RUNNING HEAD : Social networking websites in personnel selection 1

Social networking websites in personnel selection: A signaling perspective on recruiters’

and applicants’ perceptions

Nicolas Roulin

University of Lausanne

Adrian Bangerter

University of Neuchâtel

Abstract

Industry surveys and media reports suggest that recruiters increasingly use social

networking websites (SNWs) in the selection process, but corresponding scientific research is

still limited. Using signaling theory, we examine SNWs as a new way for applicants to signal

their qualities to recruiters. Results suggest that recruiters and potential applicants (students

and graduates) both perceive professional SNWs (e.g., LinkedIn) as a potential antecedent of

Person-Job fit information and personal SNWs (e.g., Facebook) as a potential antecedent of

Person-Organization fit information. When evaluating the same SNW profile, recruiters and

potential applicants focus on different sections of the profile (e.g., recruiters focus more on

job-related information), but they tend to infer similar personality traits. Implications for

using SNWs in selection are discussed.

Keywords: Personnel selection, social networking websites, signaling theory.

Published in Journal of Personnel Psychology (2013), 12(3):143-151. DOI:10.1027/1866-

5888/a000094
RUNNING HEAD : Social networking websites in personnel selection 2

Social networking websites in personnel selection: A signaling perspective on recruiters’ and

applicants’ perceptions

The use of social networking websites (SNWs) to gather information about applicants

has been reported as an increasingly common selection practice (Brown & Vaughn, 2011;

Dipboye, Macan, & Shahani-Denning, 2012; Karl, Peluchette, & Schlaegel, 2010). For

instance, the proportion of US recruiters using SNWs in selection increased from 12% in

2006 to 45% in 2009 (Careerbuilder.com, 2009). Such practice is especially relevant for

young graduates, for whom recruiters have limited background information (Du, 2007).

Recruiters can access profiles’ information, despite options available to applicants to limit

access. Many applicants fail to activate privacy settings and recruiters can create their own

profiles to infiltrate applicants’ networks or hire young graduates or students to do so

(Brandenburg, 2008). More recently, employers have started to ask applicants to provide full

access to their profiles (Brown & Vaughn, 2011). Yet, despite the growing attention devoted

to SNWs in the social sciences (Wilson, Gosling, & Graham, 2012), research in personnel

selection is still limited. For instance, there is little empirical research examining what

information recruiters actually collect from SNW profiles or potential differences between

personal SNWs (e.g., Facebook) and professional ones (e.g., LinkedIn). Also, there is little

research on applicants’ perceptions of such practices and whether or not they converge with

recruiters’ actual practices. We examine these issues using a signaling perspective on

personnel selection and presenting data from a study with professional recruiters and potential

applicants.

A signaling perspective on using SNWs in selection

Recently, Bangerter, Roulin, and König (2012) proposed a new framework for

personnel selection based on signaling theory. They suggested that organizations (and

recruiters) look for so-called honest signals of applicants’ ability and commitment to the

potential employment relationship. Honest signals constitute information about applicant’s

RUNNING HEAD : Social networking websites in personnel selection 3

characteristics that are difficult to fake because they are beyond the conscious control of the

sender (e.g., ability test scores) or because they require investment of costly resources (e.g.,

time and energy spent to obtain a university degree). Honest signals of ability include

education credentials, job experience, or letters of recommendations. Organizations can use

such signals to assess whether or not applicants possess the required abilities or skills for the

job. As such, signals of ability correspond to antecedents that can be used by organizations to

predict applicants’ level of Person-Job (P-J) fit (Kristof-Brown, 2000). Similarly, honest

signals of commitment to the employment relationship include measures of applicants’

motivations to join the organization, such as their willingness to move to another city or to

accept a lower salary. Other signals, like applicants’ personality and values, inform about the

fit between the applicant and the organization and can thus be considered as indirect signals

of potential commitment. Organizations can compare such signals with their organizational

values or culture to assess whether or not the applicant will be a good fit with the organization

and will be likely to be committed to the organization (i.e., accept an offer, and then remain in

the job for a long period). As such, signals of commitment correspond to antecedents that can

be used by organizations to predict applicants’ level of Person-Organization (P-O) fit

(Kristof-Brown, 2000). We note that the term “commitment” in this sense is similar but not

identical to organizational commitment (e.g., Meyer & Allen, 1991). Moreover, applicants try

to detect organizations’ selection criteria and adapt their behavior to send the right signals

(Bangerter et al., 2012). In turn, organizations may counteradapt to applicants’ adaptations by

changing their selection criteria or the way they interpret applicants’ signals. Over time,

cycles of adaptations and counteradaptations between organizations and applicants determine

how signals emerge, evolve, and decline, but also determine the stability of a given signal1.

