Job Seekers’ Impression Management on Facebook: Scale Development, Antecedents, and Outcomes

Many organizations rely on social media like Facebook as a screening or selection tool; however, research still largely lags behind practice. For instance, little is known about how individuals are strategically utilizing their Facebook profile while applying for jobs. This research examines job seekers’ impression management (IM) tactics on Facebook, personality traits associated with IM use, and associations between IM and job-search outcomes. Results from two complementary studies demonstrate that job seekers engage in three main Facebook IM tactics: defensive, assertive deceptive, and assertive honest IM. Job seekers lower in Honesty–Humility use more Facebook IM tactics, whereas those higher in Extraversion use more honest IM and those higher on Conscientiousness use less deceptive IM. Honest IM tactics used on Facebook are positively related to job-search outcomes. This paper therefore extends previous IM research by empirically examining IM use on Facebook, along with its antecedents and outcomes.

seeker posts professional and/or personal accomplishments on their Facebook profile. Second, assertive deceptive IM involves inventing accomplishments to create a falsified image of a good candidate. This deceptive version of IM is when a job seeker posts embellished or made-up professional and/or personal accomplishments on their profile. Finally, defensive IM involves censoring previously posted content believed to negatively impact one's professional reputation, as well as monitoring Facebook content by filtering any information that could be perceived negatively by employers. An example is when a job seeker removes controversial posts (i.e., drinking and/or partying pictures) from their profile. Roulin and Levashina (2016) also discussed other-focused IM on social media but described it as job-or organization-specific and argued it is less prevalent. It was therefore excluded from investigation.

Developing and Validating the Facebook Impression Management Scale
Building on Roulin and Levashina's (2016) framework, our first goal was to develop and empirically validate a measure of applicant IM on social media: The Facebook Impression Management Scale (FIMS). Specifically, we hypothesized that three main factors of job seekers' Facebook IM described above would emerge: Hypothesis 1: The FIMS will include three factors: (a) defensive IM, (b) assertive deceptive IM, and (c) assertive honest IM.
IM is not a new construct in the workplace or social media literatures, and several general IM measures already exist, with IM conceptualized both as a trait and as a behavior. However, these measures are either not specific to social media or not specific to job seekers. In terms of behaviors, the FIMS should demonstrate convergent validity with both general impression management measures (e.g., IMSS; Bolino & Turnley, 1999) and general self-presentation on Facebook (e.g., General Facebook Self-Presentation Scale; Rosenberg & Egbert, 2011). Yet, our measure captures different dimensions of IM than those two existing scales, because it specifically captures applicant IM behaviors on social media. Therefore, we expect the FIMS will be positively but only moderately correlated with these general measures. In addition, the FIMS should demonstrate discriminant validity with trait measures of IM, such as self-deceptive enhancement and trait-based impression management (e.g., Balanced Inventory of Desirable Responding; Hart et al., 2015). For instance, IM tactics used by applicants in the job interview context have been shown to be conceptually different from various social desirability or trait-IM measures, with correlations ranging from small but negative to small but positive (Levashina & Campion, 2007). Therefore, we expect the FIMS will only be weakly correlated with self-deceptive enhancement and trait-based impression management.

Hypothesis 2:
The FIMS will be more strongly positively correlated with the behavioral measures of IM such as the (a) general impression management and (b) general Facebook self-presentation than with the trait-IM measures such as (c) self-deceptive enhancement and (d) trait-based impression management.

