How does the job interview medium affect fairness perceptions and interview ratings? The mediating role of perceptions of social presence

Abstract

The use of video-conference interviews (VIs) has increased substantially, especially with advances in technology and the emergence of COVID-19. Although prior research has suggested that VIs are associated with a lower perception of fairness by applicants and less favorable ratings by interviewers in contrast to face-to-face (FTF) interviews, the reasons for these differences remain inadequately understood. We postulate that applicants’ and interviewers’ perceptions of the other party’s social presence may mediate the relationships among the interview medium, applicants’ perceptions of fairness and interviewers’ ratings. We also explore the moderating roles of applicants’ nonverbal communication expectations and interviewers’ construal level in these mediated relationships. We conduct a one-factor experimental design involving 232 participants (116 applicants and 116 interviewers) and find that VIs lead to lower perceptions of fairness by applicants and lower ratings by interviewers through lower perceptions by applicants and interviewers of the other party’s social presence. Additionally, applicants’ expectations for nonverbal communication during interviews moderate the relationship between the interview media and applicants’ experience of social presence. Specifically, applicants with high expectations for nonverbal communication experience greater social presence and fairness in FTF interviews than in VIs. For applicants with low expectations of nonverbal communication, however, VIs and FTF interviews do not differ in terms of social presence or fairness perceptions, and the moderating effect of interviewers’ construal level is not significant. We discuss the implications of these results for increasing applicants’ perceptions of fairness and the efficacy of interview evaluations in job interviews.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40359-026-04017-3.

Interviews are the most common employee selection tool and are typically used in the final stages of recruitment [1]. Traditional interviews are conducted as face-to-face (FTF) conversations between applicants and interviewers [2]. However, with advances in technology and the emergence of COVID-19, video-conference interviews (VIs) have been widely applied [3, 4]. VIs are an online, remote version of interviews that offer numerous advantages to organizations [5, 6], including reduced recruitment costs, time saved, and companies’ ability to obtain more global applicants [5].

However, applicants’ reactions to VIs may not be as positive. Horn and Behrend [7] reported that 89% of applicants prefer to attend FTF interviews over VIs, whereas only 9% of applicants prefer VIs. In addition, applicants’ experiences during VIs tend to be more negative than their experiences during FTF interviews. For example, participants in Straus et al.’s [8] study reported that they felt more uncomfortable, the interviewer was more annoying, and conversations were harder to understand when they were interviewed via video than when they were interviewed via the FTF method. Thus, VIs lead to a lower level of perceived fairness among applicants than the FTF approach does [9]. In addition, interviewers’ ratings vary across different interview media [10, 11]. According to Blacksmith et al.’s [6] meta-analysis, interviewers tend to rate applicants less favorably in VIs than in FTF interviews.

While these patterns are well-documented, their underlying mechanisms remain inadequately understood. A promising explanatory construct is social presence—conceptualized as “the degree of salience of the other person in the interaction and the consequent salience of interpersonal relationships” [10, 12]. Rooted in social presence theory [12], this concept help characterize key differences between communication media [10, 13, 14]. Empirical work shows that individuals experience lower social presence in online than in FTF interactions [15], a finding that extends to applicants in video-based job interviews [10]. Moreover, social presence meaningfully shapes interpersonal perceptions and evaluations. For example, in educational settings, students’ social presence in online courses positively predicts their perceived learning quality [16] and instructor ratings [17]. Similarly, in selection contexts, applicants’ social presence is a central factor influencing their fairness perceptions [9].

Notwithstanding these insights, the literature exhibits two important theoretical limitations. First, the prevailing perspective remains predominantly unilateral. Most research has focused exclusively on applicants’ perceptions and reactions, implicitly treating the interview as a one-directional process and overlooking its inherently dyadic nature [9, 10]. A comprehensive theoretical account therefore necessitates incorporating the experiences of both parties. In particular, interviewers’ social presence—which has been linked to their evaluations [18]—requires systematic inclusion in an integrated model.

Second, the boundary conditions of these effects remain unclear. Although VIs are generally associated with less favorable outcomes, it is unclear when these effects are strongest or for whom they matter most [6, 19]. A deeper theoretical understanding involves moving beyond main effects to identify moderating factors that clarify the contingencies underlying these relationships.

To address these gaps, this study proposes and tests a dual-path moderated mediation model. We argue that the interview medium shapes outcomes through the mechanism of social presence experienced by both applicants and interviewers, thereby providing a more complete account of the interpersonal dynamics involved. Furthermore, we introduce two theoretically grounded moderators to delineate the boundaries of this model.

Drawing on person–environment fit theory [20]—which posits that individuals’ experiences are most positive when their attributes align with environmental characteristics—we examine the moderating role of applicants’ nonverbal communication expectations and interviewers’ construal level. Prior research suggests that when individuals’ interaction expectations match actual interaction qualities, they report higher social presence [21] and perceive greater procedural fairness [22]. We thus propose that the fit between applicants’ nonverbal communication expectations and the richness of the interview medium will enhance their social presence and, in turn, fairness perceptions.

Similarly, construal level theory (CLT) posits that individuals evaluate objects more positively when their construal level matches the psychological distance of the target [22–25]. That is, individuals with high construal levels tend to evaluate distant objects more favorably, whereas those with low construal levels prefer proximal ones. This matching effect has been demonstrated in consumer behavior [26, 27] and persuasion [28]. Recent work by Schreurs et al. [29] extends it to interview settings, showing that congruence between an interviewer’s psychological distance and the construal level of an applicant’s self-promotion leads to higher ratings. We therefore propose that congruence between interviewers’ construal level and the psychological distance inherent in different interview media will enhance interviewers’ nonverbal social presence and, subsequently, their ratings.