The emergence of a stable signal is thus the result of a process of reciprocal

interaction between actors in a market. In this paper, we propose that information on

1
See Bangerter et al. (2012) for examples of emerging, stable, and declining signals.
RUNNING HEAD : Social networking websites in personnel selection 4

applicants’ profiles on SNWs may correspond to a potentially emerging signal of applicant

ability and commitment. (Bangerter et al., 2012). Currently, applicants post personal

information on SNWs. Some job market actors (e.g., recruiters) may realize that such

incidentally available information is revealing of some characteristic of another actor (e.g.,

applicants’ personality). Recruiters may then start to use information on SNWs to glean

information about applicants (Brandenburg, 2008). Applicants may become aware of this

trend and adapt their behavior (e.g., by censoring content or strategically posting

information). Recruiters may also adapt to applicants’ adaptations in turn. Over time, these

repeated adaptations and counteradaptations will wither lead to the stabilization of an

emerging signal or its decline and disappearance. If actors’ subsequent behaviors are mutually

reinforcing (e.g., if applicants do not post falsified information on their SNWs and if

recruiters do not develop strategies to use SNWs as a way to invade applicants’ privacy) the

signal will stabilize. SNWs allow recruiters to gather information (e.g., competencies, job

experience) they generally infer from résumés or cover letters to potentially predict P-J fit. In

addition, SNWs allow collecting information to potentially predict P-O fit. For instance,

SNWs can complement personality tests, since recruiters can attempt to infer applicants’

personality from SNW profiles (Back et al., 2010; Kluemper & Rosen, 2009). SNWs may

also yield information about applicants’ personal life that does not normally appear in

résumés or cover letters, such as interests, relationship status, and political views

(Brandenburg, 2008). But as a first step to assess whether SNWs are emerging as an honest

signal, it is necessary to demonstrate that recruiters’ and applicants’ perspectives converge

regarding the type of information that can be transmitted through SNWs.

Given the increasing variety of SNW types (e.g., personal and professional), an

optimal strategy for recruiters may be to look for different types of information in different

types of SNWs (Roberts & Roach, 2009). For instance, information to predict P-J fit (e.g.,

work experience) is more likely to be found on professional SNWs (e.g., LinkedIn) that are
RUNNING HEAD : Social networking websites in personnel selection 5

generally built like extended online résumés. Information to predict P-O fit (e.g., interests) is

more likely to be found on personal SNWs (e.g., Facebook) that have been originally created

to exchange personal information with friends or family. Thus, professional SNWs are

potential signals of abilities (i.e., to predict P-J fit) while personal SNWs are potential signals

of commitment (i.e., to predict P-O fit) and we expect recruiters to perceive them as such:

Hypothesis 1: Recruiters prefer to gather (1a) information to predict P-J fit from

professional rather than from personal SNWs, but (1b) information to predict P-O fit from

personal rather than from professional SNWs.

A second step in demonstrating the emergence of SNWs as a new signal involves

showing that applicants try to detect recruiters’ selection criteria and adapt the signals they

send. This step is especially important with SNW profiles, which allow users (e.g., applicants)

to construct a deliberate and calculated online identity (Vazire & Gosling, 2004), manipulate

the information on one’s profile (Kluemper & Rosen, 2009), and choose what information or

photographs to publish to be viewed favorably (Siibak, 2009). But in order to best adapt

signals, applicants should be aware of recruiters’ use of professional and personal SNWs as

signals of ability and commitment. This is similar to applicants preparing themselves for job

interviews or tests and corresponds to applicants’ attempts to “mind-read” organizations and

detect their selection criteria (Bangerter et al., 2012). Signaling theory suggests that

applicants’ attempts to detect selection criteria and organizations’ potential adaptation of their

selection criteria in response to these attempts are essential to the emergence and evolution of

signals. In the long run, applicants’ attempts to detect how recruiters use SNW content should

pressure organizations and recruiters to adapt their evaluation strategies, leading to one of two

possible outcomes: On the one hand, if applicants and recruiters’ behaviors are mutually

reinforcing, SNWs can reach a signaling equilibrium and become a stable signal, i.e., one that

is regularly used during the selection process. On the other hand, applicants and recruiters can

constantly adapt their behavior to each other to develop an advantage. Such cycles of
RUNNING HEAD : Social networking websites in personnel selection 6

adaptation and counteradaptation may lead to an arms race. An example would be recruiters

developing new strategies to access personal information on applicants’ SNW profiles and

applicants trying to avoid recruiters accessing such information, or adapting their presentation

of personal information to anticipate recruiters’ behavior.

The increasing use of SNWs in selection has been largely covered in the media (e.g.,

Careerbuilder.com, 2009; Du, 2007). Also, some organizations inform university career

centers that they use SNWs in selection (Roberts & Roach, 2009). And some universities have

developed training programs to help graduates strategically use SNWs (Saedi & Nguyen,

2011). Therefore, it is likely that potential applicants in many job markets are currently aware

that recruiters may be interested in using SNWs in hiring and will be able to anticipate their

interests. This in turn suggests that their perceptions will converge with recruiters’ actual

signal detection criteria:

Hypothesis 2: Potential applicants perceive recruiters to prefer gathering (2a)

information to predict P-J fit from professional than from personal SNWs, but (2b)

information to predict P-O fit from personal than from professional SNWs.

