Seminar #7 – October 28, 2020

ABSTRACT – Like many other academics, family business scholars face increasing pressures to publish in reputable journals, as an undesirable side-effect of which they might–knowingly or unknowingly–engage in p-hacking: the incremental and incompletely-disclosed adjustment of data collection, analysis, and/or reporting, until nonsignificant results turn significant. Analyses of the distribution of hypothesis-testing p-values published in three major family business journals suggest that p-hacking indeed also occurs in family business studies. Female authorship, the proportion of significant results among other hypothesis tests, and authors’ employer prestige associate negatively with p-hacking, while the negative association with employer prestige is reversed for untenured authors. Implications for family business research, as well as some suggestions going forward for authors, institutions, and journals and their editors, are discussed. While this study does not intend to discredit the achievements of the family business research community, it does hopefully inspire a more open discussion of the potential for biases to affect our work.