ArticlePDF Available

Abstract and Figures

Resume-screening by human raters is vulnerable to hiring discrimination but recruiter training as a way to overcome biased resume-screening is under-researched. The present study addresses this gap. Building on key cognitive processes that steer discriminatory decision-making in resume-screening and insights from diversity literature, we investigated the effectiveness of two cognitive training interventions (i.e., a culture-general assimilator and a structured free recall intervention) for reducing hiring discrimination against ethnic minority job applicants in the resume-screening stage. A pre-test, repeated post-test experimental study showed initial hiring discrimination (i.e., less positive evaluations of minority job applicants than majority ones), which was reduced shortly after both training interventions. Hiring discrimination, however, resurfaced 3 months later for both interventions. The culture-general assimilator also positively affected participants’ perceived ability to suppress stereotypes, both short-term and long-term. Findings are considered in the light of a comparison of these training interventions, their programme features, and their compatibility with the resume-screening task. Implications for prejudice reduction initiatives, their potential differential effects, and further research are also discussed.
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
European Journal of Work and Organizational Psychology
ISSN: (Print) (Online) Journal homepage:
Reducing ethnic discrimination in resume-
screening: a test of two training interventions
Eva Derous, Hannah-Hanh D. Nguyen & Ann Marie Ryan
To cite this article: Eva Derous, Hannah-Hanh D. Nguyen & Ann Marie Ryan (2021) Reducing
ethnic discrimination in resume-screening: a test of two training interventions, European Journal of
Work and Organizational Psychology, 30:2, 225-239, DOI: 10.1080/1359432X.2020.1756907
To link to this article:
Published online: 25 May 2020.
Submit your article to this journal
Article views: 463
View related articles
View Crossmark data
Reducing ethnic discrimination in resume-screening: a test of two training
Eva Derous
, Hannah-Hanh D. Nguyen
and Ann Marie Ryan
Department of HRM and Organizational Psychology, Ghent University, Ghent, Belgium;
Shidler College of Business, University of HawaiiatMānoa,
Honolulu, USA;
Department of Psychology, Michigan State University, East Lansing, USA
Resume-screening by human raters is vulnerable to hiring discrimination but recruiter training as a way to
overcome biased resume-screening is under-researched. The present study addresses this gap. Building
on key cognitive processes that steer discriminatory decision-making in resume-screening and insights
from diversity literature, we investigated the eectiveness of two cognitive training interventions (i.e.,
a culture-general assimilator and a structured free recall intervention) for reducing hiring discrimination
against ethnic minority job applicants in the resume-screening stage. A pre-test, repeated post-test
experimental study showed initial hiring discrimination (i.e., less positive evaluations of minority job
applicants than majority ones), which was reduced shortly after both training interventions. Hiring
discrimination, however, resurfaced 3 months later for both interventions. The culture-general assimilator
also positively aected participantsperceived ability to suppress stereotypes, both short-term and long-
term. Findings are considered in the light of a comparison of these training interventions, their pro-
gramme features, and their compatibility with the resume-screening task. Implications for prejudice
reduction initiatives, their potential dierential eects, and further research are also discussed.
Received 20 May 2019
Accepted 13 April 2020
Recruitment; resume-
screening; ethnic
discrimination; culture-
general assimilator;
structured free recall
It is hard to argue about the benets of diversity (Burrell, 2009).
Yet, worldwide ethnic minorities still have lower labour market
chances compared to their ethnic majority counterparts and
hiring discrimination has been one of the potential explana-
tions for this dierence. A recent meta-analysis on 42 corre-
spondence audit studies conducted in 18 countries around the
world (19902015) showed that ethnic minorities had to send
about 50% more resumes than equally qualied ethnic majority
applicants in order to get invited for a job interview (Zschirnt &
Ruedin, 2016). Resume-screening is still a widely popular sta-
ing tool: over 98% of companies both in the USA (Piotrowski &
Armstrong, 2006) and Germany (Diekmann & König, 2015) use
this tool as the rst hurdle to screen-in or -out applicants.
Despite its popularity and the many empirical studies show-
ing bias in resume-screening, fairness at this funneling stage
has not been reviewed systematically. Inspired by those obser-
vations, Derous and Ryan (2019) presented a three-stage model
of ethnic bias in resume-screening targeting underlying psy-
chological processes of ethnic bias (like categorization) and
contingencies of bias (like type/amount of ethnic markers).
They also recommended several interventions to reduce bias
at the level of the applicant/tool (e.g., anonymous resume-
screening), that of the decision-maker (e.g., recruiter training),
and that of the broader task/organizational/societal context
(e.g., targeted recruitment). Following from the focus of this
model, the present study considered one particular initiative to
reduce ethnic bias at the level of the decision-maker, namely
recruiter training. Employers may seek to ensure that
organizational members responsible for hiring are properly
trained to become sensitive to cultural dierences and/or to
avoid discrimination in their personnel decision making.
Training to avoid hiring biases should take place before recrui-
ters start meeting applicants (Connerley, 2014). Indeed, much
like entrepreneurial education, it may be important to include
diversity education as early as possible in ones career (e.g., in
business studentscurricula at universities) to create awareness
and sensitize future managers.
Three decades ago Sue (1991) presented a macro-level the-
oretical model for implementing cultural diversity training pro-
grammes in organizations, including in recruitment. Despite
the rise in diversity and cross-cultural training programmes
in organizations as well as research on their eectiveness over
the past decades (Bezrukova et al., 2016; Littrell & Salas, 2005),
there is a lack of systematic evaluation of training interventions
in hiring contexts in general and resume-screening in particular
(Connerley, 2014). This observation not only points to a gap in
the hiring literature but also leaves little empirical guidance
regarding how best to train recruiters.
Study aim and contributions
Following earlier calls (Connerley, 2014; Derous & Ryan, 2019;
Sue, 1991), the present study aims to advance knowledge on
training interventions to reduce hiring discrimination during
resume-screening and contributes to the existing literature on
hiring discrimination in several ways. First, this study consid-
ered two training methods chosen because of their focus on key
cognitive processes involved in decision-making, their
CONTACT Eva Derous
2021, VOL. 30, NO. 2, 225239
© 2020 Informa UK Limited, trading as Taylor & Francis
eectiveness in evaluating others, and whose programme fea-
tures t the nature of the training outcome (i.e., compatibility
principle; Ajzen, 1991), namely the Intercultural Eectiveness
Training (IET; Herfst et al., 2008) and the Structured Free Recall
intervention (SFR; Baltes et al., 2007; Bauer & Baltes, 2002).
Second, whereas diversity training typically focuses on attitu-
dinal outcomes, this study contributes by examining both post-
training attitudes (i.e., whether one feels able to suppress
stereotypes towards ethnic minorities) and the critical beha-
viour of ones actually rating ethnic minoritiesjob suitability
(Bendick et al., 2001). Third, bias in the resume-screening stage
seems to be a relatively persistent phenomenon (Zschirnt &
Ruedin, 2016). Given the tenacious nature of biased decision-
making in this stage, we investigated to what extent training
eects (if any) would last over time. Whereas diversity and
cross-cultural training typically examine training eects in less
than 1 month (Bezrukova et al., 2012; Kalinoski et al., 2013), we
employed a randomized pre-test, repeated post-test experimental
design to assess both immediate eects (i.e., end of training)
and lasting eects (i.e., delayedwith a training-post-test inter-
val of 3 months). Finally, we considered bias in resume-
screening against Arab/Moroccan applicants,
which is an
important and growing ethnic minority group in Western-
Europe but still remains a relatively less investigated group
compared to other ethnic minority groups (i.e., Blacks, Asians,
or Hispanics in the US).
In sum, the present study extends the existing literature on
discrimination during resume-screening by evaluating two cog-
nitive training interventions for reducing the impact of ethnic
bias against Arab/Moroccan applicants. Below, we rst review
research on ethnic discrimination in hiring particularly tailored
to the resume-screening stage and the cultural context of our
study. These ndings illustrate the need for interventions, like
training, to reduce hiring bias. This is followed by a discussion
of two training interventions (IET, SFR), their programme fea-
tures and potential to reduce ethnic discrimination in the
resume-screening stage.
Ethnic discrimination in resume-screening
Bias and discrimination in resume-screening
Resumes are one of the most important sources of information
when HR-managers and recruiters initially screen applicants for
jobs (Piotrowski & Armstrong, 2006). Studies (e.g., Brown &
Campion, 1994; Burns et al., 2014) consistently showed that HR-
managers and recruiters interpret biodata (like grades, work
experience, community activities) not only as indicators of job
applicantsability (based on past performance) but also as the
reection of other important attributes (like ones motivation
and personality). However, inferences might sometimes be
inaccurate (Apers & Derous, 2017). Furthermore, one can easily
infer undisclosed personal characteristics, such as job appli-
cantsethnicity, from resume characteristics (e.g., applicant
name; King et al., 2006; Gaddis, 2017b;applicant aliations;
Dovidio & Gaertner, 2000;applicant pictures; Marira & Mitra,
2013), which might also aect applicant evaluation and even
trigger preferential or discriminatory treatment of applicants of
diverse cultural backgrounds.
In resume-screening contexts, the actions of the decision-
maker i.e., the evaluations of applicants are the critical
outcome of interest. In several correspondence studies con-
ducted in both the US. (e.g., Bertrand & Mullainathan, 2004;
Thanasombat & Trasviña, 2005; Widner & Chicoine, 2011) and
Europe (e.g., Blommaert et al., 2014; Lancee, 2019; Ramos et al.,
2019; Rooth, 2010), ctitious resumes with equivalent qualica-
tions but with ethnic minority-sounding names received signi-
cantly fewer callbacks for interviews from actual employers
than those with ethnic majority-sounding names.
Correspondence audit studies are those that use hypothetical
job applicants who submit their resume information to employ-
ers via company website, telephone, or by mail (as opposed to
in-person audits where research confederates posing as legit-
imate job applicants for employment; Gaddis, 2018; Neumark,
2018; Pager, 2007). Specic to the social group and cultural
context of this study, Thanasombat and Trasviña (2005), for
instance, showed that applicants with Arab-sounding names,
especially men, received the lowest response rates to their
resumes. Similarly, Widner & Chicoine (2011) showed that
male applicants with an Arab name had to send out 2.79
more resumes than their counterparts with a White name to
receive the same number of invitations for an interview. In the
Netherlands, Blommaert et al. (2014) also found that the online
ctitious resumes of Dutch-named applicants were 60% more
likely than Arabic-named applicants to be viewed and subse-
quently contacted by employers. More recently, Ramos et al.
(2019) showed high levels of discrimination against job appli-
cants of Moroccan origin in the Netherlands when compared to
Spain. The above-mentioned studies all illustrate the persistent
and widespread nature of Arab discrimination (see for similar
ndings on Arab name discrimination: Arai et al., 2016; Derous
et al., 2012).
Research has further shown that employers may also make
negative personnel decisions against job applicants based on
other types of resume information, such as applicantsreligious,
ethnic, and political aliations (e.g., Dovidio & Gaertner, 2000),
educational level/socio-economic background (e.g., Gaddis,
2017a), and picture ID on resumes (e.g., Derous et al., 2017).
Moreover, several cues might interact. Gaddis (2017a), for
instance, found that blacknames were less likely to be per-
ceived as black when given by more educated mothers than
those given by less educated ones. Derous et al. (2009) found
that resumes with both Arab names and Arab aliations
received the lowest job suitability ratings, but that more pro-
totypical, dark-skinned pictures of Arab applicants on resumes
had even stronger negative eects than Arab-sounding names
(Derous et al., 2017).