Recruiters’ and applicants’ use and interpretation of SNW signals

Recruiters and potential applicants may well converge on what type of information

professional and personal SNWs signal about applicants’ qualities. Yet, when examining the

same SNW profile they may differ in their focus of attention (e.g., focus on different content)

or in their interpretation (e.g., infer different characteristics about the applicant). Potential

applicants are not selection experts and they sometimes differ from experienced recruiters in

the way they make hiring decisions. For instance, recruiters tend to focus on the most relevant

cues while students use all cues available, even irrelevant ones (Barr & Hitt, 1986). Also

students tend to focus more on academic qualifications when making hiring decisions, while

professionals tend to focus more on work experience (Singer & Bruhns, 1991). When

examining a SNW profile, experienced recruiters and potential applicants may also differ on
RUNNING HEAD : Social networking websites in personnel selection 7

the type of information they focus on. In exploring this issue, we only investigated personal

SNWs (e.g., Facebook) profiles. Personal SNW profiles generally include a wider range of

information about an applicant than professional SNWs. Also, because of the growing

popularity of personal SNWs like Facebook (Wilson et al., 2012), both recruiters and

(especially) potential applicants are more likely to be familiar with the structure and content

of personal SNWs than professional ones. Facebook profiles are generally structured around

main sections such as the wall (i.e., an open space where both the profile users and their

friends can post comments, pictures or videos), pictures, number of friends (people users are

connected to), personal information (e.g., birth date, relationship status), professional

information (e.g., education, jobs), and interests (e.g., sports, music, literature). Recruiters and

potential applicants may thus focus on different sections when evaluating a profile, but there

is little theoretical basis to hypothesize on what section they will focus on. Therefore, we

propose a first research question to examine this issue:

Research Question 1: Do recruiters and potential applicants focus on different sections

when evaluating an applicant’s personal SNW profile?

Personal SNWs such as Facebook were originally designed to exchange information

with friends or family. Users’ profiles may thus contain pictures or information that they do

not want employers to access, such as negative private behavior or faux pas (e.g., criminal

behavior, alcohol or drug abuse, partying) (Karl et al., 2010; Peluchette & Karl, 2008). Karl

and colleagues suggested that such private behavior may not necessarily be reproduced at

work, but they believed that recruiters may form a negative first impression and infer that,

once hired, applicants could engage in counterproductive behaviors. Such an interpretation

implies that recruiters may construe faux pas as predictive of lower P-O fit. Recent

experimental studies with student evaluators have highlighted the potentially negative impact

of such information (e.g., Bohnert & Ross, 2010; Hanley, Farabee, & Macan, 2010). Yet,

except for recent surveys (Careerbuilder.com, 2009) or media stories (Du, 2007), there is little
RUNNING HEAD : Social networking websites in personnel selection 8

evidence that recruiters actually notice faux pas when evaluating SNW profiles. On the one

hand, recruiters may notice faux pas and interpret them negatively like students do. On the

other hand, if they indeed focus only on cues that are relevant to most jobs (Barr & Hitt,

1986), they may simply consider faux pas as irrelevant. Our second research question

therefore examines whether or not recruiters and potential applicants actually focus on faux

pas in SNW profiles.

Research Question 2: Do recruiters and potential applicants differ in their focus on

faux pas when evaluating an applicant’s personal SNW profile?

Recruiters are also likely to make inferences about applicants’ qualities based on the

information collected on SNW profiles. This information could be posted by the applicant.

But information posted by third parties (i.e., friends) on someone’s page (Walther, Van Der

Heide, Kim, Westerman, & Tong, 2008) or simply the number of friends or their appearance

(Utz, 2010) can also influence inferences made by evaluators. One type of inference that

recruiters could make is related to applicants’ personality (Bohnert & Ross, 2010). People can

accurately infer applicants’ personality or intelligence based on SNW profiles (Back et al.,

2010; Kluemper & Rosen, 2009). For instance, Back et al. (2010) reported significant

correlations between the actual personality of SNWs’ profile owners (aggregated across

multiple personality reports) and personality ratings by external raters based on the SNW

profile. And perceived personality may be influenced by the content of the profile or its

orientation (Bohnert & Ross, 2010). Yet, participants in these studies were explicitly asked to

evaluate targets’ personality (e.g., using the Conscientiousness scale from the NEO FFI;

Bohnert & Ross, 2010). We thus do not know whether evaluators of a SNW profile (i.e.,

recruiters or potential applicants) would spontaneously make such inferences without being

asked to do so. This is important since spontaneous inference corresponds more closely to the

way recruiters actually evaluate applicants’ SNW profiles. Our last research question thus

explores this issue.

RUNNING HEAD : Social networking websites in personnel selection 9

Research Question 3: Do recruiters and potential applicants spontaneously infer

different personality traits when evaluating an applicant’s SNW profile?

We investigated these hypotheses and research questions in a study comparing the

responses from a sample of recruiters and potential applicants who participated in an online

survey on use of SNWs. We operationalized professional SNWs using LinkedIn and personal

SNWs using Facebook. Participants also evaluated a fictitious Facebook profile relative to a

fictitious job advertisement and commented on the profile in a free-response format. We

subsequently coded and analyzed their comments.

Method

Sample and Procedure

Our sample of recruiters was composed of 96 professional HR managers. Mean age

was 40.01 (SD = 7.69) years, 51% were women, and most of them were either French-

speaking Swiss (79.2%) or French (15.6%). Mean experience in personnel selection was 9.37

years (SD = 6.14). Recruiters were active in a variety of sectors, including banking (20%),

manufacturing (20%), or technologies and medias (13%). Our sample of potential applicants

(N = 597) was composed of 443 students and 154 graduates (i.e., currently working or

unemployed). Mean age was 24.28 (SD = 4.97), 78.4% were women, and most of them were

either French-speaking Swiss (72.2%) or French (22.3%). A large proportion of potential

applicants (i.e., 40%) studied (or were currently studying) business, economics, or social

sciences. The educational background for remaining of the sample was education sciences or

humanities (24%), law (16%), engineering or hard sciences (10%), or others (10%).

