Comparison of Subjective and Objective Judgments of Children’S Drawings

This paper presents a user-oriented technique to analyze dichotomous data arising from a double repeated measures 2 × 2 design. The design under consideration occurs most frequently in survey research in which two matched samples of subjects are measured at two points in time on a dichotomous dependent measure. Test statistics are presented that test for change on the dependent measure over time for each sample separately, as well as an interaction test statistic that compares the change rates between the two samples. An extension of the technique to higher order two-dimensional tables is also suggested.