In summary, we propose a comprehensive model (Fig. 1) to examine the mediating roles of applicants’ and interviewers’ social presence in the relationship between the interview medium, applicants’ fairness perceptions and interviewers’ ratings. We also investigate the moderating roles of applicants’ nonverbal communication expectations and interviewers’ construal level in these mediated pathways.

Fig. 1.

Model of the influence of the interview medium on applicants’ fairness perceptions and interviewers’ ratings

Theoretical background and hypotheses development

Face-to-face interviews versus video conference interviews

The interview medium refers to the tools and methods used to conduct an interview. While FTF interviews allow applicants and interviewers to interact directly [19], in VIs, applicants and interviewers communicate online [30]. Research has shown that applicants tend to experience lower levels of fairness perceptions in VIs than in FTF interviews [2, 6, 9, 10]. Fairness perceptions are generally defined as applicants’ perceptions of the overall fairness of a given selection process and selection outcomes [31, 32]. According to Gilliland’s [31] fairness model, applicants’ perceptions of the fairness of a selection procedure are related to different justice rules. These justice rules, such as allowing two-way communication, opportunities to perform and job-relevant information to assess, are less satisfactory in VIs than in FTF interviews [9, 10, 33]. Furthermore, studies have indicated that interviewers give lower evaluations to applicants’ performance in VIs than in FTF interviews [6, 10, 34, 35]. However, the underlying mechanisms by which VIs influence applicants’ perceptions of fairness and interviewers’ ratings remain unclear [6, 34].

It has been argued that the most important difference between VIs and FTF interviews is social presence [10]. In the present study, applicants’ social presence refers to the degree of salience of applicants’ feeling and perception to interviewers’ presence and their relationship with interviewers. Interviewers’ social presence refers to the degree of salience of interviewers’ perception of applicants’ presence and their relationship with applicants. Empirical evidence has demonstrated that individuals’ social presence is lower in VIs than in FTF interviews [9, 15]. First, VIs limit nonverbal communication between applicants and interviewers compared with FTF interviews [36]. For example, steady eye contact between an applicant and an interviewer, which is regarded as an important way of conveying positive affect [34], is largely limited in VIs. Both parties would have to look directly in the camera to make the other party perceive steady eye contact, but looking directly at the camera precludes the ability to look directly at the other party’s picture [36]. In addition, in VIs, the opportunity to communicate via body language and gestures (e.g., hand gestures that emphasize important comments) is limited because people are typically displayed only from the middle of the chest up [34]. Second, technological issues, including time lags, low resolution, and audio failures, may limit the quality of verbal communication in VIs. Compared with FTF interactions, online communication provides less synchronization of coordinated behaviors and less real-time feedback [37]. For example, it is more difficult for applicants and interviewers to immediately interrupt each other to clarify a statement in a VI than in an FTF interview [38]. Research on video conferencing has shown that time lags and other connection failures may hinder the quality of verbal communication by preventing fluent turn-taking between communicating parties [6]. Third, empirical evidence supports the association between the interview medium and social presence by showing that applicants tend to experience lower social presence in VIs than in FTF interviews [10, 39]. Therefore, we propose that VIs lead to a lower perception of social presence by both applicants and interviewers than FTF interviews do.

How do these factors manifest in applicants’ fairness perceptions and interviewers’ ratings? From the perspective of applicants, we suggest that a reduced perception of social presence in VIs will lead to a lower perception of fairness. Chapman et al. [2] proposed that applicants’ lower level of perceived social presence may limit their ability to monitor and adjust their behavior in real time to make a positive impression on the interviewer. This limitation in self-regulation may lead to frustration, which in turn may contribute to a reduced sense of fairness in the process [2]. Basch et al. [9] supported this notion by showing that applicants’ social presence is linked to their perceptions of fairness through the use of restricted impression management tactics. In addition, empirical evidence supports the mediating role of applicants’ social presence in the relationship between interview media (VIs vs. FTF interviews) and applicants’ perceptions of fairness [10]. Therefore, we propose the following:

Hypothesis 1

Job applicants experience lower perceptions of fairness in VIs (vs. FTF interviews), which is mediated by a lower perception of applicants’ social presence.

From the perspective of interviewers, this study aims to explore the mediating role of interviewers’ social presence in the relationship between interview media and interview ratings. Since previous research [40] has shown that interviewers’ ratings are positively associated with their perceptions of applicants’ nonverbal (rather than verbal) displays, our study focuses solely on interviewers’ nonverbal social presence—that is, the salience of interviewers’ perceptions of applicants’ nonverbal presence.

First, VIs restrict applicants’ nonverbal performance and decrease interviewers’ perception of applicants’ nonverbal presence [38, 41, 42]. Owing to media limitations, applicants in VIs exhibit fewer nonverbal behaviors, such as smiles, nods, gestures, body movements and eye contact [43]. When applicants are motivated to make a positive impression but believe that they will not succeed because of the limitations placed on them, they experience anxiety [44, 45]. Anxiety subsequently hinders nonverbal communication [46]. In addition, VIs restrict interviewers’ ability to observe applicants’ nonverbal behaviors [34, 47]. As a result, the interviewer’s perception of the applicant’s nonverbal presence tends to be more negative during VIs than during FTF interviews.

