A proposal for judging the trustworthiness of research findings.

Abstract

This paper offers a procedure for, and a description of the elements involved in judging how trustworthy a research finding is. The idea could be of value to the users of research evidence and to researchers themselves when setting out to create trustworthy evidence. The focus here is on active designs to address causal research questions, but the ideas can easily be extended to other types of research. Other than design, the elements suggested are sample size and quality, data quality, fidelity of intervention, and threats to validity. These are combined in a kind of 'sieve' to produce a judgement-based star-rating for the believability of a piece of research. Trustworthiness of research findings is currently an area with too little focus for the development of new researchers.