Introduction
Relying on subjective impressions of job applicants can block the goal of selecting the most suitable candidates. Indeed, unstructured employment interviews often have relatively low interrater reliability (Huffcutt et al., 2013), and with revised meta-analytic estimates, structured interviews still emerge as the highest ranked selection procedure; while the validity estimate for unstructured interviews is significantly lower (Sackett et al., 2022). Despite this, the unstructured interview has remained the most used method in personnel selection for over a century (Chauhan, 2019; Neumann et al., 2021).
The relatively low validity of the unstructured interview seems to apply to intuitive and holistic methods in general, as clinical approaches to data combination are equated or exceeded by simple mechanical combinations of standardized test scores (Highhouse & Kostek, 2013). Likely sources of this validity loss include a smaller consistency and a higher rate of unsystematic errors (Kuncel et al., 2013). Practitioners’ perceptions of applicants are not immune to bias (Posthuma et al., 2002), and relying on one’s intuition can lead to increased adherence to performance-irrelevant characteristics (Highhouse, 2008). Furthermore, being familiar with a domain may introduce an ‘illusion of knowledge’ where idiosyncratic experiences are over-generalized at the expense of valid cues (Hall et al., 2007). Indeed, experimental findings show that recruiters who received information from interviews suffered from more overconfidence than those who were provided with test scores only (Kausel et al., 2016). Further supporting the picture that job experience does not justify intuition-based decision-making, is research showing that recruiters do not necessarily outperform laypeople in selection-related judgments (Re & Rule, 2016; SchmidMast et al., 2011; Sinclair & Agerström, 2020). Moreover, recruiters’ use of intuition has been associated with increased adherence to the way applicants choose to present themselves, at the expense of more diagnostic information (Sinclair & Agerström, 2020).
Despite its disadvantages, recruiters often rely on their ‘gut feelings’ about applicants (Lodato et al., 2011; Miles & Sadler-Smith, 2014; Neumann et al., 2023), and they often prefer unstructured interviews (Lievens et al., 2005). Decision-makers may feel that structured and mechanical approaches have less autonomy potential (Nolan & Highhouse, 2014) or that they ‘pose a threat of technological unemployment (Nolan et al. 2016). Interestingly, job candidates also tend to perceive non-standardized selection procedures as more useful than standardized ones (Diab et al., 2011), and to regard interviews as more face valid than other selection methods (Anderson et al., 2010). Individualized selection methods may be perceived as fairer because they are more personal (Highhouse, 2008), and interviews allow for two-way communication, where both parties have the opportunity to strengthen (or weaken) their perceptions of trustworthiness (Klotz et al., 2013). Furthermore, applicants tend to distrust certain standardized methods, such as personality tests (Hausknecht et al., 2004).
We propose that an additional reason could be that applicants tend to overestimate recruiters’ intuition. While related, faith in expertise entails trusting that the recruiter is more skillful than laypeople in the area of selection. Faith in recruiter intuition specifically concerns trust in the recruiter’s ‘gut feelings’ about applicants, without the need to involve selection tools such as tests. As argued by Highhouse (2008), a belief in ‘the myth of selection expertise’ can create aversion to valid selection aids. Such aversion is also a problem in other contexts such as medicine (Giluk & Rynes, 2012), where distrust in mechanic methods is an obstacle to valid practice. In the context of personnel selection, it remains unknown whether applicants’ belief in the power of recruiter intuition explains preferences for unstructured methods.
From a recruitment perspective, selection systems need to appeal to candidates’ interest in the employer. Qualified applicants must make the decision to apply, exert effort during the selection process, accept a job offer, and be motivated to stay with the organization. Perceptions of the selection procedure can affect the applicant pool and thus the human capital of the organization (Kuncel & Klieger, 2007). It is therefore crucial to understand selection systems from the perspective of the job candidate. The aim of this study is to clarify whether applicants’ belief in the power of recruiter intuition is related to preferences for unstructured interviews and holistic methods. We expect that people have faith in recruiters’ expertise, which means that recruiters are perceived as better than laypeople at making judgments about applicants based on resume screening and hiring interviews. We further hypothesize that faith in recruiters’ intuition is positively associated with faith in recruiters’ expertise, preference for holistic over mechanical selection methods, and with process favorability for unstructured, but not structured, interviews. If anything, its relation to process favorability for structured interviews should be expected to be in the opposite direction.