Personality as an Antecedent of Facebook Impression Management
In addition to development of the FIMS, we investigated whether certain personality traits were related to the use of our three IM tactics on social media. Earlier research has examined the relationships between personality and trait-IM or social desirability (e.g., McCrae & Costa, 1983). More recently, many theoretical models of applicant IM in selection (e.g., Levashina & Campion, 2006; and empirical work on both workplace and interview IM (e.g., Bourdage et al., 2015;Bourdage et al., 2018;Melchers et al., 2020) have established personality as an important antecedent of IM behaviors. This is largely because personality traits contribute to the motivation or willingness to engage in IM. Roulin and Levashina (2016) have also proposed that personality (alongside other individual differences) should be a core antecedent of applicant IM on social media. Specifically, this literature suggests that job applicants who are high on Honesty-Humility are less likely to engage in IM, especially deceptive IM. Conversely, job seekers who are high in conscientiousness and extraversion are more likely to engage in honest IM but less likely to engage in deceptive IM and defensive IM. However, relationships for Emotionality/Neuroticism, Agreeableness, and Openness appear weaker and more inconsistent.
To examine the relationship between personality and Facebook IM, we used different personality measures in Study 1 (NEO PI-R) and Study 2 (HEXACO PI-R). In Study 1, we focused on specific facets of the NEO that were expected to either be positively or negatively related with the IM tactics: deliberation (Conscientiousness), impulsiveness (Neuroticism), and modesty (Agreeableness, which is also related to Honesty-Humility in HEXACO). In Study 2, we used the six HEXACO factors: Honesty-Humility, Emotionality, Extraversion, Agreeableness, Conscientiousness, and Openness to Experience. Hypothesis 4: Both (a) Conscientiousness and (b) the NEO facet of deliberation are positively related to assertive honest IM but negatively to assertive deceptive IM and defensive IM.
Hypothesis 5: Extraversion is positively related to assertive honest IM but negatively to assertive deceptive IM and defensive IM.
We also explore relationships between the FIMS factors and Emotionality/Neuroticism (and the impulsiveness facet), Agreeableness, and Openness, but do not propose any hypotheses given the lack of stable effects in past research.

Potential Outcomes of Facebook Impression Management
Lastly, we were interested in determining whether Facebook IM tactics were related to job-search outcomes. Prior research suggests that IM is positively related to a variety of job-search outcomes: Barrick et al. (2009) found that IM tactics used during interviews were positively related to interviewer ratings, and (to a lesser extent) job performance. More recently, Bourdage et al. (2018) found a positive relationship between IM used in job interviews (particularly honest self-promotion and ingratiation) and job offers received. In the present study, we propose to examine two potential outcomes of applicant use of IM on social media: job search self-efficacy and the number of job offers received. Recent work has positioned the "outcome" facet of job search self-efficacy (i.e., JSSE-O; Saks et al., 2015) as an important subjective outcome of applicant career planning. Indeed, it captures applicants' level of confidence regarding key outcomes of their ongoing job search (e.g., invitations to job interviews, obtaining an attractive job offer), and is associated with the number of job offers received several months later. Previous work has examined the role of stable positive self-perceptions (i.e., core self-evaluations) as an antecedent of interview IM (e.g., Roulin & Bourdage, 2017), but relationships with job search self-efficacy have not been examined. Building on Saks et al. (2015), we argue that applicant IM behaviors on social media (i.e., highlighting true qualifications, exaggerating or inventing qualifications, or deleting negative profile content) represent a strategy used to achieve career goals (i.e., secure an attractive job). Therefore, we propose that IM use on social media is positively related to both subjective (job search self-efficacy) and objective (number of job offers received) career outcomes. In addition, because our FIMS is oriented specifically toward job seekers, it should explain incremental variance in such outcomes beyond more general IM measures (e.g., General Facebook Self-Presentation).
Hypothesis 6: FIMS is positively associated with (a) job-search self-efficacy and (b) the number of job offers received.
Hypothesis 7: FIMS will explain incremental variance in (a) job-search self-efficacy and (b) the number of job offers received above and beyond general Facebook Self-Presentation.

Studies Overview
A preliminary study and two quantitative studies were conducted. The purpose of the preliminary study was to establish the conceptual framework, generate FIMS items, and perform initial content validation. The aim of Study 1 was to explore the underlying structure of the FIMS, test its convergent and discriminant validity, and examine some of its personality antecedents. Last, Study 2 aimed to confirm the factor structure of the FIMS, explore additional personality antecedents, and investigate potential outcomes (jobsearch self-efficacy and job offers received).

Item Generation
We followed Hinkin's (1998) guidelines for measure development. Based on the literature review of IM behaviors discussed, as well as an internet search on IM techniques used by Facebook users, the authors used a deductive approach to generate an initial pool of 44 items: 23 defensive; 9 assertive deceptive; 12 assertive honest. No items were reverse coded, and efforts were made to simplify the language and style used in the items to make sure they were easy to understand. Defensive IM was defined as image repair tactics, including deletion explicitly focused on removing unprofessional content. Assertive deceptive IM includes inventing accomplishments to create a falsified image of a good job applicant. Finally, assertive honest IM involves behavior such as posting positive statements to describe one's personal/professional qualities, past accomplishments or future plans.

