Affiliation:
Department of Psychology of the Faculty of Arts, Charles University
Short bio:
PhDr. Eva Höschlová, Ph.D.: is a member of the Board of the Czech Association of Work and Organizational Psychology. She focuses on applying psychology in various business spheres. As a senior consultant, she helps organizations and their leaders to thrive from the knowledge and best practices of work and organizational psychology. She also participates in international research of teams and space crews. For more than ten years, she lectured at the Department of Psychology of the Faculty of Arts, Charles University, for example, Psychodiagnostics and Psychometrics, Methods of Occupational Psychology, and Economic Psychology.
Title:
Using AI to improve user experience in personality testing with 4Elements Inventory
Abstract:
Artificial intelligence is bringing revolutionary possibilities in almost every area of human activity. Psychology is no exception, where, in recent years, we have mainly seen applications in therapeutic chatbots. The newest innovations brought many other application areas, especially in psychological testing, and we have applied some of these to the 4Elements Inventory questionnaire. This personality test is based on the ancient metaphor of four key elements describing the world around us – fire, water, air, and earth – in modern psychometrics, four factors representing interrelated clusters of personality traits. The 4Elements has been standardized and validated on a sample of over 13,000 individuals and meets all the demands of modern psychometrics, including verification of factor structure, reliability, and construct and contrast validity.
In this paper, we will introduce an artificial intelligence model combined with natural language processing which we have implemented to create better and more accurate personal report content for test users and to deliver subsequent consultation of results with a trained chatbot that can answer questions and make recommendations related to the personality profile of the test subject. We will demonstrate the developed solution’s functionality and the immediate next steps of a possible AI implementation in the test administration process. The final part of the presentation will discuss the limitations and threats of using AI in creating automated reports and other types of outputs for test takers.