“Give us your social networking site passwords”: Implications for personnel selection and personality

Highlights

• Employers are asking job applicants for their social network site passwords.

• This request may do more harm than good.

• More than half of the applicant pool could be lost.

• Adverse impact may occur.

• Any potential gains in personality scores may have minimal impact on job performance.

Abstract

Recently, employers have begun asking applicants for their social networking site (SNS) password to access private information that could be job-relevant. However, the effect that this request can have on an organization’s selection process and its selection of individual applicant traits has not been previously examined. Findings from the current study of 892 employed or previously-employed participants suggested that 57.87% of the sample would refuse the password request, thereby removing themselves from the applicant pool. Such a large reduction in the applicant pool could necessitate a drastic decrease in cutoff scores on subsequent pre-employment tests, which would lower workforce productivity and personnel selection utility. In addition, the SNS password request caused adverse impact for several minority groups, and affected the personality scores of the remaining applicant pool.

Introduction

It is becoming increasingly common among employers to look beyond traditional pre-employment evaluations to the internet as a source for collecting relevant pre-hiring information (Karl et al., 2010, Karl et al., 2010, Karl et al., 2010, Stoughton et al., 2013, Stoughton et al., 2013, Stoughton et al., 2013, Taylor, 2006, Taylor, 2006, Taylor, 2006). This trend has been increasing (CareerBuilder.com., 2009), to the extent that, as of 2011, up to 65% of organizations are estimated to use SNSs to screen job applicants (Levinson, 2011). In fact, according to the popular press, some employers have begun asking applicants for their personal social networking site (SNS) password in order to access private information that could be job-relevant (McFarland, 2012). Predictably, this practice has raised numerous legal, privacy, and ethical concerns (Levinson, 2011). Although legislation has been created in several U.S. states to ban this practice of requiring a SNS password in order to be considered for a job (Noble & Davis, 2014), it is still legal in many countries (Determann, 2012). In addition, even if only a few employers are engaging in or are considering using this practice, this is an important issue worthy of further study.

We investigated the implications that the SNS password request has regarding the individual differences of the applicants, and we investigated questions relevant to the use of the SNS password request in employee selection. In particular, could asking applicants for their SNS password, and eliminating those who refuse the password request (see Noble, 2014), substantially reduce the applicant pool? Moreover, could this practice lead to discrimination associated with applicant characteristics (e.g., gender, ethnicity, etc.)? Finally, a number of employers have adopted the practice of removing applicants from the applicant pool if they refuse the SNS password (Noble, 2014). Employers may presume that a refusal of the password request is indicative of personalities who would tend to have an abundance of undesirable material displayed in their SNS, and that this does not bode well for their future job performance. Accordingly, they might expect that those who remain in the applicant pool by complying with the password request will generally have personalities that will lead to higher job performance. Thus the question of whether the SNS password request will tend to improve the applicant pool by enhancing, in work-relevant ways, the personality scores of the remaining applicant pool, was of great interest. In particular, we looked at traits of the Five-Factor Model (FFM; McCrae & Costa, 2010), the Dark Triad (Paulhus & Williams, 2002), and Impression Management (IM; Paulhus, 1991). The hypotheses and research questions we formulated in order to address the above issues are developed below.

If a large number of applicants were be removed from the applicant pool because they refused the password request, there would ultimately be fewer candidates to choose from. As a result, a less stringent cutoff score would have to be set on the subsequent selection tests, which would reduce the bottom-line value of testing (Cascio & Aguinis, 2011). In a typical selection setting, such as the one reported by Ryan, Sacco, McFarland, and Kriska (2000), approximately 35% of applicants withdrew from the early stages of selection for a variety of reasons (e.g., negative perceptions of the organization, low commitment to obtain the job, need to relocate, less social support). Using Ryan et al.’s statistic as a baseline, if the proportion of applicants who refuse the SNS password request is significantly greater than .35, there may be grounds for inferring that the password request has caused an inordinate loss of individuals from the applicant pool. We predict that a sizable proportion of the sample will refuse the SNS password request because most individuals are concerned about the privacy of their personal information on the internet (see Joinson, Reips, Buchanan, & Schofield, 2010). Therefore, we predict:

H₁

The proportion of participants who would refuse the SNS password request will be significantly greater than .35.

Of course, if personality traits are associated with password-request compliance such that those who comply with the request tend to have noticeably superior scores on job-relevant traits, then a reduction in pool size could be beneficial. We explore this, and related implications, in a later section.

In addition to reducing the number of available applicants, the request for a SNS password may also be problematic as a screening device if it results in discrimination. The doctrine of disparate impact holds that if a selection test disproportionately and adversely affects a protected group (e.g., based on age, race, sex, marital status, or religious affiliation), its use must be discontinued unless it can be adequately demonstrated that it is necessary to ensure the success of a business (Catano, Wiesner, & Hackett, 2012). We speculate that those who are not members of the majority group may be less likely to divulge their SNS password because their SNS information could confirm and even exacerbate their outsider status or certain biases against them. For instance, women might not release their SNS password for fear of not being hired if their SNS information reveals that they are, or wish to be, pregnant.

If requiring job applicants to provide their SNS password leads to a disproportionate number of protected group members being removed from the selection process, adverse impact could result, which could have serious legal implications. Adverse impact occurs when a protected group is hired at less than 80% of the rate of another group (see the four-fifths rule in Cascio & Aguinis, 2011). Therefore:

RQ₁

Will the SNS password request result in evidence of adverse impact for gender, age, ethnicity, sexual orientation, and/or religious affiliation?

