This study adds to the literature by aligning the dual-process theory of reactions to stigma (Pryor et al., 2004) with the distinctive stages of the job interview (i.e., rapport building vs. interview stage). Specifically, Study 3.1 investigated the role of the interviewers’ willingness (i.e., motivation) and ability (i.e., cognition) during the initial impression formation of stigmatized and non-stigmatized applicants in relation to biased decisions. On te cognitive level (ability), we found that raters attended to the stigma location three times more when a stigma was present, resulting in decreased memory accuracy for content exchanged during rapport building, which in turn led to decreased hiring intentions. Put differently, we show that discrimination of stigmatized applicants resulted from the inability to engage in individuation due to visual distraction
during rapport building / initial impression formation. No such effects were found with non-stigmatized applicants.
At the motivational level, Study 3.1 investigated to what extent individual differences in need for cognitive closure (NFCC) influenced raters’ information processing. In line with the NFCC theory (Kruglanski & Webster, 1996), we found that NFCC moderated differences in the processing of visual and verbal information during rapport building. When stigmatizing information was available, raters with high NFCC “seized” on strong visual cues, such as the PWS, as this is processed almost instantaneously (Thorpe, Fize, & Marlot, 1996). Additionally, they ”froze” on this information, reducing the need to incorporate verbal information presented during rapport-building
Rapport Building (T1)
Non-stigma condition (control) Stigma condition (PWS)
Indirect effect: -0.001 [-0.06,0.03] Indirect effect: -.16 [-0.33,-0.03] Interview Stage (T2)
Non-stigma condition (control) Stigma condition (PWS)
Indirect effect: .04 [-0.01,0.15] Indirect effect: 0.01 [-0.11,0.02]
Note. Participants (n = 13) with a low eye-tracking accuracy (i.e. higher error) than 0.5˚ were omitted from the analysis. † p = .1, * p = .02. Standardized variables were used in this analysis. All models contain Need for Cognitive Closure (NFCC) as covariate
Figure 2. Mediating effect of Memory per stigma condition and interview stage
Hiring Intentions Fixations .39* .16 Memory -.37* HiringIntentions Fixations .05 .001 Memory -.02 Hiring Intentions Fixations .22† -.11 Memory .17 Hiring Intentions Fixations .09 .04 Memory -.1
78 C HAPT E R 3 a Standardized, b Proportional
Figure 3. Effect of Need For Closure on Memory and Fixations at T1 and T2
T1 T2 Fixations b Memory b 0 0.1 0.2 0.3 0.4 0.5 -3 -2 -1 0 1 2 3
Need For Cognitive Closurea
0 0.1 0.2 0.3 0.4 0.5 -3 -2 -1 0 1 2 3
Need For Cognitive Closurea
0 0.2 0.4 0.6 0.8 1 -3 -2 -1 0 1 2 3
Need For Cognitive Closurea
0 0.2 0.4 0.6 0.8 1 -3 -2 -1 0 1 2 3
when forming their impression on the stigmatized applicant. On the contrary, when strong visual cues were absent, raters with high NFCC were motivated to attend and process verbal information presented during rapport-building.
In sum, these findings provide insight into the effects of stigma on the initial impression formation process during rapport building / initial impression formation process. We show that both motivation (i.e., willingness) and cognition (i.e., inability to adjust the initial heuristic response) affect the initial impression formation process of stigmatized applicants. The relation of this process to interview outcome suggests that it anchors the interviewers’ decision- making process.
S
TUDY3.2
The main strength of Study 3.1 was the assessment of the cognitive and motivational mechanisms underlying bias against facially stigmatized applicants. Study 3.1 also differs from previous studies in that videotaped interviews (dynamic information) instead of pictures (static information) were presented to raters. Nonetheless, interviews were relatively short, there was no direct interaction between raters and applicants, raters were trained students, and there was no opportunity for raters to take notes in the eye tracker. To address these potential limitations, Study 3.2 applied a real-life setting, using face-to- face interviews conducted by professional recruiters.
