Job interviews are part of almost every selection procedure (Levashina et al., 2014). In interviews, recruiters often encourage interviewees to be authentic: “The surprising secret to interview success — be yourself, goes the typical advice” (Goldberg, 2022, in the New York Times). Initial research findings support this piece of advice, suggesting that interviewees who are perceived as being authentic are more likely to receive a job offer (Moore et al., 2017). Thus, interviewees have good reason to strive for creating an authentic impression.

However, the missing piece of information is how an interviewee should behave in order to appear authentic. This refers to what behavioral cues elicit perceptions of authenticity in interviewers. Authenticity is defined as “alignment between a person’s internal sense of self and outward behavior” (Cha et al., 2019, p. 634), but interviewers typically have to judge interviewee authenticity without having direct access to interviewees’ internal sense of self such as their cognitions and emotions. Interviewers need to make their judgments based on observable behavior. On the one side, for interviewees, it is crucial to understand what behavior will help them to create an authentic impression so that they might convince potential employers of their qualities. On the other side, for interviewers, it is important to understand what behavior might cause them to perceive one interviewee as more authentic than another so that they can make an informed decision as to whether they want to take perceived authenticity into account when evaluating interviewees.

This study aims to establish the concept of authenticity cues in interview research. We examine the relationships of authenticity cues with (a) raters’ perceptions of interviewee authenticity, (b) interviewer ratings of interview performance, and (c) supervisor ratings of job performance. In line with prior conceptualizations of authenticity (e.g., Cha et al., 2019; Chen, 2019; Harter, 2002), we define authenticity cues as verbal, paraverbal, and nonverbal behavior that can be observed in oral communication and that seemingly conveys information to interviewers about the interviewees’ inner self, that is, what the interviewee might think, feel, and behave like. In other words, authenticity cues are pieces of information that interviewers may detect and use to infer whether a person is authentic. For example, authenticity cues could include behaviors such as talking about personal topics and emotions, describing things in an expressive way, or having vivid facial expressions and speaking with your hands.

This research contributes to the literature in several ways. First, drawing from person perception theories such as the realistic accuracy model (Funder, 1995, 2012), we offer a conceptualization of authenticity cues and generate knowledge on how to operationalize this new construct. We develop a measure of authenticity cues and examine to what extent exhibiting authenticity cues results in interviewees being perceived as more or less authentic. Second, we examine the mechanism by which authenticity cues exert influence on interviewers’ ratings, such as through perceived authenticity. Along these lines, we also consider first impressions that interviewees evoke in observers (based on perceptions of warmth and competence; Fiske et al., 2007), as well as interviewees’ verbal cognitive ability, to examine whether authenticity cues have an influence beyond these factors. Third, if authenticity cues influence interview performance, it is important to understand whether this might constitute a source of error and thus a threat to interview validity (i.e., authenticity cues influencing interview performance even though they have nothing to do with job performance). To address this, we explore whether interviewees who demonstrate authenticity cues in an interview setting are more likely to perform well on the job.

Perspectives on Authenticity

Authenticity broadly means that a person’s sense of self—what is internal or private such as values, thoughts, and feelings—is aligned with their outward behavior, what is external or public such as words, gestures, and facial expressions (Cha et al., 2019). Yet, authenticity can be discussed from a variety of perspectives. The literature distinguishes between experienced authenticity (i.e., being authentic) and externally perceived authenticity (i.e., being perceived as authentic; Cha et al., 2019). Whereas experienced authenticity is thought to facilitate a positive relationship with oneself, externally perceived authenticity is considered to facilitate a positive relationship with others. In other words, experienced authenticity will likely have positive effects on internal outcomes such as subjective well-being, whereas externally perceived authenticity I thought to influence external outcomes such as performance, image, and career success (Cha et al., 2019).

Some scholars have argued that externally perceived (other-rated) authenticity is a particularly relevant perspective (Fields, 2007). This is because externally perceived authenticity influences how others will react to a person. In this regard, it does not matter whether a person deliberately or accidentally behaves authentically: A person may not deliberately behave authentically but may still create an authentic effect. In line with this, in this paper, we focus on observable behavior—authenticity cues that are observed independent of whether the person feels or intentionally acts authentic or not.

We introduce the concept of authenticity cues to be able to fully focus on the effects of observable interviewee behavior. Authenticity cues are different from actual authenticity in that they capture only behavior that seems to reveal bits and pieces about who the person is. As such, authenticity cues do not capture whether the observed behaviors are in alignment with the individual’s true self because the individual’s true self is not known to the observer. This is inherent for the observer perspective, and this perspective determines the impact that ones’ authenticity cues have on others, which is the focus of the present research.

Authenticity in Job Interviews

Job interviews are a context that is ideal for studying social interactions in general and authenticity in particular for at least three reasons. First, at its core, the interview is a social interaction in which two parties—an interviewee on the one side and at least one interviewer on the other side—meet in a short-term interaction, usually for the first time, to exchange information and start building rapport, which might be the potential starting point of a long-lasting work relationship. The interviewee’s goal is to gain a positive assessment and, ultimately, an offer for a job, and the interviewer’s goal is to gain the information that is needed to make an informed decision about the interviewee (Dipboye et al., 2012). In this exchange of information between strangers, both need to decide whether they trust the information they receive (Dipboye et al., 2012). Hence, in this type of interaction, individuals should be especially sensitive to cues signaling authenticity. This is particularly relevant because forms of inauthenticity such as faking are prevalent in job interviews (Levashina & Campion, 2006). Second, the job interview is widely used. It is a situation that practically everyone goes through at some point in their life and thus a social setting of not only high theoretical but also practical relevance. Third, the interview is a setting in which evaluation takes place such that applicants usually only get a job offer if they are evaluated positively by interviewers. Therefore, it matters how applicants are perceived and how their behavior is interpreted in this setting (Levashina et al., 2014).

There is surprisingly limited and only fragmented evidence on authenticity-related behaviors in the context of job interviews. Research on constructs that describe the intentions underlying authentic behavior has produced some positive and some mixed results. For example, previous research has shown that honest impression management—interviewees sharing true information about themselves with the intention to increase their chances of landing a job offer—can positively influence interviewer ratings (Bourdage et al., 2018). Another example is research on self-verifying behavior which refers to interviewees’ sharing of unbiased information about themselves that is in line with their self-views (Wilhelmy et al., 2020). Here, findings are mixed: Some research has found non-significant relationships between self-verifying behavior and interview performance (Charbonneau et al., 2021), while other research has found a positive relationship between self-verifying behavior and interview success (Wilhelmy et al., 2020). Honest impression management and self-verifying behavior differ from authenticity cues because they involve applicants’ personal intentions (i.e., the intention to create favorable impressions or the intention to stay true to oneself).