Content Validation
A group of eight respondents (graduate students in psychology) were given a list of the 44 items including definitions for each of the three constructs. Items were presented randomly to avoid participant bias in the sorting task. Participants were asked to match items with their corresponding definition. An acceptable agreement index for this methodology is typically 75% (e.g., Hinkin, 1985), 17 of the 44 items did not meet this threshold and were eliminated. Thus, the revised measure contained 27 items (16 items defensive; 6 assertive deceptive; 5 assertive honest).

STUDY 1 Sample
Participants were recruited on Mechanical Turk (MTurk) using the TurkPrime online recruitment platform (Litman et al., 2017). MTurk was selected as a recruiting platform because it is effective in obtaining more diverse and attentive samples and allows the screening of unmotivated individuals (e.g., Highhouse et al., 2017;Landers & Behrend, 2015). We also included an additional screening question asking participants whether they were currently searching for a job (only participants that reported they were a job seeker were permitted in our survey). 506 participants currently looking for a job, active on Facebook, and residing in North America were recruited to partici-pate. Items within each measure were presented randomly to participants. Three attention checks were also included (e.g. "I can teleport across time and space"). As a result of participants not meeting the inclusion criteria or failing an attention check, a total of 431 participants were retained for analyses. The participants were between the ages of 18-73 (M = 35.17 years, SD = 10.12), 49.6% were male, 89.4% were employed (72.5% full time), 53.3% had at least a university degree, 96% from the United States and 4% from Canada.

Measures
Facebook Impression Management. Facebook IM was measured with the new FIMS. Participants responded to 27 items on a 5-point Likert scale (1 = never to 5 = always), rating how often they engaged in the behaviors listed on Facebook. Sample items are: "I have deleted pictures of myself drinking" and "I exaggerate my professional accomplishments on Facebook". Reliabilities (α) ranged from .83 to .90.
NEO PI-R. We used 32 items from NEO-PI-R (Costa & McCrae, 1992) to measure three personality facets, with 8 items each: deliberation (α = .85; the extent to which individuals think through decisions, e.g., "I rarely make hasty decisions"), impulsiveness (α = .85; individuals' level of self-control, e.g., "I seldom give in to my impulses"), and modesty (α = .83; the extent to which individuals are humble, e.g., "I would rather praise others than be praised myself"). Participants rated each item on a 7-point Likert scale (1 = strongly agree to 7 = strongly disagree).
Balanced Inventory Desirable Responding. We used the 16 items for the Balanced Inventory of Desirable Responding (BIDR-16; Hart et al., 2015) to measure self-deceptive enhancement (8 items, α = .82, e.g., "I am very confident of my judgments") and impression management (8 items, α = .81, e.g., "I don't gossip about other people's business"). It asks participants to report how true each statement is of themselves on a 7-point Likert scale (1 = not true to 7 = very true).
General Impression Management. Participants completed a shortened 4-item version (α = .92) of the General Impression Management Scale (IMSS; Bolino & Turnley, 1999), and were asked to report how often they engaged in the listed behaviors on social media (e.g., "Make people aware of your talents or qualifications") on a 5-point Likert scale (1 = never behave this way to 5 = often behave this way).

Results and Discussion
Factor Structure. In Study 1, Hypotheses 1, 2a, 2c, 2d, 3b, and 4b were tested. Descriptive statistics and correlations for each measure are reported in Table 1.
An exploratory factor analysis (EFA) using SPSS, with a principal axis extraction and Promax (oblique rotation) method was conducted to test Hypothesis 1. The initial results supported a three-factor model, but based on Hinkin's (1998) decision rules, nine items were deleted due to low factor loadings (i.e., lower than .40) or high item cross-loadings (i.e., larger than half of the loading estimate on the appropriate factor). In support of Hypothesis 1, a second EFA with the remaining 18 items resulted in a final model with a clear three-factor structure (defensive, assertive deceptive, and assertive honest IM) explaining a total of 55.18% of variance, with good-to-excellent factor loadings (between .52 and .89), small cross-loadings, and high internal consistency reliabilities (both Cronbach's alpha and McDonald's omega coefficients between .83 and .90). Table  2 presents all factor loadings, variance explained, reliability, and means and standard deviations for each item and factor 1 .

