PUBLISHED 26 OCTOBER 2026 · METHODOLOGY
This page sets out the standards for our research programme. It is not a completed study or a claim that we have already surveyed Australian households or tested local services.
Start with a question, not an answer.
Before collecting information, we define the decision a reader faces, the question we can realistically investigate, the geography, the period and the type of evidence required. A narrow, well-supported result is more useful than a broad claim the evidence cannot establish.
Say what kind of evidence we used.
A source-based analysis uses existing statistics or published research. A public-information study records a defined sample of accessible service information. Interviews and surveys describe the experiences of their participants. First-hand testing is identified only when the service or product has actually been tested. We keep these methods distinct.
Make the comparison understandable.
Reports will identify data sources, reference periods, collection dates, locations and relevant definitions. Price comparisons need a consistent scope: advertised starting prices, complete quotes and amounts actually paid are different measures. Missing prices are recorded as missing, not treated as zero.
Where we sample services, we will explain how the sample was chosen, what was excluded and how many observations were usable. A local or self-selected sample does not automatically represent Australia. Access to a public page does not establish that a provider is available, qualified or suitable.
Separate findings from interpretation.
We will label calculations, observations, assumptions and possible explanations. Spending in dollars is not the same as the quantity purchased; travel trips are not unique people; exposure to AI is not a count of jobs lost. A change occurring alongside another change does not by itself establish a cause.
Help people make a decision.
Research should leave readers with a clearer comparison, useful questions or a practical tool. Health and care research focuses on navigating services and understanding evidence about access. Price transparency is not a measure of clinical quality. Clinical claims require appropriate expertise and review; general research does not provide individual treatment advice.
Education research identifies the service type and the meaning of any official quality measure. We do not turn incomplete information into unsupported “best provider” rankings.
Use AI with accountable review.
AI may assist with organising sources, calculations, drafts and presentation. Evidence must remain traceable, calculations must be checked and limitations must be reviewed before publication. AI-generated responses are not survey participants, interviews, customer reviews or proof of first-hand experience.
Respect people and source material.
Original research involving people needs a clear purpose and appropriate consent. We avoid collecting unnecessary personal information and publish data only when it is appropriate and permitted to do so. We link to sources and check the conditions for reusing datasets, text and images.
Keep commercial interests visible.
We currently have no paid research sponsors or provider placements. If introduced, funding, sponsorship and relevant commercial relationships will be disclosed. Paying for a listing will not purchase a favourable research finding. Information supplied by a business will be identified and checked where possible.
Publish carefully, then maintain the record.
Our editorial target is one research release every two weeks, subject to evidence and review. It is not an automated publishing schedule. Planned studies remain labelled as planned until completed. Published reports will carry their own publication date, data period, methods, sources and limitations. Material corrections will explain what changed and when, rather than silently replacing an earlier conclusion.