking a splinternet that mostly serves the powerful. Now pick ONE country, platform or instrument and defend ONE intervention. That is the difference between a survey and an argument - and it is exactly what Ch 10 drills. 高分的做法,是把三四个这样的概念串进一句分析里,然后狠狠收窄:流动的数据生态侵蚀了管辖权,于是各国通过本 地化来主张数据主权 -- 却冒着造就一个主要为强者服务的分裂的互联网的风险。现在挑出一个国家、平台或法律工 具,并为一项干预辩护。这正是概述与论点之间的差别 -- 也正是第10章反复操练的。 MECM90002 . Global Data Policy & Governance MODELS - PHILOSOPHIES MASTER GRID - THE ANALYTICAL ENGINE Three philosophies of data regulation 数据监管的三种哲学 Rights-based . market-based . state-control - the master grid you plug every case into 基于权利 · 基于市场 · 国家管控 -- 可套用任何案例的主框架 Almost every regulatory move in this subject - a law, a fine, a takedown, a localisation rule - expresses one of three underlying philosophies about what data is for and who should steer it. Holding these three apart is the single most useful analytical habit for the essay and the case study: name the philosophy, and the actor's logic, its blind spots and its likely critics all follow. Treat the grid below as a lens, not a box - most real regimes blend two of the three. 本科目中几乎每一项监管举动 -- 一部法律、一笔罚款、一次下架、一条本地化规则 -- 都表达了关于数据为何而存在以及应 由谁来掌舵的三种底层哲学之一。把这三者拆分开来,是论文与案例研究中最有用的单一分析习惯:点出哲学,行为主体的逻 辑、其盲区及其可能的批评者便随之而来。把下面的网格当作一面透镜,而非一个格子 -- 大多数现实中的体制都混合了三者 中的两种。 i How the written assessments use this 书面评估如何运用这一点 In the essay and case study, the strongest move is comparative: take one issue (say, AI scraping or content moderation) and show how a rights-based, a market-based and a state-control regime would each frame and answer it - then argue which best serves the public interest. The grid is your skeleton; the regional chapters supply the flesh. 在论文与案例研究中,最有力的做法是比较式的:取一个议题(比如 AI抓取或内容审核),展示以权利为本、以市场为 本、以国家管控为本的三种体制各自会如何框定并回应它 -- 再论证哪一种最能服务于公共利益。这个矩阵是你的骨 架;各地区章节提供血肉。 Rights-based Market-based State-control Core value the citizen / 'data subject'; human rights, privacy, free expression data as a tradable commodity; efficiency, 'free flow', consumer choice the state's authority over content & infrastructure; security, ideology Who leads supra-national + national law setting a floor of rights the private sector / market; light- touch government central government as the primary - often sole - actor Default stance regulate to protect the individual。

重要提醒：资料里明确写到本课没有考试，你“考前复习”真正要复的是：如何把题目收窄、如何写出论证、如何做3分钟政策推介、以及A2如何喂养A3（不要A2随便选题）[11]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.[15]Source: asksia-bible-mecm90002-bilingual.pdfHow A2 feeds A3 - don't waste the pitch A2 如何喂养 A3 -- 别浪费了那次陈述 1 Pitch deliberately. Choose the A2 topic knowing it must survive into a 1500-word case study. A throwaway pitch costs you twice. 有意地推介。在选择A2 题目时就要知道,它必须能存活到一篇1500 字的案例研究里。一次随手的推介会让你赔上两次。 2 Harvest the feedback. Note every objection raised in the room - those are the counter-arguments your A3 must pre-empt. 收割反馈。把现场提出的每一个反对意见记下来 -- 那些就是你的A3必须预先化解的反方论点。 3 Refine, don't restart. The brief allows only minor topic refinements between A2 and A3. Sharpen the same intervention; don't switch horses. 精炼,别推倒重来。作业说明只允许在A2与A3 之间做微小的题目调整。把同一项干预磨利;别中途换马。 MECM90002 . Global Data Policy & Governance CH 10 . CASE STUDY - CHAPTER 10 . THE CASE STUDY (A3, 25%) CASE STUDY . 1500 WORDS One bounded case, analysed in depth 一个界定清晰的案例,深入剖析 Not an essay - a rigorous 1500-word study of a single case, building on your AZ pitch 不是论文 -- 而是对单一案例的严谨1500 字研究,在你的 A2 陈述基础上展开 A3 is not a second essay. It is an in-depth look at one tightly-bounded case - the intervention you pitched in A2, now worked up with rigour. Word count is 1500 (+10%, excluding figures, tables and the reference list). Pick a lane, then run the three-part structure. A3 不是第二篇论文。