An authenticity-related behavior, which has been studied before, is revealing negative information about the self. This occurs when interviewees reveal unadorned negative information about themselves, for example, about their weaknesses or shortcomings (Wilhelmy et al., 2020). Only one study has examined effects of interviewees’ revealing negative information about themselves and found a positive association with interview success (Wilhelmy et al., 2020). A related kind of behavior is self-disclosure, which refers to verbally communicating personal information to the interviewer that is not known or not available from other sources. This construct has only been studied as a dependent variable in the interview context, so we do not know how this behavior affects interview performance (Omarzu, 2000; Wilhelmy et al., 2022). Revealing negative information and self-disclosure differ from the conceptualization of authenticity cues in the present study because these constructs describe relatively narrow and verbal behaviors only.

Scarce research has examined others’ perceptions of authenticity in the job interview as a result of interviewees’ authenticity-related behavior. Thus far, only one study has examined the influence of perceived inauthenticity on ratings of job offer likelihood and found a positive relationship (Moore et al., 2017). Perceived inauthenticity refers to the extent to which others (interviewers or observers) think that an interviewee did not reveal what they really think, feel, or behave like. Perceived (in)authenticity differs from authenticity cues as defined in the present study because it is an outcome or consequence of interviewees’ behavior during the interview (Leroy & Mor, 2015), whereas authenticity cues describe interviewee behavior during the interview.

To sum up, the limited past research that exists has studied various concepts that relate to interviewees’ authenticity but has not clearly differentiated behavior or behavioral cues—verbal, nonverbal, and paraverbal—from underlying intentions and potential consequences of the behavior and has not spoken to the mechanisms behind potential effects of authenticity cues on interview and job performance. In the present study, we bring together this thus far fragmented research by including narrower aspects of authenticity-related behavior into a broader behavioral and observational perspective.

Authenticity Cues in Job Interviews

Person perception theories such as the realistic accuracy model (Funder, 1995, 2012) offer insights into how and when interviewees are perceived as being authentic by interviewers. The realistic accuracy model describes how personality judgment happens. It proposes that personality judgment relies on the availability of relevant behavioral cues that are detected by an observer who utilizes this information to infer a judgment (Funder, 1995). Such cues can be verbal, paraverbal, or nonverbal (see, e.g., Hickman et al., 2022).

We put forward that the logic of the realistic accuracy model can be adopted to understand authenticity in job interviews. Within this logic, authenticity can be seen as a personality trait (see, e.g., Hopwood et al., 2021; Wood et al., 2008), and interviewers are observers who judge interviewees regarding this trait. To make these judgments, there must be (verbal/paraverbal/nonverbal) cues present in interviewees’ behavior that (a) are relevant to authenticity and (b) can be observed and utilized by interviewers. In other words: What interviewees say, the way they say it, and the way they engage their face and hands when speaking can be perceived as cues about who they are, for example, what they are feeling and thinking. When interviewers detect such cues, they can utilize them to infer that the interviewee is authentic; that is, having a tendency to act in line with their inner self.

Operationalizing Authenticity Cues

To operationalize authenticity cues, we need to identify a set of behaviors that are relevant to authenticity and that are observable in the interview setting. These behaviors can be either verbal, paraverbal, or nonverbal. Consulting the job interview and the authenticity literature allows to obtain such a set of behaviors. There exists a variety of measures that assess authenticity-related constructs in different contexts. This includes measures of (a) honest impression management and self-verification in job interviews, (b) authentic leadership, (c) authentic behavior in customer service, and (d) criteria for assessing the credibility of eyewitness testimony in criminal investigations (Bourdage et al., 2018; Cable & Kay, 2012; Neider & Schriesheim, 2011; Steller & Koehnken, 1989; Yagil & Medler-Liraz, 2013). We collected existing descriptions of authenticity-related behaviors from these literatures (i.e., from Bourdage et al., 2018; Cable & Kay, 2012; Kernis & Goldman, 2006; Kristof-Brown et al., 2002; Monti et al., 1984; Moore et al., 2017; Neider & Schriesheim, 2011; Steller & Koehnken, 1989; Wilhelmy et al., 2020; Wilhelmy et al., 2022; Yagil & Medler-Liraz, 2013). This broad collection of behaviors can be found in the online supplements accompanying this manuscript (see Online Supplement 1 at https://osf.io/x94y8/). From this collection, we eliminated all content that referred to either antecedents of authenticity cues instead of the behavior itself (i.e., intentions and motivations underlying behaviors) or consequences of authenticity cues (i.e., the interpretation or impression that behaviors evoke). We also excluded content that referred to behaviors that cannot be observed in the job interview setting. Based on the remaining content, we put together an initial set of authenticity cues (see Table 1). Examples for verbal authenticity cues are “This person discusses personal/private topics,” “This person clearly states their personal opinion,” and “This person answers questions vaguely and superficially” (reverse coded). Paraverbal authenticity cues include behaviors such as “This person describes things in an expressive way.” Nonverbal authenticity cues refer to behaviors such as “This person gestures/speaks with their hands.”

Table 1 Item analyses of authenticity cues
Full size table

Underlying this operationalization is the assumption that authenticity cues are behaviors that contribute to others perceiving an interviewee as being authentic. Adopting the logic of the realistic accuracy model (Funder, 1995, 2012), authenticity cues are observable behaviors, whereas perceived authenticity refers to others’ interpretation of these observed behaviors. For example, when an interviewee clearly states their opinion or describes things in an expressive way, interviewers can notice this behavior and perceive it as cues of authenticity resulting in the interviewer perceiving the interviewee as authentic. To assess whether our efforts to operationalize authenticity cues have been adequate, we start by examining whether exhibiting authenticity cues indeed results in higher levels of perceived authenticity:

Hypothesis 1: Interviewees’ authenticity cues will predict perceived authenticity in the interview as rated by teams of independent raters.

When evaluating interviewees, interviewers generally tend to seek confirmation of their first impressions of interviewees (Dougherty et al., 1994). Past research has shown that the first impressions that interviewers gain of interviewees during the opening minutes of the interview relate to how interviewers evaluate interviewees’ responses to interview questions (Barrick et al., 2010; Swider et al., 2011; Swider et al., 2016). This corresponds to assumptions from person perception theories (Evans, 2008; Fiske et al., 2007) that observers are likely to make rapid, implicit, and automatic judgments of a person that may affect their later analytical, explicit, and rational evaluations of this person.

To understand the role that authenticity cues play in the interview, a relevant question to study is whether the impact of interviewees’ authenticity cues on observers’ perceptions of interviewee authenticity goes beyond observers’ first impressions of interviewees. We expect that authenticity cues are the main driver of perceived authenticity. Therefore, authenticity cues should explain variance in perceived authenticity beyond typical artifacts that affect person perception in the interview such as first impressions. Hence, we expect that interviewees who display authenticity cues to a larger extent will be perceived as being more authentic during the interview independent of the first impressions that interviewees evoke in observers:

Hypothesis 2: Interviewees’ authenticity cues will predict perceived authenticity in the interview over and above observers’ first impressions of interviewees (i.e., first impressions of warmth and competence) as rated by teams of independent raters.

