Is the Study Valid?
Potential reasons for an article's rejection include insufficient statistical power or use of an unvalidated research instrument. Additionally, different study types bring different potential problems, some of which are summarised below.
- Was a control group (or groups) required and if controls were used were they appropriate?
- Were the randomisation, allocation concealment and/or blinding or masking adequate?
- If the trial was randomised, was it registered?
- Did the study deviate significantly from the protocol?
- Is a power calculation provided and is it sufficient?
- Were the analyses preplanned?
- Were the harms analysed?
- Were the search terms sufficiently defined?
- Was the search inappropriately limited only to recent studies or English-language research?
- Did only one researcher appraise the studies (and if so were sufficient steps taken to minimise bias)?
- Is there important confounding and bias (owing to absent or incomplete measurement of/adjustment for important factors)?
- Was the response rate robust (>65%)? If not, should a high response rate have been achievable?
- Is any important non-response bias accounted for and analysed?
- Is the study placed within a theoretical framework?
- Is the sampling strategy clearly described?
- Is the sampling driven by convenience or theory?
- Are the methods of data analysis clearly described and theoretically justified?
- Does the data analysis relate to the original research question?
- Is there a clear explanation of why the term 'pilot study' applies (see our instructions for authors)?
- Is the sample/study size justified?