它是对一个范围收得很紧的案例的深入考察 -- 就是你在A2 推介的那项干预,如今被严谨地深加工。 字数为 1500(±10%,不含图、表与参考文献列表)。先选一条赛道,再运用这套三段式结构。 Pick one of three topic lanes 从三条主题赛道中选一条 Lane What it is Worked example 1 . Company / practice A company, a social-media practice, or a tech field x one policy dimension. A platform x content moderation of misinformation (Meta / Rohingya v Facebook). 2 . Country & one policy A country and one of its national digital policies - or the absence of one (the gap itself can be the intervention). Australia's Online Safety Act and the 'duty of care' model; or a missing consumer-loT rule. 3 . International instrument A specific international policy or organisation, anywhere. The UN Global Digital Compact; the OECD AI Principles; the ITU's spectrum role.。

另外，AI 使用规则非常严格：可以用来理解术语、找/概括来源（但要声明）；不允许让 AI 写句子段落、改写润色、或翻译你的作业成英文；Turnitin 会扫描，未声明可能触发诚信调查[9]Source: asksia-bible-mecm90002-bilingual.pdf核对 AI 给出的想法是否经得起可靠来源的检验 · 用 Al 来解释或定义难懂的术语 · 帮助查找来源 · 为你自 己的理解而对学术来源进行概括或翻译 · 非 AI 的拼写 / 语法工具(关闭 AI 功能的 Grammarly / Hemingway)。所有这些都必须在你的提交中声明。 ! Not permitted - ever 绝不允许 Al writing whole sentences / paragraphs / papers . presenting AI output as your own research or analysis · citing Al output as a source . Al rewriting / rephrasing / 'improving' your text . Al translating your assignment into English. A chatbot is a next- word predictor that hallucinates - treated like paying a ghost-writer. Al 撰写整句/ 整段/整篇论文 · 把 AI 的输出当作你 自己的研究或分析呈现 · 把 AI 的输出作为来源引用 · 让 Al 改写/重述/“润色”你的文字 · 让 AI 把你的作 业翻译成英文。聊天机器人是一个会胡编的下一词预 测器 -- 会被等同于花钱雇了一个代笔人。 MECM90002 . Global Data Policy & Governance ! Late penalty is real and compounding 迟交罚分真实存在且会累积 5% of available marks per day (or part-day) without an approved extension; not accepted at all after 20 University business days. Extensions ≤10 working days are requested by email to the coordinator/tutor; longer needs Special Consideration. Submit early enough to survive an upload glitch. 未获批准延期者,每天(或不足一天按一天计)扣除 可得分数的5%;超过20个大学工作日后将完全不予 接受。≤10个工作日的延期,通过电子邮件向课程协 调员/导师申请;更长的则需 Special Consideration。要提早提交,以便挺过一次上传故 障。 ★ Declare it - every time, even for permitted use 声明它 -- 每一次都要,即便是被允许的使用 Any AI use at any stage must be declared in the submission (the University links disclosure templates). Undeclared or out-of-scope use triggers an academic-integrity investigation. The safe habit: keep a one-line log as you go ("used ChatGPT to define 'data localisation', verified against Kuner") and paste it into the declaration. Disclosure of permitted use costs you nothing; the alternative can cost you the subject. 在任何阶段对 AI的任何使用,都必须在提交中声明(大学提供了披露模板的链接)。未声明或超出范围的使用会触发学 术诚信调查。稳妥的习惯:边做边记一行日志(“用 ChatGPT 定义‘data localisation’,并对照 Kuner 核实”),再粘贴 进声明里。披露被许可的使用对你毫无代价;另一种做法则可能让你赔上这门课。 All submitted work must be your own. Permitted Al helps you understand and find - never write, rephrase, translate, or be cited. Declare every use. 所有提交的作业都必须是你本人完成的。被许可的 AI 帮你理解与查找 -- 绝不替你撰写、改写、翻译或被引用。每一次 使用都要声明。 MPF1310 . STUDENT ACADEMIC INTEGRITY MECM90002 . Global Data Policy & Governance CH 11 . NARROWING DRILLS - CHAPTER 11 . PRACTICE BANK NO EXAM . DRILL THE CRAFT Drill the writing, with model answers 操练写作,附范文 Ten AskSia-authored items in the subject's style - narrowing, models, case planning, concepts 十道由 AskSia 按本课程风格撰写的题目 -- 收窄、模式、案例规划、概念 There is no exam to cram, so these drills target the moves the three assessments reward: narrowing a huge question, naming the right model and where it breaks, planning a case study, and the concept checks that keep your analysis honest. Work each one cold, then read the model answer and the marker's-eye note. 