Authenticity Cues and Interview Performance

Engaging in behaviors that signal authenticity (i.e., authenticity cues) may help interviewees differentiate themselves from their competitors and positively stand out (Moore et al., 2017; Wilhelmy et al., 2020). This is because interviewers are constantly looking for relevant information that they can use to judge who an interviewee is as a person and as a future employee. In line with this, research has shown that interviewers tend to be attracted to applicants whom they perceive as authentic (Bangerter et al., 2012). Put another way, appearing inauthentic is likely to lead to negative assessments about the interviewees’ quality because the information gained might not seem reliable and because inauthenticity might seem inappropriate for the job (Moore et al., 2017). In contrast, displaying authenticity cues should be well received by interviewers and lead to a more positive performance assessment. When interviewers believe in what interviewees say, then interviewers will take the information they receive seriously and utilize all positive information that interviewees provide as a basis for their ratings, which should result in better ratings of interview performance. For example, authentic self-expression makes interviewers perceive their interactions with the interviewee as more authentic, which, in turn, allows them to be more confident in their assessment of them (Cable & Kay, 2012). Furthermore, authenticity cues might reassure interviewers that interviewees are confident in their abilities and can afford presenting themselves authentically.

There is initial empirical evidence supporting these considerations. Moore et al. (2017) examined interviewees’ overall linguistic style and use of words as a potential indicator of willingness to share self-knowledge in the interview. They found that interviewees’ language use was associated with both a rater’s perception of interviewee inauthenticity and the same rater’s reported job offer likelihood. Based on these arguments and findings, we suggest that interviewers will evaluate interviewees’ performance in the interview as more positively when interviewees display authenticity cues to a larger extent:

Hypothesis 3a: Interviewees’ authenticity cues will predict interview performance as rated by teams of trained interviewers.

In addition, we propose that interviewees’ authenticity cues will influence interviewers’ performance ratings via their perceptions of interviewees’ authenticity. When interviewers detect and utilize authenticity cues and subsequently perceive an interviewee as being authentic, interviewers might take interviewees’ statements in the interview more seriously and believe that the information they receive will be indicative of interviewees’ subsequent behavior in the workplace. We thus predict that interviewees’ authenticity cues will be associated with perceiving the interviewee as more authentic, ultimately leading to higher ratings of interview performance:

Hypothesis 3b: Perceived authenticity will mediate the relationship between interviewees’ authenticity cues and their interview performance.

An interesting question is whether the expected relationship between authenticity cues and interview performance might be explained by interviewees’ ability to articulate themselves (i.e., their verbal cognitive ability). Verbal cognitive ability refers to being able to understand what words mean in order to use them effectively when communicating (Salgado et al., 2003). Research has shown that individuals with particularly high levels of cognitive ability tend to be more successful in job interviews (Berry et al., 2007; Roth & Huffcutt, 2013). Extending this knowledge, we propose that authenticity cues go beyond verbal cognitive ability and that interviewees can display authenticity cues (e.g., talking about their emotions, showing vivid facial expressions) without necessarily needing high levels of verbal cognitive ability. We understand displaying authenticity cues to be a form of communication skill that could be facilitated by but does not depend on verbal cognitive ability. Thus, we expect that interviewees’ authenticity cues explain variance in interview performance beyond interviewees’ verbal cognitive ability:

Hypothesis 4: Interviewees’ authenticity cues will predict interview performance over and above interviewees’ verbal cognitive ability.

Authenticity Cues and Job Performance

We expect authenticity cues to relate to job performance. This relationship could go in either direction. On the one side, it is plausible that individuals who perform well at their job might feel safe enough to behave authentically in the interview setting. For example, an interviewee who feels confident in their work-related knowledge and skills could be more likely to display verbal authenticity cues such as clearly stating their own opinion. Similarly, an interviewee who knows that they typically perform well and are highly employable might feel more comfortable in an interview setting and therefore exhibit more paraverbal and nonverbal authenticity cues such as vivid facial expression and gesturing.

On the other side, it is possible that some people have a general tendency to display authenticity cues in social interactions and that this helps them to be perceived favorably in the interview and on the job. This speaks to assumptions from the literature that being authentic is a relatively stable person characteristic that can have positive effects in different contexts and situations in life (Cable & Kay, 2012; Moore et al., 2017). For example, research on self-verification has put forward the notion that individuals who like to present themselves authentically are more likely to end up in jobs that actually fit their personal needs and abilities which, in turn, should facilitate their performance in this well-fitting job (Cable & Kay, 2012; Moore et al., 2017). Adding to this perspective, we propose that displaying authenticity cues could be understood as a type of communication skill that helps individuals to build relationships and to interact more effectively because others perceive these individuals as authentic and trust them. This skill might come in handy in many situations in life—be it when building a relationship in a short interaction such as a job interview or building a collaborative long-term relationship with work colleagues and supervisors. Based on these different considerations, we expect that the degree to which interviewees’ show authenticity cues during the interview will be related to higher job performance:

Hypothesis 5: Interviewees’ authenticity cues will positively relate to job performance as rated by their supervisors.

Methods

Sample

We invited working adults who were interested in preparing for their next career step to participate in a job application training. A mock interview was an integral part of this training. Mok interviews were videotaped. When registering for the job application training, interviewees agreed to dress and behave as if they were in an actual selection process (similar to prior interview studies such as Barrick et al., 2012; Kleinmann & Klehe, 2011; Swider et al., 2016; Van Iddekinge et al., 2005). To participate in this study, interviewees had to be employed, and they had to ask their supervisor to agree to provide ratings of interviewees’ job performance for study purposes. As an incentive to participate in this study, interviewees received feedback on their performance in the mock interview.

The original sampleFootnote 1 consisted of 223 interviewees (i.e., working adults from different types of jobs). For the present study, we only included interviewees for whom video recordings of the mock interview were available (i.e., recordings with no missing parts and with good video and audio quality). This led to a final sample of 181 interviewees (41% female). Their mean age was 30.52 (SD = 7.49), and most of them (74%) held an academic degree. More than half of the interviewees (62%) had completed three or more job interviews in their life. About half of the interviewees (52%) had been working in their current job for 1 year or longer.

Procedure

Upon registration, interviewees provided their informed consent to participate in this study. During the job application training, interviewees completed a comprehensive simulated selection procedure, including a one-hour structured mock interview, a cognitive ability test, and several other selection devices. Interviews were administered face-to-face by a panel of interviewers. Interviewers were drawn from a pool of 78 advanced students majoring in psychology who were on average 28.96 (SD = 7.89) years old. They had previously completed a face-to-face frame-of-reference interview training (Roch et al., 2012; Woehr & Huffcutt, 1994) which took a full day.

