Personality tests are a popular tool in recruitment processes to gain deeper insights into candidates. But what happens when applicants adjust their answers to present themselves in the best possible light? This behavior, known as socially desirable responding — also referred to as self-presentation, distortion, or faking — is a well-known issue that can distort the results of personality assessments. We have already outlined traditional solutions to this challenge in a previous article.
Now, a team of researchers from the University of Mannheim and HR Diagnostics has developed a new method for addressing self-presentation, leveraging advanced techniques from statistical modeling and probabilistic test theory. The core innovation lies in simultaneously considering both the targeted personality traits and each candidate's individual tendency to distort their responses. This distortion tendency is measured using statement-specific desirability characteristics, tailored to the specific application context, allowing for highly precise modeling.
This new approach provides two major advantages:
- Detection of Strong Self-Presentation: Recruiters and hiring managers can identify candidates exhibiting high levels of self-presentation and can address this during subsequent stages of the selection process through targeted follow-up questions.
- Correction of Test Results: By modeling the impact of self-presentation directly, the method corrects personality trait measurements for distortion, leading to more accurate and trustworthy assessment results.
The researchers tested their method both through statistical simulations and on data from over 3,000 applicants. The results are highly promising: Not only did the new model provide a better fit to the empirical data, but it also allowed for clearer differentiation between different personality traits. As expected, the data showed that candidates exhibited stronger tendencies to distort their responses when the stakes of the application were higher.
Beyond improving personality diagnostics, the new method also provides insights into the psychology behind self-presentation, examining links to other personality traits and intelligence.
In conclusion, this innovative method represents a significant advancement for personality measurement in recruitment settings. The adjustment for social desirability is already applicable to traditional Likert-scale personality tests and will soon also be extended to alternative formats like forced-choice tests. Additionally, the objectivity of this approach has the potential to enhance the fairness of hiring decisions. Ongoing studies are now investigating how this method affects the prediction of professional success.
Further details on the new method and its empirical validation in recruitment settings are available in the following article, published in the journal Educational and Psychological Measurement:
Seitz, T., Spengler, M., & Meiser, T. (2025). “What if applicants fake their responses?”: Modeling faking and response styles in high-stakes assessments using the multidimensional nominal response model. Educational and Psychological Measurement. Advance online publication. https://doi.org/10.1177/00131644241307560