Abstract
Because of contradictions among the various methods, sample size selection in multiple regression has been problematic. For example, how does one reconcile the difference between a 15: 1 subject-to-variable rule and a 30: 1 rule? The purpose of this paper is to analyze the advantages and disadvantages of the various methods of selecting sample sizes in regression. A discussion of the importance of cross-validity to prediction studies will be followed by descriptions of the three categories of sample size methods: cross-validation approaches, rules-of-thumb, and statistical power methods. A rationale will then be developed for the application of precision power to multiple regression, leading to the presentation, through multiple examples, of the precision power method for sample size selection in prediction studies.
Recommended Citation
Brooks, Gordon P. and Barcikowski, Robert S.
(1996)
"Precision Power and Its Application to the Selection of Regression Sample Sizes,"
Mid-Western Educational Researcher: Vol. 9:
Iss.
4, Article 5.
Available at:
https://scholarworks.bgsu.edu/mwer/vol9/iss4/5
