NCJ Number
52818
Journal
Psychological Bulletin Volume: 84 Issue: 2 Dated: (1977) Pages: 212-219
Date Published
1977
Length
8 pages
Annotation
AN ANALYSIS OF VARIANCE PROCEDURE IS DEVELOPED FOR EVALUATING THE RESULTS OF EXPERIMENTS WHERE THERE IS AN INTERACTION BETWEEN THE DESIGN TYPE AND THE SUBSTANTIVE INDEPENDENT VARIABLE.
Abstract
THE ASSUMPTION IS OFTEN MADE THAT ONE MAY PERFORM AN EXPERIMENT BY MANIPULATING THE INDEPENDENT VARIABLE, EITHER WITHIN OR BETWEEN SUBJECTS, AND THAT THE EFFECT OF THE INDEPENDENT VARIABLE IS THE SAME IN BOTH TYPES OF EXPERIMENTS. EXPERIMENTS IN DIFFERENT AREAS OF PSYCHOLOGY DEMONSTRATE THAT THIS ASSUMPTION IS OFTEN FALSE. IN THESE EXPERIMENTS, THE DESIGN TYPE IS TREATED AS AN EXPLICIT INDEPENDENT VARIABLE IN A FACTORIALLY DESIGNED EXPERIMENT. THE OUTCOME OF THE EXPERIMENTS IS THAT THERE IS AN INTERACTION BETWEEN THE DESIGN TYPE AND THE SUBSTANTIVE INDEPENDENT VARIABLE. SUCH EXPERIMENTS GENERALLY ARE DESIGNED IN AN INEFFICIENT MANNER, OR THE RESULTS ARE POORLY ANALYZED. THE PROPOSED ANALYSIS OF VARIANCE TECHNIQUE IS APPROPRIATE FOR THESE EXPERIMENTS. A GENERALIZED VERSION OF THE EXPERIMENTS IS DEALT WITH, ONE IN WHICH MORE THAN MERELY TWO LEVELS OF THE SUBSTANTIVE INDEPENDENT VARIABLE EXIST. THERE ARE M + 1 GROUPS IN THE EXPERIMENT, EACH GROUP HAVING N SUBJECTS. THE SUBSTANTIVE INDEPENDENT VARIABLE HAS M LEVELS. THERE ARE M GROUPS IN THE BETWEEN-SUBJECTS DESIGN TYPE, ONE FOR EACH LEVEL OF THE INDEPENDENT VARIABLE. THERE IS ONE GROUP IN THE WITHIN-SUBJECTS DESIGN TYPE, AND THE N SUBJECTS IN THIS GROUP RECEIVE THE TREATMENT OF ALL M LEVELS OF THE INDEPENDENT VARIABLE. SUPPORTING EQUATIONS AND DATA FOR A COMPUTATIONAL EXAMPLE ARE PROVIDED. (DEP)