Validity Study of the MAPP™ Career Test and Strong Interest Inventory

Executive Summary
April 2001

A study was undertaken by Gilbert (1997) to compare the performance of the MAPP to the Strong Vocational Interest Inventory. While the results of that study were very encouraging, they were not as compelling as perhaps they could have been, given that the Gilbert (1997) study employed a very small sample (n=30). As such, a second study was undertaken to further compare the performance of the MAPP to the Strong Vocational Interest Inventory using a substantially larger sample. The results of this study found that occupational preferences predicted by the MAPP were substantially similar to those predicted by the Strong Vocational Interest Inventory.

Research Assumptions

This study anticipated that occupational-match predictions made by the MAPP would be similar to those made by the Strong (positively correlated) for like or similar occupations. It was also anticipated that the occupational-match predictions made by the MAPP would be dissimilar to the predictions made the Strong (negatively correlated) for unlike or dissimilar occupations.

Method

To examine MAPP-Strong relationships, a sample of 133 people, primarily employed in technology positions, completed both the MAPP and the Strong Vocational Interest Inventory. Occupational-match correlations were subsequently examined for all MAPP and Strong occupations, based on the occupational-match predictions reported in the Gilbert (1997) study.

Findings

In order to assert that the MAPP and the Strong measure substantially similar things, one would expect to find a substantial number of positive and negative correlations. One would also expect to find the average of those correlations to be approximately zero. As anticipated, a substantial number of positive and negative correlations were found when all MAPP and Strong occupational-match predictions were correlated, with a median correlation close to zero.

Additionally, if the MAPP and the Strong measure substantially similar things, one would also expect to find significant correlations among occupational matches. As expected, virtually all of the correlations were significant and in the expected direction when only the 88 Strong and corresponding MAPP occupations were examined. Correlations ranged for a high of +.78 and a low of -.25. Of the 149 correlations, three were in the wrong direction (negative) and three were essentially zero. The median correlation was .56.

Further examination of the major MAPP occupations across Strong occupations grouped within thematic area produced an interpretable pattern of correlation. That is, similar occupations tended to be correlated positively and dissimilar occupations tended to be negatively correlated.

Construct Validation Appendix Tables.xls