Recruiters received an email message containing an invitation to participate in the

study and a link to an online questionnaire. A total of 216 HR professionals were originally

contacted, 133 went to the study website and 96 completed the questionnaire (i.e., a 44%

response rate). Potential applicants were contacted by email (approximately 2,000 students

were originally contacted, 193 went to the study website and 146 completed the
RUNNING HEAD : Social networking websites in personnel selection 10

questionnaire, i.e., a 7.5% response rate) or through a Facebook announcement containing a

similar invitation which also encouraged people to forward the invitation to members of their

social network (snowball sampling; the invitation was visible for 2231 persons, 610 went to

the study website and 451 completed the questionnaire, i.e., a 20% response rate). All

participants completed a short online questionnaire in French (average completion time was

13.3 minutes for recruiters and 12.1 for potential applicants). The first part of the

questionnaire included measures of use/perceived use of SNWs in selection. The second part

of the questionnaire included a short job description for a Junior Hiring Manager at a bank,

the Facebook profile of a mock applicant for the job, and a free-response format question

about the content they considered to be noteworthy in this profile.

Measures

Recruiters reported their use of Facebook (i.e., a personal SNW) and LinkedIn (i.e., a

professional SNW) to gather information about applicants to predict their level of P-J fit and

P-O fit, while potential applicants reported their perceptions of recruiters’ use. For instance,

items for recruiters were all phrased as “When I search an applicant’s Facebook/LinkedIn

profile, I look for information about his/her…” A 3-item scale (professional experiences,

competencies, and recommendations; α = .74 for Facebook and α = .76 for LinkedIn) was

used as information to potentially predict P-J fit. A 3-item scale (personality, values,

interests; α = .70 for Facebook and α = .82 for LinkedIn) was used as information to

potentially predict P-O fit. All items were developed based on the antecedents of P-J/P-O fit

from Kristof-Brown (2000) and measured on 5-point Likert scales, where 1 = strongly

disagree and 5 = strongly agree. Moreover, we also measured participants’ frequency of use

of Facebook/LinkedIn in their private life using a 5-point Likert scales, where 1 = never and

5 = very often.

Coding
RUNNING HEAD : Social networking websites in personnel selection 11

In the second part of the study, both recruiters and potential applicants read a short

mock job description for a Junior Hiring Manager in a bank and examined the Facebook

profile of a male applicant for this position. It contained all traditional sections of such

profiles, including pictures (e.g., main profile picture, photographs of sport activities or

parties with friends), number of Facebook friends, wall information (e.g., recent activities or

friends’ postings), personal information (e.g., name, birth date, relationship status, political

orientation, favorites quotes), professional information (e.g., business education, internship in

HR), and interests (e.g., interest in French literature, extracurricular activities). Participants

were then asked to openly describe the content they considered to be noteworthy in a free-

response format. We did not specify “noteworthy” with respect to personnel selection. Yet,

because participants were informed that they would have to examine the Facebook profile of

an applicant for the target job and answer some question about this applicant, it seems likely

that this question was interpreted in relation to the selection. Response length was equivalent

for the two types of participants. The average number of characters used was 108.8 (SD =

134.5) for recruiters and 109.9 (SD = 117.5) for potential applicants, F(1,692)=.007, p = .93.

One of the authors content-analyzed descriptions according to the presence or absence of

various features. Interrater agreement statistics (Cohen’s kappa) for all features coded were

computed based on the independent coding of 70 participants by two different coders (i.e., the

authors) and showed sufficient to perfect agreement (.70-1.00).

Sections of Facebook profile mentioned. We coded whether or not participants

mentioned information from the six sections of the profile mentioned above (i.e., pictures,

wall, number of friends, personal information, professional information, interests). For

instance, the description “he is engaged, cultivated, has a lot of friends, likes to do things

well, is active in sports, and enjoys life” was coded as mentioning number of friends (“a lot of

friends”), personal information (“engaged”), and interests (“active in sports”).

RUNNING HEAD : Social networking websites in personnel selection 12

Faux pas. We coded whether or not participants mentioned faux pas in applicants’

profile (e.g., excessive partying, drug or alcohol abuse). For instance, the description “[…] the

picture where he is sitting on the floor with friends drinking alcohol and (maybe) smoking

joints” was coded as mentioning a faux pas.

Inferences about applicant’s personality. We coded participants’ inferences about

personality traits of the applicant, using the NEO PI-R model (Costa & McCrae, 1992) as a

coding basis. We coded any personality descriptors mentioned into 10 categories defined by

the two extreme ends of the continuum for each of the Big Five personality traits (i.e., open,

low openness, extroverted, introverted, conscientious, impulsive, agreeable, disagreeable,

stable, neurotic) using the definitions and adjectives from the NEO PI-R. For instance, the

description “a young man, sociable, ambitious […]” was coded as mentioning traits

pertaining to extraversion ("sociable") and conscientiousness ("ambitious").