这里没有需要临时抱佛脚的考试,所以这些练习针对的是三项评估所奖励的招式:收窄一个宏大问题、点出正确的模型及其失 效之处、规划一份案例研究,以及那些让你的分析保持诚实的概念检查。每一道都先冷启动独立做完,再去读范例答案与评卷 人视角的批注。 ★ How to use this bank 如何使用这个题库 Cover the answer. These are skill drills, not facts to memorise. Write or speak your response first, then compare. The narrowing drills are the highest-value - if you can reliably cut a planet-sized question to one defensible claim, you have the whole subject. These items are AskSia's own, built in the subject's style; they paraphrase and transform the real prompts - never reproduce them. Always answer the official brief on Canvas. 把答案盖住。这些是技能演练,不是要背的事实。先 写出或说出你的回答,再对照。收窄类演练是价值最 高的 -- 如果你能可靠地把一个庞大无边的问题切到 一个可辩护的主张,你就掌握了整门课。这些题目是 AskSia 自创的,按本课风格构建;它们对真实题目做 了转述与改造 -- 绝不照搬。请始终回答 Canvas 上 的官方说明。[12]Source: asksia-bible-mecm90002-bilingual.pdfWeight Length Due / timing A1 . Argumentative essay 50% 2500w 19 Apr A2 . Oral presentation 25% 3 min / 3 slides Wk 10-11 A3 . Written case study 25% 1500w 7 Jun A1: pick one of three set questions (national policy rationale · UN- style transnational brief . platform responsibility); ~ 20 references expected, peer-reviewed + policy/report/news. A2: step into a policymaker's shoes and pitch a policy intervention - it directly feeds A3. A3: develops that same intervention into an in-depth, single-case study (a company/practice, a country's policy or its absence, or one international instrument). A1: 从三道既定题目中选一道(国家政策依据 · UN 风格的 跨国简报 · 平台责任);预期约 20 篇参考文献,须包含同 行评审文献+政策 /报告/新闻。A2:设身处地成为一名决 策者,推介一项政策干预 -- 它直接为 A3 提供素材。A3: 将同一项干预深化为一份深入的单一案例研究(一家公司/ 一种做法、某国的政策或政策的缺位,或一项国际工具)。 The twohurdles&the AIrule 两道门槛与 AI 规则 ! Hurdles - clear both or you cannot pass 门槛 -- 两道都过,否则无法通过 Attendance: a minimum 80% of tutorials/seminars, monitored; falling below without documented evidence means you cannot pass. Submission: every piece must be submitted as a genuine attempt. Late work loses 5% of available marks per day without an approved extension. 出勤:导修课/研讨课出勤率至少为 80%,并受监 控;若无书面证明而低于此线,将无法通过。提交: 每一项作业都必须作为真实尝试提交。未获批准延期 的迟交,每天扣除可得分数的5%。 MECM90002 . Global Data Policy & Governance i AI policy in brief (MPF1310) AI 政策速览 (MPF1310) All submitted work must be your own. AI is permitted to define hard terms, locate or summarise sources for comprehension, and check spelling/grammar - with disclosure. It is not permitted to write or rephrase your sentences, to be cited as a source, or to translate your assignment. Turnitin auto-scans for AI; undeclared use triggers an integrity investigation. 所有提交的作业都必须是你本人完成的。AI 可以用来 解释难懂的术语、为帮助理解而查找或概括来源、检 查拼写与语法 -- 但必须声明。Al 不可以用来撰写或 改写你的句子、作为来源被引用,或翻译你的作业。 Turnitin 会自动扫描 AI 痕迹;未申报的使用会触发学 术诚信调查。 The one strategic insight behind every task ★ 贯穿每项任务背后的那一个战略洞见 The whole subject is a tour of a huge, global debate - but you are never asked to cover it. You are asked to narrow it: take one borderless problem and compress it into a single, sharp, evidence-backed policy argument. Essay = argue one position; presentation = pitch one intervention; case study = analyse one bounded case. Markers reward a tight scope and a defended claim, and punish broad description. Every chapter that follows gives you the concepts, regions and cases to make that one argument land.。

为了把你的“复习地图”做得更贴合你，我想先问你 4 个问题（请按编号回答即可）：

1. 你现在最急的是哪一项作业/环节：A1 论文、A2 3分钟展示、还是 A3 案例研究？截止时间离你还有多久？[11]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.
