Recruiters prefer expert recommendations over digital hiring algorithm: a choice-based conjoint study in a pre-employment screening scenario

“This preference for human recommendations emerges even if the recommendation is strongly subjective (Oberst et al., 2020).”

“Organizations use AI-based systems because they expect from them both efficient and impartial recommendations for selection decisions (Oberst et al., 2021; Lacroux and Martin-Lacroux, 2022).”

“…in this systematic review pertain to perceptions surrounding the use of hiring with AI (Suen et al. 2019; Newman et al. 2020; Lee 2018; Langer et al. 2018, 2019; Kaibel et al. 2019; Kodapanakkal et al. 2020; Oberst et al. 2020; Acikgoz et al. 2020; Noble et al. 2021; Warrenbrand 2021; Bigman 2020).”

“…to utilize healthcare algorithms (Longoni et al. 2019), to rely on algorithms for tasks that seem subjective (Castelo et al. 2019), and are hesitant to trust algorithms for recommending jokes (Yeomans et al. 2019), hedonic products (Longoni and Cian 2022), and job candidates (Oberst et al. 2021).”

“Despite the high potential ascribed to AI use in organizations (Bongard, 2019; Oberst et al. , 2021; Langer et al. , 2021), the question of how AI can be successfully integrated into personnel selection is one of the main challenges in the digital transformation of HRM (Prikshat et al. , 2023).”

“Moreover, gender bias in employee selection can be detected and reduced as well with the application of advanced technologies like AI-based digital hiring (Oberst et al., 2021; Palomar et al., 2021).”

“…recruiting was unrealized due to persistent challenges in achieving unbiased matching and candidate searching (e.g., Furtmueller et al., 2012; Oberst et al., 2020). as previously mentioned, there is an international trend toward female domination in the hR occupation (scarborough, 2017).…”

“With regard to acceptability, Oberst et al. (2020) found that recruiters prefer expert recommendations over those given by algorithms.”

“To stimulate deeper and more tangible discussions about algorithm use and fairness issues [38, 52, 67, 84, 106] while protecting sensitive company information, we introduced hypothetical scenarios in the second and third focus groups (see the appendix in Section A for a details).”

“…can be increased (Sithambaram and Tajudeen, 2023; Yadav et al., 2023) by freeing up HR professionals from repetitive administrative tasks allowing them to focus on strategic and human-centric activities (Horodyski, 2023b; Islam et al., 2022; Oberst et al., 2021; Pandya and Al Janahi, 2021).”

“Companies are adopting algorithms, AI, and other machines to aid in hiring decisions under the “assumption that hiring recommendations made by this highly advanced technology is both efficient and impartial—at least more so than any human recruiter” (Oberst et al., 2020, p. 627).”

“They believe human experts possess specialised skills and can grasp the complexities and nuances of an issue (Oberst et al. 2021).”

“Drawing from experimental studies of decision making in complex tasks involving human and ADSS experts, we consider that the human expert is generally considered to be more trustworthy than an ADSS (Langer et al., 2022; Oberst et al., 2020; Rieger et al., 2022).”

“There have been calls for greater regulation of AI in recruitment, with suggestions that companies should audit their algorithms regularly to ensure they are not inadvertently discriminating against certain groups (Oberst, De Quintana, Del Cerro, & Chamarro, 2021).”

“Literature also suggests that AI is increasingly used to conduct pre-employment assessments (Oberst et al., 2021.”

“A recent study further supports this insight, demonstrating that recommendations remain the primary recruitment channel, preferred even over digital algorithms in the digital age [25].”

“Furthermore, the emerging themes, such as technological change, highlight the need for research into AI-human collaboration to ensure human-centric, inclusive hiring practices (Niehueser W et al., 2020; Wuisan DSS et al., 2023; Oberst U et al., 2021).”

“Given heuristic procedure also uses keyword occurrence frequency or inverse document frequency (IDF), the resulting population is more meaningful and adaptable [8].”

“…ceux liés au contrôle de gestion social (HR scorecards) et d’autres chantiers en développement rapide, comme le recrutement prédictif (Ajunwa et al., 2016; Oberst et al., 2020) ou les modèles d’apprentissage supervisé appliqués à l’étude des déterminants de la fidélisation (Atef et al., 2022).”

“This is then presented to the sample population to determine the preferences of the consumers (35).”