JOUR90028 Fact-checking, Misinformation and AI
The Complete Study & Assessment Guide · S2 2026JOUR90028 Overview
- Semester Two journalism study at Melbourne
- Graduate fact checking and verification practice
- Three published assessment components in total
- Video explainer completes the assessment sequence
Reliable fact-checking depends on reproducible evidence routes rather than a single verification tool
Fact-checking, Misinformation and AI treats verification as an evidence practice rather than a verdict generated from one tool.
- Choose a checkable claim Separate a verifiable assertion from opinion, prediction or a bundle of claims.
- Preserve the verification trail Record original context, searches, sources and uncertainty so another reader can retrace the work.
- Verify media in layers Combine provenance, content, location, time and source checks instead of trusting one detector.
- Match verdict to evidence Use language no stronger than the record and correct visibly when the record changes.
How JOUR90028 is assessed
| Component | Weight | Format |
|---|---|---|
| Assignment One: In-class written assessment | 40% | Individual in-class short answers |
| Assignment Two: Pitch | 20% | Individual fact-check pitch |
| Assignment Three: Final assignment | 40% | Video explainer and reflection |
The current assessment summary publishes three weighted components totalling 100%. The 2026 subject guide separately states a minimum attendance requirement of 80%; it does not label an assessment component as a hurdle in the captured text.
Assessment structure
JOUR90028 segment widths reproduce the supported published weights; consult the notes for conflicts or unconfirmed details.
Current JOUR90028 dates
| Date | Item | Control |
|---|---|---|
| 19 August 2026 | In-class written assessment | Current assessment summary |
| 18 September 2026 | Pitch | Current assessment summary |
| 2 November 2026 | Video explainer | Current assessment summary |
Dates are as published in captured 2026 course materials. Confirm exact deadlines and submission settings in the live LMS.
What JOUR90028 covers
Fact-checking, Misinformation and AI treats verification as an evidence practice rather than a verdict generated from one tool. The subject moves from information disorder and claim selection through source, image and video checks to transparent explainers and corrections.
Information Disorder and Checkworthiness
misinformation · disinformation · malinformation · propaganda · claims worth checking02Claim Triage and Source Verification
atomic claims · primary evidence · lateral reading · source motive · triangulation03Images, Video and Media Provenance
reverse search · provenance · geolocation · temporal verification04Synthetic Media and AI Verification
synthetic media · disclosure · provenance · corroboration05Verdicts, Explainers and Public Correction
evidence matrices · verdict scales · explanatory scripts · attribution · correction practiceThe subject moves from information disorder and claim selection through source, image and video checks to transparent explainers and corrections. This fact-checking guide follows the captured Semester 2 or Spring 2026 teaching package and separates verified course facts from independently authored practice.
JOUR90028 does not infer missing assessment rules, current dates or official questions from neighbouring courses.
Intent distinguishes categories conceptually but is often difficult to establish from a single post, so reporting should separate observed content from inferred motive
Information Disorder and Checkworthiness
Information Disorder and Checkworthiness connects Misinformation, Disinformation, Malinformation and Checkworthiness.
Intent distinguishes categories conceptually but is often difficult to establish from a single post, so reporting should separate observed content from inferred motive. Misinformation fixes the starting object, Disinformation carries a relation or operation, and Checkworthiness to test the reach of the conclusion.
A category should describe the evidence currently available; absence of proven intent is not evidence that intent was harmless or deceptive.
Claim Triage and Source Verification
Claim Triage and Source Verification connects Atomic claim, Primary source, Lateral reading and Triangulation.
Claim triage separates names, dates, quantities, causes and attributed statements so one true fragment cannot validate the full narrative. Atomic claim fixes the starting object, Primary source carries a relation or operation, and Triangulation to test the reach of the conclusion.
Multiple links do not constitute independent confirmation when they all inherit the same unverified source, dataset or quotation.
Images, Video and Media Provenance
Images, Video and Media Provenance connects Reverse image search, Media provenance and Geolocation. An earlier appearance can reveal that genuine media has been relabelled, cropped or detached from its original event.
Reverse image search fixes the starting object, Media provenance carries a relation or operation, and Geolocation to test the reach of the conclusion.
Image matching and geolocation support a provenance claim only for the media examined; they do not establish the uploader's intent or every caption attached to later copies.
Synthetic Media and AI Verification
Synthetic Media and AI Verification connects Synthetic media and Media provenance.
Detection tools can produce unstable scores across compression and editing; source history, content consistency and corroborating evidence remain essential. Synthetic media fixes the starting object, Media provenance carries a relation or operation, and Media provenance to test the reach of the conclusion.
A detector score is one diagnostic signal, not proof of generation or authenticity; the verdict remains bounded by provenance, disclosure and corroborating evidence.