Upon registration, interviewees provided the contact details of their direct supervisors and allowed us to collect supervisor ratings of job performance. We obtained job performance ratings for each of the 181 interviewees. In total, 179 supervisors completed the survey; of these, 177 supervisors rated one interviewee, and two supervisors rated two interviewees. Supervisors’ mean age was 44.28 years (SD = 9.84). Most supervisors reported that they had been working with the interviewee they evaluated for more than half a year (86.0%) and that they interacted with the respective interviewee on a daily or weekly basis (86.6%). Supervisor ratings were kept confidential and not shared with interviewees. Vice versa, information on interviewees’ performance in the mock interview was not made available to supervisors.

We used the videotaped mock interviews to collect data specifically for the present study. We extensively trained eight observers (i.e., raters) who watched the videotaped interviews and provided individual ratings of what they observed and perceived during the interviews. Raters were on average 28.94 years old (SD = 7.12), and all of them were graduate students pursuing a Master’s degree in psychology. Raters evaluated three different types of variable categories. Variables in Category 1 were assessments of authenticity cues; variables in Category 2 referred to ratings of perceived authenticity; and variables in Category 3 were raters’ first impressions of interviewees. To avoid any confounds between the ratings of variables in these three categories, each rater assessed each interview only once. Thus, ratings of the three variable categories were independent so that each rater would evaluate each interviewee only once. Table 7 in the Appendix section provides an overview of all ratings obtained from interviewers, video raters, and supervisors.

Measures

Interview Performance

The simulated job interview consisted of 30 interview questions, including 15 past-behavior interview questions (comparable with patterned behavior description interview questions as introduced by Janz, 1982) and 15 future-behavior interview questions (comparable with situational interview questions introduced by Latham et al., 1980). Each interviewee was asked the same interview questions. To prevent order effects, the presentation of interview questions was randomized so that half of the sample completed the past-behavior interview questions first and the other half of the sample completed the future-behavior interview questions first. Interview questions were designed to measure manifestations of interviewees’ personality that are relevant to the work context. We decided for this approach because personality traits are frequently assessed constructs in job interviews (Huffcutt, 2011). All the interview questions related to general work-related situations could occur in almost any type of job. An example interview question is, “Sometimes we notice that someone else makes a mistake. Think of a situation when a co-worker made a mistake and you pointed the mistake out to them. Please describe exactly how you perceived this situation and what you did in this situation when you spoke to your co-worker.”

A pair of two trained interviewers administered the face-to-face mock interview. Interviewers were provided with an interview guide with the 30 questions and behaviorally anchored rating scales to evaluate each interview question. Each interviewer rated interviewees’ response to each interview question individually and independently form the other interviewer. After all interview questions had been completed, interviewers were allowed to discuss their ratings with each other and to adjust their individual ratings if their ratings were discrepant by two points or more on a 5-point scale (as commonly done in interview research; Ingold et al., 2015). We calculated a one-way random effects ICC for every interview question to assess interviewers’ interrater reliability. Across all interview questions, the mean interrater reliability for the interviewer panel was ICC(1,2) = .80, which is comparable to previous interview studies (Thorsteinson, 2018). We averaged ratings across interviewers and interview questions to calculate an overall score for interview performance. The internal consistency for this measure was Cronbach’s alpha α = .85.

Authenticity Cues

Based on an extensive literature review, we first developed an initial set of 15 items describing either verbal, paraverbal, or nonverbal authenticity cues that can be observed in a job interview (for the literature review see Online Supplement 1 at https://osf.io/x94y8/). We pretested this initial set of items. For the pretest, we used video recordings of 15 interviewees that were not included in the main study because supervisory performance ratings had not been available for these interviewees. Two video raters watched the 15 video recordings independently and rated interviewees’ authenticity cues using the newly developed items. Next, we analyzed interrater reliabilities and item statistics. The mean interrater reliability was ICC(2,2) = .78. We decided to exclude two of the initial items, because the raters of the pre-study stated that they could hardly observe item-relevant behavior based on the video material due to low frequency. In line with this, the means for these two items were relatively low (M = 1.54 and 1.58 on a 5-point Likert scale). The excluded items were “This person corrects themselves during the interview” (verbal authenticity cue, reverse coded) and “This person says that they have difficulty to remember details” (verbal authenticity cue, reverse coded). Based on this decision, our operationalization of authenticity comprised 13 items. These items are presented in Table 1.

Another result of the pretest was that video raters reported that it was difficult for them to rate authenticity cues after watching the response to only one interview question. They suggested basing their ratings on more video material (i.e., watching responses to several interview questions in a row) in order to observe more of the relevant behavior and be able to evaluate authenticity cues. On the basis of this feedback, we decided to have raters watch three interview questions in a row before completing the measure of authenticity cues.

Next, we developed a rater manual to allow for a clear understanding of the generated items. For each item, the rater manual provided behavioral examples. For example, for the item “This person discusses personal/private topics,” valid behavioral examples were “(S)he Mentions topics outside of the job context”; “My mom always says... ”; and “I am approaching situations like this similar to my marriage/when I am cooking/doing sports, etc.”. The rater manual is presented in the online supplements (see Online Supplement 2 at https://osf.io/x94y8/). Finally, in the main study, an independent group of video raters assessed authenticity cues using the 13 pre-tested items that are presented in Table 1. Items were to be rated on a 5-point scale ranging from 1 = disagree to 5 = agree. Raters were provided with the rater manual containing written behavioral examples for each item. In addition, we provided raters with video sequences as behavioral examples for those items that had been identified as difficult to rate, as indicated by lower interrater reliabilities in the pre-study (i.e., ICCs below .70).

Raters assessed the authenticity cues for each interviewee four times: Once for the first three interview questions and once for the last three interview questions of both the future-behavior and past-behavior interview questions. This implies that video recordings of responses to twelve (out of the original 30) interview questions were rated. We chose not to rate the video recording of the entire interview because most interviews are shorter than the interview in the present study, as indicated by meta-analytic results (Thorsteinson, 2018). We decided to rate future-behavior and past-behavior interview questions from the beginning and the end of the interview to allow for observations of a range of behavioral examples from each part of the interview.Footnote 2

Each interviewee was evaluated by a randomly assigned pair of raters. Each rater watched their assigned video recordings on their own and provided their individual ratings. Afterwards, video raters would be allowed to discuss and adjust their ratings if discrepancies between a pair of raters were larger than two points (on a 5-point scale).

Across all items, the mean interrater reliability for the measure of authenticity cues was ICC(1,2) = .86. Regarding subscales, the mean interrater reliabilities were ICC(1,2) = .86 for verbal authenticity cues, ICC(1,2) = .84 for paraverbal authenticity cues, and ICC(1,2) = .88 for nonverbal authenticity cues. This is comparable to other studies in which raters assessed behaviors in the job interview (e.g., Feiler & Powell, 2016). To obtain final ratings of authenticity cues, we averaged ratings across raters and items. Internal consistencies were α = .89 for overall authenticity cues, α = .76 for verbal authenticity cues, α = .69 for paraverbal authenticity cues, and α = .82 for nonverbal authenticity cues.

