The work is devoted to the question of identifying the relationship between the results of the R. Cattell test (16-factor test) and user subscriptions in a social network. A vector approach with space reduction is used to represent user subscales, and some machine learning algorithms are used to test the relationship. The results of this study showed that factors A, B, C, F, G, N, O, Q1, Q3, Q4 of R. Cattell test. The results of this study showed that the relationship between subscales and R. Cattell test results is possible to predict better than random.