For decades, researchers have studied the relationship between personality and job performance. Taking a new perspective, researchers (Song et al., 2024) use advanced machine learning techniques to examine whether personality has a more complicated relationship with job performance than previously believed. They explore how broad personality factors (e.g., conscientiousness) and more narrow facets (e.g., self-discipline) have nuanced relationships with job performance that machine learning is uniquely suited to uncover.
MACHINE LEARNING AND PERSONALITY
The researchers used machine learning, which is a type of algorithm that can explore complicated relationships otherwise undetected by more traditional statistical methods. They obtained data from over 1,000 employees from a pharmaceutical company in South Korea. Employees were given a personality test during the application process, which was compared to their job performance a few years into the role.
The results revealed that in terms of broad personality factors, conscientiousness was the most predictive of job performance. However, conscientiousness and agreeableness showed curvilinear relationships with job performance – that is, they were related up to a certain point, but beyond that, increasingly higher levels did not continue to improve performance.
Among more specific facets, three types of conscientiousness – self-discipline, order, and deliberation – were the most important in predicting job performance. Notably, for specific facets, machine learning models outperformed more traditional statistical models, such as regression. This emphasizes the value that these newer techniques can add to the discussion of personality and job performance.
PRACTICAL IMPLICATIONS
The researchers demonstrated how machine learning techniques can assist organizations in various HR practices, especially in employee selection. Using these techniques may result in better predictions of who will excel in certain roles. Even though this method may seem daunting, there are certain machine learning techniques that are relatively easier to interpret than others, making them more accessible to researchers and practitioners.
Song, Q. C., Oh, I.-S., Kim, Y., & So, C. (2025). Revisiting the nature and strength of the personality–job performance relations: New insights from interpretable machine learning. Journal of Applied Psychology, 110(1), 1–26.
Image credit: Unsplash+