Perceived Authenticity

Independent raters indicate their perception of interviewees’ authenticity. Each interviewee was evaluated by a randomly assigned pair of raters. Raters watched the same sets of videos and followed the same approach as the one that was used to measure authenticity cues (see description above). Raters rated their perceptions of interviewees’ authenticity using an eight-item measure developed by Leroy and Mor (2015). This measure has successfully been used by Moore et al. (2017) to assess perceived inauthenticity in job interviews. Example items are “This person seems fake,” “This person proclaims A, but seems to really think B,” “This person does not seem to reveal what he or she really thinks,” and “This person seems not genuine.” The full set of items can be found in Cha et al. (2019). Items were rated on a 5-point scale ranging from 1 = disagree to 5 = agree. All items were originally developed to assess perceived inauthenticity and therefore reverse coded to measure perceived authenticity in the present study. Across all items, interrater reliability was ICC(1,2) = .82 and thus comparable with the interrater reliabilities for authenticity cues (see above). The internal consistency was Cronbach’s α = .96, which is similar to the reliability of this scale reported in Moore et al. (2017). To obtain a final score of perceived authenticity, we averaged ratings across raters and items.

First Impressions of Warmth and Competence

Independent raters provided evaluations of their first impressions of interviewees’ warmth and competence. To this end, we extracted the first two minutes of each video-recorded interview (similar to Ingold et al., 2018). We asked two raters to watch the short video sequences and to provide individual ratings of their first impressions. Consistent with prior research on first impressions (Barrick et al., 2010; Ingold et al., 2018; Swider et al., 2016), raters were not specifically trained, and they were not allowed to discuss their ratings with each other. Following the line of research on person perception (see Fiske et al., 2007, for a summary), we constructed adjective lists to measure first impressions of interviewees’ warmth and competence. We used six items each to assesses impressions of warmth (“warm,” “friendly,” “good-natured,” “empathetic,” “caring,” and “sincere”) and competence (“competent,” “capable,” “efficient,” “intelligent,” “clever,” and “proficient”) based on prior studies (Abele et al., 2016; Rosenberg et al., 1968). For each adjective, raters indicated how strongly they agreed that the given adjective was suited to describe a given interviewee on a 7-point scale ranging from 1 = strongly disagree to 7 = strongly agree. All interviewees were evaluated by the same two raters. Across all items, interrater reliability was ICC(2,2) = .61 for warmth and ICC(2,2) = .60 for competence, which is similar to the ICCs reported in previous studies measuring first impressions (e.g., Ingold et al., 2018). Internal consistencies for the six-item scales were Cronbach’s alpha α = .94 for warmth and α = .92 for competence. We averaged ratings across raters and items to obtain final scores.

Verbal Cognitive Ability

To assess verbal cognitive ability, interviewees completed the verbal reasoning module of a larger cognitive ability test (IST 2000 R; Amthauer et al., 1999). Meta-analytic findings support the criterion-related validity of this specific cognitive ability test (Hülsheger et al., 2007). As part of the verbal reasoning module, interviewees had to (1) complete sentences, (2) understand analogies, and (3) find similarities in wordings. Each of these three verbal abilities was tested with 20 items rated on a dichotomous scale from 0 = wrong answer to 1 = correct answer. For the present sample, the internal consistency for this test was α = .82. Following the instructions of the test manual, we computed sum scores across all items and then transformed the sum scores to IQ values.

Job Performance

We asked interviewees to provide the contact details of their current supervisors so that we could send a link to an online survey to supervisors to gather information about interviewees’ job performance. Job performance was measured with nine items that have been used in previous studies and demonstrated high levels of reliability (Jansen et al., 2013). The nine items are based on scales of in-role performance from Williams and Anderson (1991) and task-based performance from Bott et al. (2003). An example item is “S/he achieves the objectives of the job.” Supervisors rated each item on a 7-point scale ranging from 1 = not at all to 7 = absolutely. Internal consistency for the present sample was Cronbach’s alpha α = .92.

Control Variables

In our analyses, we took age, sex, and extraversion into account as control variables when predicting perceived authenticity in the interview, interview performance, and job performance. Best-practice recommendations on the use of control variables in correlational research suggests that controls should only be included when expecting meaningful relationships between controls and main study variables (Becker et al., 2016). In the present study, we considered age, sex, and extraversion as potential confounds because ample evidence suggests that all three variables can affect interview evaluations and/or job performance ratings (Bourdage et al., 2020; De Groot & Gooty, 2009; Hora et al., 2021; Levashina et al., 2014; Morgeson et al., 2008; Ng & Feldman, 2008; Roth et al., 2012; Salgado & Moscoso, 2002). We could not control for race, given that there was no variation in race in this sample because the study was conducted in central Europe (i.e., all interviewees were of European descent).

Extraversion was measured with ten items from the International Personality Item Pool (IPIP; Goldberg, 1990, 1992). An example item is “I feel comfortable around people.” Each item was rated on a scale from 1 = very inaccurate to 5 = very accurate. The reliability of this scale has been reported to be α = .89 (Heimann et al., 2022).

Results

Measurement Model of Authenticity Cues

Before testing our hypotheses, we first investigated the measurement model of the newly developed scales that were designed to assess (1) verbal, (2) paraverbal, and (3) nonverbal authenticity cues. Using the lavaan package for the R environment (Rosseel, 2012), we conducted a set of confirmatory factor analyses testing six measurement models (CFAs). Table 2 shows the fit indices for all six measurement models.

Table 2 Fit statistics for confirmatory factor analyses of authenticity cues
Full size table

Measurement Models 1, 2, and 3 included the initial set of 13 items. Model 1 represented a three-factor model specifying verbal, paraverbal, and nonverbal authenticity cues as separate factors. Model 2 was a two-factor model distinguishing only between verbal authenticity cues and a compound of para/nonverbal authenticity cues. Model 3 was a one-factor model not distinguishing between verbal, paraverbal, and nonverbal authenticity cues. As can be seen in Table 2, neither of the three models performed on the initial set of 13 items showed acceptable fit when compared to recommendations for model evaluation (Schermelleh-Engel et al., 2003).

Measurement Models 4, 5, and 6 were performed on a reduced item set. Specifically, we removed all items with low factor loadings (i.e., factor loadings below .40) in the previous models. This led to the exclusion of three items (Item 1, Item 7, and Item 13, all shown in Table 1). With this reduced set of ten items, we again calculated a three-factor model (Model 4), a two-factor model (Model 5), and a one-factor model (Model 6) of authenticity cues following the same approach as described above for Models 1 to 3.