Theoretical rationales
Several social psychological theories address why outgroup
derogation may occur in hiring contexts. According to the
realistic group conict theory (Esses et al., 1998), actual threats
to ingroup members (e.g., economic or monetary in nature,
such as employment, promotion, etc.) may result in negative
intergroup reactions and outgroup derogation. Research has
shown that a combination of perceived actual/symbolic threats
and social group identity may even be more impactful than
realistic conicts of interests. Specically, according to the
social identity theory (Tajfel & Turner, 2004), people may derive
a sense of belonging and self-esteem from favourably compar-
ing the social group to which they belong (e.g., Dutch majority
employees) with other groups to which they do not (e.g., Arab/
Moroccan minority workers); therefore, they will favour mem-
bers of their own group over those from outgroups. In a similar
vein, the social dominance theory (Sidanius & Veniegas, 2000)
posits that the dominant group (i.e., Dutch majority employees)
may be resistant to a loss of the status-quo which aords
ingroup members power and privileges, and thus will engage
in outgroup derogation (i.e., to protect their own power and
When we meet a new person for the rst time, we typically
tend to initially categorize that individual on the basis of group-
level characteristics, such as gender or ethnicity (Fiske, 1998).
As a result, we may think, feel and behave towards that person
in very specic ways. Cognitive models of impression formation
(Brewer & Harasty-Feinstein, 1999; Fiske et al., 1999) suggest
that category-based information processing will be particularly
strong when limited individualized information is available
about that person. In resume-screening, job applicants are
typically judged on the basis of a one- or two-page resume
that only provides limited individuating information. Hence,
from a theoretical point of view, screening resumes may be
very vulnerable to stereotyping and biased decision-making.
This is attested by empirical ndings showing hiring discrimi-
nation to be very prevalent in the rst selection hurdle (i.e.,
resume-screening) compared to later selection stages (Zschirnt
& Ruedin, 2016).
In sum, resume-screening practices seem vulnerable to dis-
criminatory behaviours, especially when human raters are con-
sidered. Hence, interventions are needed that avert hiring
discrimination in this rst selection hurdle. In the following
section, we discuss recruiter training as one such intervention
that addresses this need. We rst explain how we selected the
training interventions, describe each intervention, and end with
a systematic comparison of both interventionsprogramme
features and our study hypotheses.
Training interventions on biased hiring
Because recruiters have an important gatekeeping role in orga-
nizations, training may be critical to enhancing diversity within
organizations (Derous & Ryan, 2019). In an extensive overview
covering 50 years of recruitment research, Connerley (2014)
showed the lack of recruiter training research and more
particular the impact that recruiter training might have on
the organizations ability to attract a diverse workforce. The
author concluded that recruiters should be made aware of
their own biases and that more research is needed that com-
bines insights from recruitment with those of diversity training
literature, which is our aim here.
Several meta-analyses support the eectiveness of training
interventions in increasing cross-cultural sensitivity levels in
organizations (Bezrukova et al., 2016; Kalinoski et al., 2013;
Mendenhall et al., 2004), but their usefulness in hiring con-
texts in general and resume-screening, in particular, has not
been explored (Derous & Ryan, 2019). Training eectiveness in
this area, however, largely depends on training design
(Bezrukova et al., 2012) and when selecting suitable training
interventions one rst needs to consider an intervention that
ts with the nature of the task (i.e., resume-screening) and
outcomes of interest (i.e., unbiased screening). Because social
categorization has been acknowledged as a key cognitive
process in forming (biased) impressions of others and discri-
minatory actions (Fiske & Taylor, 2013) and because resume-
screening is vulnerable to categorization by its very nature,
interventions that target categorization and associated stereo-
types are well suited for lessening the role of categorization in
Even when training interventions have a similar focus (e.g.,
reducing categorization and biased decision-making) they
may still dier in the kind of information to which they draw
traineesattention, how such information is generated,and
what trainees have to do with it. For instance, interventions
may draw traineesattention to one particular ethnic minority
group (e.g., consider information about Arab/Moroccan per-
sons) or consider information about persons with dierent
ethnic origins (Mendenhall et al., 2004), and trainees can be
asked to draw attention to group-level information (e.g., when
the focus is on stereotypes about Arab/Moroccans) or indivi-
dual-level information (e.g., when the focus is on one parti-
cular Arab/Moroccan applicant). Training information may
further be provided by the trainers (e.g., when trainees get
[counter-] stereotypical information about Arab/Moroccan
applicants in vignettes or cases) or may be generated by the
trainees (e.g., when trainees are asked to elicit [counter-]
stereotypical information about Arab/Moroccan applicants).
Also, trainees may be asked to reect on their own compe-
tencies (e.g., how they would best react towards Arab/
Moroccan applicants) or on otherscompetencies (e.g., how
well others reacted towards Arab/Moroccan applicants).
Hence, when evaluating training initiatives, one should
rst of all consider training programme features and their
potential eects on training outcomes (Kalinoski et al., 2013).
Moreover, it has been suggested by Kulik and Roberson (2008)
that Researchers need to extend the traditional no-training
control group design to include comparisons with groups
receiving dierent diversity education content or receiving
the same content in dierent formats (p. 317).
Based on our review of the cross-cultural and diversity
training literatures, we identied a cultural assimilation (i.e.,
the Intercultural Eectiveness Training; Van Oudenhoven,
2004) and a structured free recall intervention (Baltes et al.,
2007), as two methods that address social categorization in
dierent ways and might address the call of Kulik and
Roberson (2008). Specically, the IET is a well-regarded repre-
sentative of cultural-assimilator training method (or intercul-
tural sensitizer), one of the most accepted methods of cross-
cultural training (Mendenhall et al., 2004). The SFR has been
increasingly adopted in organizational research (as in perfor-
mance appraisal) because it also aims at reducing stereotypes.
Furthermore, both have been subject to empirical evaluation
in the past but have not been compared to each other. While
they both focus on social categorization, they dier in terms
of training features (i.e., type of training information and
approach). Finally, these interventions have not yet been
investigated in the context of hiring although from
a theoretical and practical point of view they seemed amen-
able to being used in training those involved with making
hiring decisions. Hence, this study contributes to the literature
by investigating the eectiveness of two diversity training
interventions (i.e., the IET and SFT).
Culture-general assimilator: intercultural eectiveness
Culture assimilators are among the most commonly used and
accepted methods of cross-cultural training interventions
(Bhawuk & Brislin, 2000; Fowler & Blohm, 2004; Mendenhall
et al., 2004). Culture assimilators can be either culture-specic
(e.g., focused solely on one target group or culture) or culture-
general (e.g., about dierent cultures in general; Herfst et al.,
2008). While both have value, organizations with multicultural
workforces and applicant pools may be more likely to invest in
culture-general assimilators to maximize the use of resources.
We thus opted to investigate a culture-general assimilator,
namely the Intercultural Eectiveness Training (IET; Van
Oudenhoven, 2004,2014).
The IET programme shows a series of cross-cultural incidents
in a multimedia format (i.e., short video-clips) each of which
represents a problem or misunderstanding between members
from the traineesown culture (ingroup members) and those
from another culture (outgroup members). Critical incidents
embody ingroup members perceptions and interpretations of
outgroup membersbehaviours that may lead to misunder-
standings or negative evaluations of those from other cultures
(i.e., group-level information). After each critical incident is
shown, trainees consider a xed set of pre-recorded responses
(i.e., externally generated information) and indicate how they
would personally react to the incident (i.e., self-focused
reection). Table 1 presents an example of one critical incident
from the IET intervention and the accompanying xed response
set (Herfst et al., 2008). In the end, trainees receive feedback
about their own responses, i.e., what it is that makes the chosen
action alternative more or less eective.
In their review, Bhawuk and Brislin (2000) concluded that
cultural assimilators have a positive impact on cognitive as well
as behavioural and aective criteria. As regards the IET in
particular, Herfst et al. (2008) showed that traineesperfor-
mance on the critical incidents related to their self-reported
intercultural behaviour (see also Van Oudenhoven, 2014).
Structured free recall interventions
Structured free recall interventions (SFR; Baltes et al., 2007;
Baltes & Parker, 2000a,2000b; Bauer & Baltes, 2002) build on
impression formation theories (Fiske et al., 1999) and social
cognition research as a way to reduce prejudiced attitudes
and stereotyping in performance assessment. Baltes and
Parker (2000a,2000b) specically expected that ratersreliance
on person categorization, or stereotypes when making beha-
vioural evaluations, diminishes when the accessibility of both
cue-consistent and cue-inconsistent behaviours in memory is
When judging others, raters typically conduct a limited
search for information that is suciently diagnostic to provide
a response. The SFR might mitigate potential bias due to cate-
gorization by increasing the accessibility of a range of relevant
events in memory (Baltes et al., 2007). Specically, in SFR, raters
are presented with vignettes showing information about
a target person, being an individual member of a social-
cultural group (i.e., culture-specic). Raters are subsequently
asked to recall both positive and negative behaviours that
they have observed before (i.e., other-focused reection) and
Table 1. Example of a critical incident from the IET-intervention (see Herfst et al., 2008).
Customer service
You are working for the customer service department of a Dutch company. For a while, you have been getting complaints from foreign callers.
They are complaining about not being able to understand the Dutch customer service phone tapes. What do you do?
A. You understand the helplessness of the foreigners and you listen to their suggestions.
B. You tell the foreigners that the company is mainly focused on the Dutch market.
C. You think that you are personally not able to solve their complaints. Therefore you refer them to the head oce of the company.
D. You write a letter to the manager of the company about this problem.
Table 2. Example of a vignette from a script of the SFR-intervention (see Baltes et al., 2007).
Manager: Yousef Haoud
Mr. Haoud, 53, is a
manager of a large
production company.
Below you see a
between him and one
of his subordinates
called Jip.
Jip: There is a problem with Hans. He isnt coming in on time, he is not doing the list of things I have
for him to do, hes not focusing on the things he should be doing.
Yousef: Jip, I think a real good strategy for you in dealing with Hans would be to come up with a list of
you are having problems with, and present those to him.
to rely on these observations when completing their assess-
ment of this target (i.e., focus on individual-level information).
Recalling both negative and positive information about the
target person (i.e., internally generated information) is expected
to reduce the ethnic majority assessorsover-reliance on nega-
tive, stereotypical information about the target. Table 2 shows
a sample vignette from a script of the SFR-intervention (Baltes
et al., 2007).
In a series of lab experiments, researchers provided empiri-
cal evidence for the eectiveness of SFR in reducing gender,
ethnic, and bodyweight-based stereotype endorsement on
performance ratings (Baltes et al., 2007; Bauer & Baltes, 2002;
Rudolph et al., 2012).
Testing the training interventions
When evaluating training interventions, critical components of
the training design such as training objectives/outcomes and
training programme features (Bezrukova et al., 2012) need to
be considered. In what follows, we discuss both in the context
of the interventions chosen to reduce bias in resume-screening.
Training objectives/outcomes
As shown by both eld (e.g., Gaddis, 2018; Zschirnt & Ruedin,
2016) and lab studies (Derous et al., 2012), ethnic discrimination
in hiring and resume-screening, in particular, is a prevalent,
persistent, and wide-spread phenomenon. Research ndings
illustrate the need to assist HR-managers and recruiters in
evaluating resumes in a fair and non-discriminatory manner in
initial screening stages. Since behavioural change is generally
considered the ultimate goal of diversity training (Bendick et al.,
2001), we rst evaluated traineesjob suitability ratings as
a behavioural outcome in the context of resume-screening.
Secondly, we measured ratersperceived ability to suppress
stereotypes, dened as ones self-ecacy in suppressing
stereotypical thinking and reducing prejudiced reactions
towards outgroup members (Plant & Devine, 1998). Because
self-ecacy has been noted as a useful predictor of skill transfer
or skill maintenance (Kraiger et al., 1993) we measured per-
ceived ability to suppress stereotypes as an indicator of inter-
vention eects on evaluatorscondence.