Second, interviewers’ nonverbal social presence is an important source of their overall evaluation. Basch and Melchers [18] reported that interviewers’ social presence increases their interview ratings. In addition, DeGroot and Motowidlo [35] demonstrated that nonverbal cues have the greatest impact on interviewers’ judgment of applicants. Similarly, Cuddy et al.’s [40] study of power poses showed that interviewers’ perceptions of applicants’ nonverbal (rather than verbal) displays mediate the effects of power poses on interview ratings. On this basis, we propose the following:

Hypothesis 2

Interviewers give lower performance ratings to applicants in VIs (vs. FTF interviews), which is mediated by lower levels of interviewers’ nonverbal social presence.

The moderating role of applicants’ expectations of nonverbal communication in the relationship between the interview medium and applicants’ perceptions of social presence

Person-environment fit is generally defined as the congruence between individuals and the environment. Person-environment fit theory suggests that when there is congruence between individuals and the environment, individuals tend to perceive their experience more positively [19, 48]. This theory has received empirical support across different domains, including occupational health and well-being [49, 50], work adjustment [51], vocational choice [52], and public service [53]. Drawing upon this theory, our study aims to explore whether the fit between the interview medium and applicants’ expectations of nonverbal communication influences applicants’ social presence and subsequently affects their perceptions of fairness.

Before applicants participate in an interview, they form expectations about the process on the basis of the information they receive about the interview [54, 55]. In previous studies, researchers have focused primarily on two types of applicants’ expectations, namely, hiring expectations [19, 31, 56] and justice expectations [22, 54, 57], both of which have been linked to applicants’ perceptions [56, 58]. Although most studies have focused on the main effect of applicants’ expectations on their perceptions of selection and the organization, researchers have suggested that applicants’ expectations may also play a moderating role in the relationship between selection characteristics and applicants’ perceptions [19, 31]. The dynamic model of organizational justice suggests that applicants’ perceptions of selection procedures depend on whether a procedure is deemed consistent or inconsistent with applicants’ prior expectations [22]. Ryan and Ployhart’s [19] heuristic model posits that when applicants’ expectations of the selection procedure match the type of selection test (e.g., video vs. paper-and-pencil versions of the situational judgment test), applicants’ perceptions of the selection procedure are more positive. Empirical studies support these propositions by showing that when applicants’ expectations of being hired are congruent with the actual hiring decision, their perception of fairness increases [31, 59].

Proost et al. [60] were the first to examine applicants’ nonverbal communication expectations in VIs. This concept refers to the extent to which applicants expect to use and interpret nonverbal signals during the interview process. Even though VIs are generally expected to offer fewer opportunities to use nonverbal communication than FTF interviews, applicants differ in their expectations for nonverbal communication with respect to VIs. A recent study revealed that some applicants have high expectations for nonverbal communication even after they know that they will be interviewed by video [60]. On the basis of the above literature, we propose that the consistency between applicants’ expectations for nonverbal communication in an interview and the richness of nonverbal cues contained in certain interview media may increase applicants’ perceptions of fairness. More specifically, given the limitations of nonverbal communication in VIs, high expectations for nonverbal communication are more consistent with FTF interviews, whereas low expectations for nonverbal communication are more consistent with VIs.

Moreover, we believe that the consistency or inconsistency between applicants’ nonverbal communication expectations for an interview and the richness of nonverbal cues in an interview may influence fairness perceptions through applicants’ social presence. Research has shown that when individuals’ expectations for nonverbal communication in computer-mediated interactions are sustained, these individuals tend to perceive a higher level of social presence. For example, in Li et al.’s [21] study, participants were asked to communicate with avatars in virtual environments while they observed the nonverbal communication behavior of the avatar. These authors found that the participants perceived a higher level of social presence when their expectation of the avatar’s nonverbal behavior was consistent with its actual behavior. Therefore, we propose the following moderation hypothesis:

Hypothesis 3

The relationship between the interview medium and applicants’ social presence is moderated by applicants’ expectations for nonverbal communication such that applicants with high (low) nonverbal communication expectations will perceive more social presence in FTF interviews (VIs).

Hypothesis 4

The indirect effect of the interview medium on applicants’ fairness perceptions via applicants’ social presence is moderated by applicants’ expectations for nonverbal communication such that applicants with high (low) nonverbal communication expectations report greater perceptions of social presence and, ultimately, a higher perception of fairness in FTF interviews (VIs).

The moderating role of interviewers’ construal level in the relationship between the interview medium and interviewers’ perceptions of nonverbal social presence

According to construal level theory (CLT) [61, 62], when an individual’s construal level matches the psychological distance to a certain object, event or person, the individuals tends to give this object, event, or person a higher evaluation. Individuals with a high construal level tend to use more abstract mental models and extract relatively simple, decontextualized, and coherent information from a stimulus, whereas individuals with a low construal level tend to use more concrete mental models and extract relatively complex, contextualized, and incidental information [63, 64]. Psychological distance is a subjective experience in which something is close to or far from the self, here, and now [62, 65]. An event is more psychologically distant if it takes place farther in the future, if it occurs in more remote locations, if it happens to people less like oneself, or if it is less likely to occur. The more psychologically distant an event is, the more likely it is to be represented at a high construal level, whereas the less psychologically distant an event is, the more likely it is to be represented at a low construal level. Since an individual with a high (low) construal level is congruent with a psychologically distant (near) object, the evaluation of a psychologically distant (near) object by an individual with a high (low) construal level will be more positive [23]. Sungur et al. [28] simultaneously manipulated individuals’ construal levels and the spatial distance depicted in the news. The results revealed that the congruence between the participants’ construal level and the psychological distance (i.e., a high construal level with far distance or a low construal level with near distance) increased the participants’ ratings of the believability of a message. Similarly, Connors and Spangenberg’s [26] study of consumer behavior showed that when participants’ construal levels matched their temporal psychological distance to a particular brand, they tended to evaluate the brand more positively. In the context of job interviews, Schreurs et al. [29] reported that the congruence between an interviewer’s spatial distance (near vs. distant) from an applicant and the construal level of the applicant’s self-promotion (direct self-promotion vs. indirect self-promotion) led to higher interview ratings.