Overall, the models we performed on the reduced item set fit the data better than the models with the full item set. Both the three-factor model (Model 4) and the two-factor model (Model 5) showed acceptable fit; that is, they met common criteria for model evaluation (Schermelleh-Engel et al., 2003). For the three-factor model (Model 4), fit indices were χ2(32) = 83.13 (p < .001), χ2/df = 2.60, RMSEA = .09, SRMR = .05, and CFI = .96, TLI = .94. For the two-factor model (Model 5), fit indices were χ2(33) = 83.27 (p < .001), χ2/df = 2.52, RMSEA = .09, SRMR = .05, CFI = .96, and TLI = .95.Footnote 3 There was no significant difference between the fit of the two models, Δχ2(1) = 0.14, p = .709, but both models fit the data significantly better than did the one-factor model (Model 6), with Δχ2(3) = 96.45, p < .001 (Model 4) and Δχ2(2) = 96.31, p < .001 (Model 5).

Despite the acceptable fit of both the two-factor model (Model 5) and the three-factor model (Model 4), we decided to focus on the two-factor model (Model 5) for two reasons. First, the latent correlation between paraverbal and nonverbal authenticity cues in the three-factor model was extremely high (.99), indicating that even trained observers cannot distinguish between paraverbal and nonverbal cues. Second, extant interview research only differentiates between interviewees’ verbal versus nonverbal cues as broader categories (e.g., Barrick et al., 2009; Cuddy et al., 2015). The factor loadings for the final measurement model (two-factor model with reduced item set) are presented in Table 3. In the following, we test all hypotheses using this two-factor model (Model 5). Analogous results for (a) the one-factor model (Model 6) and (b) the three-factor model (Model 4) can be found in the online supplements (see Online Supplement 3https://osf.io/x94y8/).

Table 3 Factor loadings and factor intercorrelations for two-factor model of authenticity cues with reduced item set
Full size table

Authenticity Cues Predicting Perceived Authenticity

Hypothesis 1 predicted that interviewees who exhibit more authenticity cues in the interview would be perceived as more authentic by independent raters (i.e., they would achieve a higher score on perceived authenticity). Table 4 shows intercorrelations for all study variables. In support of Hypothesis 1, both verbal authenticity cues (r = .34, p < .001) and para/nonverbal authenticity cues (r = .34, p < .001) correlated positively and significantly with perceived authenticity. As supplementary analyses (see Online Supplement 4 at https://osf.io/x94y8/), we examined stable person characteristics (i.e., interviewees’ age, sex, extraversion, verbal cognitive ability) as boundary conditions for the relationships of verbal and para/nonverbal authenticity cues with perceived authenticity. We only found significant interactions between verbal authenticity cues and interviewees’ age and sex, suggesting that interviewees who display verbal authenticity cues are more likely to be perceived as authentic in the interview if they are younger and if they are male.

Table 4 Means, standards deviations, and intercorrelations of study variables (N = 181)
Full size table

In addition, we explored the relationships of verbal and para/nonverbal authenticity cues with stable person characteristics (interviewees’ age, sex, extraversion, verbal cognitive ability) to gain a better understanding of the nomological network of authenticity cues. As can be seen in Table 4, age was not related to verbal (r = .03, p = .732) or para/nonverbal (r = -.01, p = .937) authenticity cues. In contrast, sex correlated significantly with both verbal (r = .20, p = .008) and para/nonverbal (r = .19, p = .009) authenticity cues, indicating that female interviewees are generally more likely to send authenticity cues. Extraversion was not significantly correlated with verbal authenticity cues (r = .10, p = .178) but correlated positively and significantly with para/nonverbal authenticity cues (r = .25, p < .001), suggesting that extraversion has little to no influence on what people say in the interview, but rather on how they say it. Vice versa, verbal cognitive ability correlated positively and significantly with verbal authenticity cues (r = .24, p < .001) but was not significantly correlated with para/nonverbal authenticity cues (r = .13, p = .124), indicating that verbal cognitive ability influences what people say in the interview but has little impact on how they say it.

Hypothesis 2 posited that interviewees who exhibit more authenticity cues in the interview would be perceived as more authentic—independent of observers’ (i.e., raters) first impressions of interviewees (i.e., first impressions of warmth and competence). To test this hypothesis, we conducted hierarchical regression and relative weights analyses using the relaimpo package for the R environment (Grömping, 2006). As can be seen in Table 5, authenticity cues explained a significant proportion of variance in perceived authenticity over and above first impressions of interviewees, ΔR = .11, F(2, 175) = 11.12, p < .001, lending support for Hypothesis 2. In the full regression model (see Model 2 in Table 5), verbal authenticity cues were a significant predictor (β = .20, p = .031) and accounted for 40.6% of the variance explained in perceived authenticity, whereas para/nonverbal cues were a non-significant predictor (β = .18, p = .063) and accounted for 37.9% of the explained variance. Hence, relative weights analyses suggest that verbal authenticity cues were slightly more relevant to predicting perceived authenticity as compared to para/nonverbal authenticity cues. These analyses were conducted without control variables (age, sex, extraversion). We reran the analyses with control variables and the pattern of results remained the same (see Online Supplement 3 at https://osf.io/x94y8/).

Table 5 Hierarchical regression and relative weights analyses of authenticity cues predicting perceived authenticity, interview performance, and job performance (N = 178)
Full size table

Authenticity Cues Predicting Interview Performance

Hypothesis 3a predicted that interviewees who exhibit more authenticity cues in the interview would perform better in the interview (i.e., they would achieve a higher score in the interview as rated by a team of trained interviewers). As can be seen in Table 4, both verbal authenticity cues (r = .25, p < .001) and para/nonverbal authenticity cues (r = .34, p < .001) correlated positively and significantly with interviewers’ ratings of interview performance, providing support for Hypothesis 3a. As supplementary analyses (see Online Supplement 4), we examined stable person characteristics (i.e., interviewees’ age, sex, extraversion, verbal cognitive ability) as boundary conditions for the relationships between authenticity cues and interview performance. We did not find any significant interactions between verbal or para/nonverbal authenticity cues and stable person characteristics, suggesting that authenticity cues relate to interview performance independent of these characteristics of the interviewee.

Hypothesis 3b stated that the relationship between interviewees’ authenticity cues and their interview performance would be mediated by their perceived authenticity. We tested this hypothesis with the mediation package for the R environment (Tingley et al., 2014) and computed bootstrapping mediation analyses with 20’000 simulations. Figure 1 shows that indirect effects were small but significant for verbal authenticity cues (B = 0.04, 95% CI = 0.01, 0.07, p = .008) and for para/nonverbal authenticity cues (B = 0.02; 95% CI = 0.00, 0.04, p = .023) predicting interview performance mediated through perceived authenticity in separate analyses. At the same time, direct effects remained significant for verbal authenticity cues (B = 0.10, 95% CI = 0.03, 0.18, p = .009) and for para/nonverbal authenticity cues (B = 0.11, 95% CI = 0.06, 0.16, p < .001) predicting interview performance. Supporting Hypothesis 3b, these results imply that perceived authenticity contributes to but does not fully explain the relationships between verbal and para/nonverbal authenticity cues and interview performance.