Results

We examined our hypotheses simultaneously using a 2 X 2 X 2 X 2 mixed-model

ANOVA (see Figure 1). Type of information (P-J fit vs. P-O fit) and type of SNW (Facebook

vs. LinkedIn) were entered as within-subjects variables, whereas type of participant (recruiters

vs. potential applicants) was entered as a between-subject variable. Moreover, because of the

large proportion of women in the potential applicant sample, we included gender as a

between-subject control variable. Only participants using (for recruiters) or knowing (for

potential applicants) both types of SNWs were included in this analysis (i.e., N = 30 recruiters

and 226 potential applicants). We found no main effect of gender, F(1,252) = .76, p = .38, η2

=.00, partial η2 =.00, and none of the interactions including gender were significant. We found

no effect of the type of participant, F(1,252) = .04, p = .84, η2 =.00, partial η2 =.00. Thus, we

cannot reject the hypothesis that recruiters’ use of SNWs to gather information about

applicants and potential applicants' perceptions of such use are the same. Moreover, results

showed a type-of-information by type-of-SNW interaction, F(1,252) = 178.85, p < .001, η2

RUNNING HEAD : Social networking websites in personnel selection 13

=.22, partial η2 =.42. Hypotheses 1a and 1b stated that recruiters preferred to gather (1a)

information to predict P-J fit from professional than from personal SNWs, but (1b)

information to predict P-O fit from personal than from professional SNWs. Mean

comparisons suggest that recruiters preferred to gather information about antecedents of P-J

fit from LinkedIn (M = 4.16, SD = .64) rather than from Facebook (M = 2.70, SD = .83), but

preferred to gather information about antecedents of P-O fit from Facebook (M = 3.54, SD =

.83) rather than from LinkedIn (M = 3.36, SD = 1.01). Yet, only the difference for antecedents

of P-J fit was significant (at p < .05), providing full support for Hypothesis 1a but not for 1b.

Moreover, Hypotheses 2 stated that potential applicants perceive recruiters to prefer gathering

(2a) information to predict P-J fit from professional than from personal SNWs, but (2b)

information to predict P-O fit from personal than from professional SNWs. Mean

comparisons suggest that potential applicants perceived recruiters to prefer gathering

information about antecedents of P-J fit from LinkedIn (M = 4.06, SD = .72) rather than from

Facebook (M = 2.47, SD = .83), but perceived recruiters to prefer gathering information about

antecedents of P-O fit from Facebook (M = 4.12, SD = .55) rather than from LinkedIn (M =

3.20, SD = .76)2. Both differences were significant (at p < .05), providing full support for both

Hypotheses 2a and 2b3.

[Figure 1]

We used chi-square tests to examine Research Questions 1, 2, and 3 (Table 1). We

first investigated potential differences between recruiters and potential applicants in the

2
Although only 30 recruiters used both Facebook and LinkedIn in their selection process, more recruiters used
only one SNW. Means computed on the full data available were similar to those obtained with n = 30 for the use
of LinkedIn to gather antecedents of P-J fit (M = 3.95, SD = .62, n = 65) and P-O fit (M = 3.02, SD = .97, n =
65), and the use of Facebook to gather antecedents of P-J fit (M = 2.63, SD = .87, n = 36) and P-O fit (M = 3.54,
SD = .82, n = 36). Similarly, although only 226 potential applicants knew both SNWs (i.e., knew LinkedIn), all
of them but one knew Facebook, And means computed on the full data available were similar to those obtained
with n = 226 for the use of Facebook to gather antecedents of P-J fit (M = 2.61, SD = .87, n = 596) and P-O fit
(M = 4.07, SD = .59, n = 596).
3
Participants who used LinkedIn more in their private life seem to use LinkedIn (or perceived recruiters to use
it) more to gather information about antecedents of P-J fit (r = .25, p < .01) but not information about
antecedents of P-O fit (r = -.06, p = .31). Participants who used Facebook more in their private life seem to use
Facebook (or perceived recruiters to use it) more to gather information about antecedents of P-0 fit (r = .08, p <
.05) but not information about antecedents of P-J fit (r = .01, p = .76).
RUNNING HEAD : Social networking websites in personnel selection 14

sections of the mock Facebook profile they mentioned after examining it (Research Question

1). We found that recruiters mentioned the section of the profile containing potentially job-

related information (i.e., the professional information section) more often than potential

applicants, χ2(1, N=668) = 10.18, p < .01. Recruiters also mentioned the personal information

section, χ2(1, N=668) = 4.92, p < .05, and the number of friends, χ2(1, N=668) = 6.35, p < .05,

more often than potential applicants. Potential applicants mentioned pictures and the wall

more often, χ2(1, N=668) = 7.28, p < .01, and χ2(1, N=668) = 4.34, p < .05, respectively.

These features correspond to the sections of the profile containing less-job-related

information. There was no significant difference between potential applicants and recruiters

regarding the interests section, χ2(1, N=668) = .40, p = .52. We then investigated potential

differences between recruiters and potential applicants regarding faux pas (e.g., alcohol

abuse) mentions (Research Question 2). Potential applicants noticed content related to faux

pas more often than recruiters, χ2(1, N=668) = 4.12, p < .05. We finally investigated potential

differences between recruiters and potential applicants in the personality inferences made

based on the SNW profile (Research Question 3). About one-third of participants made

spontaneous references to personality traits in their response. Both recruiters and potential

applicants mostly mentioned characteristics that were coded as conscientious and extroverted.

We found no significant difference between recruiters and potential applicants for open, low

in openness, extroverted, introverted, agreeable, disagreeable, conscientious, and stable. But

recruiters mentioned more often than potential applicants characteristics that were coded as

impulsive, χ2(1, N=668) = 10.85, p < .01, and neurotic, χ2(1, N=668) = 17.72, p < .01.