2. 你更想主打哪个“赛道/对象”：一个国家、一个平台/公司、还是一个国际组织工具（例如 OECD/UN 等）？[15]Source: asksia-bible-mecm90002-bilingual.pdfHow A2 feeds A3 - don't waste the pitch A2 如何喂养 A3 -- 别浪费了那次陈述 1 Pitch deliberately. Choose the A2 topic knowing it must survive into a 1500-word case study. A throwaway pitch costs you twice. 有意地推介。在选择A2 题目时就要知道,它必须能存活到一篇1500 字的案例研究里。一次随手的推介会让你赔上两次。 2 Harvest the feedback. Note every objection raised in the room - those are the counter-arguments your A3 must pre-empt. 收割反馈。把现场提出的每一个反对意见记下来 -- 那些就是你的A3必须预先化解的反方论点。 3 Refine, don't restart. The brief allows only minor topic refinements between A2 and A3. Sharpen the same intervention; don't switch horses. 精炼,别推倒重来。作业说明只允许在A2与A3 之间做微小的题目调整。把同一项干预磨利;别中途换马。 MECM90002 . Global Data Policy & Governance CH 10 . CASE STUDY - CHAPTER 10 . THE CASE STUDY (A3, 25%) CASE STUDY . 1500 WORDS One bounded case, analysed in depth 一个界定清晰的案例,深入剖析 Not an essay - a rigorous 1500-word study of a single case, building on your AZ pitch 不是论文 -- 而是对单一案例的严谨1500 字研究,在你的 A2 陈述基础上展开 A3 is not a second essay. It is an in-depth look at one tightly-bounded case - the intervention you pitched in A2, now worked up with rigour. Word count is 1500 (+10%, excluding figures, tables and the reference list). Pick a lane, then run the three-part structure. A3 不是第二篇论文。它是对一个范围收得很紧的案例的深入考察 -- 就是你在A2 推介的那项干预,如今被严谨地深加工。 字数为 1500(±10%,不含图、表与参考文献列表)。先选一条赛道,再运用这套三段式结构。 Pick one of three topic lanes 从三条主题赛道中选一条 Lane What it is Worked example 1 . Company / practice A company, a social-media practice, or a tech field x one policy dimension. A platform x content moderation of misinformation (Meta / Rohingya v Facebook). 2 . Country & one policy A country and one of its national digital policies - or the absence of one (the gap itself can be the intervention). Australia's Online Safety Act and the 'duty of care' model; or a missing consumer-loT rule. 3 . International instrument A specific international policy or organisation, anywhere. The UN Global Digital Compact; the OECD AI Principles; the ITU's spectrum role.