Fig. 1
figure 1

Bootstrapping mediation analysis of interview performance regressed on authenticity cues and perceived authenticity (N = 180) CI = confidence interval. *p < .05, **p < .01

Full size image

Hypothesis 4 posited that interviewees who exhibit more authenticity cues would perform better in the interview—independent of their verbal cognitive ability. As can be seen in Table 5, the results of hierarchical regression and relative weights analyses show that authenticity cues explained a significant proportion of variance in interview performance over and above interviewees’ verbal cognitive ability, ΔR = .11, F(2, 175) = 11.48, p < .001, supporting Hypothesis 4. In the full regression model (see Model 2 in Table 5), verbal authenticity cues were a non-significant predictor (β = .01, p = .911) and accounted for 20.4% of the variance explained in interview performance, whereas para/nonverbal cues were a significant predictor (β = .33, p < .001) and accounted for 61.4% of the explained variance. Thus, relative weights indicate that para/nonverbal authenticity cues have a greater impact on interview performance than verbal authenticity cues. Further analyses taking control variables into account (age, sex, extraversion) produced comparable results (see Online Supplement 3).

Authenticity Cues and Job Performance

Hypothesis 5 predicted that interviewees who exhibit more authenticity cues in the interview would perform better in their jobs (i.e., a correlation between authenticity cues and supervisory ratings of job performance). As can be seen in Table 4, only verbal authenticity cues correlated positively and significantly with supervisor ratings of job performance (r = .24, p = .001), but para/nonverbal authenticity cues did not (r = .09, p = .237). Hence, Hypothesis 5 found support only for verbal authenticity cues.

Supplementary Analyses

To further explore the link between authenticity cues and job performance, we investigated whether authenticity cues in the interview explain variance in job performance over and above interview performance. Results indicated that authenticity cues explained a significant proportion of variance in supervisor ratings of job performance that was not explained by interview performance (ΔR = .06, F(2, 176) = 5.73, and p = .004; see Table 5). Thus, independent of interview performance, there seems to be a connection between displaying authenticity cues and performing well on the job. Relative weights analyses further revealed that verbal authenticity cues were a significant predictor (β = .32, p < .001) and accounted for 55.4% of the variance explained in job performance, whereas para/nonverbal cues predicted job performance negatively and non-significantly (β = -.19, p =.051) and accounted for only 10.3% of the explained variance. Analyses including control variables yielded a similar pattern of results (see Online Supplement 3).

To gain more insight on the relationships between authenticity cues and job performance, we conducted a series of exploratory post-hoc analyses that are presented in Online Supplement 4 (https://osf.io/x94y8/). First, we examined whether it makes a difference whether authenticity cues are measured at the beginning or end of a block of past-behavior or future-behavior interview questions. Results point in the direction that authenticity cues displayed and rated in past-behavior interview questions may contain more information relevant to job performance than authenticity cues displayed and rated in future-behavior questions, but more research is needed to examine the stability of these findings. Second, we examined the relationships of each authenticity cue item with job performance to gain knowledge about the relevance of individual authenticity cues. We found three verbal authenticity cues that drive the relationship with job performance (“This person clearly states their personal opinion,” “This person answers questions vaguely and superficially,” reverse coded,” and “This person talks about their emotions”). Third, we explored stable person characteristics (i.e., interviewees’ age, sex, extraversion, verbal cognitive ability) as boundary conditions for the relationships between authenticity cues and job performance. We did not find significant interactions between verbal or para/nonverbal authenticity cues and any of the examined person characteristics in the prediction of job performance.

Discussion

It is often propagated that applicants should behave authentically in job interviews (Cable & Kay, 2012; Moore et al., 2017; Wilhelmy et al., 2020). Despite this, we know little about observable behaviors that evoke impressions of authenticity in the job interview and the effects of those behaviors. Using multisource data, this study set out to measure behaviors that signal authenticity (i.e., authenticity cues) through behavioral observations to generate knowledge on how such authenticity cues relate to interview and job performance. Our findings reveal that observers can meaningfully differentiate between verbal and para/nonverbal authenticity cues and that both types of cues matter for interview performance. Informing person perception theories, this implies that it is relevant to distinguish different channels of communication (verbal, paraverbal, nonverbal) that can be used to exhibit cues that observers will detect and utilize to infer authenticity judgments. In addition, findings showed that only verbal authenticity cues—but not para/nonverbal authenticity cues—were linked to supervisor ratings of job performance. This relationship remained stable even after controlling for interview performance.

Theoretical Implications

The present findings add to the way we have thought about interview functioning. What we have learned from past research is that job interviews are the most criterion-valid selection tool when compared to many others (Sackett et al., 2022). An often stated explanation for the high levels of interviews’ criterion-related validity is that interviews assess various specific skills and characteristics that interviewees need to perform well at their job (Huffcutt, 2011). Our findings suggests that the interviews’ validity may not just be due to the job-specific competencies that structured interviews typically assess but be partially driven by verbal authenticity cues that influence interview ratings. Going beyond what is formally measured in a structured interview, interviewers seem to detect and utilize cues of verbal authenticity from interviewees, and these cues tend to contain some information about how well interviewees are doing at their job. Conceptually, this finding suggests that an implicit process of detecting and utilizing authenticity cues takes place on the part of the interviewer during the interview and that this process runs parallel to the explicit evaluation of interview performance and influences its outcomes.

Furthermore, the present findings enrich our understanding of how authenticity cues influence performance in different contexts. We found that para/nonverbal authenticity cues seem to be more relevant to performance in the interview, whereas verbal authenticity cues seem to be more relevant to performance on the job. One explanation could be that para/nonverbal cues are more crucial to shaping others’ impressions in short-term interactions, whereas verbal cues are more relevant for influencing others’ impressions in long-term interactions. In line with this, research from other contexts (e.g., leadership emergence) has found that different factors contribute to how individuals are perceived in short-term versus long-term interactions (Badura et al., 2018). In short-term interactions, such as job interviews, the interacting parties do not know each other well and thus do not know whether to trust and rely on each other’s words. Therefore, individuals might be more sensitive to para/nonverbal cues that indicate authenticity and trustworthiness. In contrast, in long-term interactions, the interacting parties have already established a relationship and might be paying less attention to para/nonverbal cues from each other because they already feel safe in interacting with each other. Instead, they may rely more on what the other party expresses verbally. The relationship between authenticity cues and performance can therefore not be viewed without also considering the context (i.e., short-term vs. long-term interaction).