[Table 1]

Discussion

In recent years, both media (Careerbuilder.com, 2009; Du, 2007) and the scientific

literature (Brown & Vaughn, 2011; Dipboye et al., 2012) have described a growing interest of
RUNNING HEAD : Social networking websites in personnel selection 15

recruiters in SNWs as a new source of information in the selection process. But empirical

research is still relatively scarce. The present study was an attempt to fill this gap.

We found that that recruiters use professional SNWs to infer applicants’ ability (i.e., to

assess characteristics predictive of P-J fit) and personal SNWs to infer applicants’

commitment to the potential employment relationship (i.e., to assess characteristics predictive

of P-O fit). Moreover, our analyses could not reject the hypothesis that potential applicants'

and recruiters' perceptions are the same. These results are based on a small sample of

recruiters and thus limited statistical power. But this constitutes initial evidence that

applicants’ perceptions of recruiters’ use of these websites may correspond to recruiters’

actual use, thus suggesting that SNWs may be emerging as a signal. However, our results

suggest that potential applicants (i.e., students and graduates) and recruiters tend to differ in

the way they analyze an applicant’s Facebook profile. First, they seem to focus on different

sections of the profile. Recruiters focus more than potential applicants on professional

information (i.e., education and experience) and personal information (i.e., demographics,

marital status) sections, which correspond to information they traditionally find on applicants’

résumés. But they also focus on the number of friends, which may indicate the scope of

applicants’ network. As one respondent wrote: “He has a lot of friends, which may suggest a

good network”. On the other hand, potential applicants focused more than recruiters on

pictures and the wall. Overall, recruiters focus more than potential applicants on sections

containing potentially job-related information, and less than potential applicants on sections

less-related to the job. Differences between experienced recruiters and potential applicants in

hiring decisions (Barr & Hitt, 1986) may thus also apply to SNW profile evaluations. Second,

potential applicants focused more than recruiters on faux pas (Karl et al., 2010). As such, faux

pas may not be perceived as the most relevant information by recruiters, who may focus more

on the personal or professional information available in an applicant’s profile. Finally, and

despite focusing on different sections of SNW profiles, recruiters and potential applicants
RUNNING HEAD : Social networking websites in personnel selection 16

tend to infer similar personality traits from the profile. Both mainly perceived the mock

applicant as conscientious or extraverted. Yet, recruiters perceived the mock applicant as

impulsive and neurotic more frequently than potential applicants. Therefore, when receiving

the same signal (i.e., the same SNW profile), experienced recruiters and potential applicants

tend to have different analytical strategies (i.e., focus on different sections) but end up making

similar interpretations about an applicant’s commitment (i.e., inferring similar personality

traits) to potentially predict P-O fit.

Our findings advance research on signaling theory by highlighting that SNWs may be

currently emerging as a new signal in personnel selection (Bangerter et al., 2012). Yet, it is

not clear whether signals of ability and commitment sent by applicants through their profiles

can be considered as honest signals. Honest signals should be associated with investment cost

(i.e., be costly to send for applicants) and cheating cost (i.e., involve risks in situation of

falsification) (Bangerter et al., 2012). On the one hand, applicants can easily select and

manipulate the information they post of their profile (Kluemper & Rosen, 2009; Siibak,

2009), making SNWs susceptible to honest impression management (e.g., online self-

promotion), but also to deceptive impression management (i.e., cheating). Such behavior

would be difficult to detect by recruiters and undermine the value of the signals. On the other

hand, applicants’ profiles are primarily designed for contact with friends (or family members),

can be edited by friends (e.g., by posting pictures or comments; Walther et al., 2008) and

generally correspond to the actual qualities (e.g., personality) of the applicant (Back et al.,

2010), thus making the signals more difficult to falsify without consequences. But future

research is needed to examine potential applicant manipulation of their profile and the actual

validity of recruiters’ inferences made from SNW profiles.

Our findings also have implications for research on SNWs in selection. Since

experienced recruiters and non-professionals (i.e., potential applicants) differ in the way they

analyze and (to some extent) interpret SNW profiles, results from previous studies involving
RUNNING HEAD : Social networking websites in personnel selection 17

students as evaluators of SNW profiles (e.g., Bohnert & Ross, 2010; Hanley et al., 2010;

Kluemper & Rosen, 2009) may not generalize to recruiters. For instance, Hanley et al. (2010)

stressed that negative pre-interview impressions based on evaluating a Facebook profile

including faux pas may impact post-interview hiring decisions. Yet, their sample was

composed of students. And since less than 20% of recruiters actually mentioned such faux pas

when examining SNW profiles, their impact may be lower than previously suggested.

Therefore, further research examining how the content of SNW profiles actually influences

selection outcomes should be conducted with actual recruiters. Moreover, the importance of

faux pas may also depend on the availability of such information as compared to more

positive and/or job-related information (e.g., job experience, positive recommendations).

Future experimental studies could manipulate the content of profiles (e.g., the proportion of

faux pas as compared to professional information) to assess their impact on recruiters’

perceptions. Future research may also try to replicate the above results about inferences made

by profile evaluators using a professional SNW (e.g., LinkedIn).

Our results also have practical implications for organizations and applicants.