Method
Participants
The participants were recruited at Swedish universities and various workplaces. A priori power calculation showed that obtaining 95% statistical power to detect a small correlation coefficient of 0.20 (two-tailed) would require a minimum of N = 319. The total sample was 345 (192 women, 149 men, 2 nonbinary, 1 ‘other’; Mdn age = 28 years, M = 33.8, SD = 13). About half of the participants (50.3%) were employed, while 43.3% were university students, 2.6% were unemployed, and 3.8% selected ‘other’ occupation status. The average participant had applied to 17.8 jobs, and 95.1% had previously attended job interviews.
Procedure and methods
The study was introduced as a project about selection procedures, and key terms (recruiter, applicant, selection methods) were defined in the paper-pencil survey that consisted of five measures plus filler items (see online Appendix). Following a procedure developed by Steiner and Gilliland (1996; also see Anderson & Witvliet, 2008; Hoang et al., 2012), participants were asked to think about a job that they were likely to apply for in the future, and to write down what type of job it was. Of the jobs that the participants imagined, 72.5% were high-skilled, 22.3% were low-skilled, and the remaining 5.2% were not clearly described.
They were then asked to consider different selection methods (structured and unstructured interview, followed by six others, such as personality tests and work-sample tests) in light of the job that they had in mind. The hypotheses concerned interviews only; the remaining selection methods were filler items. The nature and purpose of each selection technique was briefly described, followed by a two-item measure of its process favorability: ‘How would you rate the effectiveness of this method for identifying qualified people for the job that you specified above?’ (1 = not at all effective, to 7 = very effective) and ‘If you did not get the job based on this selection method, what would you think of the fairness of this procedure?’ (1 = very unfair, to 7 = very fair; Cronbach’s α = 0.82 for unstructured and 0.71 for structured interviews).
It was then explained that there are two general ways to combine information about applicants during the selection procedure, and the holistic and the mechanical method were briefly described. The participants responded to three items: which one of these methods they believed results in the selection of the most competent applicant, which one is most fair, and which one they would personally prefer to be evaluated from when applying for work (1 = mechanical methods, 4 = no difference, 7 = holistic methods; Cronbach’s α = 0.90).
Next, the participants responded to whether they believed that recruiters are better than laypeople at making judgments about applicants based on resume screening and hiring interviews (1 = much worse than people in general, 4 = neither better nor worse, 7 = much better than people in general). This was measured regarding judgments about applicants’ personality, intelligence/cognitive skills, honesty, craftsmanship/professional skills, and leadership abilities. These five items were averaged into the faith in recruiters’ expertise scale (Cronbach’s α = 0.75).
The participants were then told that many recruiters use their intuition (‘gut feeling’), which is based on previous experiences, when making judgments about candidates. They rated the extent to which they thought that this is a good and appropriate idea. These two items served to measure faith in recruiters’ intuition (Cronbach’s α = 0.94). Finally, the participants answered demographic questions.
Screening of the data (e.g. through boxplots and Mahalanobis distance) revealed no indications of careless responding, and the analyses were based on the entire sample. Descriptive statistics and correlations are presented in Table 1.
Table 1
Descriptive statistics and Pearson correlations for the study variables (N = 319).
| MEAN | SD | 1. | 2. | 3. | 4. | |
|---|---|---|---|---|---|---|
| 1. Faith in recruiter intuition | 4.17 | 1.53 | – | |||
| 2. Faith in recruiter expertise | 4.59 | 0.72 | 0.27*** (0.17, 0.37) | – | ||
| 3. Preference for holistic methods | 5.31 | 1.56 | 0.28*** (0.18, 0.37) | 0.06 (–0.05, 0.17) | – | |
| 4. Process favorability for unstructured interviews | 4.10 | 1.42 | 0.35*** (0.25, 0.44) | 0.13* (0.02, 0.23) | 0.17** (0.07, 0.27) | – |
| 5. Process favorability for structured interviews | 4.86 | 1.15 | –0.10 (–0.21, 0.00) | –0.00 (–0.11, 0.10) | –0.12* (–0.23, 0.02) | –0.18** (–0.28, –0.07) |
[i] Note: Numbers in parentheses are 95 % CI. All variables were measured on a scale from 1–7.
