The applicability of the Rasch model to data from a typical personality test, the High School Personality Questionnaire (HSPQ), was studied. The data were gathered on Junior High and High School students in the Louisville Public Schools (Kentucky). Item easinesses and person abilities were estimated and compared by age group, within each age group, and across two points in time for the older age group. In addition, certain results from the Rasch analyses were compared with those of factor analysis. A sample of 1,000 students was taken from each of the groups (Junior High, 7th graders; Senior High, 11th and 12th graders). Results of the study are related to five questions considered. The first question was whether or not there were patterns of fit to the Rasch model when responses are dichotomized in different ways. The results indicated that no single key was superior to others in producing fit. The second question was concerned with fit of the model for the data considered; it was found that frequently there was lack of fit, but it is noted that the test statistic was conservative. The third question related to the stability of item easinesses estimates within a group and across two points in time for that group. The conclusion was that different item easinesses are obtained when different degrees of possession of the trait are focused upon. The fourth question was how stable the tests of fits results are across time; pre- and post-comparisons of fit found 55% in agreement. The fifth question concerned how the item mean squares are related to factor loadings; in almost all cases, the item with the highest mean square was also the item with the lowest loading. (Author/DB)