The Five-Factor Model of personality. FFM traits are important and useful predictors of job performance (Ones, Dilchert, Viswesvaran, & Judge, 2007) and personality could play a role in complying with or refusing a SNS password request. Thus, it is important to examine whether employers who limit their choice of applicants to those who comply with the password request may be systematically affecting the FFM scores of the remaining applicant pool. Because compliance with the SNS password request could be associated with personality, and because personality scores are related to job performance, organizations could be improving or worsening the personality scores of the subsequent applicant pool when they require one’s SNS password as a precondition for hiring.

Junglas, Johnson, and Spitzmüller (2008) found a significant, weak, but positive (standardized path coefficient of .11) relation between Openness to Experience and concern for privacy. Junglas et al. described open individuals as being more aware, and therefore, more sensitive to threatening situations. However, other characteristics of Openness, such as the Actions facet (see McCrae & Costa, 2010), might make high scorers more willing to release their information because they are more open to trying and experiencing new things. Thus, we propose this non-directional hypothesis:

H₂

Individuals who would agree to divulge their SNS password will have a significantly different mean level of Openness to Experience than those who would refuse.

Individuals high in Agreeableness are more likely to trust and less likely to appraise others’ actions as harmful (Junglas et al., 2008). Also, agreeable individuals may reveal their SNS password as a way of avoiding conflict and maintaining harmony (see McCrae & Costa, 2010). Moreover, Agreeableness had a weak but significant negative standardized path coefficient (−.22) with concern for privacy in Junglas et al. (2008). Therefore:

H₃

Individuals who would agree to divulge their SNS password will have a significantly higher mean level of Agreeableness than those who would refuse.

On the one hand, Junglas et al. (2008) reported that Conscientiousness had a significant, positive, albeit weak (.12), standardized path coefficient with one’s concern for privacy. On the other hand, conscientious individuals are likely to be more careful about the content they allow on their SNS accounts in the first place, meaning that they have less to fear by allowing access to their SNSs. Thus, as with Openness, we propose a non-directional hypothesis:

H₄

Individuals who would agree to divulge their SNS password will have a significantly different mean level of Conscientiousness than those who would refuse.

Based on the definition of Extraversion (McCrae & Costa, 2010), and the fact that Exhibition can be an important component of Extraversion (Jackson, Paunonen, Fraboni, & Goffin, 1996), extraverted individuals should tend to display their social, outgoing nature via their SNS content. Therefore:

H₅

Individuals who would agree to divulge their SNS password will have a significantly higher mean level of Extraversion than those who would refuse.

Neurotic individuals tend to experience more negative affect, including fear and anxiety (McCrae & Costa, 2010). Therefore, they may have heightened concerns about employers accessing their SNS content. Accordingly:

H₆

Individuals who would agree to divulge their SNS password will have a significantly lower mean level of Neuroticism than those who would refuse.

The Dark Triad. The Dark Triad is a commonly-studied set of personality traits that falls outside the FFM, but is linked to job performance and counterproductive work behaviors (O’Boyle, Forsyth, Banks, & McDaniel, 2012). Machiavellianism describes individuals who are manipulative and cold (Paulhus & Williams, 2002). Psychopathy describes individuals who lack empathy for others, engage in antisocial behaviors, engage in risk-taking, and are callous (Williams, Paulhus, & Hare, 2007). Finally, Narcissism describes individuals who are egotistic, self-enhancing, and who believe they are entitled and superior to others (Paulhus & Williams, 2002). We could find no compelling conceptual or empirical linkages between Machiavellianism and the likelihood of complying with the SNS password request, so we focused our attention on Psychopathy and Narcissism.

Psychopathy has been linked to rebelling against authority and engaging in illegal activities (O’Boyle et al., 2012). As a result, those who score high on Psychopathy may tend to have evidence of illegal activities and rebellion against authority within their SNS accounts. Also, refusing the SNS password request may be a form of rebellion against authority. Consequently,

H₇

Individuals who would agree to divulge their SNS password will have a significantly lower mean level of Psychopathy than those who would refuse.

The more narcissistic people are, the more willing they are to have others find out all about them because of their strong need for self-affirmation (Morf & Rhodewalt, 2001). Therefore, we predict:

H₈

Individuals who would agree to divulge their SNS password will have a significantly higher mean level of Narcissism than those who would refuse.

Impression Management. Those who score high on Impression Management try to actively manage or distort the perception of their image (Paulhus, 1991). Such individuals should be less concerned about employers accessing their SNS because they should be more likely to censor their SNS content in the first place as part of their self-presentation strategy. Thus:

H₉

Individuals who would agree to divulge their SNS password will have a significantly higher mean level of Impression Management than those who would refuse.

In order to investigate these issues, we decided to present a hypothetical question about a SNS password request by a hiring manager to participants who had work experience but were not actually applying for a job. We did not use actual applicants because if they were asked whether they were no longer in contention for a prospective job as a result of a SNS password request, we presumed they would not readily admit this. Such an admission would imply that they had something to hide, and this runs counter to the pervasive tendency to respond in a socially desirable manner (Jackson, 1999) Consequently, they might be more inclined to identify other factors (e.g., poor fit with the organization, other job prospects, relocation, etc.) as the reasons for them no longer being in contention for the job, in much the same way that employees who are fired often do not report “firing” as the cause of their job loss (Giacalone & Duhon, 1991). Consequently, using actual applicants as participants could result in a drastic underestimate of the true rate of refusing the SNS password request. Also, we avoided asking participants whether they would have complied with the request for their SNS password if it had been part of the selection process for their present job, because we believed that cognitive dissonance (Wicklund & Brehm, 2013) would push them towards rationalizing their current job tenure by saying that they would have complied with the SNS password request, whether that were true or not. Therefore, we felt that asking a hypothetical question about compliance with the SNS password request to individuals who have work experience, and who will likely be in the job market in the near future, would provide an effective approach to testing our research question and hypotheses.