BUSS6002 · Data Science In Business
Organisational Data & Analytical Capabilities
Week 2 of BUSS6002 is the “business and information-systems” framing of data science, and it is where quant-anxious students leave easy marks on the table. It covers how data flows through an organisation as a lifecycle (collect → store → process → analyse → act → feedback) that closes into a learning loop, and how a data-science team is structured along a continuum from the business domain to the IT domain.
It then introduces three classification frameworks that the mid-semester and final test as MCQ and short-answer items: the four decision contexts (simple, complicated, complex, chaotic), the four analytical capabilities (descriptive, diagnostic, predictive, prescriptive), and the statistical data-type taxonomy (categorical vs numerical, with their sub-types). Master the one-line triggers and you bank the cheapest marks in the unit.
What this chapter covers
- 011. Data lifecycle — collect → store → process → analyse → act → archive/feedback, run as a continuous cycle
- 022. Learning loop — sense → decide → act → feedback, so plans are checked against reality and adjusted
- 033. Data-science team continuum — roles from the business domain through analytics to the IT domain; always a team effort
- 044. Decision contexts (Cynefin) — simple, complicated, complex, chaotic, each defined by its cause–effect relationship
- 055. Decision framework — diagnoses which context you are in so you avoid a misaligned management style
- 066. Four analytical capabilities — descriptive (what), diagnostic (why), predictive (what next), prescriptive (what to do)
- 077. Advanced flavours — social physics, visual analytics, and deep learning as a limited-interpretability black box
- 088. Data fundamentals — variable vs value; categorical {nominal, ordinal} vs numerical {discrete, continuous}
- 099. Formats and documentation — CSV vs JSON vs Excel, and the data dictionary needed before EDA
Match each scenario to its analytics type
- +1A reports what happened (a summary of the past) → Descriptive analytics.
- +1B explains why it happened (root cause of a past outcome) → Diagnostic analytics.
- +1C estimates a future value (next week's demand) → Predictive analytics.
- +1D recommends an action (the reorder quantity). Even though a forecast sits inside it, the whole-system output is a decision → Prescriptive analytics.
Key terms
- Data lifecycle
- The repeatable path data takes through an organisation — collect, store, process, analyse, act, then archive and feed back — run as a continuous cycle. The advantage compounds only when acting on data generates new data that re-enters the cycle.
- Learning loop
- The decision-level loop sense → decide → act → feedback. The organisation senses from data, decides a response, acts, then measures the result and uses it to improve the next decision; a firm that never acts or never feeds back has broken the loop.
- Decision context
- A situation classified by its cause–effect relationship, which determines the appropriate response. The four contexts (from Snowden & Boone's Cynefin) are simple, complicated, complex and chaotic.
- Decision framework
- A tool that helps a decision-maker diagnose which decision context they are in, so they apply a context-appropriate response and avoid a misaligned management style.
- Analytical capabilities
- The four types of analytics used individually or together: descriptive (what happened), diagnostic (why it happened), predictive (what is likely to happen) and prescriptive (what to do about it). They form a ladder of rising decision value.
- Deep learning (black box)
- A powerful analytics flavour whose internal workings are not transparent, giving limited interpretability and poor transferability across contexts, and requiring situational and behavioural stability — a key limitation the unit flags.
- Statistical data type
- The classification of a variable as categorical (nominal = unordered, e.g. state; ordinal = meaningfully ordered, e.g. grades) or numerical (discrete = countable/listable; continuous = describable only by intervals). It governs which techniques are valid.
- Data dictionary
- A document listing each variable with its data type, possible values (if categorical) and a description. It is essential before exploratory data analysis because it tells you what each column is and what its codes mean.
Organisational Data & Analytical Capabilities FAQ
Is Week 2 actually examinable or just background?
It is directly examinable and is among the cheapest marks in BUSS6002. It shows up as 1-mark MCQ classification items (analytics type, decision context, data type) and short-answer definition or justify questions in both the 25% mid-semester and the cumulative 45% final. Students who over-focus on linear algebra routinely drop these easy marks.
How do I tell predictive from prescriptive analytics?
Predictive analytics estimates what is likely to happen (a forecast or probability); prescriptive analytics recommends what to do about it (an action or decision). Classify by the whole system's output: if it delivers a recommended action — even with a forecast inside — it is prescriptive. The highest rung the system reaches wins.
What is the difference between a complicated and a complex decision context?
Complicated problems have a right answer that experts or analysis can find (intricate but solvable, such as tuning a supply chain). Complex problems have cause–effect that is only clear in hindsight, so you must probe and experiment first (such as launching into an untested market). Examiners deliberately place these two side by side.
What is the difference between ordinal and nominal data?
Both are categorical, but ordinal categories have a meaningful order (grades FA < PS < CR < DI < HD) while nominal categories do not (home state, marital status). The distinction depends on whether order is meaningful, not on whether the values look like text — grades are stored as text yet are ordinal.
Why does the course ask for both the statistical type and the Python type?
Because they are different labelings and a drag-and-drop item often asks for both. For example an ordinal grade is statistically ordinal but stored as a Python str; a count of orders is discrete numerical but a Python int; a balance is continuous numerical but a float. Knowing both shows you understand how a concept maps to code.
Is this guide official or affiliated with the University of Sydney?
No. This is an independent AskSia study resource for BUSS6002. It is not produced, endorsed by or affiliated with the University of Sydney; always confirm definitions, dates and assessment details against your official Canvas unit outline.
Exam move
Treat Week 2 as guaranteed, low-effort marks and bank them early by memorising four crisp checklists: the data lifecycle (collect → store → process → analyse → act → feedback), the four decision contexts with their cause–effect triggers (simple = known, complicated = expert-knowable, complex = hindsight-only, chaotic = no link and no time), the analytics ladder (descriptive → diagnostic → predictive → prescriptive), and the data-type taxonomy (categorical {nominal, ordinal} vs numerical {discrete, continuous}). Practise classifying short business scenarios under each framework, because that is exactly how the MCQ and short-answer items are set. In short-answer decision-context questions, always justify the label by naming the cause–effect signature — the justification carries its own mark. Finally, when tagging data, give both the statistical type and the natural Python type, and remember the classic traps: a recommendation is prescriptive, order (not text) decides ordinal vs nominal, and discrete can still be infinite.