On the basis of these prior studies, we examine the effect of the congruence between interviewers’ construal levels and the psychological distance between interviewers and applicants in different media on interviewer ratings. Specifically, compared with FTF interviews, the spatial distance and social distance between interviewers and applicants in VIs are much greater [23, 66]. Therefore, interviewers with high construal levels are more congruent with VIs, whereas interviewers with low construal levels are more congruent with FTF interviews. When these fit conditions are met, the ratings of interviews may be increased.

In addition, we believe that this congruence may affect interviewers’ ratings via interviewers’ nonverbal social presence. Schreurs et al. [29] speculated that enhanced nonverbal cues (e.g., politeness) may explain the effects found in their study of the congruence between an interviewer’s psychological distance to applicants and the construal level of applicants’ self-promotion on the increased interview rating. Basch and Melchers’ [18] experimental study showed that interviewers’ social presence mediated the relationship between eye contact in interviews and interviewers’ ratings. Therefore, we propose the following moderation hypothesis:

Hypothesis 5

The relationship between the interview medium and the interviewer’s nonverbal social presence is moderated by the interviewer’s construal level such that an interviewer with a high (low) construal level perceives greater nonverbal social presence in VIs (FTF interviews).

Hypothesis 6

The indirect effect of the interview medium on interviewers’ ratings via interviewers’ nonverbal social presence is moderated by interviewers’ construal levels such that interviewers with high (low) construal levels give higher ratings in VIs (FTF interviews) due to greater perceptions of nonverbal social presence.

Method

Participants and procedure

A convenience sampling method was used to recruit participants in both online and offline recruitment processes. We posted participant recruitment ads on WeChat Moment (the social media feature of WeChat) and the campuses of several universities in northern China. Potential participants were invited to register and participate in our simulated job interview experiment and then complete a packet of questionnaires. We used G*Power (v. 3.1 [67]), to calculate the minimum required sample size. With a small to medium sized effect of 0.10, an α of 0.05 and a power of 0.80, we estimated that the minimum required sample size should be larger than 114. A total of 256 participants were recruited, the majority of whom were undergraduate students. All the participants had prior experience with job interviews, which ensured their familiarity with the interview process. Informed consent was obtained from the participants. The consent procedures were approved by the ethics committee of the Faculty of Psychology at the corresponding author’s university. Participants who completed the interviews and answered all the questionnaires were compensated for their time with 20 RMB ($3 USD).

After they agreed to participate in the experiment, the participants were randomly assigned to one of two conditions, namely, a face-to-face (FTF) interview (N = 128) or a video-conference interview (VI) (N = 128). Within each condition, participants were randomly assigned the role of interviewer (N = 64) or applicant (N = 64) and randomly paired one-on-one with each applicant matched with a single interviewer (64 dyads in each condition). Subsequently, all the applicants were asked to complete questionnaires that measured their nonverbal communication expectations and demographics. The interviewers were also asked to complete a questionnaire that measured their construal levels and demographics. Dyads were excluded from the analysis if either the applicant or the interviewer failed to complete the questionnaires or if the interview process encountered disruptions (i.e., technical issues in VIs). In both the FTF and VI conditions, six dyads were removed, resulting in a valid sample of 58 dyads in each condition. Thus, a total of 116 applicants (37.9% males, 62.1% females) and 116 interviewers (37.1% males, 62.9% females) were retained for the experiment. The age distribution for the applicants was as follows: 21.6% were under 20 years old, 66.4% were between 20 and 25 years old, and 12.1% were between 26 and 30 years old. For the interviewers, the age distribution was 2.6% under 20 years old, 66.4% aged 20–25, 19% aged 26–30, and 12.1% over 30 years old.

To ensure the quality of the interview ratings, all the interviewers received several hours of frame-of-reference training before the experiment [68]. After the training, an experimenter invited all the participants to the laboratory for the simulated interview. In both the FTF interviews and the VIs, the experimenter was present to oversee the experimental process, and the experiment was performed in a dyad-by-dyad manner. First, the experimenter provided a brief overview of the experiment without revealing its specific aims to avoid introducing bias. Then, the interviews started.

In the FTF condition, interviews were conducted in a controlled environment. Both the interviewer and the applicant were seated in a designated interview room [2] approximately 1.5 m apart to facilitate natural interaction. In the VI condition, the applicants and interviewers were seated in different rooms, and the interviews were conducted via Tencent Conference (an online conferencing platform, https://meeting.tencent.com/). Both rooms were quiet and distraction-free to replicate the controlled atmosphere of the FTF condition as closely as possible [10]. All video interviews utilized the same equipment and settings, including camera angles, lighting, and audio quality. Each dyad interacted only through Tencent Conference to ensure that their experience of the interview was mediated entirely by the computer.

