The Downside of Using Computer Algorithms To Score Job Interviews

Topic(s): artificial intelligence, fairness, selection
Publication: Journal of Occupational and Organizational Psychology
Article: Applicant reactions to algorithm- versus recruiter-based evaluations of an asynchronous video interview and a personality inventory
Authors: J.K. Oostrom, D. Holtrop, A. Koutsoumpis, W. van Breda, S. Ghassemi, R.E. de Vries
Reviewed by: Tyler Cowley

In recent years, asynchronous video interviews (AVIs) have taken the personnel selection process by storm. These interviews offer applicants the flexibility to record their responses on their own time, eliminating scheduling headaches. Recruiters can then evaluate these recordings, assigning scores based on candidate answers. However, a growing trend involves using algorithms to score candidate answers to AVIs, potentially saving organizations significant resources. While this algorithmic approach is gaining traction, one crucial element remains largely unexplored: applicant reactions to these new evaluation methods. This study aimed to bridge that gap by investigating how applicant perceptions differ when their AVI is assessed by algorithms versus recruiters.


To investigate applicant reactions, the researchers (Oostrom et al., 2023) recruited working adults for two online studies. The findings revealed several negative reactions towards algorithm-based evaluations. Compared to human reviewers, applicants perceived algorithms as less fair, less accurate in evaluating their interview performance, and less helpful in providing feedback. Additionally, the use of algorithms triggered feelings of unease and uncertainty, described as “emotional creepiness.” As such, applicants disliked algorithm-based evaluations across the board and preferred recruiter-based evaluations.


These findings suggest that job applicants might be wary of algorithms used during the hiring process. While companies see advantages in using algorithm-based evaluations to streamline hiring, these tools could backfire by deterring qualified candidates. As such, the authors recommend the following:

  • Prioritize human raters over algorithms. Organizations that are focused on attracting a strong candidate pool or are concerned about legal implications might consider minimizing or even eliminating algorithm-based evaluations.
  • If organizations choose to abandon algorithms altogether, a standardized human review process becomes essential. For example, organizations should focus on training human raters on scoring criteria, such as what makes a good vs. bad candidate. This will ensure fairness, effectiveness, and minimize negative applicant reactions.
  • If organizations must use algorithms, they should be transparent with applicants about the reasons for their use. Explain the purpose of the algorithm and how it benefits applicants. Highlight advantages like consistent evaluations, quicker processing times, and reduced bias. This transparency can help applicants feel more comfortable with these tools.


Oostrom, J. K., Holtrop, D., Antonis Koutsoumpis, Ward van Breda, Ghassemi, S., & de, E. (2023). Applicant reactions to algorithm‐ versus recruiter‐based evaluations of an asynchronous video interview and a personality inventory. Journal of Occupational and Organizational Psychology.

Image credit: istockphoto/Moor Studio