Thinking About Data, Research Methods, and Statistical Analyses: Commentary on Sijtsma's (2014) "Playing with Data"

Psychometrika. 2016 Mar;81(1):16-26. doi: 10.1007/s11336-015-9447-z.

Abstract

We comment on Sijtsma's (2014) thought-provoking essay on how to minimize questionable research practices (QRPs) in psychology. We agree with Sijtsma that proactive measures to decrease the risk of QRPs will ultimately be more productive than efforts to target individual researchers and their work. In particular, we concur that encouraging researchers to make their data and research materials public is the best institutional antidote against QRPs, although we are concerned that Sijtsma's proposal to delegate more responsibility to statistical and methodological consultants could inadvertently reinforce the dichotomy between the substantive and statistical aspects of research. We also discuss sources of false-positive findings and replication failures in psychological research, and outline potential remedies for these problems. We conclude that replicability is the best metric of the minimization of QRPs and their adverse effects on psychological research.

Keywords: minimizing false positives; questionable research practices; replicability.

Publication types

  • Comment

MeSH terms

  • Humans
  • Psychometrics*
  • Research Design*