Although training eects typically decline over time (Blume
et al., 2010), research on diversity and cross-cultural training has
not paid much attention to long-term eects. Only a few studies
go beyond immediate and end-of-training evaluations but they
typically examined the training eect in less than 1 month
(Bezrukova et al., 2016; Kalinoski et al., 2013). Yet, it is also
important for organizations to know how long training eects
last in order to re-evaluate, change, or plan recurrent initiatives;
therefore, we examined training eects immediately after the
training (i.e., end-of-training) as well as 3 months later.
A training-post-test interval of 3 months was chosen in the
present study because that is a suciently long period of
time in a recruiting or hiring season in which retraining would
be unlikely to occur. This time frame has also been used by Hill
and Augoustinos (2001) as a reasonable time frame to investi-
gate extended training eects in the context of cross-cultural
training interventions. Although one might hope training
eects to last longer than 1 month, a recently published meta-
analysis on diversity training (Bezrukova et al., 2016) showed
that behavioural and attitudinal learning tend to subside over
time, thereby supporting earlier ndings from Blume et al.
(2010). For both the IET and the SFR, we thus hypothesized:
Hypothesis 1. Ethnic majoritiesratings of ethnic minority (i.e.,
Arab/Moroccan) applicantsjob suitability (H1a) and their per-
ceived ability to suppress stereotypes (H1b) will be higher
immediately after the training than prior to the training, but
these eects will subside over time.
Training programme features
Despite the volume of research on diversity training, including
meta-analyses on training eectiveness (see Bezrukova et al.,
2016) there is only limited research that considered how train-
ing programme features might aect training eectiveness (i.e.,
less biased resume screening). Based on diversity training lit-
erature as well as basic social-cognitive research on categoriza-
tion, Table 3 summarizes four general programme features that
may explain dierences in training eectiveness. Training inter-
ventions in the present study (i.e., IET and SFR) have been
established as eective in the broad literature (e.g., perfor-
mance appraisal in organizations); therefore, we expected
they would also to be eective in resume-screening. However,
both the IET and the SFR dier from each other in programme
Arst research question is whether SFR would be more
successful than IET in changing the target behaviour in this
study because the nature of the resume-screening task is
more aligned with the SFR technique than with the IET techni-
que. Because the SFR involves internal generation of information
and is more culture-specicthan the IET, it might be more
suitable for reducing bias in information processes.
Specically, self-generated information is valuable because
actively generating materials during encoding improves later
memory (the generation eect; Slamecka & Graf, 1978).
Further, impression formation theories (Fiske et al., 1999) sug-
gest that providing more (culture-) specic information
increases individualization, which counters stereotypical infor-
mation processing. Moreover, in the SFR, one evaluates the
past and current performance of ethnic minority workers as
represented in personal records (vignettes), which resembles
the screening of applicantspast and future performance based
on resumes (i.e., compatibility principle; Ajzen, 1991). Hence,
given these similarities, the SFR might be more eective than
the IET when job suitability ratings of particular applicants are
the behavioural outcome variables. We hypothesized:
Hypothesis 2.Therewillbeadierence in the magnitude of train-
ing eects between the SFR and the IET in terms of job suitability
ratings such that greater eects will be found for the SFR.
On the other hand, the IET might be more successful than the
SFR in changing ratersperceived ability to suppress stereotypes.
The rationale here is that the IET asks individuals to reect upon
their own competencies (self-focused) instead of the targets
competencies (other-focused); self-focused reection in training
interventions may be more eective than other-focused
reection when a self-ecacy measure, such as perceived ability
to suppress stereotypes, is considered the outcome variable.
According to the theory of self-schemata and self-reference vs.
other-reference(Markus, 1977;Symons&Johnson,1997), pro-
cessing of self-focused information results in greater elaboration
and better recall because of individualsrich network associated
with the self. Martin et al. (2011) further showed that other-
focused imagery is overall less eective at weakening stereo-
types than self-focused imagery because of the greater elabora-
tion and heightened accessibility of positive associations that
come with self-focus. Furthermore, the IET focuses on how one
thinks about social groups, and groups instead of individuals are
the referent in measures of stereotype suppression (group-level
information; mutual intergroup dierentiation model;see
Hewstone & Brown, 1986). Given these ndings as well as the
compatibility regarding the nature of the IET interventions and
this outcome variable (i.e., perceived ability to suppress stereo-
types), we hypothesized:
Hypothesis 3. There will be a dierence in the magnitude of
eects between the SFR and the IET in terms of ability to
suppress stereotypes, such that greater eects will be found
for the IET.
We used a 2 (Applicant ethnicity: minority vs. majority) x 3
(Training intervention: IET vs. SFR vs. control) pre-test, repeated
post-test experimental design (see Shadish et al., 2002) with
applicant ethnicity as the within-subjects variable and training
intervention as the between-subjects variable. Ratersability to
suppress stereotypes against ethnic minorities and job suitabil-
ity ratings were the studys main outcomes.
A total of N= 264 economic/business students in their nal year
at a large university in The Netherlands participated voluntarily
in a broader research programme (i.e., development centre)
that also included the training intervention manipulation;
they then performed a resume-screening task (see below).
Only ethnic majority (i.e., 100% Dutch/White) students partici-
pated in this study at Time 1 (pre-test, n= 259), Time 2
(immediate post-test, n= 251), and Time 3 (three-month inter-
val post-test, n= 223). Approximately half were males (47.9%)
Table 3. Cross-cultural training programme features and potential eects on training outcomes.
No. Focus Program Feature Potential Eect
1. Focus on culture-specic
vs. culture-general level
of information
Whether information is culture-specic (focused on one cultural
group only) or whether information is culture-general
(focused on a mix of dierent cultures).
Programmes providing more specic information might
increase processes of individualization and counter
stereotypical information processing (i.e., impression
formation theories; Fiske et al., 1999)
2. Focus on internal vs.
external gathering of
cultural information
Whether cultural information is generated internally (by trainee)
or gathered and presented externally (by trainer or another
external source) to the trainee.
Programmes involving actively, internally generated
information might improve encoding and retrieval/recall
from memory (i.e., generation eect; Slamecka & Graf,
3. Focus on group vs.
individual level of
cultural information
Whether cultural information about a group of people/
applicants is being evaluated (group level) or whether an
individual person/applicant is being evaluated (individual
Programmes presenting individual-level information (one
person) might only generalize to outgroup members if social
group information is made salient during the encounter (i.e.,
mutual intergroup dierentiation model; Hewstone &
Brown, 1986)
4. Self- vs. other-focused
reection on cultural
Whether reection on cultural information is self-focused
(trainee has to reect upon how he/she would react) or
other-focused (trainee has to imagine how others would
Programmes including self-focused imaginary might be more
eective at weakening stereotypes because of greater
elaboration and heightened accessibility of positive
associations with the self (Martin et al., 2011; Symons &
Johnson, 1997).
Table 4. Means, standard deviations, intercorrelations and internal consistencies.
Variable MSD 123456789101112
1. Job suitability Dutch applicant (T1) 4.50 .62 .70
2. Job suitability Moroccan applicant (T1) 3.75 .84 .15* .81
3. Job suitability Dutch applicant (T2) 4.38 .75 .34** .06 .80
4. Job suitability Moroccan applicant (T2) 4.11 .86 .29** .32** .29** .82
5. Job suitability Dutch applicant (T3) 4.10 .87 .18* .24** .32** .23** .82
6. Job suitability Moroccan applicant (T3) 3.60 .96 .13 .17* .05 .23** .36** .85
7. Ability to suppress stereotypes (T1) 3.06 .90 .00 .19** .02 .12 .00 .03 .87
8. Ability to suppress stereotypes (T2) 3.03 .83 .17** .18* .12 .07 .11 .05 .10 89
9. Ability to suppress stereotypes (T3) 3.03 .79 .07 .09 .00 .02 .04 .05 .04 .52** .80
10. Age 23.57 2.21 .01 .08 .14* .00 .05 .09 .04 .03 .08
11. Gendera 1.49 .50 .14* .07 .02 .03 .10 .04 .02 .0 .05 .05
12. Work experience 2.20 1.19 .07 .02 .02 .11 .07 .01 .14 .0 .0 .10 .03
Note. N = 209 (listwise deletion). Cronbachs alphas are italicized and on the diagonal. T1 = Time 1 (baseline pre-test); T2 = Time 2 (post-test, immediately after the
intervention); T3 = Time 3 (post-test, 3 months after the intervention). Spearmans Rho, with 0 = male; 1 = female.*p <.05. **p <.01.
and participants had some relevant work experience (i.e., about
M= 2.2 jobs per participant) and were familiar with recruiting.
Training interventions
Intercultural eectiveness training
The IET developed by Van Oudenhoven (2004;seealso
Herfst et al., 2008; Van Oudenhoven, 2014)presentspartici-
pants with two sets of 14 cross-cultural critical incidents
(see Table 1 above for an example) in a multimedia format
(i.e., video clips on a computer). Incidents are based upon
experiences of several hundreds of sojourners and immi-
grantsinTheNetherlands(Herfstetal.,2008). Assimilators
are most eective when critical incidents are further dis-
cussed using lectures, small group discussions, and role
plays (Bhawuk & Brislin, 2000). Therefore, the second set of
14 cross-cultural incidents was used as a post-test after an
intercultural competency session that discussed intercultural
interactions with various cultures (i.e., Antillean/Surinamese,
Asian, Turkish, and Arab/Moroccan) in professional settings.
We discussed the incidents from the rst set and also pre-
sented one extra incident on dealing with Arab/Moroccan
employees. The training lasted 4 hours, including the
assessment of the computerized critical incidents (1 h for
each set).
Structured free recall intervention
We adapted and pilot-tested two scripts from the SFR materials
(courtesy of Baltes), which referred to the performance of two
ethnic minority (i.e., Arab/Moroccan) managers, Jamil
Abdelkimand Yousef Haoud.
The scripts introduced the
manager by means of his photograph, name and main respon-
sibility, followed by eight vignettes of critical incidents (see
Table 2 for an example). Each vignette consisted of a short
description of that manager interacting with a subordinate.
As in the Bauer and Baltes (2002;2007), we instructed partici-
pants in the SFR condition to (a) study the rst script for 5 min,
(b) recall as many positive behaviours of the manager as pos-
sible for 5 min (those behaviours that are the most relevant to
the performance dimensions on which they had to rate the
manager), and (c) recall as many negative behaviours as possi-
ble for another 5 min. (The order of positive and negative
behaviour recall was counter-balanced.) Participants had to
record those behaviours under three dimensions of manager
performance (i.e., employee motivation, employee develop-
ment, and establishing and maintaining rapport). Finally, parti-
cipants had to evaluate the manager on a 7-point behaviourally
anchored rating scale (1 = low performance;7=high perfor-
mance). Participants were also encouraged to refer back to their
recall sheets when they completed the performance evaluation
of that manager. This training intervention was brief, lasting
less than 1 h in total.
The initial procedure was similar for both training inter-
vention conditions and the control conditions. Two weeks
before the intervention (Time 1), dependent variables (i.e.,
job suitability ratings; ability to suppress stereotypes) and
biographical information were measured to establish
a baseline. Job suitability was measured through a resume-
screening task in which participants were asked to evalu-
ate a ctitious job applicant (i.e., based on Derous et al.,
2009). Specically, participants were asked to role-play
a recruiter for an international company. They reviewed
an advertisement for a semi-skilled customer service posi-
tion with a great deal of client contact. Subsequently, they
received two (randomized) applicant resumes to review
and reported whether they would hire each applicant
(i.e., job suitability rating). Beside these two target
resumes, they also reviewed distractor resumes that were
not further taken into account. Study materials for the
resume-screening task (i.e., at Time 1, 2 and 3) were
adapted from Derous et al. (2009). All (pilot-tested)
resumes were job-relevant and equivalent in qualications,
except for ethnicity (Arabic/Moroccan vs. Dutch). Gender
was kept constant (male applicants only). After the
resume-screening task, participants completed the ability
to suppress stereotypes measure.
