A Novel Approach to Personalized Personality Assessment with the Attachment-Caregiving Questionnaire (ACQ): First Evidence in favor of AI-Oriented Inventory Designs

TL;DR

This paper proposes a new AI-oriented approach to personality assessment that considers multiple interpretations of questionnaire items, potentially improving clinical profiling and decision-making over traditional factor analysis methods.

Contribution

It introduces a personalized, meaning-sensitive questionnaire design and demonstrates how AI can enhance interpretation beyond traditional factor analysis.

Findings

Question interpretation varies among individuals, affecting assessment accuracy.

AI can mimic expert clinicians in interpreting personality questionnaires.

Personalized assessment impacts clinical decision-making.

Abstract

Background. Personality is a primary object of interest in clinical psychology and psychiatry. It is most often measured using questionnaires, which rely on Factor Analysis (FA) to identify essential domains corresponding to highly correlated questions/items that define a (sub)scale. This procedure implies the rigid assignment of each question to one scale - giving the item the same meaning regardless of how the respondent may interpret it - arguably affecting the assessment capability of the instrument. Methods. To test this hypothesis, we use the Attachment-Caregiving Questionnaire (ACQ), a clinical and personality self-report that - through extra-scale information - allows the clinician to infer the possible different meanings subjects attribute to the items. Considering four psychotherapy patients, we compare the scoring of the ACQ provided by expert clinicians to the detailed information gained from therapy and the patients. Results. Our analysis suggests that a question can be interpreted differently - receiving the same score for different (clinically relevant) reasons - potentially impacting personality assessment and clinical decision-making. Moreover, accounting for multiple interpretations requires a specific questionnaire design and a more advanced pattern recognition than FA - which Artificial Intelligence (AI) could provide. Conclusion. Our results indicate that a meaning-sensitive, personalized read of a personality self-report can affect profiling and treatment. Since a machine learning model can mimic the interpretative performance of an expert clinician, our results also imply a novel, AI-oriented approach to inventory design, of which we envision the first implementation steps. More evidence is required to support these preliminary findings.