The interview process lasted approximately 15 min. The four structured interview questions developed by Dong [69] were used, including a self-introduction question (“Would you please introduce yourself?”), two behavioral questions (e.g., “Have you ever been in an emergency at work or in life? How did you feel at that time? What did you do? Please give an example to illustrate”) and one situational question (“Suppose that you have performed well in your work and have made outstanding achievements, but you have not been able to win the trust of your manager. Instead, those who are not as capable as you but have the ability to talk well have obtained the trust of the manager. You feel very distressed about this. What would you do?”).

After the interview, the applicants were asked to complete measures of their perceptions of social presence and fairness, and the interviewers were asked to complete a measure of their perception of nonverbal social presence and rate the performance of the applicant. After all the scales were completed, the participants were paid and debriefed.

Measures

Following the translation/back-translation procedure [70], we translated the English-language scales into Chinese scales. Variables were assessed using a 5-point Likert scale (“1” = “strongly disagree” to “5” = “strongly agree”) unless indicated otherwise (The complete set of scale items used in this study is provided in Supplementary Material 1).

Applicant’s perception of social presence

The Chinese adaptation [71] of Short et al.’s [12] social presence measure was used (α = 0.74), which includes three dimensions, namely, coexistence presence (e.g., “During the interview, I felt like I was coexisting in the same time and space with the interviewer”), communication presence (e.g., “During the interview, I had an exchange of ideas with the interviewer”), and emotional presence (e.g., “During the interview, the interviewer’s emotion affected me”).

Interviewer’s nonverbal social presence perception

Tsai et al.’s [72] nonverbal behavior assessment scale was used to measure interviewers’ perceptions of nonverbal social presence (α = 0.84). A sample item is “Applicants display very friendly nonverbal messages, such as smiles and nods”.

Nonverbal communication expectations

Proost et al.’s [60] measure of applicants’ expectations of nonverbal communication during interviews was used (α = 0.87). A sample item included “During this interview, I expect that I will be able to make eye contact with the interviewer”.

Construal level

The construal level was measured using Vallacher and Wegner’s [73] behavior identification form (α = 0.72). The participants were asked to choose between two alternative descriptions of 25 different actions (e.g., reading). One described the means of the action (e.g., following lines of print), whereas the other described the purpose of the action (e.g., gaining knowledge). Choices for the former were coded as 0, whereas choices for the latter were coded as 1. A median split was used to identify two construal levels [25, 74]. Individuals who scored 14 or above were classified as having a high construal level, while individuals with scores of 13 or below were classified as having a low construal level.

Fairness perceptions

Following Basch et al. [10], four subscales from Bauer et al.’s [32] Selection Procedural Justice Scale were used to measure applicants’ fairness perceptions (α = 0.89). These four subscales represented predictive job-relatedness (e.g., “Doing well in such an interview means I could do well on the job”), the chance to perform (e.g., “In such an interview, I can really show my skills and abilities”), two-way communication (e.g., “There is enough communication during such an interview”), and global fairness (e.g., “I think that such an interview is a fair way to select people”).

Interview rating

The interviewers used four structured interview questions developed by Dong [69] to rate applicants on six dimensions, namely, planning and organization, flexibility and adaptability, interpersonal skills, overall impression, verbal communication, and logical thinking (The complete interview questions and scoring rules are provided in Supplementary Material 2). Following Dong [69], the scoring reference was based on a 1–10 behaviorally anchored rating scale, and the overall evaluation was synthesized from these six dimensions (α = 0.88).

Control variables

We controlled for demographics, including the applicants’ and interviewers’ gender, age, education and employment status, because of their potential effects [75].

Results

Preliminary analyses

For applicants, preliminary analyses by means of t tests revealed that the two interview medium groups did not differ with regard to age, gender, education or working conditions (ts < 1.15, ps > 0.15). We obtained similar results for the interviewers (ts < 1.75, ps > 0.10). The means, standard deviations, and intercorrelations for all the study variables can be seen in Table 1.

Table 1.

Descriptive information and correlations for all applicants and interviewers

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15
1. Interview medium	-
2. Gender (interviewer)	− 0.02	-
3. Age (interviewer)	− 0.25**	0.13	-
4. Education (interviewer)	0.14	− 0.21*	0.03	-
5. Employment status (interviewer)	− 0.01	0.02	0.33**	− 0.21*	-
6. Gender (applicant)	0.04	− 0.16	− 0.15	0.28**	− 0.13	-
7. Age (applicant)	0.20*	0.09	0.07	− 0.05	0.16	− 0.22*	-
8. Education (applicant)	0.09	0.07	0.00	0.15	− 0.03	− 0.16	0.48**	-
8. Employment status (applicant)	0.01	0.05	0.06	− 0.03	− 0.02	− 0.10	0.41**	0.13	-
10. Social presence (applicant)	− 0.35**	− 0.04	− 0.08	0.09	− 0.01	0.04	0.07	0.24**	− 0.03	(0.74)
11. Nonverbal social presence (interviewer)	− 0.34**	− 0.05	0.10	0.08	0.09	0.00	− 0.04	− 0.11	0.08	0.11	(0.84)
12. Nonverbal communication expectation (applicant)	− 0.52**	0.00	0.13	− 0.03	0.02	0.08	− 0.06	0.01	0.00	0.41**	0.25**	(0.87)
13. Construal level (interviewer)	− 0.05	0.01	− 0.19*	0.13	− 0.36**	0.11	− 0.04	0.01	0.02	0.14	− 0.15	0.05	(0.72)
14. Fairness perception (applicant)	− 0.22*	− 0.11	− 0.12	− 0.03	0.06	− 0.03	0.09	0.06	− 0.03	0.62**	0.10	0.42**	0.08	(0.89)
15. Interviewer rating (interviewer)	− 0.19*	0.13	− 0.01	− 0.26**	− 0.06	− 0.09	− 0.05	0.04	− 0.01	0.06	0.51**	0.09	− 0.02	− 0.04	(0.88)
M	1.50	1.63	2.41	3.52	1.56	1.62	1.91	3.32	1.12	3.73	18.5	3.98	1.62	3.58	41.95
SD	0.50	0.49	0.73	0.74	0.89	0.49	0.58	0.68	0.46	0.54	3.69	0.64	0.49	0.58	7.69

