Slides of TAT cards 1 and 2 were shown to 87 college students asked to write stories about them. Subjects also ranked the importance of 5 attributes in understanding their personalities. Attribute scores from projections and centrality scores from rankings, for achievement and autonomy, were regressed onto college GPA, trail-making, and autonomy by life history. Predictions failed to reach suitably low significance levels. Subscaling the life history questionnaire by factor analysis and subsequently regressing toward subscales resulted in multiple correlation significant at p < .01. Important to the model was the significant (p < .02) improvement in using the attribute by centrality interaction over either variable alone. Results were discussed in terms of a cognitive model for projectives.