* p < .05, ** p < .01, *** p < .001, two-tailed.
Results
First, a one-sample t-test confirmed that recruiters were perceived as better than laypeople at making judgments about applicants based on resume screening and hiring interviews, t (342) = 15.19, p < .001, mean difference = 0.59, 95% CI [0.51, 0.66]. The effect size was large (Cohen’s d = 0.82).
Faith in recruiters’ intuition was positively associated with faith in recruiter expertise (r = 0.27), and with preference for holistic over mechanical selection methods (r = 0.28; see Table 1). Lending further support for the hypothesis, faith in recruiters’ intuition was positively associated with process favorability toward unstructured interviews (r = 0.35), while its association with process favorability toward structured interviews was weak and in the negative direction (r = –0.10). These two correlation coefficients were statistically different from each other, ZH = 5.59, p < .001 (calculated with a formula by Hoerger, 2013, which is an updated version of Steiger’s Z; Steiger, 1980).
Exploratory analyses
We proceeded with a multiple regression analysis to examine the relative contribution of faith in recruiter intuition for process favorability for unstructured interviews, when the effects of process favorability of structured interviews, preference for holistic methods, and faith in recruiter expertise are included in the model. The regression model was significant (p < .001), explaining 14.5 % of the variance in the outcome measure (see Table 2).
Table 2
Model coefficients of the multiple regression analysis predicting process favorability for unstructured interviews (N = 319; model R2 = 0.145***).
| PREDICTOR | β | 95% CI LOWER | 95% CI HIGHER |
|---|---|---|---|
| Faith in recruiter intuition | 0.31*** | 0.21 | 0.42 |
| Faith in recruiter expertise | 0.04 | –0.07 | 0.14 |
| Preference for holistic methods | 0.06 | –0.04 | 0.17 |
| Process favorability for structured interviews | –0.12* | –0.22 | –0.02 |
[i] Note: * p < .05, ** p < .01, *** p < .001, two-tailed.
In line with previous research, holistic methods were preferred over mechanical ones, t (343) = 15.63, p < .001, mean difference = 1.31, 95% CI [1.15, 1, 48]. The effect size was large (Cohen’s d = 0.84). However, unexpectedly, a paired samples t-test showed that structured interviews received higher ratings than unstructured ones, t (344) = 7.17, p < .001, mean difference = 0.76, although the effect size was small (Cohen’s d = 0.38).
Discussion
The choice of selection method has a nontrivial impact on selection decisions, with consequences for organizations’ performance levels as well as for applicants’ employment opportunities and their right to fair treatment (Lodato et al., 2011). However, practitioners’ and applicants’ perceptions of a selection method are not always aligned with its predictive validity. Understanding why people tend to prefer unstructured methods and oppose evidence-based practice, is both practically and theoretically important. Past research has shed light on some explanations for practitioners’ preferences for unstructured methods (e.g., Nolan & Highhouse, 2014), but less is known about applicants’ preferences. Therefore, the aim of this study was to clarify whether applicants’ belief in the power of recruiter intuition is related to preferences for unstructured interviews and holistic methods. As expected, faith in recruiter intuition had a moderate positive association with process favorability for unstructured interviews, while its relationship with process favorability of structured hiring interviews was in the opposite direction (and weaker). This suggests that faith in recruiter intuition helps explain process favorability for unstructured interviews in particular, rather than interviews in general. Moreover, faith in recruiter intuition predicted preference for unstructured hiring interviews even when accounting for other variables, suggesting that it is beliefs about recruiter intuition per se rather than general perceptions of skills that explain why some applicants like unstructured interviews. In line with previous research (Diab et al., 2011), we also found that holistic methods were much preferred over mechanical ones. The reasons for this could be many, for example, applicants might feel that holistic methods are more personal or that they provide opportunities to get to know the prospective employer and get a sense of the person-organization fit (Cable & Judge, 1996).