Convergent and Discriminant Validity.
To explore the nomonological network of the FIMS and test Hypothesis 2a, c and d, which predicted that the FIMS would be positively related to behavioral measures of IM but only weakly correlated with trait-IM measures, we examined the correlation between our three IM factors and the general IM and trait-based IM measures. In line with Hypothesis 2a, the general IM measure was positively associated with all three FIMS factors, with moderately-strong correlations (r ranging from .26 to .58, p < .01). Such relationships with an established (but broader) IM measure provide initial evidence for the convergent validity of the FIMS. Importantly, these correlations suggest that the two measures are not redundant and that the FIMS captures different dimensions of IM (i.e., specific to job seekers' behaviors on social media).
Next, we found only weak negative correlations between deceptive self-enhancement and both defensive IM (r = -.11, p < .05) and assertive deceptive IM (r = -.10, p < .05), but no significant correlation with assertive honest IM (r = .09, p = .08). Similar patterns were found for traitbased IM, with small correlations with both defensive IM (r = -.16, p < .01) and assertive deceptive IM (r = -.19, p < .01) and no correlation with assertive honest IM (r = -.07, p = .16). Consistent with Hypotheses 2c-d, the weak and negative correlations between the FIMS and the two sub facets of the BIDR confirm the trait versus state-based distinctions between the measures (i.e., FIMS capturing IM behaviors whereas the BIDR capturing stable traits). Such small correlations are also similar to those found in interview IM research (e.g., Levashina & Campion, 2007).
We also used the Fisher r-to-z transformation to further test whether the correlations between our three FIMS 1 We conducted an alternative set of analyses where we randomly split Study 1 data into two subsamples: Subsample 1 (N=216) was used to conduct an EFA, whereas subsample 2 (N=215) was used to conduct a CFA. The EFA findings were almost identical to those reported in Table 2 (i.e., the same three-factor structure emerged, with similar loadings/cross-loadings, similar internal consistencies, etc.). The CFA results also largely confirmed the three-factor structure (Χ2/df = 2.87, RMSEA 90% CI = .08-.10, CFI = .89). Detailed results are provided in the OSF online supplement: https://osf.io/a2jyx/?view_ only=249d2b6247d942cc8ff1535e565a7470.  To test Hypotheses 3b and 4b, which predicted that the FIMS would be related to the personality facets of modesty and deliberation, the correlations between the FIMS and the NEO sub-facets were inspected. Modesty was negatively correlated with all three FIM factors: defensive IM (r = -.12, p < .05), assertive deceptive IM (r = -.31, p < .01), and assertive honest IM (r = -.29, p < .01), supporting Hypothesis 3b. Deliberation was negatively correlated with both assertive deceptive IM (r = -.20, p < .01) and defensive IM (r = -.16, p < .01), but unrelated to assertive honest IM (r = -.03, p = .55), which provides partial support for Hypothesis 4b. In addition, impulsivity was positively correlated with only the defensive IM factor (r = .10, p < .05).

ReseaRch aRticles
Overall, these results are consistent with Bourdage et al.'s (2015;2018) findings that individuals high in Honesty-Humility and Conscientiousness are less likely to engage in "negative" or deceptive IM practices. This supports the idea that job seekers high in modesty, (characterized by humility and a self-effacing nature; Costa & McCrae, 1992) are less prone to bragging about their achievements on Facebook or engaging in deceptive activities (e.g., posting false images). Job seekers low in modesty believe that they are superior to others and may be perceived as conceited and arrogant (Costa & McCrae, 1992), potentially leading to more IM on social media to support their self-image. Our findings also suggest that job seekers high in deliberation may be weighing the cost of engaging in IM practices on Facebook, particularly deceptive IM and defensive IM, and are more careful and discriminant in how they apply IM tactics. Finally, our results indicate that job seekers high in impulsivity are more likely to engage in defensive IM on Facebook. According to Costa and McCrae (1992), such individuals struggle to resist their cravings and urges. They may engage in actions and activities on Facebook (e.g., posting pictures of themselves drinking or making derogative comments) that could be considered "red flags" by employers (e.g., Hartwell & Campion, 2020), later requiring them to use tactics to repair their image.