N = 116. Sex was coded as 0 = female and 1 = male. The interview condition was coded as 0 = face-to-face and 1 = video conference. Education was coded as 1 = high school, 2 = professional training college, 3 = university, and 4 = postgraduate and above. The values on the diagonal represent coefficient alphas. The construal level was coded as 1 = low construal level or 2 = high construal level

^*p < .05; ^**p < .01

Hypothesis testing

To test Hypotheses 1 and 2, we first utilized ordinary least squares regression and then conducted a bootstrapping-based mediation test using the PROCESS macro to construct 95% confidence intervals on the basis of 5000 replications [76]. For Hypothesis 1, we found that the interview medium (coded as 0 = face-to-face, 1 = video conference) was negatively associated with applicants’ social presence (b = − 0.43, SE = 0.09, p < .01; Table 2, Model 1), and applicants’ social presence was positively associated with their fairness perceptions (b = 0.70, SE = 0.09, p < .01; Table 2; Model 2). The bootstrapping results revealed that the indirect effect of the interview medium through applicants’ social presence on their perceptions of fairness was significant (indirect effect = − 0.30, SE = 0.07, 95% CI [-0.45, − 0.16]), supporting Hypothesis 1. For Hypothesis 2, we found that the interview medium was negatively associated with interviewers’ nonverbal social presence (b = − 0.48, SE = 0.14, p < .01; Table 3, Model 1), and interviewers’ nonverbal social presence was positively associated with the interview rating (b = 0.91, SE = 0.15, p < .01; Table 3, Model 2). The bootstrapping results revealed that the indirect effect of the interview medium through the interviewer’s nonverbal social presence on the interview rating was significant (indirect effect = − 0.43, SE = 0.15, 95% CI [-0.76, − 0.18]), thus supporting Hypothesis 2.

Table 2.

Regression analyses of applicants’ social presence and fairness perceptions

Variable	Step 1: Nonconditional Regression		Step 2: Conditional Regression
	Applicant’s social presence	Fairness perception	Applicant’s social presence
	Model 1	Model 2	Model 3	Model 4
Control variables
Gender	0.12 (0.10)	− 0.06 (0.09)	0.08 (0.10)	0.08 (0.09)
Age	0.07 (0.10)	0.14 (0.09)	0.05 (0.10)	0.05 (0.10)
Education	0.21** (0.08)	− 0.14 (0.07)	0.20** (0.08)	0.18* (0.07)
Employment Status	− 0.10 (0.11)	− 0.06 (0.10)	− 0.09 (0.11)	− 0.13 (0.11)
Predictors
Interview medium	− 0.43** (0.09)	0.00 (0.09)	− 0.26* (0.11)	− 0.23* (0.11)
Applicant’s nonverbal communication expectation			0.24** (0.08)	0.47** (0.13)
Mediator
Applicant’s social presence		0.70** (0.09)
Interaction
Interview medium × Applicant’s nonverbal communication expectation				− 0.37* (0.17)
R 2	0.22	0.41	0.27	0.30

N = 116. The estimates are unstandardized regression coefficients. The values in the brackets are standard errors (SEs)

^*p < .05; ^**p < .01

Table 3.

Regression analyses of the interviewer’s nonverbal social presence and rating

Variable	Step 1: Nonconditional Regression		Step 2: Conditional Regression
	Interviewer’s nonverbal social presence	Interviewer’s rating	Interviewer’s nonverbal social presence
	Model 1	Model 2	Model 3	Model 4
Control variables
Gender	− 0.12 (0.14)	0.33 (0.21)	− 0.12 (0.14)	− 0.11 (0.14)
Age	− 0.01 (0.10)	− 0.02 (0.15)	− 0.02 (0.10)	− 0.03 (0.10)
Education	− 0.11 (0.09)	− 0.32* (0.15)	− 0.10 (0.09)	− 0.09 (0.10)
Employment status	0.12 (0.08)	− 0.26* (0.12)	0.09 (0.08)	0.09 (0.08)
Predictors
Interview medium	− 0.48** (0.14)	0.02 (0.22)	− 0.50** (0.14)	− 0.76 (0.45)
Interviewer’s construal level			− 0.18 (0.14)	− 0.26 (0.20)
Mediator
Interviewer’s nonverbal social presence		0.91** (0.15)
Interaction
Interview medium × Interviewer’s construal level				0.16 (0.27)
R 2	0.15	0.34	0.17	0.17

N = 116. The estimates are unstandardized regression coefficients. The values in the brackets are standard errors (SEs)

^*p < .05; ^**p < .01

To test Hypothesis 3, we followed a two-step procedure in line with established practices [76, 77]. First, we used hierarchical multiple regression analysis to test the significance of the interaction term. The control variables (gender, age, education and employment status), interview medium, and applicants’ nonverbal communication expectations were entered in the first step to predict applicants’ social presence. The interaction between the interview medium and the applicants’ nonverbal communication expectations was entered in the second step. The results showed that the interview medium was negatively associated with applicants’ social presence (b = − 0.23, SE = 0.11, p < .05; Table 2, Model 4) and that applicants’ nonverbal communication expectations were positively associated with their social presence (b = 0.47, SE = 0.13, p < .01; Table 2, Model 4). Most importantly, the interaction effect between the interview medium and applicants’ nonverbal communication expectations was significant (b = − 0.37, SE = 0.17, p < .05; Table 2, Model 4) for applicants’ social presence.