3. 你目前最不稳的能力是哪一个：把题目收窄、段落“分析而非描述”、找够并用好约20个来源、还是反方段落怎么写？[11]Source: asksia-bible-mecm90002-bilingual.pdfCH 10 . THE WRITING GAME - CHAPTER 10 . THE WRITING GAME ESSAY 50 . PITCH 25 . CASE 25 No exam - you are graded on what you write 没有考试 -- 你的成绩取决于你写的东西 Three tasks, one craft: narrow a huge debate into one defensible argument 三项任务,一门手艺:把庞大的辩论收窄为一个可辩护的论点 This subject is assessed entirely through writing and one short talk. There is no exam. So this chapter and the practice bank that follows are the most practically valuable pages in the bible: master the moves here and you have mastered the grade. Every task rewards the same core skill - narrowing a borderless, planet- sized debate into one specific, evidence-backed claim - and punishes broad description. 本科目完全通过写作与一次简短的口头展示来评估。这里没有考试。所以这一章以及随后的练习题库,是本宝典中最具实用价 值的几页:掌握这里的招式,你便掌握了分数。每一项任务都奖励同一项核心技能 -- 把一场无边界、行星尺度的争论收窄为 一个具体、有证据支撑的主张 -- 并惩罚宽泛的描述。 50% ESSAY . 2500W . 19 APR 论文 · 2500 字 · 4月19日 25% PITCH . 3 MIN / 3 SLIDES 推介 · 3 分钟 /3 张幻灯片 25% CASE STUDY . 1500W . 7 JUN 案例研究 · 1500 字 · 6月7日 80% TUTORIAL ATTENDANCE HURDLE 导修出勤硬性要求 The tasks are a pipeline, not three islands 三项任务是一条流水线,而非三座孤岛 The smartest move in the subject is to treat A2 and A3 as one project in two stages, and to let the reading you do for A1 seed both. They run in this order: 本科目中最聪明的做法,是把 A2 与 A3 当作分两个阶段的 同一个项目,并让你为 A1 所做的阅读为二者播下种子。它 们按如下顺序进行: 1 A1 . Argumentative essay (50%). Pick one of three set questions, take a position, defend it across 2500 words with ~20 sources. Mid- semester (due 19 Apr). A1 · 论辩型论文(50%)。从三个设定问题中挑一个,提出 一个立场,用2500 字、约20条来源为之辩护。学期中 (4月19日截止)。 2 A2 · Oral pitch (25%). Step into a policymaker's shoes; in 3 minutes / 3 slides name one problem and pitch one intervention. Early-stage ideas are fine - the point is to test the idea and harvest peer feedback (Weeks 10-11). A2· 口头推介(25%)。站到一位政策制定者的位置上;在 3 分钟/3张幻灯片里点明一个问题并推介一项干预。早期 阶段的想法没问题 -- 重点是检验这个想法并收集同伴反馈 (第 10-11周)。 ★ The one insight behind all three 三者背后的同一个洞见 You are never asked to cover global data governance - you are asked to narrow it. Essay = argue one position. Pitch = propose one intervention. Case study = analyse one bounded case. A marker can feel the difference between a paper that surveys "everything about data policy" and one that says one sharp, defended thing. Be the second paper. 从来没有要求你去覆盖全球数据治理- -要求你做的 是收窄它。论文=论证一个立场。推介=提出一项干 预。案例研究=分析一个边界清晰的案例。评分者能 感觉到:一篇通盘罗列“关于数据政策的一切”的论 文,和一篇只说一件犀利、有辩护之事的论文,二者 之间的差别。要做那第二种论文。 MECM90002 . Global Data Policy & Governance 3 A3 . Written case study (25%). Take the same intervention from A2 and work it up into a rigorous, in-depth 1500-word study of one bounded case (due 7 Jun). A2 is the rough cut; A3 is the polished film. A3 · 书面案例研究(25%)。取来A2 的同一项干预,把它 打磨成一项严谨、深入、1500字、针对一个边界清晰案例 的研究(6月7日截止)。A2是粗剪,A3 是成片。 ! Description vs argument - the single biggest mark-sink 描述对阵论证 -- 最大的失分黑洞 Description tells the reader what happened (GDPR was passed in 2018; it created the data subject). Argument tells them what to think and why (GDPR's extraterritorial reach is what makes it a de facto global standard - and that is precisely why a single bloc should not set the world's privacy floor). Every paragraph must earn its place by advancing a claim, not by reporting a fact.