Two weeks later (Time 2), participants were randomly
assigned to either of the training intervention conditions (IET
or SFR) or the control condition. In the training intervention
conditions, all participants were retested on the dependent
variables after the intervention took place. Dierent but equiva-
lent resumes (as pilot-tested) were used for pre- and post-
training intervention resume-screening tasks. In the control
condition at Time 2, participants were simply retested on the
dependent variables.
In a follow-up session 3 months later (Time 3), the depen-
dent variables (job suitability; ability to suppress stereotypes)
were measured again (both control and training intervention
groups) with equivalent but dierent materials (as pilot-tested).
We ended the study with an open-ended probe to ask for any
suspicion regarding the study purpose.
Job suitability rating
This scale was adapted from existing studies on hiring discri-
mination (Derous et al., 2009) and consisted of ve items mea-
suring ratersoverall impression of the candidate, such as to
what extent they would invite an applicant for an interview (i.e.,
the next stage of the hiring process), using a 7-point Likert-type
scale (1 = not at all;7=very much/absolutely). The internal
consistency levels of the scale ranged from adequate to satis-
factory, Cronbachsα=.70 to .85.
Ability to suppress stereotypes
Five items were selected and adapted from prior research (i.e.,
behavioural intentions towards minorities; Wessel et al., 2007;
motivation to respond without prejudice; Plant & Devine, 1998)
to measure the perceived ability to disregard stereotypes about
ethnic minorities of Arab/Moroccan descent. A sample item is I
believeIhavetheabilitytodisregard stereotypes regarding ethnic
minorities of Arab/Moroccan descent(1 = strongly disagree/not at
all applicable;5=strongly agree/totally applicable). The internal
consistency levels were satisfactory: Cronbachsα= .80 to .89.
Preliminary analyses
Manipulation checks
The manipulation checks at the end of the resume-screening
task showed the correspondence between applicantsethnic
identiers (i.e., names and aliations) and the perceived ethni-
city of the applicants (i.e., Dutch vs. Arab/Moroccan) as inferred
from resumes at Time 1, Time 2, and Time 3. Specically, only
one of the Chi-squares was signicant at Time 3 (χ2 = 436.08,
p< .001), indicating that all but two respondents perceived
applicantsethnic identiers in an expected way at each time
period. (Those two respondents were deleted from further
analyses at Time 3.)
Ethnicity eect
We tested the main eect of applicantsethnicity on job suit-
ability ratings at Time 1 to see whether we could replicate the
ethnicity eects on hiring discrimination found in previous
studies. Overall, job suitability rating was signicantly lower
for the ethnic minority (Moroccan) applicants than for the
majority (Dutch) applicants, F(1, 258) = 137.73, p< .001.
Table 4 presents the descriptive statistics per condition.
Hypothesis testing
A series of ANCOVAs and repeated measures AN(C)OVAs was
conducted to test the hypotheses with applicant ethnicity as
the within-subjects variable and training intervention (IET, SFR,
or control) as the between-subjects variable.
Hypothesis 1a investigated whether job suitability ratings of
the ethnic minority applicants were higher after the training
interventions. At Time 2, there was a signicant two-way inter-
action between ethnicity (minority vs. majority) and training
intervention (IET vs. SFR vs. control): Job suitability ratings were
lower for the ethnic minority (Moroccan) than the majority
(Dutch) applicant in the control condition, but not in the train-
ing intervention conditions, F(1, 219) = 7.34, p < .001,η2 = .06.
Table 5. Descriptive statistics of job suitability ratings for Dutch and Moroccan applicant by training intervention, condition, and time.
Applicant Ethnicity
Dutch Moroccan
Time Condition MSDMSD F pη
Time 1 Control 4.47 .62 3.63 .80 F(1, 109) = 76.12 .00 .41 110
IET 4.62 .53 3.70 .79 F(1, 93) = 71.99 .00 .44 94
SFR 4.41 .74 4.11 .94 F(1, 54) =5.88 .02 .10 55
Time 2 Control 4.44 .72 3.77 .80 F(1, 105) = 55.31 .00 .35 106
IET 4.41 .67 4.29 .80 F(1, 92) = 1.50 .22 .02 93
SFR 4.26 .97 4.54 .76 F(1, 51) =4.16 .05 .08 52
Time 3 Control 4.04 .91 3.56 .86 F(1, 98) = 22.37 .00 .19 99
IET 4.18 .75 3.64 1.07 F(1, 85) = 23.83 .00 .22 86
SFR 4.22 1.00 3.63 .96 F(1, 37) = 10.39 .00 .22 38
Note. IET = Intercultural Eectiveness Training; SFR = Structured Free Recall Intervention.
Figure 1. Job suitability ratings of Arab/Moroccan and Dutch applicants before
and after the IET and the SFR interventions.
We further investigated whether any eects persisted by not
only considering immediate eects at Time 2 (end of training)
but also considering eects at Time 3 (i.e., lasting eects 3
months later). Results showed that the interaction between
applicantsethnicity and the type of training intervention
depended on the time of measurement (i.e., immediately or 3
months after the interventions): A signicant three-way
was found between ethnicity, training interven-
tion, and time of testing (Time 1 vs. Time 2 vs. Time 3),
F(3.80, 416.33) = 7.01, p<.001, η2 = .06. Further analyses
showed that post-training intervention job suitability at Time
2 (immediate post-training intervention) was signicantly
higher for ethnic minority applicants when compared to the
control condition and at Time 1 (pre-training intervention), F(2,
247) = 7.71, p< .001, η2 = .06, and when compared to Time 3
(delayed post-training intervention: T2 + 3 months), F(2,
219) = 12.55, p< .001, η2 = .10. Post-training intervention job
suitability ratings for the ethnic minority applicant at Time 3 did
not dier signicantly from that at Time 1, F(2, 220) = 2.86,
p= .06, η
= .03. Therefore, Hypothesis 1a was supported for
both the IET and SFR: Job suitability ratings (a behavioural
outcome) were higher when immediately measured after the
intervention but this eect disappeared three months later (see
Table 5 and Figure 1).
Hypothesis 1b investigated whether participantsability
to suppress stereotypes increased after training. Overall,
there was a signicant two-way interaction between train-
ing intervention and time of testing after a three-month
interval, F(4, 416) = 6.75, p< .001, η
immediately after the training intervention, F(2,
212) = 7.97, p< .001, η
Specically, ability to suppress stereotypes was signi-
cantly higher immediately after the IET intervention
(Time 2) when compared to the control condition and at
Time 1, F(1, 177) = 10.45, p< .001, η
= .06. At Time 3,
a similar pattern emerged: Ability to suppress stereotypes
was signicantly higher for the IET intervention when com-
pared with the control condition and at Time 1, F(1,
178) = 7.37, p= .01, η2 = .04. (Note that, at Time 3, partici-
pantsability to suppress stereotypic patterns did not dier
signicantly from Time 2, F(1, 178) = .75, p= .39).
Ability to suppress stereotypes signicantly dropped
when measured immediately after the SFR intervention
(Time 2), when compared to that at Time 1, F(1,
129) = 9.22, p< .001, η
= .07, to a level similar to the
control, F(1, 129) = 1.31, p= .25. Also, at Time 3, the ability
to suppress stereotypic patterns did not dier from that at
Time 2, F(1, 131) = 1.07, p= .30.
Therefore, Hypothesis 1b was partially supported for the IET:
Ability to suppress stereotypes (an attitudinal outcome) was
higher when immediately measured after the intervention, but
contrary to what we expected regarding attitudinal outcomes
(Bezrukova et al., 2016), there was no decline in perceived
ability to suppress stereotypes after 3 months. For the SFR,
Hypothesis 1b was not supported (see Table 6 and Figure 2).
Hypotheses 2 and 3 further investigated whether the two
training interventions diered in the magnitude of eects on
job suitability ratings (Hypothesis 2) and ability to suppress
stereotypes (Hypothesis 3). Job suitability ratings of the ethnic
minority applicants were only higher after training intervention
at Time 2; therefore, we investigated Hypothesis 2 at Time 2.
Interestingly, the three-way interaction of ethnicity by training
intervention and time of testing (Time 2 vs. Time 1) showed
medium eect sizes (according to Cohen, 1992) for the IET-
intervention, F(1, 196) = 17.32, p< .001, η
= .08, with slightly
smaller eect sizes for the SFR, F(1, 155) = 3.80, p= .05; η
= .02.
The dierence in eect sizes of the IET and the SFR, however,
was not signicant (z= 1.36, p= .08). Hypothesis 2 was not
supported since both training interventions seemed equally
eective in reducing hiring discrimination immediately after
at Time 2. As already illustrated, when we tested Hypothesis
1b, the eects on the ability to suppress stereotypes were much
greater for the IET than for the SFR, z= 2.2, p= .01 (see
Table 6. Descriptive statistics of ability to suppress stereotypes by training intervention, condition and time.
Ability to Suppress Stereotypes
Condition Control IET Control vs. IET SFR Control vs. SFR
Time M SD M SD F-values p M SD F-values p
Time 1 3.11 .85 2.88 .94 F(1, 199) = 3.27 .07 3.32 .83 F(1, 162) = 3.99 .05
Time 2 2.90 .71 3.27 .82 F(1, 183) = 8.85 .00 2.86 1.03 F(1, 133) =.19 .66
Time 3 2.97 .73 3.23 .79 F(1, 179) = 4.71 .03 2.75 .84 F(1, 133) = 2.23 .14
Note. IET = Intercultural Eectiveness Training; SFR = Structured Free Recall Intervention.
Figure 2. Ability to suppress stereotypes against Arab/Moroccan applicants
before and after the IET and the SFR interventions.
Hypothesis 1b for F-values). Therefore, Hypothesis 3 was
Ethnic discrimination in resume-screening is still prevalent
(Zschirnt & Ruedin, 2016) and ethnic majority raters gave over-
all lower job suitability scores to applicants of Arab/Moroccan
descent compared to equally qualied majority applicants in
the baseline condition of this study. Such ndings mirror
reports on labour market discrimination against ethnic mino-
rities (e.g., Gaddis, 2018; Rooth, 2010) and are troublesome.
Dierential selection rates are unethical, unlawful, and may
lead to a more homogeneous workforce, which in the long
run may result in lower organizational and nancial perfor-
mances (Gompers & Kovvali, 2018). Hence, initiatives are
needed to counter ethnic bias in early screening stages.
Notwithstanding the widespread use of resume-screening
and the plethora of studies showing its vulnerability to ethnic
discrimination, relatively little empirical attention has been
given to initiatives that might counter recruitersethnic bias
in general (Connerley, 2014; Sue, 1991) and during resume-
screening specically (Derous & Ryan, 2019). Inspired by
insights from the recruitment and diversity training literature,
we aimed to advance knowledge of the selection/evaluation of
training interventions to reduce hiring discrimination during
resume-screening. Specically, steered by social-cognitive the-
ories on social categorization (e.g., Fiske, 1998; Fiske & Taylor,
2013) and information-processing features of diversity inter-
ventions (e.g., Bezrukova et al., (2016)), we selected two cogni-
tive-based diversity training methods (i.e., IET and SFR) based
on several premises: (a) that they address categorization in
dierent ways given their programme features (see Table 3;
Bezrukova et al., 2016; Kalinoski et al., 2013), (b) that they have
both been subject to empirical evaluation in the past but have
not yet been compared with each other (Kulik & Roberson,
2008), (c) that they were successful in related but other areas
than recruitment (like performance appraisal; Baltes et al.,
2007), and (d) that they seemed amenable to being used in
training those involved with hiring decisions but have not yet
been applied in this way.
