The study also examined whether combining predictor variables increased the strength of the correlation with the length of stay. Previous research had suggested that inmates who remain in therapy at Grendon for 18 months or more seem to make the greatest changes across a range of measures. However, previous analyses usually compared individual, separate measures rather than identifying particular typologies that could simplify interpretation. The data for the present study covered age, the number of previous convictions, the Ravens Progressive Matrices, the Hostility and Direction of Hostility Questionnaire, and the Eysenck personality Questionnaire. The research produced a correlation matrix between the assessment variables and time at Grendon and subjected it to principal component analysis. Results revealed four typologies of Grendon inmates. The correlations between all four components and length of stay were of a low order of magnitude but were highly statistically significant. Findings were consistent with the results of earlier studies conducted at Grendon. The study’s main contribution was to confirm that certain personality characteristics could be grouped together to describe the population at Grendon and that these dimensions are associated with differing lengths of stay during the period studied, even though the differences accounted for a relatively small proportion of the variance in the length of stay. Tables and 42 references (Author abstract modified)