Investigating the Reliability and Factor Structure of Kalichman's Survey 2: Research Misconduct Questionnaire: A Post Hoc Analysis Among Biomedical Doctoral Students in Scandinavia

A precondition for reducing scientific misconduct is evidence about scientists' attitudes. We need reliable survey instruments, and this study investigates the reliability of Kalichman's Survey 2: research misconduct questionnaire. The study is a post hoc analysis of data from three surveys among biomedical doctoral students in Scandinavia (2010-2015). We perform reliability analysis, and exploratory and confirmatory factor analysis using a split-sample design as a partial validation. The results indicate that a reliable 13-item scale can be formed (Cronbach's alpha = .705), and factor analysis indicates that there are four reliable subscales each tapping a different construct: (a) general attitude to misconduct (alpha = .768), (b) attitude to personal misconduct (alpha = .784), (c) attitude to whistleblowing (alpha = .841), and (d) attitude to blameworthiness/punishment (alpha = .877). A full validation of the questionnaire requires further research. We, nevertheless, hope that the results will facilitate the increased use of the questionnaire in research.