Study ndings
Using a pre-test, repeated post-test experiment, we specically
investigated whether ethnic majority ratings gave higher job
suitability ratings (i.e., a behavioural outcome) and felt more
able to suppress stereotypes (i.e., an attitudinal outcome)
regarding ethnic minority applicants of Arab/Moroccan origin
after an Intercultural Eectiveness Training (IET) or a Structured
Free Recall intervention (SFR). As expected, the eectiveness of
such initiatives depended on the nature of the target outcomes
(i.e., behavioural/job suitability ratings vs. attitudinal/perceived
ability to suppress stereotypes) and the time interval of train-
ing-post-test measures (end-of-training vs. 3 months later). The
further discussion of the main ndings is structured around the
outcome measures: job suitability and ability to suppress
Job suitability
As expected, for both the IET and SFR, job suitability ratings
were higher when measured immediately after the training
interventions but they subsided when measured 3 months
later. When evaluating diversity training initiatives, Kalinoski
et al. (2013) also showed the strongest eects for immediate
evaluations (end-of-training) and the weakest eects for
delayed evaluations (i.e., longer than 1 month after the training
intervention). The latter is in line with more recent meta-
analytic ndings of Bezrukova et al. (2016) who explain this
general decline with prompting theory (Sitzmann et al., 2010).
First, trainees may use dierent cues from their environment to
self-regulate learning after training. If initial attitudes and beha-
viours are not reinforced after training, training outcomes may
easily gravitate back to their original state. Rudolph et al. (2012)
also suggest that the best real-world applications (of the SFR
programme) might be situations with minimal delays between
observing and rating individualsperformances. Secondly,
a culture-general approach like the IET provides information
about dierent social categories, which may result in less reten-
tion of learning when applicants from specic cultural groups
are evaluated. Finally, since ethnic minoritiesjob suitability
ratings increased immediately after the SFR, we have evidence
that the structured free recall technique can actually change
subsequent rating behaviour in another rating context even
when raters are not explicitly told to engage in structured free
recall. Three months later, however, a rebound eect occurred
also for the SFR, perhaps because of a lack of self-managed use
of trained techniques.
We further anticipated that the SFR could be slightly more
eective than the IET because of the compatibility of what
trainees had to do during the resume-screening tasks (i.e.,
evaluating individual minoritiesfuture job performance for
hiring) with the nature of the screening task during the SFR
training (i.e., evaluating individual minoritiespast job perfor-
mance for appraisal) and because the information provided
in the SFR was culture-specic and internally generated by
the trainees. However, the latter was not the case: Job suit-
ability ratings were similarly aected immediately after the
training by both training programmes. One potential expla-
nation might be training duration: training might be more
eective when it is longer. In Kalinoski et al.s(2013)meta-
analysis, diversity training duration of less than 4 h typically
yields a much smaller eect size (d=.09,SD =.19)than
training duration of 48h(d=.49,SD =.22).Thismight
explain why the SFR (lasting less than 1 h) did not result
into greater eectiveness than the IET (lasting 4 h) despite
the formers higher compatibility with the training outcome
(i.e., resume-screening task).
Ability to suppress stereotypes
We found dierential training eects on stereotype suppres-
sion across the programmes. In the IET condition, participants
reported being able to suppress stereotypes about ethnic min-
ority applicants (i.e., Arab/Moroccans) for both short term and
long term. Contrary to what we expected, the SFR did not
signicantly improve reports of ability to suppress stereotypes
at either time point. As expected, the IET was more eective
than the SFR (both end-of-trainingand delayed) and one
reason might be the compatibility regarding the nature of the
IET intervention and this outcome variable. These results might
be explained by dierences in the information-processing tasks
of the two programmes in terms of the foci of information
generation (self-focus vs. other-focus), the level of training infor-
mation (group vs. individual level), and individualization of cul-
tural information (culture-general vs. culture-specic). Because
the IET led participants to reect on their own biases against
groups (i.e., self-focused reection on group-level information), it
may have made the concept of stereotypes of other groups and
the need to suppress them very salient to trainees, whereas the
SFR did not discuss stereotypes of social categories at all but
rather focused on traineesreection on individual targets
competencies (i.e., other-focused reection on individual-level
information). The latter is less compatible with the self-
focused reection, which is inherent to ratersreecting about
their own self-ecacy or ability to suppress stereotypes.
Moreover, ability to suppress stereotypes deals with sup-
pressing negative thoughts, which might be dierent from the
nature of the SFR, in which people are not forced to control
their negative thoughts but forced to generate both negative
and positive thoughts about individuals. Hence, the task that
trainees have to engage in when trained with the SFR ts less
well with this outcome variable (i.e., ability to suppress stereo-
types) than the task in which trainees engage with the IET. In
other words, there is a mismatch between the training task (i.e.,
generating both negative and positive thoughts about Arab/
Moroccans) and the training outcome (i.e., ability to suppress
stereotypes against Arab/Moroccans) in the SFR training,
whereas the training task in the IET ts that outcome well.
Research should also take into account the compatibility of
information-processing features with outcome characteristics
when designing exercises or choosing interventions. That is,
tasks associated with hiring (such as resume-screening and
interviewing) possess certain features that aect how indivi-
duals are able to process information and the extent to which
categories will be activated and relied upon in decision-making
(see Kulik et al., 2007, for a review). Matching interventions to
reduce bias in hiring more closely with the context is likely to
yield greater success than adopting diversity training designed
for broader organizational purposes (Mores, 2016).
Further, it is interesting that IET attendeesdiscriminatory
behaviour resurfaced after 3 months despite their perceived
ability to suppress stereotypes. The broad literature of general
self-determination and maintenance of positive behaviours
(e.g., Deci & Ryan, 2000), however, suggests that intrinsic moti-
vation, high self-ecacy, and lack of barriers are some of the
factors that may prevent a rebound/relapse of an old behaviour
or maintain a newly learned behaviour. Immediately after the
IET-intervention, trainees might have been motivated and able
to perform as trained (i.e., bias-free resume-screening) but they
might not have had the required intrinsic motivation for that
behaviour to maintain. Therefore, and regardless of their per-
ceived ability to suppress stereotypes, the intended beha-
vioural eects could not be maintained over time. These
ndings also illustrate the need for careful evaluation of train-
ing interventions: Evaluating interventions solely on the basis
of Kirkpatricks Level 1 criteria (e.g., attitudinal reactions like
perceived ability of being non-discriminatory; see Kirkpatrick,
1998) might not be sucient and even misleading. To reduce
hiring discrimination, one should focus on the corresponding
outcome variable (i.e., job suitability ratings), which is the cri-
tical outcome of interest in the context of resume-screening.
Strengths, limitations, and research opportunities
To the best of our knowledge, this is one of the rst few studies
that investigated the eectiveness of recruiter training to
reduce hiring discrimination in resume-screening. In doing so,
we address Sues call (1991) to investigate ways to avert dis-
crimination at the recruitment stage. Moreover, we considered
two dierent training interventions (IET vs. SFR), given Kulik
and Roberson (2008) suggestion that more systematic com-
parisons are needed that identify the unique value of dierent
elements of diversity education(p. 316). Secondly, the success
of training programmes largely depends on how compatible
training programmes are with criteria of successful task perfor-
mance (i.e., criterion relevance; Goldstein & Ford, 2002). To
date, there is limited research that considers and discusses
cross-cultural training programmes features that might aect
training eectiveness (Kalinoski et al., 2013), and no research
that targets resume-screening (Derous & Ryan, 2019). We pre-
sented four training programmes features on which the two
training programmes of interest (i.e., IET and SFR) diered and
we discussed their potential dierential eects on training out-
comes. As a next research step, one could use a modular
approachby disentangling these training characteristics in
a more systematic way to get a more ne-grained picture of
diversity training eects (e.g., using culture-specic instead of
culture-general vignettes in an assimilator). Furthermore, we
replicated some of Baltes and colleaguesearlier ndings
(Rudolph et al., 2012) and we successfully adapted the SFR to
reduce bias against Arab/Moroccan individuals in a subsequent
resume-screening task, which is a novel result. Follow-up
research might investigate the task focusing on other poten-
tially stigmatized social groups.
As with any study, some limitations are noteworthy.
Respondents in our study were business students preparing
for the labour market, which might be considered as subopti-
mal for the ecological validity of the design, but was deemed
appropriate for several reasons. First, for ethical reasons, we
could not conduct our research among real recruiters nor
applicants as we would potentially impact actual hiring deci-
sions. Secondly, business students are commonly considered
future managers in the literature of business ethics training
(e.g., Ballantine et al., 2018; Grünbaum, 1997), entrepreneurship
(e.g., Sansone et al., 2019), and work-family research (e.g., Bu &
McKeen, 2000). Similarly, business students in our sample are
future managers who will make personnel decisions in their
future career. While we acknowledge that context characteris-
tics (e.g., organizational culture, tie pressure) might be dierent
for professionals versus students, the basic psychological and
cognitive processes regarding information processing (e.g.,
Fiske, 1998) should be the same. There is indeed evidence
from studies on hiring biases that similar eects are found
with college students as with professional recruiters (Hosoda
et al., 2003; Kalinoski et al., 2013). However, the job interview
literature also shows that more experienced interviewers (like
HR-professionals) are more prone to intuitive decision-making
than less experienced interviewers (Dipboye & Johnson, 2013;
Highhouse, 2008). From this point of view, one can expect that
more experienced recruiters might benet even more from
diversity training initiatives. Meta-analytic results indeed
showed somewhat (but not signicantly) larger eects of diver-
sity training initiatives for employee participants in organiza-
tional settings (g= .42) than for student participants in
educational settings (g= .36; Kalinoski et al., 2013), and for
eld than for lab settings (see also Deshpande et al., 1994).
Hence, student samples may provide similar or even more
conservative estimates of training intervention eects.
We kept the gender of the (ctitious) applicants constant in
order not to make our design too complex. We selected male
applicants because ethnic minority males may experience high-
(est) levels of hiring discrimination in many Western-European
countries (i.e., subordinate male target hypothesis; Arai et al.,
2016;Derousetal.,2012; Thanasombat & Trasviña, 2005).
Typically, discriminatory resume-screening seems not to be
much aected by the gender of the rater (see Watkins &
Johnston, 2000, for an example in the context of resume-
screening and applicantsphysical attractiveness), unless gender
roles are violated (see Waung et al., 2015, for an example in the
context of the video resumes-screening and impression man-
agement tactics). Because our design did not incorporate such
gender role violations, we did not expect gender dierences and
additional analyses also showed that results were not aected
by participantsgender. In terms of future research, we did not
investigate other trainee characteristics, such as prejudiced atti-
tudes, that might moderate training outcomes (e.g., hiring dis-
crimination; Derous et al., 2015). As mentioned by Baltes et al.
(2007): If individuals are consciously attempting to discriminate,
instructions to recall positive behaviours probably will not have
any eect, and the consciously discriminating person will still
only recall negative behaviours(p. 152). Further research, there-
fore, could count in individual dierences as potential boundary
conditions of training eectiveness (Berry, 2004).
Finally, the training intervention itself took place at one
point in time. To prevent potential behavioural relapse and
increase the transfer of training, future research could evaluate
recurrent training initiatives. For example, Saks and Burke
(2012) recently found that the organizational-level initiative of
conducting training evaluation (completed with feedback and
traineesaccountability for training outcomes) on a frequent
basis was positively related to a higher rate of training transfer.