Verdicts, Explainers and Public Correction
Verdicts, Explainers and Public Correction connects Evidence matrix, Verdict scale, Fact check explainer and Correction record.
A verdict scale should distinguish falsehood, missing context, unsupported claims and genuine uncertainty without implying precision the evidence cannot bear. Evidence matrix fixes the starting object, Verdict scale carries a relation or operation, and Correction record to test the reach of the conclusion.
A verdict applies to the exact claim and evidence date stated; it should not be generalised to a broader actor, event or intention.
The current assessment summary publishes three weighted components totalling 100%
The current assessment summary publishes three weighted components totalling 100%.
The 2026 subject guide separately states a minimum attendance requirement of 80%; it does not label an assessment component as a hurdle in the captured text.
JOUR90028 planning translates each published task into deliverable, evidence, process and verification requirements. JOUR90028 operations remain controlled by the live site's submission format, extensions, collaboration, AI use and dates.
JOUR90028 evidence gaps remain visible instead of becoming a tidy but unsupported table.
Without Misinformation or Reverse image search as a control
Study JOUR90028 by retrieving one chapter relation, applying it to an independent case and changing one controlling fact.
JOUR90028 comparison uses the source-defined limit and records the exact assumption, population, record, timescale or design that prevents a broader claim.
JOUR90028 transfer therefore retains its discipline rather than ending in a universal recommendation template.
A fact-check must include a reproducible verification trail, a verdict calibrated to the available evidence and a clear account of what remains unknown before publication
End a fact-check with a reproducible verification trail, a verdict calibrated to the available evidence and a clear account of what remains unknown before publication.
Integrated media and communications transfer
- 2Define the source-supported starting object.
- 2Trace the relation and test a changed case.
- 2State the discipline-specific evidence boundary.
Key terms
- Misinformation
- False or misleading information shared without the defining purpose of causing harm.
- Disinformation
- False or misleading information created or shared deliberately to deceive or cause harm.
- Malinformation
- Genuine information used out of context or strategically disclosed to cause harm.
- Checkworthiness
- The public significance, factual verifiability and potential harm that make a claim worth prioritising.
- Atomic claim
- One bounded factual assertion that can be checked without silently changing the proposition.
- Primary source
- Evidence produced by a person, record or institution directly connected to the event or data in question.
- Lateral reading
- Leaving a page to investigate its publisher, reputation, evidence and outside coverage in other sources.
- Triangulation
- Comparison of independent evidence routes to test whether they converge, conflict or leave a gap.
- Reverse image search
- A search using image content to locate earlier appearances, variants and contextual pages.
- Media provenance
- The origin, custody, transformations and publication history of an image, audio file or video.
JOUR90028 FAQ
What verification problem connects the fact-checking chapters?
Within JOUR90028, concepts connect to disciplined evidence and a bounded conclusion. Begin with the decision or explanatory object, show the relation carrying the analysis and identify the source, design or condition that limits transfer.
How is the current fact-checking assessment weighted?
Within JOUR90028, the captured current table contains 3 weighted components whose published weights sum to 100%. Exact submission settings, permitted resources and later amendments remain controlled by the live learning site.
Are the verification cases official assignment questions?
Within JOUR90028, the cases and point allocations are independently written study aids, not official questions. They rehearse course concepts and evidence moves without reproducing a current university prompt, rubric or confidential solution.
How should provenance terms support retrieval?
Within JOUR90028, each glossary term is a retrieval cue for a noun concept connected to an observation, relation and limiting condition. A memorised definition opens the analysis; application determines whether the concept fits the case.
What makes a changed-claim verdict defensible?
Within JOUR90028, a changed case alters one controlling fact, holds unrelated conditions stable and traces the first consequence. The answer states whether the result remains, narrows or reverses and identifies the evidence responsible.
What minimum attendance requirement applies?
The 2026 subject guide states a minimum attendance requirement of 80%. It also explains that one in-class assignment contributes 40% of the final mark, so attendance and the weighted assessment table describe separate parts of the subject requirements.
How should a published fact-check conclude?
Within JOUR90028, use this discipline-specific final check: End a fact-check with a reproducible verification trail, a verdict calibrated to the available evidence and a clear account of what remains unknown before publication.
How to prepare for the assessments
JOUR90028 revision moves chapter by chapter: define the starting concept, trace the relation, work one independent counter-case and state the supported boundary. JOUR90028 finishes with this discipline-specific control: End a fact-check with a reproducible verification trail, a verdict calibrated to the available evidence and a clear account of what remains unknown before publication.
Your AI Media tutor for JOUR90028
Stuck on a hard JOUR90028 question? Sia is AskSia’s AI Media and Communications tutor — ask any JOUR90028 Fact-checking, Misinformation and AI question and get a clear, step-by-step explanation grounded in how the course is actually taught and assessed. Read this whole study guide free, then take your hardest questions to Sia.