Second, for the significant interaction, we conducted simple slope analyses to interpret its nature. Following Cohen et al. [78], the continuous moderator was conditioned at high (+ 1SD) and low (-1SD) levels to estimate the effect of the interview medium on the dependent variable at these specific values. Simple slope analyses revealed that the interview medium had a negative association with applicants’ social presence when the level of applicants’ expectations of nonverbal communication was high (+ 1SD; b = − 0.46, SE = 0.14, p < .01). However, this association was not significant when the level was low (-1SD; b = 0.01, SE = 0.16, p > .05) (see Fig. 2). This finding partially supports Hypothesis 3.

Fig. 2.

The moderating effect of nonverbal communication expectations on the interview medium and social presence

We followed the same procedure to test Hypothesis 5, with the interviewer’s nonverbal social presence as the dependent variable. As shown in Table 3, the effects of the interview medium (b = − 0.76, SE = 0.45, p > .05; Table 3, Model 4) and the interviewer’s construal level (b = − 0.26, SE = 0.20, p > .05; Table 3, Model 4) on the interviewer’s nonverbal social presence were both nonsignificant. Moreover, the interaction effect between the interview medium and the interviewer’s construal level was not significant (b = 0.16, SE = 0.27, p > .05; Table 3, Model 4) on the interviewer’s nonverbal social presence. Thus, Hypothesis 5 was not supported.

Next, we calculated the conditional indirect effects of applicants’ social presence or interviewers’ nonverbal social presence at different levels of the moderator (i.e., applicants’ nonverbal communication expectation or interviewers’ construal level) utilizing the methods of Hayes [76] to test Hypotheses 4 and 6. The bootstrapping results revealed that the indirect effect of the interview medium on fairness perceptions through applicants’ social presence was significant when nonverbal communication expectations were high (indirect effect = − 0.32, SE = 0.12, 95% CI [-0.58, − 0.09]) but was not significant when these expectations were low (indirect effect = 0.004, SE = 0.11, 95% CI [-0.20, 0.22]). The index of moderated mediation was also significant (index = − 0.26, SE = 0.13, 95% CI [-0.52, − 0.01]), providing partial support for Hypothesis 4. We also found that the indirect effects of the interview medium on interviewers’ ratings through interviewers’ nonverbal social presence were significant both when the interviewer’s construal level was high (indirect effect = − 0.39, SE = 0.17, 95% CI [-0.76, − 0.10]) and when it was low (indirect effect = − 0.54, SE = 0.22, 95% CI [-0.99, − 0.14]). The index of moderated mediation was not significant (index = 0.15, SE = 0.24, 95% CI [-0.35, 0.63]). Therefore, Hypothesis 6 was not supported.

Discussion

This study was designed to address two key theoretical limitations in the literature on technology-mediated interviews: the predominantly unilateral focus on applicants, and the unclear boundary conditions of interview medium effects. By testing a dual-path moderated mediation model, our findings not only corroborate the central role of both applicants’ and interviewers’ social presence but, more critically, clarify the scope and operation of person-environment fit and CLT in the context of personnel selection. The results provide nuanced theoretical insights that move beyond documenting main effects.

Our experimental design, by randomly assigning participants to conditions, extends previous studies on the effects of different interview media [6, 10, 79] in establishing a clear causal sequence: the VI format directly causes a reduction in social presence for both applicants and interviewers. Furthermore, we provide robust evidence that these diminished perceptions subsequently explain the declines in applicants’ fairness perceptions and interviewers’ ratings. This model advances beyond a unilateral focus on applicants [10] by demonstrating that the interview is a dyadic interaction in which the communication medium fundamentally shapes the interpersonal experience for both parties. Thereby, our study establishes a more complete theoretical foundation for social presence theory in selection contexts.

A key contribution of this study lies in exploring the boundary conditions of the interview medium’s effects, responding to calls for such research [6]. For applicants, we discovered a theoretically instructive asymmetrical matching pattern. The beneficial effect of a rich medium (FTF) was pronounced for applicants with high nonverbal communication expectations, whereas a match between low expectations and a lean medium (VI) did not yield a comparable boost in social presence or fairness.

This finding directly informs an ongoing theoretical debate regarding the essence of person-environment fit. One perspective contends that any alignment between individual and environmental traits leads to positive effects [48, 80–82]. For example, employees with a high need for initiating structure from leaders, a type of leadership behavior in which leaders clarify tasks and provide direction, exhibit more organizational citizenship behaviors when their leaders display a high level of this behavior, while employees with a low level of this need also demonstrate more organizational citizenship behaviors when their leaders exhibit a minimal initiating structure [82]. Conversely, an alternative view posits that fit occurs when the environment fulfills individuals’ high needs [83–85]. For example, employees with high expectations for the use of certain job skills at work are more engaged when they are provided with ample opportunities to use these skills. Conversely, for employees with low expectations, the absence of an opportunity to use these skills does not lead to increased work engagement [85]. Our results therefore align with the latter “needs-supplies” perspective. The mere alignment of low needs with a poor environment was insufficient to produce a positive state in the interview context, as demonstrated by the non-significant effect for low-expectation applicants. This moves beyond a simplistic “congruence is good” postulate and provides a more precise understanding of how applicant expectations interact with selection technologies.