Practical implications
Training may be an eective initiative to avert hiring bias in the
resume-screening stage and our ndings may help to shape the
training of future managers/recruiters. According to Connerley
(2014), recruiters should be made aware of their own biases
before they have contact with potential applicants. Hence,
providing anti-discrimination training as early as possible in
educational curricula might help changing discriminatory atti-
tudes and behaviour in the long run.
Practitioners, however, will not only factor in training eec-
tiveness but also more pragmatic criteria before making pro-
gramme adoption decisions. The IET is a computerized,
multimedia training intervention, which may be ecient but
also relatively costly. The SFR is a less cost-intensive interven-
tion, much shorter and relatively easier to organize, that more
closely resembles the type of evaluation typically done in selec-
tion settings. While the SFR may have somewhat more utility
value compared with the IET in this regard, organizations may
still be more likely to engage in training interventions with
a more general focus (e.g., the IET) when their desire is to
prevent discrimination against applicants across ethnic minor-
ity groups.
As illustrated, training eects may not last that long.
Moreover, in organizations, individuals may be trained to eval-
uate others but they seldom go through a documented set of
activities each time they evaluate a resume as that is often not
feasible to do. Therefore, researchers could develop recurrent
training initiatives that recruiters have access to when needed.
For instance, organizations can utilize training-related remin-
ders or nudges(i.e., something that changes an individuals
behaviour in a predictable fashion; Thaler & Sunstein, 2008),
which has been a new trend to boost economic behaviours and
even eectiveness in organizations (Ilieva & Drakulevski, 2018;
Tikotsky et al., 2019). E-learning modules and webinars may
demonstrate rating techniques in both asynchronous, self-
paced ways (i.e., facilitated by online resources so that learning
is not constrained by time and/or place) and synchronous ways
(i.e., when learning takes place at a certain timeslot and/or
place) for supporting recruiters in rating applicants in a more
ecient way.
Training human raters is just one initiative to reduce hiring
bias in resume-screening; organizations could also alter the
environment in which decisions are made to facilitate more
rational decision-making, such as making decision-makers
more accountable, anonymous resume-screening, or targeted
recruitment (see Derous & Ryan, 2019, for an overview of other
interventions). Some also argue for replacing human raters by
algorithms, which are considered to be more ecient, eective,
and bias-free than humans (Cowgill, 2018; Kleinberg et al.,
2017). Mechanical data combination and machine judgement
can be more accurate and predictive than clinical, human
judgements (Kleinberg et al., 2017; Kuncel et al., 2013).
However, recent cases, such as Amazons eliminating their
automated recruitment system because of its discrimination
tendency against women applicants (Meyer, 2018), also show
that, when trained on historical data, machine learning might
be biased as well. Hence, as long as human input is needed,
either for traditional or automatic resume-screening tasks,
training human raters might remain useful.
Finally, the focus of this study was on averting discrimina-
tion during resume-screening. Future studies might extend the
application of our training initiatives to other domains where
discrimination against minority groups, such as Arabs, is pre-
valent, such as in the rental housing market. In several US.
metropolitan areas, Gaddis and Ghoshal (2015) found that the
enquiries made by ctitious Arab American women to room-
mate-wanted ads received 40% fewer replies than those by
white-sounding women, particularly in neighbourhoods with
mosques. In Sweden, Ahmed and Hammarstedt (2008) found
that Swedish landlords were less likely to call, respond to
enquiries, and/or showing their rental apartments to male
renters with distinctive Arabic-sounding names than to
Swedish male renters. Professional associations of real estate
agents and online platforms hosting rental, room-share ser-
vices of private owners might also incorporate (web-based)
versions of the cognitive-based anti-bias training programme,
aiming at reducing any ethnic biases of real estate agents or
room/apartment owners who advertise on those websites.
Resume-screening is vulnerable to biased-decision making,
particularly when ethnic minority applicants are evaluated.
Addressing earlier calls for diversity-related recruiter training
(e.g., Connerley, 2014), we investigated whether training can be
an eective way to reduce ethnic bias during resume-
screening. Some promising eects were found and can be
useful for the training of future HR managers. However, specic
programme features should be further investigated for a better
understanding of the combined eects of recruiter training
interventions and participant characteristics to limit ethnic dis-
crimination upon organizational entry.
1. Cross-cultural training typically refers to training focused on under-
standing appropriate behaviour across national boundaries,
whereas diversity training programmes more typically focus on
a wide array of social categories (ethnicity, gender, sexual orienta-
tion) which can include cross-cultural topics. Cross-cultural training
is therefore considered a subsection of diversity training. The litera-
ture on both diversity training and cross-cultural training will be
referred to.
2. Morocco belongs to the Arab world as one of the Arab states of
North-Africa that established the Arab Maghreb Union to promote
cooperation and economic integration. Moroccansocial lan-
guage and cultural heritage is Arabic.
3. We also piloted scripts for two ethnic majority (i.e., Dutch) man-
agers, Mark van den Heuveland Dennis Breedtveld. However,
since we were interested in using an adapted version of the SFR for
reducing bias against ethnic minorities (i.e., Moroccans), and since
the pilot study on SFR-eectiveness showed satisfactory results in
this regard, we only used the scripts for the ethnic minority (i.e.,
Moroccan) managers in the main study.
4. Mauchlystest indicated that the assumption of sphericity had been
slightly violated for the two-way interaction of ethnicity and time.
Therefore, the degrees of freedom were corrected using the Huynh
and Feldts estimate of sphericity.
The authors want to thank Jan Pieter van Oudenhoven and Boris Baltes for
their support.
Disclosure statement
No potential conict of interest was reported by the authors.
Eva Derous
Ahmed, A. M., & Hammarstedt, M. (2008). Discrimination in the rental
housing market: A eld experiment on the Internet. Journal of Urban
Economics,64(2), 362372.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior
and Human Decision Processes,50(2), 179211.
Apers, C., & Derous, E. (2017). Are they accurate? Recruiterspersonality
judgments in paper versus video resumes. Computers in Human
Behavior,73(2017), 919.
Arai, M., Bursell, M., & Nekby, L. (2016). The reverse gender gap in ethnic
discrimination: Employer stereotypes of men and women with arabic
names . International Migration Review,50(2), 385412.
Ballantine, J. A., Guo, X., & Larres, P. (2018). Can future managers and
business executives be inuenced to behave more ethically in the work-
place? The impact of approaches to learning on business students
cheating behavior. Journal of Business Ethics,149(1), 245258. https://
Baltes, B. B., Bauer, C. B., & Frensch, P. A. (2007). Does a structured free recall
intervention reduce the eect of stereotypes on performance ratings
and by what cognitive mechanism? The Journal of Applied Psychology,92
(1), 151164.
Baltes, B. B., & Parker, C. P. (2000a). Clarifying the role of memory in the
performance cue eect. Journal of Business and Psychology,15(2),
Baltes, B. B., & Parker, C. P. (2000b). Reducing the eects of performance
expectations on behavioral ratings. Organizational Behavior and Human
Decision Processes,82(2), 237267.
Bauer, C. C., & Baltes, B. B. (2002). Reducing the eect of gender stereotypes
on performance evaluations. Sex Roles,47(9/10), 465476. https://doi.
Bendick, M., Egan, M. L., & Lofhjelm, S. M. (2001). Workforce diversity
training: From anti-discrimination compliance to organizational
development. Human Resource Planning,24(2), 1035. http://www.bend-
Berry, J. W. (2004). Fundamental psychological processes in intercultural
relations. In D. Landis, J. M. Bennett, & M. J. Bennett (Eds.), Handbook of
intercultural training (pp. 166184). Sage Publications.
Bertrand, M., & Mullainathan, S. (2004). Are Emily and Greg more employ-
able than Lakisha and Jamal? A eld experiment on labor market
discrimination. American Economic Review,94(4), 9911013. https://doi.
Bezrukova, K., Jehn, K. A., & Spell, C. S. (2012). Reviewing diversity training:
Where we have been and where we should go. Academy of Management
Learning & Education,11(2), 207227.
Bezrukova, K., Spell, C. S., Perry, J. L., & Jehn, K. A. (2016). A meta-analytical
integration of over 40 years of research on diversity training evaluation.
Psychological Bulletin,142(11), 12271274.
Bhawuk, D., & Brislin, R. (2000). Cross-cultural training: A review. Applied
Psychology,49(1), 162191.
Blommaert, L., Coenders, M., & van Tubergen, F. (2014). Discrimination of
Arabic-named applicants in the Netherlands: An internet-based eld
experiment examining dierent phases in online recruitment
procedures. Social Forces,92(3), 3.
Blume, B. D., Ford, J. K., Baldwin, T. T., & Huang, J. L. (2010). Transfer of
training: A meta-Analytic review. Journal of Management,36(4),
Brewer, M. B., & Harasty-Feinstein, A. S. (1999). Dual processes in cognitive
representation of persons and social categories. In S. Chaiken & Y. Trope
(Eds.), Dual process theories in social psychology (pp. 255270). Guilford
Brown, B. K., & Campion, M. A. (1994). Biodata phenomenology: Recruiters
perceptions and use of biographical information in resume screening.
Journal of Applied Psychology,79(6), 897908.
Bu, N., & McKeen, C. A. (2000). Work and family expectations of the future
managers and professionals of Canada and China. Journal of Managerial
Psychology,15(8), 771794.
Burns, G. N., Christiansen, N. D., Morris, M. B., Periard, D. A., & Coaster, J. A.
(2014). Eects of applicant personality on resume evaluations. Journal of
Business and Psychology,29(4), 573591.
Burrell, L. (2009, August). We just cant handle diversity a research roundup.
Harvard Business Review,90(78), 7074.
Cohen, J. (1992). A power prime. Psychological Bulletin,112(1), 155159.
Connerley, M. L. (2014). Recruiter eects and recruitment outcomes. In
K. Y. T. Yu & D. M. Cable (Eds.), The Oxford handbook of recruitment (pp.
2134). Oxford University Press.
Cowgill, B. (2018). Bias and productivity in humans and algorithms: Theory
and evidence from resume screening [Working paper].
Deci, E. L., & Ryan, R. M. (2000). The whatand whyof goal pursuits:
Human needs and the self-determination of behavior. Psychological
Inquiry,11(4), 227268.
Derous, E., Nguyen, -H.-H., & Ryan, A. M. (2009). Hiring discrimination
against Arab minorities: Interactions between prejudice and job
characteristics. Human Performance,22(4), 297320.
Derous, E., Pepermans, R., & Ryan, A. M. (2017). Ethnic discrimination during
résumé screening: Interactive eects of applicantsethnic salience with
job context. Human Relations,70(7), 7.
Derous, E., & Ryan, A. M. (2019). When your resume is (not) turning you
down: Modelling ethnic bias in resume screening. Human Resource
Management Journal,29(2), 113130.
Derous, E., Ryan, A. M., & Nguyen, H. H. D. (2012). Multiple categorization in
resume screening: Examining eects on hiring discrimination against
Arab applicants in eld and lab settings. Journal of Organizational
Behavior,33(4), 544570.
Derous, E., Ryan, A. M., & Serlie, A. W. (2015). Double jeopardy upon
resumé screening: When Achmed is less employable than Aïsha.
Personnel Psychology,68(3), 659696.
Deshpande, S. P., Joseph, J., & Viswesvaran, C. (1994). Does use of student
samples aect results of studies in cross-cultural training? A
meta-analysis. Psychological Reports,74(3), 779785.
Diekmann, J., & König, C. (2015). Personality testing in personnel selection:
Love it? Leave it? Understand it! In I. Nikolaou & J. K. Oostrom (Eds.),
Employee recruitment, selection, and assessment: Contemporary issues for
theory and practice (pp. 117135). Psychology Press.