Practical Implications

This study offers implications for organizations on how to set up job interviews. Organizations could explicitly analyze to what extent being perceived as authentic by others is relevant for the success in a particular job (i.e., whether displaying authenticity cues is deemed to be relevant for the job at hand). If being perceived as authentic is considered to be relevant, a practical implication of this research is to explicitly assess authenticity cues in job interviews to (a) make transparent and raise awareness that authenticity cues and perceived authenticity might affect interview ratings (if wanted or not) and/or (b) use authenticity cues as an additional selection criterion.

The present findings also provide implications for individuals searching for a job. Even in highly structured interviews, interviewers seem likely to pick up on behavioral cues that make an interviewee appear as being authentic. Hence, interviewees seem to be well advised to display both verbal and para/nonverbal authenticity cues to achieve favorable interview evaluations. Specifically, the present results suggest that exhibiting simple authenticity cues such as discussing personal/private topics or gesturing and using your hands can potentially improve interview performance ratings.

Relatedly, we suggest that the authenticity cues identified in the present work are trainable. For example, career counselors could leverage this knowledge and offer trainings for job seekers to receive video-feedback on the authenticity cues they display to better understand how interviewers might perceive them. It must be noted that such a training (i.e., learning how to display authenticity cues) would just increase authenticity as perceived by others, but not necessarily authenticity as experienced by the individual. This can involve the risk of encouraging trainees to engage in inauthentic behavior. A solution to this problem could be to develop a training that targets both authenticity as perceived by others and as experienced by the individual (i.e., training authenticity cues that are aligned with the inner self). To achieve this, for example, trainees could learn how to prepare for an interview with the objective of clearly expressing their opinion or train to be mindful of their own emotions on different job-relevant topic so that they can share them when appropriate.

Limitations

A limitation of this study was that we used a mock interview. Some interviewees might show less authenticity cues in a real job interview when stakes are very high for them as compared to a mock interview. Still, interviewees in the present study indicated that they behaved as if they were in an actual selection process (M = 3.92, SD = 0.82, Mdn = 4.00 on a scale from 1 = strongly disagree to 5 = strongly agree), which mitigates possible concerns.

In addition, the interview in this study was highly structured. In a less structured interview, authenticity cues might have a stronger effect on interview performance. Structured interviews focus on strictly rating the specific content of applicants’ responses to interview questions (Campion et al., 1997), whereas less structured interviews focus more on evaluating the applicant and their behavior as a whole. Most job interviews that organizations conduct are probably less structured than the one used for this study, suggesting that the effects of authenticity cues on interview performance could be underestimated in this study. Yet, we observe a tendency towards more structured interviews with the rise of using new technologies in interviewing. For example, “automated” interviews are becoming more popular in practice (Dunlop et al., 2022). In such automated interviews, applicants are videotaped, and artificial intelligence is used to score applicants’ answers to pre-defined interview questions.

Another limitation of this study is that raters only rated the extent that interviewees engaged in different authenticity cues, but not considered the quality of these cues. For example, the identified verbal cue “this person clearly states their personal opinion” does not differentiate between an interviewee stating a lot of dubious versus a lot of benign personal opinions. The actual content of interviewee responses to interview questions is more likely to be captured in traditional interview performance ratings, but not in the current measure of authenticity cues.

Finally, this study used a concurrent validation design, meaning that behavior in the mock interview and supervisory job performance ratings were assessed almost at the same time (i.e., with a time lag of approximately one to two weeks). Given the concurrent study design, we could not study the causality of the relationship between authenticity cues and job performance. As discussed earlier, the relationship between (verbal) authenticity cues and job performance could go in either direction: It is plausible that higher levels of job performance allow interviewees to be more confident in the interview and express this through verbal authenticity cues (e.g., clearly stating their opinion, discussing personal topics, or talking about their emotions). Vice versa, it is possible that displaying verbal authenticity cues is a type of communication skill that applicants show in the interview, but that also helps to build positive relationships at work and evokes favorable impressions in supervisors, subsequently affecting performance evaluations.

Directions for Future Research

This study opens several avenues for future research on authenticity. As a starting point, a relevant question is the extent to which authenticity cues observed by others are valid indicators of authenticity as experienced by the individual. Future studies could explore this question, for example, by having applicants (or participants in the role of applicants) watch videotapes of their interviews and using think aloud protocols in which they comment on when they were more or less authentic and the extent to which their authenticity cues reflected their inner selves at that moment.

A related question for future research is how authenticity and inauthenticity relate to faking in the job interview. For example, does a lack of authenticity represent faking? Definitions of faking imply that it is a strategic behavior used to gain an advantage in the selection process (Levashina & Campion, 2006). In contrast to faking, inauthenticity does not need to involve any strategic motives. For instance, an interviewee might act cagey or artificial, but without any strategic thoughts or even without any awareness of their behavior. Another unanswered question is whether the tendency to fake can be a component of one’s authentic self. In this case, faking (outward behavior) would be in alignment with the person’s internal sense of self, which would make the person act authentically while faking (causing a misalignment between how a person authentically describes their abilities to others and their actual abilities). Thus, future research may benefit from assessing interviewees’ strategic thinking and genuine tendencies to clearly differentiate inauthenticity from faking, and potential forms of “authentic faking.”

Future research is also necessary to investigate the appropriateness of authenticity cues. The measure we developed only captures whether and to what extent interviewees engage in different types of authenticity-related behaviors, but this measure does not take into account whether the respective behavior is appropriate for the interview context or not. For example, we identified “a person discussing personal topics” as a verbal authenticity cue. But the specific topic that the person is discussing could either be more appropriate (e.g., discussing their personal journey towards finding the right career) or less appropriate (e.g., discussing their personal journey towards finding a romantic partner), which most likely would have different effects on interview performance. Another example would be a person who communicates expressively and emotionally in a situation where more direct, professional communication is expected (e.g., when discussing project management or budgeting). As a result, even if two equally qualified candidates interview for the same job and display the same authenticity cues, the two candidates may elicit different reactions from interviewers depending on how situationally appropriate their authenticity cues appear (e.g., talking about emotions when it fits the interview question versus when it does not).

Finally, more knowledge is needed to understand what makes applicants more likely to demonstrate authenticity cues. For example, the present study—unexpectedly—found that individuals who are (a) female, (b) extraverted, and/or (c) have higher scores in verbal cognitive ability are more likely to exhibit certain authenticity cues. Building on this, systematic research is needed to find out to what extent observable authenticity cues are driven by applicants’ abilities (e.g., communication skills, situation perception), traits (e.g., honesty-humility), or intentions (e.g., self-verification striving or the intent to conform to certain gender stereotypes). As another example, past findings indicate that interviewers’ own authenticity may affect interviewees’ authenticity. So far, a qualitative study showed that interviewers often have the intention to act and appear authentic (Wilhelmy et al., 2016). In addition, an experimental study showed that when interviewers openly share downsides of the job, interviewees were less willing to open up during the interview, especially when interviewees perceived high competition for the job (Wilhelmy et al., 2022). Thus, research is needed that clarifies the potential dynamic between interviewers’ and interviewees’ authenticity-related behaviors.