Using the multifaceted Rasch model to improve the TAT/PSE measure of need for achievement

We used the Rasch (1980) model to develop new pictures for the Thematic Apperception Test (C. D. Morgan & Murray, 1938; McClelland, Atkinson, Clark, & Lowell, 1953) or picture story exercise to measure need for achievement (nAch). In Experiments 1 and 2, we analyzed stories to assess the difficulty level of a total of 8 pictures using the multifaceted Rasch model with picture difficulty, story probe difficulty, and participant ability as facets with a partial credit model (FACETS; Linacre, 2005). A total of 6 pictures were retained and 4 new ones added for Experiment 3 in which 201 participants wrote 6 stories to a random set of the 10 pictures. FACETS analysis revealed improved person separation reliability. In Experiment 4, 206 participants wrote 1 story to the Studying picture either before or after filling out a battery of achievement-related questionnaires. The 2 experimental groups did not differ in the amount of nAch in their stories. The coder facet was demonstrated with 2 independent coders using the revised coding system for nAch.