[17]Source: asksia-bible-mecm90002-bilingual.pdfback to the thesis and on to the next paragraph. MECM90002 . Global Data Policy & Governance i Worked CETAL (Q1, China localisation) CETAL 示范(Q1,中国本地化) C: Data-localisation is a defensible response to genuine sovereignty loss. E: China's 2017 Cybersecurity Law requires citizens' data to be stored onshore ("data comes in but cannot get out"). T: read through Couture & Toupin's national digital sovereignty. A: because the fluid ecology lets data drift to foreign servers beyond any court's reach (the jurisdiction 'breaking point'), onshoring is the only instrument that restores enforceable control - that is the analytical link description misses. L: but the same tool also enables censorship, which the next paragraph weighs. C:数据本地化是对真实主权流失的一种可辩护的回 应。E:中国 2017年的《网络安全法》要求公民数据 存储在境内(“数据进得来,但出不去”)。T:透过 Couture & Toupin 的国家数字主权来解读。A:因为 流动的生态让数据漂向任何法院都触及不到的外国服 务器(管辖权的“断点”),本地化是唯一能恢复可执行 控制的工具 -- 这正是描述所漏掉的分析环节。L:但 同一工具也使审查成为可能,下一段将对此加以权 衡。 ! The A in CETAL is where marks live CETAL 中的 A 才是分数所在 Most students nail C, E and T - topic sentence, a case, a citation - then stop. The Analysis beat (the explicit because linking evidence to claim through the theory) is what separates a 70 from an 85. If a paragraph has no sentence beginning "This matters because . . . " or "What this shows is . . . ", you have described, not argued. 多数学生都能搞定 C、E、T -- 主题句、一个案例、 一处引用 -- 然后就停了。那一拍分析(即把证据经 由理论显式地因为连到主张上)才是把 70分与 85 分 区分开来的东西。如果一个段落里没有一句以“这之所 以重要,是因为 . . . ”或“这所表明的是 …… ”开头的 话,那你就是在描述,而不是在论证。 ✓ Counter-argument earns critical-engagement marks 反驳论证能赢得批判性参与的分数 Devote at least one body paragraph to the strongest objection and answer it. For Q1, that is the 'splinternet' / over-reach critique; for Q3, 'platforms are too big to govern. ' Conceding and rebutting a real counter is the clearest signal of a postgraduate-level argument. 至少用一个主体段落去处理最强的反对意见并予以回 应。对于 Q1,那是“分裂的互联网”/过度触达的批 评;对于 Q3,则是“平台大到无法治理”。承认并驳斥 一个真实的反方,是研究生水准论证的最清晰信号。 Paragraph budget for 2500 words ★ 2500 字的段落预算 Intro ~280 . four body paras at ~420 = ~ 1680 . one counter-argument para ~330 . conclusion ~210. That is ~2500 with one paragraph spent on the objection. If you are writing a sixth body paragraph, your scope was too broad - go back to the funnel on the previous page. 引言约 280 字 · 四个主体段各约 420 字=约 1680 字 · 一个反方论点段约 330 字 · 结论约 210 字。这样就是约 2500 字,其中一段用于处理反对意见。如果你在写第六个主体段,那你的范围就太宽了 -- 回到上一页的漏斗去。 MECM90002 . Global Data Policy & Governance CH 10 . THE PITCH CHAPTER 10 . THE PITCH (A2, 25%) ORAL . 3 MIN / 3 SLIDES Three minutes, three slides, one policymaker 三分钟、三张幻灯片、一位决策者 Step into the role, name one problem, pitch one intervention - out loud, not read 进入角色,点明一个问题,提出一项干预 -- 开口说,而非照念 A2 asks you to become a policymaker and pitch a policy intervention in exactly three minutes across three slides. It is low-stakes by design: early-stage ideas are welcome, the room gives peer feedback, and - crucially - whatever you pitch becomes the seed of your A3 case study. Choose a topic you would happily live with for 1500 more words. A2 要求你成为一名决策者,并在恰好三分钟、三张幻灯片内推介一项政策干预。它在设计上低风险:欢迎早期阶段的点子, 现场会给予同伴反馈,而且 -- 至关重要的是 -- 你所推介的内容会成为你 A3 案例研究的种子。选一个你乐意再相处 1500 字的题目。 The three-slide architecture 三张幻灯片的结构
4. 你希望我接下来怎么帮你复习：做一页式“速记清单”、还是带你把一个具体选题从 0 到 1 规划出 thesis + 结构 + 反方 + 干预？

你随时可以告诉我：有哪些内容希望我以后也记住（例如你常犯的写作错误、你选定的案例/地区、你想用的理论模型），这样下次我们可以直接接着推进。

现在把你最想做的作业（A1/A2/A3）和一个你感兴趣的案例方向发我，我们就从“收窄成一句可辩护的 thesis”开始。