SACE Stage 2 Cross-disciplinary Studies cheat sheet

Inquiry cycle: define a question that no single discipline can answer → map contributing disciplines → gather evidence per discipline → integrate → evaluate → communicate.

A cross-disciplinary question must be open, contestable, and require at least two distinct knowledge bases; if one discipline suffices, the task is disciplinary, not cross-disciplinary.

Multidisciplinary = disciplines stay separate and parallel; interdisciplinary = concepts and methods are blended into a new synthesis; transdisciplinary = non-academic stakeholders co-produce the framing.

Method selection rule: choose the method matched to the evidence type, not the discipline you prefer - quantitative for measurable variables, qualitative for meaning and lived experience, mixed when the question has both parts.

Evidence hierarchy for inquiry: primary data > peer-reviewed secondary synthesis > expert commentary > popular media; record the tier of every source used.

Triangulation rule: a claim gains strength when at least two independent sources, using different methods, converge on it; convergence is not proof, only reduced uncertainty.

Ethics gate: any inquiry involving human participants needs informed consent, anonymity, right to withdraw, and a review of potential harm before data collection.

Scope control: state boundaries (time, place, population) at the start; unbounded questions cannot be answered with available time and evidence.

Inquiry log should record the question, source, date, method, and how each piece changed the direction of the investigation.

Each discipline has a paradigm: its own assumptions, accepted evidence, key concepts, and limits; naming the paradigm is the first step in using it.

Disciplinary boundary test: ask what a discipline would count as proof and what it would exclude; disagreement often comes from different evidence standards, not from bad reasoning.

Concept-transfer check: a concept borrowed across disciplines (e.g. 'system', 'resilience', 'cost') must be redefined in the new context, or the argument commits equivocation.

Scale rule: causes identified at one level (individual) may not explain outcomes at another (institutional, global); state the level of analysis for every claim.

Correlation–causation rule: a statistical association supports a causal claim only with a mechanism, temporal order, and control of confounders.

Model limits: every model simplifies; list what it omits and state how those omissions could change the conclusion.

Data literacy: distinguish sample from population, absolute from relative risk, and measured from modelled values before citing a figure.

Source evaluation: check author expertise, publication venue, date, funding, and whether the claim is replicated; a single study is provisional.

Integration rule: a synthesis should show how disciplines modify each other's findings, not merely place them side by side.

Argument anatomy: claim + reasons + evidence + warrant (the assumption linking evidence to claim); identify the warrant to find where an argument is vulnerable.

Common fallacies to detect: straw man, false dilemma, slippery slope, ad hominem, appeal to authority, circular reasoning, hasty generalisation, post hoc.

Counterargument protocol: state the strongest opposing view fairly, then respond with evidence or concede a limited point; a weak counterargument does not strengthen your case.

Assumption audit: list what must be true for the conclusion to hold; rank assumptions by how much damage their failure would cause.

Creative ideation rules: defer judgement during generation, aim for quantity, combine unrelated concepts, and reverse the problem to find new framings.

Divergent then convergent: generate broadly, then screen ideas against explicit criteria (feasibility, evidence, ethics, impact) rather than intuition alone.

Analogy check: an analogy is an illustration, not evidence; test whether the relevant structure matches, not just the surface similarity.

Uncertainty language: match wording to evidence strength - 'demonstrates' for strong replicated evidence, 'suggests' for single studies, 'may' for modelled projections.

Team roles should be assigned by task need (research, synthesis, editing, presenting) and rotated so no member is excluded from the intellectual work.

Group norms set at the start - meeting cadence, decision rule, file conventions, conflict procedure - reduce later coordination cost.

Decision rules: consensus for shared ownership, majority for time-pressured choices; record the decision, the rationale, and any dissent.

Conflict escalation ladder: clarify the disagreement → return to shared criteria → seek third-party mediation; personal attack is never a valid step.

Contribution tracking: log who did what and when; this supports fair assessment and makes individual accountability visible.

Communication design: identify audience, purpose, and channel; adjust register, jargon level, and format to the audience's prior knowledge.

Visual and data presentation: label axes and units, keep scales honest, cite sources on the figure, and avoid decorative elements that distort comparison.

Reflection model: describe the experience → analyse what caused the outcome → evaluate against criteria → plan a specific changed action.

Feedback quality: specific, behavioural, timely, and tied to criteria; 'good job' is not feedback because it cannot be acted on.

Placeholder - this section is intentionally omitted; the four domains above are the only ones required.

Do not use.

Do not use.

Do not use.

Do not use.

Do not use.