STUDY 2 Sample
A total of 371 MTurk participants were recruited. We eliminated participants for failing the screening for country (i.e., residents of U.S. and Canada only), not currently looking for a job, not being regular Facebook users, failing one  Exploratory Factor Analysis of the attention checks mentioned in Study 1, or potential "bots." The final sample included 166 U.S. MTurk users with an average age of 33.8 (SD = 10.0), 42% were female, 79% were Caucasian, and 66% were employed.

General Facebook Self-Presentation.
Facebook self-presentation was assessed through 16 items (Rosenberg & Egbert, 2011), which included three factors: damage control (α = .88; e.g., "I apologize on Facebook when I have done something wrong"), self-promotion (α = .90; e.g., "I tell others about my positive qualities on Facebook"), and role model (α = .80; e.g., "I try to set an example for others to follow on Facebook"). Participants were asked to reflect on their behaviors and indicate how often they have engaged in these behaviors from 1 (never) to 5 (every time).
Job-Search Self-Efficacy. We used the 10-item JSSE-O scale (α = 95; Saks et al., 2015). Participants were asked to indicate their confidence regarding outcomes from their current job search (e.g., "be invited to job interviews," "be successful in your job search") from 1 (not at all confident) to 5 (totally confident).
Job offers. Participants were asked to report how many jobs they have recently applied to, as well as the number of job offers they received. The latter was also used as our measure of job search success.

Results and Discussion
Factor Structure and IM Use. In Study 2, Hypotheses 1, 2b, 3a, 4a, 5, 6, and 7 were tested. Descriptive statistics and correlations between key study variables can be found in Table 3.
A confirmatory factor analysis showed that a three-factor structure (defensive, assertive deceptive, and assertive honest) was associated with superior fit with the data than a one-or two-factor structure, with χ 2 /df = 2.36, CFI = .91, TLI = .90, and RMSEA 90% CI = .08-.11 (see Table 4 for detailed fit indices). These findings provide additional support to Hypothesis 1. We also conducted an additional six-factor CFA (i.e., three FIMS and three Facebook Self-Presentation Scale). This six-factor model showed strong factor loadings and generally acceptable fit indices (e.g., χ 2 /df = 1.93, CFI = .88, RMSEA 90% CI = .07-.08). In addition, the observed latent factor covariances (between the FIMS and Facebook Self-Presentation Scale factors) were similar to the correlations reported in Table 4 and discussed below. 2 Out of the three FIM factors, job seekers engaged in more assertive honest tactics (M = 2.68, SD = 1.08), followed by defensive tactics (M = 2.54, SD = 1.56), with assertive deceptive being the least common type of tactic (M = 1.56, SD = .75).

Convergent and Discriminant Validity.
To test Hypothesis 2c, which predicted that the FIMS would be related to the general Facebook self-presentation, we examined the correlations between the subfacets of each scale. The three FIMS factors showed good convergent validity with the three Facebook self-presentation factors, ranging from r = .26, p < .001 to r = .77, p < .001, supporting Hypothesis 2c. Although the two scales are related, the majority of the correlations are moderate, with the exception of the Facebook self-promotion factor and the FIMS assertive honest factor (r = .77) having the highest correlation. These results provide further support that the General Facebook Self-Presentation scale and the FIMS are likely measuring different aspects of IM.