Organizations may want to use SNWs as part of their selection process, but may be concerned

about potential legal issues (Brandenburg, 2008) or negative applicants’ reaction (Stoughton,

Thompson, Meade, & Wilson, 2012). Previous studies stressing potential ethical and legal

issues (e.g., invasion of privacy) related to such practice have focused mainly on personal

SNWs (e.g., Facebook) (Brandenburg, 2008; Brown & Vaughn, 2011; Clark & Roberts,

2010). Professional SNWs (e.g., LinkedIn) may represent a practical alternative for recruiters.

Such websites may limit ethical or legal issues and negative reactions, because applicants

build their profiles for professional use (including job search) and expect employers to view

them (Stoughton et al., 2012). But more research is required to examine the advantage (e.g.,

cost reduction, quality and quantity of information available) and disadvantages (risk of

applicants manipulating information, potential discrimination) of using professional SNWs in

RUNNING HEAD : Social networking websites in personnel selection 18

selection. Moreover, when entering the job market, applicants may want to spend time

constructing a deliberate and calculated online identity (Vazire & Gosling, 2004) and make a

good e-impression on recruiters (Hanley et al., 2010). Our results suggest that they should

worry less about smaller faux pas, such as pictures of parties with friends, and focus more on

developing the professional information section in their profiles (i.e., the section most often

mentioned by recruiters).

This study has limitations. Our results are based on sample composed of Swiss and

French participants and should be replicated in other countries. Our sample of recruiters was

also small for some analyses, resulting in limited statistical power (e.g., when comparing

recruiters’ and applicants’ perceptions). As such, our results should thus be replicated in

future studies. We also note that, as compared to standard response rates for online surveys

(e.g., Cook, Heath, & Thompson, 2000; Deutskens, de Ruyter, Wetzels, & Oosterveld, 2004),

our response rate was relatively high for recruiters but relatively low for potential applicants.

In the first part of our study, we examined the type of information recruiters were looking for

in professional vs. personal SNWs in general (and potential applicants’ perceptions of it).

Future research could examine information-gathering strategies on SNWs that are job-

specific. Moreover, observed differences between recruiters and potential applicants in their

way of analyzing a Facebook profile may also be due to generational differences. Recruiters

and potential applicants were 40 and 24 years old on average respectively. Potential

applicants may also be more used to Facebook profiles than recruiters. Indeed, 84% of

potential applicants reported using Facebook “often” or “very often” in their personal life, as

compared to 25% for recruiters. Also, results of the second part of the study are based on the

Facebook profile of only one mock applicant for only one specific job (i.e., junior hiring

manager at a bank). We do not know to what extent participants paid attention to the job

description and thus if their responses (e.g., sections of the profile mentioned, personality

inferences) were specific to this job or reflected a more general way of analyzing SNWs
RUNNING HEAD : Social networking websites in personnel selection 19

profiles. Moreover, the educational background of a large proportion of our sample of

potential applicants was not in HR, management, or banking. Future research should replicate

these results with other types of SNW profiles, other types of jobs or actual applicants for

specific jobs.

In conclusion, SNWs represent a promising way for applicants to send signals about

their ability or commitment to recruiters. Professional SNW profiles may work as an extended

online résumé, allowing applicants and recruiters to exchange detailed job-related

information, with low cost and without the legal or ethical issues associated with private

SNWs. This study is only a first step in investigating the potential of SNWs in selection and

calls for more research including professional recruiters.

References

Back, M. D., Stopfer, J. M., Vazire, S., Gaddis, S., Schmukle, S. C., Egloff, B., & Gosling, S.

D. (2010). Facebook profiles reflect actual personality, not self-idealization.

Psychological Science, 21, 372-374.

Bangerter, A., Roulin, N., & König, C. J. (2012). Personnel selection as a signaling game.

Journal of Applied Psychology, 97, 719-738.

Barr, S. H., & Hitt, M. A. (1986). A comparison of selection decision models in manager

versus student samples. Personnel Psychology, 39, 599-617.

Bohnert, D., & Ross, W. H. (2010). The influence of social networking web sites on the

evaluation of job candidates. Cyberpsychology, Behavior, and Social Networking, 13,

341-347.

Brandenburg, C. (2008). The newest way to screen job applicants: A social networker’s

nightmare. Federal Communication Law Journal, 60, 597-626.

Brown, V., & Vaughn, E. (2011). The writing on the (Facebook) wall: The use of social

networking sites in hiring decisions. Journal of Business and Psychology, 26, 219-

225.
RUNNING HEAD : Social networking websites in personnel selection 20

Careerbuilder.com. (2009). Nearly half of employers use social networking sites to screen job

candidates. CareerBuilder.com Survey Finds. Retrieved from

http://thehiringsite.careerbuilder.com/2009/08/20/nearly-half-of-employers-use-social-

networking-sites-to-screen-job-candidates/

Clark, L., & Roberts, S. (2010). Employer’s use of social networking sites: A socially

irresponsible practice. Journal of Business Ethics, 95, 507-525.

Cook, C., Heath, F., & Thompson, R. L. (2000). A meta-analysis of response rates in web- or

internet-based surveys. Educational and Psychological Measurement, 60, 821-836.

Costa, P. T., & McCrae, R. R. (1992). Revised neo personality inventory (neo pi-r) and neo

five-factor inventory (neo-ffi). Odessa, FL: Psychological Assessment Resources.

