Statistical Power in Contemporary Management Research

Statistical Power and the Testing of Null Hypotheses: A Review of Contemporary Management Research and Recommendations for Future Studies

“…As such, leading journals that publish a mix of management and/or applied psychology research were excluded from the sample. Such a distinction is in line with previous power assessments, whose authors have chosen to include strictly management journals in their assessments of management research (e.g., Mazen, Graf, et al, 1987;Schwenk & Dalton, 1991). This is in contrast to the power assessment by Mone et al (1996), which included journals that published a mix applied psychology and management research.…”

“…Previous surveys of statistical power in abnormal and social psychology, education, sociology, applied psychology, social work, and marketing have traditionally reviewed studies published in one or a few volumes of one journal in their respective disciplines. Notable exceptions to this trend include Mazen, Graf, et al (1987), Rossi (1990), Mone et al (1996), and Ferguson and Ketchen (1999), who incorporated multiple journals and/or multiple years in their power assessments. Our study followed the lead of these studies by drawing on pub- These journals were chosen for this study because they reflect current research in the management discipline, contain empirical studies, are peer reviewed, have high rejection rates, and have scholarly orientations.…”

“…As a result, Type I errors are usually focused on more frequently and guarded against more stringently by researchers (Baroudi & Orlikowski, 1989;Brewer, 1972;Chase & Chase, 1976;Cohen, 1977;Cowles & Davis, 1982;Greenwald, 1993;Mazen, Graf, et al, 1987;Myers & Melcher, 1969;Mone et al, 1996;Nickerson, 2000;Rossi, 1990;Sedlmeier & Gigerenzer, 1989). However, it is important to note that several authors (e.g., Mazen, Graf, et al, 1987;Sauley & Bedeian, 1989) advocate distributing the risk between Type I and Type II errors because the risks/consequences associated with committing a Type II error may be extremely costly. In other words, the ratio of β:α should be sensitive to the relative risks inherent in Type I and Type II errors for every test conclusion (Mazen, Graf, et al, 1987;Sauley & Bedeian, 1989).…”

“…However, it is important to note that several authors (e.g., Mazen, Graf, et al, 1987;Sauley & Bedeian, 1989) advocate distributing the risk between Type I and Type II errors because the risks/consequences associated with committing a Type II error may be extremely costly. In other words, the ratio of β:α should be sensitive to the relative risks inherent in Type I and Type II errors for every test conclusion (Mazen, Graf, et al, 1987;Sauley & Bedeian, 1989).…”

Behavioral Assembly Required: Particularly for Quantitative Courses

“…The emphasis on collaboration, creativity, process, and validation distinguishes metaphors from mere examples, which teachers and researchers use as well. For instances, Mazen, Graf, Kellog, and Hemmasi (1987) used the Challenger disaster to illustrate Type I and Type II errors in an examination of statistical power in management research, but the example was not intended to create the holding environment of mutual discovery outlined above.…”