Although previous research has shown that recruiters often do not outperform laypeople in judgments of applicants (e.g., Re & Rule, 2016; Sinclair & Agerström, 2020), the results further suggest that they are generally perceived as much better than laypeople at making such judgments. In other words, applicants may significantly overestimate the extent that recruiters can accurately assess qualities such as personality, intelligence, honesty, or leadership abilities based on simply screening resumes and conducting interviews. Taken together, these findings contribute new knowledge to the literature on resistance to decision aids, by pointing to the importance of focusing on not only how applicants perceive specific selection methods, but also on how they perceive the skills of practitioners.
Limitations
The most obvious limitation of this study is that it did not involve a real-life selection situation, and future research should ideally replicate the findings in a real setting. However, as participants rated their perceptions of selection methods in light of a job that was relevant to them personally, the scenario should be easy for them to imagine and provide a relevant context that has worked well in past research. Moreover, the measure of process favorability included how fair they would perceive the method if they were not hired for the job. This is crucial, considering that the hiring outcome (being selected vs. rejected) tends to influence favorability (Anderson et al., 2010), and omitting this aspect may thus prevent to capture true favorability. Still, future research is encouraged to expand the dependent variables by including measures of perceived procedural justice according to Gilliland’s (1993) model (also see Bauer et al., 2001).
The measures were not counterbalanced across participants, which entails a risk of order effects influencing the results. Furthermore, as some of the measures were self-constructed, it is possible that the description of the selection methods provided to participants might explain the unexpected finding that structured interviews received somewhat higher ratings than unstructured ones. For example, telling participants what practices recruiters use most often might influence their appraisals of those practices. The dichotomization between unstructured and structured interviews might also be regarded as sub-optimal, as structure can be viewed as a continuum (Levashina et al., 2014). Future studies may thus want to include attitudes to semi-structured interviews.
Finally, although the sample was heterogeneous regarding age, gender, and vocation, it is unclear how well the results generalize to specific occupational groups. Furthermore, the findings may not generalize to all cultures; however, favorability among applicants has been found structurally similar across different countries (Anderson et al., 2010).
Future research
As applicants who trust recruiters’ intuition tend to prefer unstructured selection procedures, a question for future research to pursue is whether enlightening applicants about the soundness of using selection aids compared to intuition will alter their preference. Moreover, in-depth interviews might prove useful for shedding more light on the reasons behind applicants’ beliefs about the use of intuition in the selection process. An increased understanding of the resistance to standardized hiring practices among applicants should be useful when designing and implementing employee selection procedures that are more attractive to applicants, while retaining the predictive validity of standardized approaches (Nolan et al., 2016). Additionally, it would be interesting to examine whether patients’ or clients’ beliefs about practitioners’ intuition is of similar importance in other fields where distrust in mechanical methods is an obstacle to valid practice, for example in medicine (Giluk & Rynes, 2012) or law (Eastwood et al., 2012).
Conclusions
Our findings tentatively suggest that applicants who rely on recruiters intuition tend to prefer selection methods that leave more room for decision-makers idiosyncratic beliefs, as well as for increased opportunities for job candidates to use impression management tactics (Arthur et al., 2006). Applicant reactions can have important consequences for organizations (McCarthy et al., 2017), and considering that this was the first attempt to study applicants faith in recruiters intuition, and that the study has certain limitations, more research, perhaps using other methods, is needed.
Additional File
The additional file for this article can be found as follows:
Online Appendix
Online appendix with survey items and the results from an Exploratory Factor Analysis. DOI: https://doi.org/10.16993/sjwop.183.s1
Competing Interests
The author has no competing interests to declare.