Changing the means of managerial work: effects of automated decision support systems on personnel selection tasks

Abstract

To enhance the quality and efficiency of information processing and decision-making, automation based on artificial intelligence and machine learning has increasingly been used to support managerial tasks and duties. In contrast to classical applications of automation (e.g., within production or aviation), little is known about how the implementation of automation for management changes managerial work. In a work design frame, this study investigates how different versions of automated decision support systems for personnel selection as a specific management task affect decision task performance, time to reach a decision, reactions to the task (e.g., enjoyment), and self-efficacy in personnel selection. In a laboratory experiment, participants (N = 122) were randomly assigned to three groups and performed five rounds of a personnel selection task. The first group received a ranking of the applicants by an automated support system before participants processed applicant information (support-before-processing group), the second group received a ranking after they processed applicant information (support-after-processing group), and the third group received no ranking (no-support group). Results showed that satisfaction with the decision was higher for the support-after-processing group. Furthermore, participants in this group showed a steeper increase in self-efficacy in personnel selection compared to the other groups. This study combines human factors, management, and industrial/organizational psychology literature and goes beyond discussions concerning effectiveness and efficiency in the emerging area of automation in management in an attempt to stimulate research on potential effects of automation on managers’ job satisfaction and well-being at work.

Trust in Artificial Intelligence: Comparing trust processes between human and automated trustees in light of unfair bias

“…In this study, we capture four facets of trustworthiness that trustors may consider in relation to human and automated trustees: ability, flexibility, integrity, and benevolence (Höddinghaus et al, 2020;Wang & Benbasat, 2005). In the case of personnel selection, ascribing high ability means that trustors believe that trustees can successfully select suitable applicants (Langer et al, 2020). Flexibility reflects whether trustors assume that trustees can react flexibly to changes in a given task (Höddinghaus et al, 2020).…”

“…In these contexts, what trustors evaluate when assessing trustworthiness is focused on classical performance measures associated with effectiveness and efficiency (e.g., prediction accuracy). In comparison, there is a paucity of research investigating trust processes in contexts where automated systems support decisions that affect the fate of individuals (e.g., personnel selection; Langer et al, 2020). In such contexts, practitioners, researchers, and policy-makers are commonly concerned about ethical issues when using automated systems as decision support (Jobin et al, 2019;Martin, 2019).…”

Automation of Psychological Selection Procedures for Personnel to Specific Activities

“…Automation based on artificial intelligence and machine learning has increasingly been used to support managerial tasks and duties to enhance the quality and efficiency of information processing and decision-making (Langer et al, 2020). Solving the problem of determining professional suitability (PS) leads to the determination of a person's inclination to a particular category of professions, where his potential, psychophysiological reactions to external conditions, the ability to make adequate decisions depending on the circumstances are most fully revealed (Prykhodko, 2008).…”

“I” feel(s) left out: The importance of information and communication technology in personnel selection research

“…As another example, with recent advancements in artificial intelligence (AI), the use of automated systems to support or even automate selection processes has received increasing attention (Hickman et al, 2021; Langer et al, 2019, Langer et al, 2021, Newman et al, 2020). For instance, in addition to having human hiring managers assess asynchronous video interviews, several companies provide fully automated assessments of applicant recordings that can be used as an additional criterion or even to automatically filter the most viable applicants from an applicant pool.…”

“…Although there is less research for hiring managers and this work has not yet assessed actual hiring managers’ experiences and behavior, initial research indicates that implementing automated decision-support systems can change decision makers’ work-design-related experiences in personnel selection decisions (Langer et al, 2020). Although this kind of research is nascent in I-O psychology, research in this area (e.g., Langer et al, 2021) draws from classical work in human–automation interaction (e.g., Sheridan & Parasuraman, 2005), considers findings from human–computer interaction research (e.g., Lee, 2018), and has much potential for further interdisciplinary integration (e.g., considering legal implications of AI in personnel selection). Therefore, this work is a prime example of a field that is working to achieve the interdisciplinary connection that Hu et al see generally lacking in I-O psychology and ICT-oriented research.…”