Dipboye, R. L., & Johnson, S. K. (2013). Understanding and improving
employee selection interviews. In K. F. Geisinger, B. A. Bracken,
J. F. Carlson, J. I. C. Hansen, N. R. Kuncel, S. P. Reise, &
M. C. Rodriguez (Eds.), APA handbook of testing and assessment in
psychology; Vol. 1: Test theory and testing and assessment in industrial
and organizational psychology (pp. 479499). American Psychological
Dovidio, J. F., & Gaertner, S. L. (2000). Aversive racism and selection deci-
sions: 1989 and 1999. Psychological Science,11(4), 315319. https://doi.
Esses, V. M., Jackson, L. M., & Armstrong, T. L. (1998). Intergroup
Competition and Attitudes Toward Immigrants and Immigration: An
Instrumental Model of Group Conict. Journal of Social Issues,54(4),
Fiske, S. T. (1998). Stereotyping, prejudice, and discrimination. In
D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology
(pp. 357411). McGraw-Hill.
Fiske, S. T., Lin, M., & Neuberg, S. (1999). The continuum model, ten years
later. In S. Chaiken & Y. Trope (Eds.), Dual process theories in social
psychology (pp. 231254). Guilford Press.
Fiske, S. T., & Taylor, S. E. (2013). Social cognition: From brains to culture. Sage
Fowler, S. M., & Blohm, J. M. (2004). An analysis of methods for intercultural
training. In D. Landis, J. M. Bennett, & M. J. Bennett (Eds.), Handbook of
intercultural training (pp. 3784). Sage Publications.
Gaddis, S. M. (2017a). How Black are Lakisha and Jamal? Racial perceptions
from names used in correspondence uudit studies. Sociological Science,4
(19), 469489.
Gaddis, S. M. (2017b). Racial/ethnic perceptions from Hispanic names:
Selecting names to test for discrimination. Socius: Sociological Research
for a Dynamic World,3(1), 111.
Gaddis, S. M. (2018). An introduction to audit studies in the social sciences.
In Gaddis, S. M. (Ed). Audit studies: Behind the scenes with theory, method,
and nuance (pp. 344). Springer.
Gaddis, S. M., & Ghoshal, R. (2015). Arab American housing discrimination,
ethnic competition, and the contact hypothesis. The Annals of the
American Academy of Political and Social Science,660(1), 282299.
Goldstein, I. L., & Ford, K. J. (2002). Training in organizations.Wadsworth
Gompers, P., & Kovvali, S. (2018). The other diversity dividend. Harvard
Business Review, August,7277.
Grünbaum, L. (1997). Attitudes of future managers towards business ethics:
A comparison of Finnish and American business students. Journal of
Business Ethics,16(4), 451463.
Herfst, S. L., van Oudenhoven, J. P., & Timmerman, M. E. (2008). Intercultural
eectiveness training in three western immigrant countries: A
cross-cultural evaluation of critical incidents. International Journal of
Intercultural Relations,32(1), 6780.
Hewstone, M., & Brown, R. J. (1986). Contact and conict in intergroup
encounters. Basil Blackwell.
Highhouse, S. (2008). Stubborn reliance on intuition and subjectivity in
employee selection. Industrial and Organizational Psychology,1(3),
Hill, M. E., & Augoustinos, M. (2001). Stereotype change and prejudice
reduction: Short- and long-term evaluation of a cross-cultural awareness
programme. Journal of Community & Applied Social Psychology,11(4),
Hosoda, M., Stone-Romero, E. F., & Coats, G. (2003). The eects of physical
attractiveness on job-related outcomes: A meta-analysis of experimental
studies. Personnel Psychology,56(2), 431462.
Ilieva, V., & Drakulevski, L. (2018). Applying behavioral economics insights at
the workplace. Journal of Human Resource Management,21(2), 4048.
Kalinoski, Z. T., Steele-Johnson, D., Peyton, E. J., Leas, K. A., Steinke, J., &
Bowling, N. A. (2013). A meta-analytic evaluation of diversity training
outcomes. Journal of Organizational Behavior,34(8), 10761104. https://
King, E. B., Mendoza, S. A., Madera, J. M., Hebl, M. R., & Knight, J. L. (2006).
Whats in a name? A multiracial investigation of the role of occupational
stereotypes in selection decisions. Journal of Applied Social Psychology,
36(5), 11451159.
Kirkpatrick, D. L. (1998). Evaluating training programs (2nd ed. ed.). Berrett-
Kleinberg, J., Lakkaraju, H., Leskovec, J., Ludwig, J., & Mullainathan, S. (2017).
Human decisions and machine predictions. The Quarterly Journal of
Economics, 133(1), 237293.
Kraiger,K.,Ford,K.,&Salas,E.(1993). Application of cognitive, skill-based, and
aective theories of learning outcomes to new methods of training
evaluation. Learning,78(2), 311328.
Kulik, C. T., & Roberson, L. (2008). Common goals and golden opportunities:
Evaluations of diversity education in academic and organizational
settings. Academy of Management Learning and Education,7(3), 309331.
Kulik, C. T., Roberson, L., & Perry, E. L. (2007). The multiple-category pro-
blem: Category activation and inhibition in the hiring process. Academy
of Management Review,32(2), 529548.
Kuncel, N. R., Klieger, D. M., Connelly, B. S., & Ones, D. S. (2013). Mechanical
versus clinical data combination in selection and admissions decisions: A
meta-analysis. Journal of Applied Psychology,98(6), 10601072. https://
Lancee, B. (2019). Ethnic discrimination in hiring: Comparing groups across
contexts. Results from a cross-national eld experiment. Journal of Ethnic
and Migration Studies,3,111.
Littrell, L. N., & Salas, E. (2005). A review of cross-cultural training: Best
practices, guidelines, and research needs. Human Resource
Development Review,4(3), 305334.
Marira, T. D., & Mitra, P. (2013). Colorism: Ubiquitous yet understudied.
Industrial and Organizational Psychology,6(1), 103107.
Markus, H. (1977). Self-schemata and processing information about the self.
Journal of Personality and Social Psychology,35(2), 6378.
Martin, B. A. S., Lee, M. S. W., & Lacey, C. (2011). Countering negative country
of origin eects using imagery processing. Journal of Consumer
Behaviour,10(2), 8092.
Mendenhall, M. E., Stahl, G. K., Ehnert, I., Oddou, G., Osland, J. S., &
Kühlmann, T. M. (2004). Evaluation studies of cross-cultural training
programs: A review of the literature from 19882000. In D. Landis,
J. M. Bennett, & M. J. Bennett (Eds.), Handbook of intercultural training
(pp. 129144). Sage Publications.
Meyer, D. (2018). Amazon reportedly killed an AI recruitment system because
it couldnt stop the tool from discriminating against women. Fortune.
Mores, G. (2016). Designing a bias-free organization. Harvard Business
Review,94(78), ISSN 0017-8012.
Neumark, D. (2018). Experimental research on labor market discrimination.
Journal of Economic Literature,56(3), 799866.
Pager, D. (2007). The use of eld experiments for studies of employ-
ment discrimination: Contributions,critiques,anddirectionsforthe
future. The Annals of the American Academy of Political and Social
Science,609(1), 104133.
Piotrowski, C., & Armstrong, T. (2006). Current recruitment and selection
practices : A national survey of Fortune 1000 rms. North American
Journal of Psychology,8(3), 489496.
Plant,E.A.,&Devine,P.(1998). Internal and external motivation to respond
without prejudice. Journal of Personality and Social Psychology,75(3),
Ramos,M.,Thijssen,L.,&Coenders,M.(2019). Labour market discrimination
against Moroccan minorities in the Netherlands and Spain: A cross-national
and cross-regional comparison. Journal of Ethnic and Migration Studies,
Rooth, D. O. (2010). Automatic associations and discrimination in hiring:
Real world evidence. Labour Economics,17(3), 523534.
Rudolph, C. W., Baltes, B. B., Zhdanova, L. S., Clark, M. A., & Bal, A. C. (2012).
Testing the structured free recall intervention for reducing the impact of
bodyweight-based stereotypes on performance ratings in immediate
and delayed Contexts. Journal of Business and Psychology,27(2),
Saks, A. M., & Burke, L. A. (2012). An investigation into the relationship
between training evaluation and the transfer of training. International
Journal of Training and Development,16(2), 118127.
Sansone, G., Battaglia, D., Landoni, P., & Paolucci, E. (2019). Academic spin-
os: The role of entrepreneurship education. International
Entrepreneurship and Management Journal.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimental and quasi-
experimental designs. Houghton Miin Co.
Sidanius, J., & Veniegas, R. C. (2000). Gender and race discrimination: The
interactive nature of disadvantage. In S. Oskamp (Ed.), Reducing prejudice
and discrimination (pp. 4769). Erlbaum.
Sitzmann, T. E., Ely, K., Brown, K. G., & Bauer, K. N. (2010). Self-assessment of
knowledge: A cognitive learning or aective measure? Academy of
Management Learning & Education,9(2), 169191.
Slamecka, N. J., & Graf, P. (1978). The generation eect: Delineation of a
phenomenon. Journal of Experimental Psychology. Human Learning and
Memory,4(6), 592604.
Sue, D. W. (1991). A model for cultural diversity training. Journal of
Counseling & Development,70(1), 99105.
Symons, C. S., & Johnson, B. T. (1997). The self-reference eect in memory: A
meta-analysis. Psychological Bulletin,121(3), 371394.
Tajfel, H., & Turner, J. C. (2004). The social identity of intergroup behavior. In
J. T. Jost & J. Sidanius (Eds.), Key readings in social psychology (pp.
276293). Psychology Press.
Thaler, R., & Sunstein, C. (2008). Nudge Improving decisions about health,
wealth, and happiness. Yale University Press.
Thanasombat, S., & Trasviña, J. (2005). Screening names instead of quali-
cations: Testing with emailed resumes reveals racial preferences. AAPI
Nexus Journal: Policy, Practice, and Community,3(2), 105115. https://doi.
Tikotsky, A., Peer, E., & Feldman, Y. (2019). Which nudges do businesses
like? Managersattitudes towards nudges directed at their business or at
their customers. Journal of Economic Behavior & Organization, 190(2020),
Van Oudenhoven, J. P. (2004). Trainen van interculturele eectiviteit: De
Interculturele EectiviteitsTraining [Training of intercultural eective-
ness: The Intercultural Eectiveness Training]. In P. W. J. Thijssen &
J. G. L. Schramade (Eds.), Handboek eectief opleiden (pp. 114). Delwel.
Van Oudenhoven, J. P. (2014). Crossculturele psychologie: De zoektocht naar
verschillen en overeenkomsten tussen culturen. Uitgeverij Coutinho.
Watkins, L. M., & Johnston, L. (2000). Screening job applicants: The impact of
physical attractiveness and application quality. International Journal of
Selection and Assessment,8(2), 7684.
Waung, M., Hymes, R., Beatty, J. E., & McAuslan, P. (2015). Self-promotion
statements in video resumes: Frequency, intensity, and gender eects
on job applicant evaluation. International Journal of Selection and
Assessment,23(4), 345360.
Wessel, J. L., Hebl, M. R., & King, E. B. (2007). Diversity training eectiveness:
An examination of training and trainee eectiveness.Paper presented at
the Annual American Psychological Association (APA),San Francisco, CA.
Widner, D., & Chicoine, S. (2011). Its all in the name: Employment discrimi-
nation against Arab Americans. Sociological Forum,26(4), 806823.
Zschirnt, E., & Ruedin, D. (2016). Ethnic discrimination in hiring decisions: A
meta-analysis of correspondence tests 1990-2015. Journal of Ethnic and
Migration Studies,42(7), 119.