Regarding the non-significant moderating role of interviewers’ construal level, our findings offer a critical theoretical insight that helps delineate the boundary conditions of CLT in selection contexts. We posit that the potent situational forces inherent in a high-stakes, structured interview—such as standardized questions, explicit scoring rubrics, and pressures for objectivity—may attenuate or overwhelm the influence of an interviewer’s idiosyncratic cognitive style. This suggests that the beneficial effects of a match between construal level and psychological distance, well-documented in domains like consumer behavior [26, 27], might not readily generalize to the complex, socially charged, and standardized environment of a personnel selection interview. Thus, our null finding serves as a valuable theoretical caution; it indicates that the predictive power of CLT is likely context-dependent and may be limited in situations where situational strength is high. Future research is needed to further investigate these potential boundary conditions and explore the specific contexts in which cognitive styles like construal level do and do not significantly impact professional evaluations.

Limitations and future research directions

The present study has limitations, several of which point toward new directions for future research. First, our sample consisted mainly of college students, who tend to have more positive attitudes toward technology-mediated communication [86]. Basch et al. [10] noted that college students’ experience with VIs may be more positive in general, which may lead to an underestimation of the negative impact of VIs. Additionally, college students’ experience of being applicants or interviewers may be rather limited. Future research should investigate these effects among actual applicants and interviewers.

Second, although our experimental design effectively controlled variables and established causal relationships, the simulated interviews lacked the high-stakes context of actual job interviews, which may have affected the applicants’ and interviewers’ motivation, engagement and, ultimately, perceptions and evaluations. Future research should adopt diverse research designs to improve external validity.

Third, while our study focused on the moderating effects of applicants’ nonverbal communication expectations and interviewers’ construal level, other individual characteristics may also play a moderating role in the relationships among the interview medium, applicants’ fairness perceptions and interviewers’ ratings. For example, applicants’ familiarity with and preference for modern technology-mediated communication or interviewers’ expertise and experience may play a moderating role [6]. Future research can further investigate these potential moderators.

Practical implications

VIs present significant advantages by reducing recruitment costs and broadening applicant pools in globalized and remote work environments [2, 5], and they have enormous application prospects in recruitment processes [3, 4]. However, the effective use of VIs and their impact on applicants’ reactions and interview ratings for organizations remain unclear. The findings of the present study offer several recommendations.

First, consistent with prior studies, our study revealed that compared with FTF interviews, VIs lead to lower levels of fairness perceptions among applicants and lower interviewer ratings [10]. Therefore, it is important for managers to be aware of these effects and not regard VIs as an equivalent substitute for FTF interviews.

Second, the current study highlights the critical role of applicants’ and interviewers’ experiences of social presence in shaping applicants’ fairness perceptions and interviewers’ ratings in VIs. When VIs are used, strategies such as high-resolution video conferencing systems, providing better lighting and audio setups, and ensuring real-time interaction to reduce communication delays can be adopted to increase the social presence experience of applicants and interviewers [87, 88].

Third, the current study shows that aligning applicants’ nonverbal communication expectations with the interview medium may elicit more favorable reactions from applicants. Organizations should consider the expectations of applicants when choosing an interview medium [22, 31]. Specifically, for applicants with high expectations for nonverbal communication, FTF interviews may be preferable because they tend to enhance applicants’ social presence and fairness perceptions. For applicants with low nonverbal communication expectations, however, VIs serve as an excellent alternative to FTF interviews because they can provide flexibility and efficiency without decreasing applicants’ social presence or fairness perceptions. Therefore, knowledge about applicants’ nonverbal communication expectations is beneficial for organizations to select an appropriate interview medium and obtain better reactions from applicants [60]. In addition, applicants’ expectations about interviews can be shaped through explanations and descriptions of the forthcoming interview process, and highlighting the advantages of the selected medium can improve applicants’ reactions [89–91]. Therefore, explaining the flexibility of the VI approach before it is used can have a positive effect on applicants’ reactions [92].

Abbreviations

VIs

Video-conference Interviews

FTF

Face-to-Face

COVID-19

Corona Virus Disease 2019

CLT

Construal Level Theory

SE

Standard Error

CI

Confidence Interval

SD

Standard Deviation

Authors’ contributions

QG and YH contributed equally to this work. RB and RY contributed to the study conception and design. RY and YH were responsible for material preparation, data collection and analysis. The first draft of the manuscript was written by QG and YH, and all authors commented on and revised previous versions. All authors read and approved the final manuscript.

Funding

No funding was received for this study.

Data availability

The datasets used in this study are available from the corresponding author upon request. We confirm that the dataset has not been published elsewhere.

Declarations

Ethics approval and consent to participate

This study was performed in accordance with the ethical standards of the Declaration of Helsinki. Ethics approval was granted by the Ethics Committee of the Faculty of Psychology, Beijing Normal University (Approval No. 202110010101). Written informed consent was obtained from all individual participants included in the study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.