Deutskens, E., de Ruyter, K., Wetzels, M., & Oosterveld, P. (2004). Response rate and

response quality of internet-based surveys: An experimental study. Marketing Letters,

15, 21-36.

Dipboye, R. L., Macan, T., & Shahani-Denning, C. (2012). The selection interview from the

interviewer and applicants perspective: Can't have one without the other. In N. Schmitt

(Ed.), The Oxford handbook of personnel selection and assessment (pp. 323-352).

New York: Oxford University Press.

Du, W. (2007). Job candidates getting tripped up by facebook. MSNBC. Retrieved from

http://www.msnbc.msn.com/id/20202935/from/ET/

Hanley, K., Farabee, A., & Macan, T. (2010). Generation Y hide your secrets? The

e.impression and interview ratings. Paper presented at the 25th Conference of the

Society for Industrial and Organizational Psychology, Atlanta, GA.

Karl, K., Peluchette, J., & Schlaegel, C. (2010). Who's posting Facebook faux pas? A cross-

cultural examination of personality differences. International Journal of Selection and

Assessment, 18, 174-186.

RUNNING HEAD : Social networking websites in personnel selection 21

Kluemper, D. H., & Rosen, P. A. (2009). Future employment selection methods: evaluating

social networking web sites. Journal of Managerial Psychology, 24, 567-580.

Kristof-Brown, A. L. (2000). Perceived applicant fit: Distinguishing between recruiters'

perceptions of person-job and person-organization fit. Personnel Psychology, 53, 643-

671.

Meyer, J. P., & Allen, N. J. (1991). A three-component conceptualization of organizational

commitment. Human Resource Management Review, 1, 61-89.

Peluchette, J. V., & Karl, K. A. (2008). Social networking profiles: An examination of student

attitudes regarding use and appropriateness of content. Cyberpsychology and

Behavior, 11, 95-97.

Roberts, S. J., & Roach, T. (2009). Social networking web sites and human resource

personnel. Business Communication Quarterly, 72, 110-114.

Saedi, B., & Nguyen, H.-H. (2011). Training students to increase employment opportunity

using social networking websites. Paper presented at the 26th Conference of the

Society for Industrial and Organizational Psychology, Chicago, IL.

Siibak, A. (2009). Constructing the self through the photo selection-Visual impression

management on social networking websites. Cyberpsychology: Journal of

Psychosocial Research on Cyberspace, 3.

Singer, M. S., & Bruhns, C. (1991). Relative effect of applicant work experience and

academic qualification on selection interview decisions: A study of between-sample

generalizability. Journal of Applied Psychology, 76, 550-559.

Stoughton, J. W., Thompson, L. F., Meade, A. W., & Wilson, M. A. (2012). Reactions to

using social networking websites in pre-employment screening. Paper presented at the

27th Annual Meeting of the Society for Industrial and Organizational Psychology, San

Diego, CA.
RUNNING HEAD : Social networking websites in personnel selection 22

Utz, S. (2010). Show me your friends and I will tell you what type of person you are: How

one's profile, number of friends, and type of friends influence impression formation on

social network sites. Journal of Computer Mediated Communication, 15, 314-335.

Vazire, S., & Gosling, S. D. (2004). e-Perceptions: Personality impressions based on personal

websites. Journal of Personality and Social Psychology, 87, 123-132.

Walther, J. B., Van Der Heide, B., Kim, S. Y., Westerman, D., & Tong, S. T. (2008). The role

of friends’ appearance and behavior on evaluations of individuals on Facebook: Are

we known by the company we keep? Human Communication Research, 34, 28-49.

Wilson, R. E., Gosling, S. D., & Graham, L. T. (2012). A review of Facebook research in the

social sciences. Perspectives on Psychological Science, 7, 203-220.

RUNNING HEAD : Social networking websites in personnel selection 23

Table 1.

Percentage of Recruiters and Potential Applicants Mentioning SNW Profile Sections,

Mentioning Faux Pas, and Making Inferences about Personality Traits

Potential
Personality trait Recruiters Total χ2
applicants
SNW profile sections
Picture 33.3 49.6 47.7 7.28**
Wall 5.1 13.4 12.5 4.34*
Number of friends 16.7 8.0 9.0 6.35*
Personal information 56.4 43.1 44.7 4.92*
Professional information 39.7 23.1 25.0 10.18**
Interests 28.2 31.7 31.3 .40

Faux pas 19.2 30.3 29.0 4.13*

Personality trait inferences

Open 1.3 3.4 3.1 1.01
Low openness 0 0.7 .6 .47
Extroverted 14.1 14.1 14.1 .00
Introverted 0 0.3 .3 .26
Agreeable 2.6 2.0 2.1 .09
Disagreeable 5.1 2.5 2.8 1.66
Conscientious 16.7 19.2 18.9 .29
Impulsive 9.0 2.2 3.0 10.85**
Stable 1.3 3.2 3.0 .03
Neurotic 5.1 0.3 .9 17.72**

Note: N = 668 (78 recruiters and 590 potential applicants). * p < .05; ** p < .01
RUNNING HEAD : Social networking websites in personnel selection 24

Figure 1 : Information gathered by recruiters and perceived as being gathered by potential

applicants on two types of SNWs to potentially predict P-J fit and P-O fit. Bars represent

standard errors.