How job seekers react to AI in the application process
More and more HR departments are using artificial intelligence (AI) to screen the applications they receive. While the advantages for companies are obvious, this practice seems to have a negative impact on applicants. BFH researcher Andreas Sonderegger and work psychologist Jenny Wesche from the FU Berlin have investigated this in experiments.
According to a report by the career portal Monster, the share of digital selection systems in the application process has doubled between 2017 and 2019, and around 70 % of companies use an applicant tracking system, according to Handelsblatt. Around 90 % of companies forecast effective use of their career website, applicant management system and IT in HR, according to Monster‘s HR Report 2021. With the help of AI, HR departments save valuable resources and time. However, recruiters have to mention that they use AI in the advertisement. It is still unclear how this affects the attractiveness of the advertised job and the likelihood of applying. “However, initial findings indicate that job seekers tend to be critical of such use of technology,” says Dr Andreas Sonderegger, lecturer and researcher at BFH Wirtschaft. Together with Dr Jenny Wesche from the FU Berlin, he has conducted various experimental studies on this topic. In these, the researchers systematically varied the information in job advertisements regarding the use of AI in the different phases of the selection process. That is, the study participants were exposed to AI at each stage:
- Preselection by a human
- Preselection by AI
- No information on who carries out the pre-selection
Study 2 & 3:
- Human evaluates the dossiers and conducts the interview
- AI evaluates the dossiers, but a human conducts the interview
- AI evaluates the dossiers and conducts the interview.
Pre-selection by AI is ok, interview rather not
“Our results show that the applicants hardly have a negative attitude towards the use of AI in the pre-selection. But they perceive it very negatively when AI is also used in the job interview,” says Sonderegger. According to this, the study participants distrust the abilities of AI, especially if they have had no previous experience with the technology in interviews. In addition, the candidates have less control and influence over the decision-making process, and they do not feel sufficiently valued if they cannot interact with humans. “On the other hand, the study participants had fewer problems with AI when they had the feeling that the technology was established and worked well,” Sonderegger sums up.
Companies should check this
The researchers advise companies to carefully consider whether and, if so, at which stage of the selection process they use automation. The way companies recruit has a negative impact on the perception of attractiveness and also on the application intentions of job candidates. Companies and organisations should not only consider the costs and validity of selection procedures, but also the reactions of job seekers and applicants, Sonderegger and Wesche emphasise. This is proven by many results from current research in the field of automated selection procedures. It is true that it is demanding and complex to evaluate the advantages and disadvantages – financial costs, time, validity, applicant reactions – in the different procedures in all phases of the recruitment process. “But we firmly believe that the efforts are worthwhile and that both sides can win,” say the study authors.
- Wesche, J. S. & Sonderegger, A. (in press). Repelled at first sight? Expectations and intentions of job-seekers reading about AI selection in job advertisements. Computers in Human Behavior. https://doi. org/10.1016/j.chb.2021.106931
Dr. Jenny S. Wesche teaches and researches at the Department of Social, Organisational and Economic Psychology at the Free University of Berlin.
Prof. Dr Andreas Sonderegger teaches and researches economics at BFH and is a lecturer at the University of Fribourg. His research interests include cognitive ergonomics, human-computer interaction and work and organisational psychology.