How the Credibility Score works.

Can a reader find the method and repeat it?

A high score means the report publishes its full methodology, its questionnaire or coding rules, and enough detail to reproduce the rank. A low score means the method is a paragraph or two, key choices are unexplained, or the data and code are not available. Where a report relies on a survey, respondent selection is part of the method: was it truly random, or targeted to known people for easy access, and are the criteria published? A method that cannot be checked cannot be trusted.

Can the data be traced to a named, checkable source, and was it measured soundly?

Data transparency is scored strictly. The report must publish its raw data, sampling frame, and processing steps (cleaning, weighting, imputation) so an independent researcher can reproduce the result. Sample quality is scored explicitly too. For any report that relies on survey or expert-coding data, is the sample large enough for the claims made, does it cover the population across geography, demographics, and viewpoint, and was selection truly random? A report that does not publish its sample size, diversity, or randomness cannot prove its sample was not gamed, and that opacity is punished severely. Expert surveys are not automatically penalised: coding with a public codebook, named coders, and inter-coder reliability can outscore modelled or un-audited administrative data.

Does India’s rank make sense against the report’s own numbers?

We check India’s rank against the five countries immediately above and below it on the published sub-indicators, and count the reversals: countries ranked above India that score worse than India on the report’s own sub-indicators. A high score means the rank follows from the sub-indicators with few or no reversals. A low score means the headline rank contradicts the underlying data.

Is the concept what the India rank claims, and is India actually observed or regionally imputed?

A high score means the index measures what it says it measures, and India’s figure comes from data collected in India. A low score means the label promises one thing and the measurement delivers another, or India’s rank rests on imputed data. A rank with no error bars is false precision, and it loses points.

Who pays for the report, and does the funder have a stake in the outcome?

Independence requires that all funding sources and amounts are disclosed, structural governance insulates researchers from funders, and the publisher holds no advocacy stake in the ranking. A high score means the publisher is independent, its funding is fully disclosed, and no funder stands to gain from a particular ranking. A low score means the funding is undisclosed or concentrated, or the publisher is funded by an organisation with a stated position on the countries it ranks. Every report page also carries a funding and independence line stating the institution’s funding sources and governance, so the reader sees the independence picture directly.

Does the publisher admit mistakes and fix them?

A high score means the report publishes corrections, versions its data, and responds to substantive criticism. A low score means errors go uncorrected, older editions disappear without note, or the publisher ignores documented challenges. Five points is small, but it separates an institution that treats its ranking as a living measurement from one that treats it as a press release.