Personality Research and Assessment in the Era of Machine Learning

The increasing availability of high-dimensional, fine-grained data about human behaviour, gathered from mobile sensing studies and in the form of digital footprints, is poised to drastically alter the way personality psychologists perform research and undertake personality assessment. These new kinds and quantities of data raise important questions about how to analyse the data and interpret the results appropriately. Machine learning models are well suited to these kinds of data, allowing researchers to model highly complex relationships and to evaluate the generalizability and robustness of their results using resampling methods. The correct usage of machine learning models requires specialized methodological training that considers issues specific to this type of modelling. Here, we first provide a brief overview of past studies using machine learning in personality psychology. Second, we illustrate the main challenges that researchers face when building, interpreting, and validating machine learning models. Third, we discuss the evaluation of personality scales, derived using machine learning methods. Fourth, we highlight some key issues that arise from the use of latent variables in the modelling process. We conclude with an outlook on the future role of machine learning models in personality research and assessment.     

INTROMISSION FREQUENCY AND INTROMISSION DURATION IN THE MALE RAT MATING BEHAVIOR

C arlsson , S. G. & L arsson , K. Intromission frequency and intromission duration in the male rat mating behavior. Scand, J. Psychol ., 1962, 3 , 189–191.—The male rat attains ejaculation after repeatedly mounting and penetrating the receptive female. By artificially prolonging the intervals of rest separating the intromissions the male is made to ejaculate after fewer intromissions than occurs when he copulates ad libitum . It was found that the reduction in intromission frequency was accompanied by an increase in intromission duration.     

THE FALSE PROMISES OF RISK ANALYSIS

The relatively new discipline of risk analysis promises to provide objective guidance in some of the most controversial issues in modern high-technology societies. Four conditions are discussed that must be satisfied for this promise to be fulfilled. Since none of these conditions is satisfied, risk analysis does not keep its promise. In its attempts to reduce genuinely political issues to technocratic calculations, it neglects many of the factors that should influence decisions on risk acceptance. A list of tentative guidelines is given for numerical decision support that encourages democratic decision-making instead of trying to evade it.     

Inclusion in a superordinate category, in-group prototypicality, and attitudes towards out-groups

We hypothesized that group members’ attitudes towards an out-group are negatively related to the in-group’s perceived relative prototypicality for a superordinate category, but only if both the in-group and out-group are included in this superordinate category. In Experiment 1 (N=40), Germans’ attitudes towards Poles were negatively correlated with the relative prototypicality of Germans when “Europe” (including Poles), but not when “West-Europe” (excluding Poles), was the superordinate category. In Experiment 2 (N=63), female single parents’ attitudes about the competence of single parents to raise children depended on the in-group’s relative prototypicality for “single parents” (including fathers), but not on their relative similarity to “mothers” (excluding fathers). Both experiments showed that inclusion in a superordinate category had a more negative influence on attitudes towards the out-group when relative in-group prototypicality is high rather than low.     

Correlates of aggressive and delinquent conduct problems in adolescence

The primary purpose of the study was to test a hypothetical model for aggressive and delinquent conduct problems in adolescence. A total of 168 adolescents from 121 families were studied using several questionnaires and a semi‐structured interview. The following factors were considered: obstetric complications, temperament (novelty‐seeking), self‐esteem, family influences (perceived parenting, alcohol abuse/dependence of parents, antisocial personality disorder of the father) and peer‐group characteristics (peer rejection and membership in a deviant peer group). The evaluation methods applied included correlation analyses and testing of two hypothetical models using structural equation modeling. The correlation analyses revealed significant relationships between adolescent aggressive and delinquent behavioral problems and parental antisocial behavior; perceived parental rejection and low emotional warmth; adolescent novelty‐seeking, self‐esteem, peer rejection and peer deviance. The two empirical models, separately for aggressive and delinquent behavior problems, revealed direct relationships between paternal antisocial behavior, parental rejection, adolescent novelty seeking, peer deviance, peer rejection, and offspring aggression and delinquency. There were, however, two differences with respect to the relationships between peer rejection, peer deviance, aggression and delinquency. First, peer rejection was more strongly associated with aggressive behavior and only moderately linked to delinquency. Second, deviance in the peer‐group was found to be closely related with delinquency but only moderately with aggression. Our findings suggest that several pathways for aggressive and delinquent conduct problems are comparable while others were not. Regarding the findings of the empirical models, we conclude that only intervention measures that include parents, peers and individual adolescents may help decrease the incidence of aggressive and delinquent conduct problems. Aggr. Behav. 31:24–39, 2005. © 2005 Wiley‐Liss, Inc.