The primary aim of Study 3.2 was also to further investigate whether anchoring-and-adjustment drives interview bias against facially stigmatized applicants. We do so by investigating whether the interference of stigma during rapport building / initial impression formation, found in Study 3.1, anchors the interviewer’s decision-making process (i.e., insufficient adjustment), and results in bias. In order to assess anchoring and adjustment, and in line with previous studies, Study 3.2 assessed job suitability at different moments: following rapport building (T1), as a measure of initial impressions, and following the behavioral interview stage (T2), as a measure of interview outcome (Barrick et al., 2010; Cable & Judge, 1997; Higgins & Judge, 2004; Stevens & Kristof,
1995). We expected to replicate findings of bias against facially stigmatized applicants such that:
Hypothesis 1. Stigmatized applicants will be judged less suitable for the job compared to non-stigmatized applicants.
Because Study 3.1 showed that a bias against facially stigmatized applicants originates from reflexive reactions during the rapport building stage (i.e., initial impression formation) we first investigated whether this affected the initial impressions. Specifically, dual-process theory proposes that upon perception of the applicant’s stigma, Type 1 processes immediately activate emotions and stereotypical thoughts (Pryor et al., 2004; Wheeler & Petty, 2001). The general perception is that these stereotypes are largely negative in nature, and besides activated they are expected to be applied and in doing so negatively affect the interviewer’s initial impression of the stigmatized applicant to ultimately result in bias (Landy, 2008; Levashina et al., 2013). However, stereotype activation does not imply stereotype application (Krieglmeyer & Sherman, 2012), and dual-process theory proposes that stereotype activation provides additional information that facilitates the initial impression formation process resulting in a strong heuristic output (Nordstrom, Hall, & Bartels, 1998; Thompson, Prowse Turner, & Pennycook, 2011; Thompson et al., 2013). Such a strong initial impression is expected to render further elaboration, or individuation, unnecessary, as was shown in Study 3.1. Given this contradiction in expected effects of the applicant’s stigma on the interviewer’s initial impression of the applicant, we will investigate this. We hypothesize:
Hypothesis 2. Observation of the applicant’s stigma will negatively affect the interviewer’s initial impression of the applicant following rapport building.
One critical assumption driving the validity of the interview is that interview judgments are based, or at least influenced, by job-relevant information exchanged during the interview stage. Moreover, the initial impressions following rapport building (T1) are largely based on small-talk, and therefore can be described as superficial. However, final ratings (T2) are made
following a discussion of job-relevant topics during the behavioral interview stage, and are therefore characterized as content-laden, representing a more valid judgment of the applicants’ job suitability. Hence, as more job-relevant topics are discussed the impressions should positively evolve over time (T1T2). Yet, the Study 3.1 findings suggest that raters’ hiring intentions of stigmatized applicants were anchored in their initial impressions, formed during rapport building, and insufficiently adjusted following the exchange of job-relevant information during the structured interview stage. Therefore we hypothesized:
Hypothesis 3. Overall, ratings of job suitability will be adjusted such that initial impressions following rapport building (T1) will be lower compared to final ratings following the interview stage (T2).
Hypothesis 4. Adjustments in judgments of job suitability from initial impressions to final ratings should be stronger for non-stigmatized applicants, whereas judgments of stigmatized applicants are anchored in the initial impressions and characterized by a lack of adjustment.
If the reflexive reactions to stigma during impression formation, found in Study 3.1, indeed anchor the decision-making process, then eliminating this interference during the initial impression formation process (i.e., during rapport building) should mitigate the anchoring effect. To test this, we introduced partially-blind interviews. Typically, applicants and interviewers have visual contact from beginning to end of the interview (i.e., further referred to as ‘traditional interviews’). Yet, in partially-blind interviews applicant and interviewer are visually separated during the rapport building stage of the interview, and visibility is restored at the beginning of the structured interview stage (see Figure 4). This procedural change was inspired by the music industry, where blind auditions were found to reduce bias against female musicians (Goldin & Rouse, 2000).
Because blind rapport building prevents visual attention towards the stigma, and therefore also the disruption of initial impression formation, we expect that this should mitigate the anchoring mechanism described above. Therefore, we hypothesize:
Hypothesis 5. The interaction effect of stigma with the adjustment in ratings of job suitability throughout the interview (T1 T2) should occur in traditional interviews, but not in partially-blind interviews.