FIMS Personality Antecedents.
To test Hypotheses 3a, 4a, and 5, which predicted that the FIMS would be related to Honesty-Humility, Conscientiousness, and Extraversion, the correlations between the FIMS and the HEXACO factors were inspected. All three types of IM were moderately correlated (r ranging between -.20 to -.44, p < .01) with Honesty-Humility. This indicates that job seekers higher on Honesty-Humility were less likely to engage in IM on Facebook, supporting Hypothesis 3a. Conscientiousness was only significantly (and negatively) correlated with assertive deceptive IM. This partly aligned with findings from Study 1, and only partially supports Hypothesis 4a. Thus, defensive IM appears to be more strongly associated with the specific facet of deliberation than Conscientiousness more generally. Additionally, those higher on Extraversion were more likely to engage in honest assertive tactics (r = .27, p < .001), but there was no relationship with the other two IM types, partially supporting Hypothesis 5. Finally, although the other five personality factors were correlated to at least one (or two in the case of openness) of the FIMS factors, Emotionality was not significantly correlated with any. This finding is aligned with Bourdage et al.'s (2015) results, which found emotionality correlated with only one of their IM tactics (i.e. supplication), which is not part of the FIMS.
Potential FIMS Outcomes. We examined Hypothesis 6 using the correlations in Table 3. Results suggest that both defensive IM (r = .16, p = .04) and assertive honest IM (r = .32, p < .001) were positively associated with jobsearch self-efficacy, but assertive deceptive IM was not (r = -.00, p = .96). Moreover, only assertive honest IM (r = .24, p < .001) was positively associated with the number of job offers received, whereas assertive deceptive IM (r = .15, p = .06) and defensive IM (r = .06, p = .44) were not. These findings provide partial support for Hypotheses 6a and b.
Hierarchical regression analyses were conducted to test Hypothesis 7, which predicted that the FIMS would be related to job-search outcomes (job-search self-efficacy and number of job offers received) over and above the general Facebook Impression Management scale (Tables 5-6). We included age, ethnicity, education, and employment as control variables in Step 1. For the outcome job offers received, we also included number of jobs applied for as a control variable in Step 1. General Facebook self-presentation was added in Step 2. The three FIMS measures were added in Step 3. Results revealed that FIMS accounted for incremental variance in job-search self-efficacy, over and above what was explained by Facebook self-presentation (ΔR 2 = .06), and this change was significant, supporting Hypothesis 7a. Specifically, in Step 3, assertive honest IM tactics were pos-

Personnel Assessment And decisions
Job seekers' imPression mAnAgement on FAcebook     Note. CFA Analysis conducted using JAMOVI 0.9.6.7 solid and current version. a Factor 1 = defensive IM; Factor 2 = assertive deceptive IM and assertive honest IM combined. **p ≤ .001 Multiple Regression for Job Search Self-Efficacy

Predictor
Step 1 Step 2 Step  itively (b = .24, SE = .11, p = .04) and assertive deceptive tactics negatively (b = -.29, SE = .12, p = .02) associated with job-search self-efficacy. FIMS also accounted for incremental variance in job offers received, beyond what was explained by Facebook self-presentation (ΔR 2 = .03); however, this change failed to reach significance, thus not supporting Hypothesis 7b. Assertive honest IM was positively (but not significantly) associated with job offers received in Step 3 (b = .34, SE = .20, p = .10). Assertive deceptive and defensive IM were unrelated to job offers.

Theoretical and Practical Contributions
The purpose of the present study was to (a) develop and validate a measure of job seekers' IM on social media, (b) explore the types of tactics job seekers engage in the most, (c) investigate the personality antecedents of IM tactics, and (d) examine whether the use of IM tactics is positively associated with job-search outcomes.
Both an EFA in Study 1 and CFA in Study 2 confirmed a three-factor structure for the FIMS, with defensive, assertive deceptive, and assertive honest IM. Although Facebook IM tactics are relatively common among job seekers, some types of IM tactics are more prevalent than others. Out of the three categories of IM tactics, participants engaged in more assertive honest and defensive tactics with assertive deceptive tactics being the least common. This pattern of IM use by applicants on social media is consistent with the prevalence of both honest and deceptive IM behaviors in job interviews (e.g., Bourdage et al., 2018;Melchers et al., 2020). Interestingly, we observed slightly higher use of all three types of IM in our second (vs. first) study, perhaps because of the more restrictive screening criteria used. Additionally, we also observed varying base rates for behaviors within the three types of IM tactics. For example, within the assertive honest IM tactics, job seekers were more likely to post about their professional goals or experiences than about personal accomplishments.
Convergent and discriminant validity analyses demonstrated that our three Facebook IM tactics were related to measures of general IM behaviors and general Facebook self-presentation. However, correlations were moderate, supporting the value of the FIMS as a more specific measure of job seekers' IM behaviors online. The only exception is the strong (.77) correlation between our assertive honest factor and the Facebook self-promotion factor. Although this suggests some conceptual overlap, the FIMS items largely focus on posts around professional experiences and accomplishments, whereas the self-promotion items are more general (i.e., about doing "well at tasks" or doing "positive things"). FIMS relationships with trait IM (i.e., self-deceptive enhancement and trait-based impression management) were much smaller. Additionally, both the FIMS and the general Facebook self-presentation scale were positively correlated with job-search self-efficacy and the number of jobs offers received. Yet, regression results revealed that the FIMS explains unique variance in jobsearch self-efficacy beyond Facebook self-presentation. These results demonstrate that although these two scales are related, they measure two independent constructs.
Relationships between personality traits and the use of IM tactics with our FIMS measure were similar to those observed in the workplace or in job interviews (e.g. Bourdage et al., 2015;2018;Melchers et al., 2020). Specifically, individuals high in Honesty-Humility or modesty were less likely to engage in IM on Facebook. Conversely, individuals high on Extraversion were more likely to engage in IM tactics, especially honest tactics. Last, we found that the use of IM on Facebook is associated with job-search outcomes. For instance, assertive honest tactics were positively related to job seekers confidence during their current job search and to job offers received (in correlations but not in regressions). These results demonstrate that Facebook IM tactics, specifically assertive honest tactics (e.g., posting positive statements about one's personal/professional qualities), are associated with job seekers confidence throughout the job search and potentially related to the chance of receiving job offers, which aligns with interview IM research (Barrick et al., 2009;Bourdage et al., 2018).

