| Sign In to gain access to subscriptions and/or personal tools. |
A Quantitative Analysis of Experimentwise Error Rates in Applied Behavioral Science ResearchCenter for Health Sciences in the School of Allied Health Professions at the University of Wisconsin-Madison, 1300 University Avenue, Madison, Wisconsin 53706. This article discusses theproblem of Type I error in behavioral science research and describes three measures of error rate: experimentwise, per-experiment, andpercentage error rates. The author computes error rates for a sample of behavioral science literature using these three procedures, with the sample comprising seven articles published in The Journal of Applied Behavioral Science. The results of the analysis suggest the presence of a substantial incidence of Type I error in the research. The author presents possible solutions to these problems.
The Journal of Applied Behavioral Science, Vol. 22, No. 4,
495-a-501 (1986) |
|||