Through a multimethod lens of artificial intelligence and qualitative content analyses: Effects of intelligence and gender on response patterns and elements of the cognitive process of faking

• Applicant: Dr. Jessica Roehner
• Subject Area: Personality Psychology, Clinical and Medical Psychology, Methodology
• Term: since 2024
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 538784935

Faking is a serious problem in psychological assessment, but little is known about how people fake and what elements are associated with the cognitive process of faking. However, understanding faking is a first step toward detecting or preventing it. There is evidence that crystallized and fluid intelligence (gc and gf) influence faking. Moreover, there are indications that men and women differ in their faking behavior. But because previous studies have focused on sex instead of gender, the psychological process (i.e., gender role self-concept) behind this difference is not clear. Recently, research has also shown that faking occurs at the item level, thus pointing to the need to analyze response patterns. The current multimethod project will be aimed at answering the following questions: First, where specifically in the response patterns can differences in faking related to intelligence and gender be found? Machine learning will be used to answer the question because our preliminary work showed that machine learning was useful for identifying differences in response patterns. Second, how do the cognitive elements involved in the faking process (e.g., comprehension, retrieval, judgment, response) vary with intelligence and gender? Thus, the quantitative approach (machine learning) will be complemented by a qualitative approach that will be based on participants’ comments to gain insight into the cognitive elements involved in faking behavior. The project will include three studies. In Study 1, ideal profiles concerning two jobs and two measures will be established in order to be able to score the quality and quantity of faking. In Study 2, response patterns when faking good and faking bad for these two jobs will be analyzed in relation to intelligence and the gender role self-concept on the basis of machine learning. In Study 3, participants will be asked about their thoughts when faking to gain insight into the elements involved in the cognitive process of faking. The results will add to prior research by combining analyses of response patterns and the cognitive process of faking with individual differences in faking associated with intelligence and gender. It is expected that these in-depth insights will help to promote insights in how to detect and prevent faking.

• DFG Programme: Research Grants
• Co-Investigators: Professorin Dr. Ute Schmid; Professorin Dr. Astrid Schütz