... Finally, an automatic resume matching system can be significant in filtering relevant candidates during the recruitment process. Moreover, resume screening is a sensitive subject in biased decision making i.e., ethnic minority application [7]. Since machine learning models are trained using data, and if the data focuses on specific features, then machine learning models will make biased predictions that can have detrimental effects. ...
... Recently, a variety of data and model versioning tools have appeared to support data engineers and scientists [26]. Popular tools comprise DVC [4], MLFlow [5], Pachy-derm [6], ModelDB [7] and Quilt Data [8]. They typically combine the ability to specify data and/or model pipelines, with advanced versioning support for data/models, and the ability to define and manage model experiments. ...
... -The evaluation metrics. 7 to mitigate the fact that we considered only compendex databse ...
In the last few decades, companies are interested to adopt an online automated recruitment process in an international recruitment environment. The problem is that the recruitment of employees through the manual procedure is a time and money consuming process. As a result, processing a significant number of applications through conventional methods can lead to the recruitment of clumsy individuals. Different JD/Resume matching model architectures have been proposed and reveal a high accuracy level in selecting relevant candidatesfor the required job positions. However, the development of an automatic recruitment system is still one of the main challenges. The reason is that the development of a fully automated recruitment system is a difficult task and poses different challenges. For example, providing a detailed matching explanation for the targeted stakeholders is needed to ensure a transparent recommendation. There are several knowledge bases that represent skills and competencies (e.g, ESCO, O*NET) that are used to identify the candidate and the required job skills for a matching purpose. Besides, modernpre-trained language models are fine-tuned for this context such as identifying lines where a specific feature was introduced. Typically, pre-trained language models use transfer-based machine learning models to be fine-tuned for a specific field. In this proposal, our aim is to explore how modern language models (based on transformers) can be combined with knowledge bases and ontologies to enhance the JD/Resume matching process. Our system aims at using knowledge bases and features to support the explainability of the JD/Resume matching. Finally, given that multiple software components, datasets, ontology, andmachine learning models will be explored, we aim at proposing a fair, ex-plainable, and traceable architecture for a Resume/JD matching purpose.
... Alternatively, training for selection personnel to counter prejudice and bias could be another promising route to explore (e.g. Derous et al., 2020;Isaac et al., 2009). More concretely, knowledge of the burnout stigma could be implemented in the development of such programmes to foster participant's understanding of burnout and stigmatisation and ultimately reduce hiring discrimination. ...
Full-text available
Hiring discrimination towards (former) burnout patients has been extensively documented in the literature. To tackle this problem, it is important to understand the underlying mechanisms of such unequal hiring opportunities. Therefore, we conducted a vignette experiment with 425 genuine recruiters and jointly tested the potential stigma against job candidates with a history of burnout that were mentioned earlier in the literature. We found candidates revealing a history of burnout elicit perceptions of requiring work adaptations, likely having more unpleasant collaborations with others as well as diminished health, autonomy, ability to work under pressure, leadership capacity, manageability, and learning ability, when compared to candidates with a comparable gap in working history due to physical injury. Led by perceptions of a reduced ability to work under pressure, the tested perceptions jointly explained over 90% of the effect of revealing burnout on the probability of being invited to a job interview. In addition, the negative effect on interview probability of revealing burnout was stronger when the job vacancy required higher stress tolerance. In contrast, the negative impact of revealing burnout on interview probability appeared weaker when recruiters were women and when recruiters had previously had personal encounters with burnout.
... Clearly, burnout syndrome's negative career impact remains a threat in the mid to long term. Despite burnout patients' poor promotion prospects, interventions aimed at reducing stigmatic perceptions found to fuel discriminatory behaviour (Corrigan & O'Shaughnessy, 2007) have had success in counteracting other sources of workplace discrimination (Derous, Nguyen & Ryan, 2020). These methods might be effective to combat promotion discrimination; however, this strategy remains a matter to be addressed in future research. ...
Full-text available
Recent studies have explored hiring discrimination as an obstacle to former burnout patients. Many workers, however, return to the same employer, where they face an even more severe aftermath of burnout syndrome: promotion discrimination. To our knowledge, we are the first to directly address this issue in research. More specifically, we conducted a vignette experiment with 406 genuine managers, testing the potential of the main burnout stigma theoretically described in the literature as potential mediators of promotion discrimination. Estimates reveal that compared to employees without an employment interruption, former burnout patients have no less than a 34.4% lower probability of receiving a promotion. Moreover, these employees are perceived as having low (1) leadership, (2) learning capacity, (3) motivation, (4) autonomy and (5) stress tolerance, as well as being (6) less capable of taking on an exemplary role, (7) having worse current and (8) future health, (9) collaborating with them is regarded more negatively, and (10) managers perceive them as having fewer options to leave the organisation if denied a promotion. Four of these perceptions, namely lower leadership capacities, stress tolerance, abilities to take on an exemplary role and chances of finding another job explain almost half the burnout effect on promotion probabilities.
Full-text available
New ventures depend on the capability of entrepreneurs to transform an idea or a technology into a successful company. The literature on entrepreneurship has recognized that Entrepreneurship Education (EE) plays a key role in this process, but the literature on academic spinoffs has focused on other determinants (e.g., Technology Transfer Offices - TTO, and university research expenditures). This research investigates the role of EE in the creation of academic spinoffs by using a new dataset built around 1262 entrepreneurship courses offered between 2011 and 2014 by 80 US universities included in the Licensing Survey by the Association of University Technology Managers - AUTM). Adopting a Poisson panel regression model, we show that, in addition to TTO size and university research expenditures, EE favours the creation of academic spinoffs. Moreover, we find that practical – rather than theoretical - entrepreneurship courses favour the creation of academic spinoffs. We conclude discussing the theoretical and practical implications for universities, students and scholars interested in entrepreneurship.
Full-text available
This paper examines discrimination against job candidates of Moroccan origin in Spain and the Netherlands. By drawing on insights from group threat theory, we specifically examine how latent ethnic conflicts regarding economic goods or cultural values at the regional level affect discrimination rates of Moroccan minorities in both countries. To this aim, we make use of a cross-nationally standardised field experiment with fictitious candidates applying for real job vacancies in Spain and the Netherlands (n = 3681). We find higher levels of discrimination against job applicants of Moroccan origin in the Netherlands. Whereas job candidates of Moroccan origin are six percentage points less likely to receive a positive response from an employer in Spain, the predicted ethnic gap in call-back rates is fourteen percentage points in the Netherlands. Furthermore, while regional differences in discrimination are not related to regional unemployment figures, we do find some evidence that in the Netherlands a larger share of Moroccans in the region exacerbates discrimination against Moroccan minorities. Altogether, the findings point to the need to give greater weight to the impact of widespread negative beliefs about ethnic minorities and how these beliefs can have a profound adverse impact on the integration of disadvantaged ethnic groups within the labour market.
Full-text available
Behavioral economics insights are widely applied by policymakers worldwide, but the private sector seems to be still not fully embracing them. This opens up a new research field and there is a salient call among researchers and practitioners for accumulation of evidence-based nudge interventions at the workplace. This paper reviews studies that apply behavioral economics insights in an organizational setting. The reviewed workplace interventions are based on reminders, default nudges, implementation intentions and priming. There are important take-home messages for human resource practitioners from this relatively novel research stream which has already helped policymakers improve individual and societal welfare worldwide.
Full-text available
Resume screening is the first hurdle applicants typically face when they apply for a job. Despite the many empirical studies showing bias at the resume‐screening stage, fairness at this funnelling stage has not been reviewed systematically. In this paper, a three‐stage model of biased resume screening is presented. We first discuss relevant theoretical perspectives (e.g., job market signalling and impression formation theories) to explain why resume screening is vulnerable to biased decision‐making and ethnic discrimination in particular. On the basis of the best available evidence, we consider contingencies of ethnic discrimination in the applicant, the decision‐maker, and the broader context (e.g., organisation), as well as the effectiveness of interventions that might counter ethnic bias in resume screening. The paper ends with a critical agenda for further research and practice.
Behaviorally informed interventions are growingly used to nudge consumers for various goals. While consumers are usually the target of nudges, businesses are involved in nudges in two manners: either they are the target of the nudges (i.e., government-to-business, or G2B nudges), or they serve as “nudging agents" on behalf of government regulation (i.e., government-to-business-to-consumer, or G2B2C nudges). Businesses’ support of nudges might be critical for nudges' success, but the perceptions of managers towards nudges have never been directly assessed. We surveyed owners and senior managers of small to medium-sized businesses and discovered an overall high level of support for nudges, and in particular for G2B nudges, as compared to G2B2C nudges. We also find variations between types of nudges and their domain. Specifically, the highest levels of support were found for reminders and for nudges that provide information, and for nudges that promote self-regulation. Lower levels of support were found for nudges that involve incentive design or enhanced consumer protections. Policy makers should thus consider the perceptions of business leaders (both as targets of nudges and as nudging agents) to ensure nudges' success.
In today’s California, Asian Americans and Arab Americans have diminished employment opportunities because employment agencies focus on their names, not qualifications. The Discrimination Research Center has documented the response rates to resumes submitted on behalf of men and women who have equal qualifications and ethnically identifiable names of Asian American, Arab American, Latino, African American and white backgrounds. Although potentially illegal and certainly unacceptable, results that showed that individuals with Arab or South Asian names, especially men, received the lowest response rates to their resumes were not particularly surprising in the aftermath of September 11, 2001 and subsequent changes in world affairs. Local animosity and antagonism ranging from discrimination to violence in response to events in the Middle East are well known and fit a historic pattern. Other statistically significant results showing Asian Americans receiving far fewer responses than white women applicants despite their comparable resumes suggest the persistence of long-held perceptions of Asian Americans as “foreigners”, not capable of “fitting in”, and reluctant to complain when wronged. Asian American community organizations and leaders may wish to replicate DRC testing in other parts of the United States or utilize these research results as a basis for workplace advocacy and litigation.
Existing field experimental research unequivocally shows the existence of ethnic discrimination in the labour market. Furthermore, studies have documented considerable variation in discrimination rates across countries. However, while the field of discrimination research is rapidly expanding, there are at present no harmonised comparative studies. This is unfortunate, as we do not know why there are cross-national differences in discrimination. In this paper, I present the GEMM study (N = 19,181), a harmonised cross-national field experiment on hiring discrimination. The GEMM study contains 53 ethnic minority groups and is carried out in six countries: Germany, Norway, The Netherlands, Spain, United Kingdom and the United States. Furthermore, I discuss the need and potential for a comparative analysis of discrimination and outline the methodological challenges of carrying out a cross-national field experiment. The special issue presents results for the major ethnic minority groups in six countries and compares discrimination rates across national contexts.
Understanding whether labor market discrimination explains inferior labor market outcomes for many groups has drawn the attention of labor economists for decades—at least since the publication of Gary Becker’s The Economics of Discrimination in 1957. The decades of research on discrimination in labor markets began with a regression-based “decomposition” approach, asking whether raw wage or earnings differences between groups—which might constitute prima facie evidence of discrimination—were in fact attributable to other productivity-related factors. Subsequent research—responding in large part to limitations of the regression-based approach—moved on to other approaches, such as using firm-level data to estimate both marginal productivity and wage differentials. In recent years, however, there has been substantial growth in experimental research on labor market discrimination—although the earliest experiments were done decades ago. Some experimental research on labor market discrimination takes place in the lab. But far more of it is done in the field, which makes this particular area of experimental research unique relative to the explosion of experimental economic research more generally. This paper surveys the full range of experimental literature on labor market discrimination, places it in the context of the broader research literature on labor market discrimination, discusses the experimental literature from many different perspectives (empirical, theoretical, and policy), and reviews both what this literature has taught us thus far, and what remains to be done.