The execution of experiments in human factors studies can be improved by thoughtful choice of dependent variables, summary statistics, and statistical techniques. Deficiencies in current practice of experimental design are discussed, and some specific suggestions for improvement are made. In particular, rationale for a choice of criteria and dependent variables should be documented as fully as possible; summary statistics other than amplitude measures may sometimes be appropriate; and multivariate analysis should be used when more than one dependent measure is used. Multivariate and univariate analysis of variance techniques are compared.
Get full access to this article
View all access options for this article.
References
1.
CochranW. G.CoxG. M.Experimental designs. New York: Wiley, 1957.
2.
CooleyW.LohnesP.Multivariate data analysis. New York: Wiley, 1971.
3.
KirkR. E.Experimental design: Procedures for the behavioral sciences. Monterey, California: Brooks-Cole, 1968.
4.
PoultonE. C.Tracking skill and manual control. New York: Academic Press, 1974.