Limitations and Future Research Directions
This research has some limitations. First, both studies relied on MTurk samples. Although previous studies have shown that MTurkers provide reliable data (e.g., Highhouse et al., 2017), there could be some concerns with the generalizability of the findings. For instance, even though our studies inclusion criteria required participants to be active job seekers, it is possible that some MTurkers were not truthful with their reported answers. Future research should endeavor to include actual applicants in order to further confirm our findings. Second, 70% of our participants in Study 2 reported having received a job offer in the last year, suggesting that we have perhaps oversampled individuals who were successful in their job search.
Third, our data were cross-sectional and thus could be impacted by common method variance bias. We conducted several additional analyses that suggested that the significant relationships between assertive honest IM (or defensive IM) and job search self-efficacy in Study 2 were not (or at least not strongly) impacted by common method variance. 3 Yet, our findings should be replicated with longitudinal approaches, or using different sources of data. Fourth, the majority of the participants were from the United States. Therefore, further validation of the FIMS measure and rep-lication of our findings should include more diverse/international samples.
Fifth, we only examined a very distal objective outcome with the number of job offers received. However, cybervetting can be used as a screening device early in the selection process (e.g., before applicants are invited for interviews or tests) or as alternative to a background check closer to the end of the selection process (Berkelaar, 2017). Therefore, engaging in IM on social media might theoretically be associated with getting more job interview invitations (when cybervetting is used for screening) and/or more job offers (when used as background check and indirectly for screening). But the latter outcome is also impacted by how well job applicants perform in the selection process. As such, future studies could examine more proximal outcomes, such as the number of invitations to initial screening tests or job interviews.
Finally, our measure is focused exclusively on IM tactics used on Facebook. However, it is likely that the frequency of several behaviors captured by the FIMS vary by social media platform. For instance, most of the defensive IM behaviors included in our measure involve removing content (e.g., posts, comments, pictures) from one's profile that might be perceived negatively by potential employers. Such a behavior can be seen as strategic and effective because hiring managers do indeed focus on negative elements when assessing job applicants' social media (e.g., Hartwell & Campion, 2020). However, such potentially problematic posts are more likely to appear on personal social media (e.g., Facebook, Instagram, Twitter) than professional ones (e.g., LinkedIn) in the first place. As such, defensive IM tactics are likely more prevalent on personal social media. Nonetheless, future research should consider the implications of IM on other social media platforms such as Twitter, Instagram, and LinkedIn. The parameters and primary functions of these platforms may influence the types and frequency of IM behaviors used. Much like the differential role played by honest and deceptive IM tactics in the interviewing process (Bourdage et al., 2018;Levashina & Campion, 2007), employers would benefit from an understanding of the potential different IM tactics used on various platforms.

Conclusion
Overall, the present research provides initial empirical support for a multidimensional measure of job seekers' Facebook IM. This measure will help researchers to investigate IM from both applied and theoretical perspectives, thus furthering our knowledge of applicant IM. This research not only provides further support for the associations between personality and IM use but extends it to the use of IM in job search and the realm of social media. In conclusion, the current study will hopefully inspire further theoretical and applied research in this area.