Direct vs. projective: measures of children's perceptions of the parent-child relationship as predictors of future relationship quality

Abstract

The predictive validity of a projective measure of childrens perceptions of the parent-child relationship was compared with the predictive validity of more direct measures. The criterion variable was the quality of a newly formed relationship between children and their mentors. Subjects were 61 children in the second and third grades who were nominated as aggressive by their teachers. The direct and projective measures were not correlated with each other. Generally, neither the direct nor the projective measures were individually associated with quality of the mentor-child relationship. In addition, the projective measure failed to explain a significant percentage of the variance in the quality of the mentor-child relationship above and beyond the contribution of the direct measures. The projective measure was subjected to a cluster analysis based on factor scores, but valid clusters failed to emerge. Results are discussed in terms of difficulties with the use of projective measures in general and with the scoring system used with this particular projective measure.