Social, Emotional, and Behavioral (SEB) Skills in the Workplace

Despite the growing attention towards SEB skills, this literature has long encountered challenges such as fragmentation, terminology confusion, and a lack of consensus on how to define and assess these skills (Abrahams et al., 2019; Napolitano et al., 2021; Olderbak & Wilhelm, 2020). With the aim to bring order to chaos, Soto and colleagues (2022) proposed a framework for integrating the conceptualizations and taxonomies of SEB skills. They defined SEB skills as functional capacities that reflect how well an individual can enact a particular behavior when the situation calls for it. This definition explicitly distinguishes SEB skills from personality traits, which reflect how frequently an individual tends to perform the behavior, averaged across situations (John & De Fruyt, 2015; Shiner, 2021). The definitional distinction between traits and skills implies that the level of skills and the level of traits do not always go together. For example, an individual may be shy in general (low trait) but is capable of expressing ideas and actively engaging with others when selected as a group leader (high skill). On the contrary, an individual may enjoy spending time in a crowd and sharing their thoughts and feelings (high trait) but may have poor social skills to maintain high-quality social interactions (low skill).

Soto and colleagues’ (2022) integrative framework includes five general skill domains: (1) self-management skills, (2) social engagement skills, (3) cooperation skills, (4) emotional resilience skills, and (5) innovation skills. Specifically, self-management skills are defined as capacities used to pursue goals and complete tasks, social engagement skills as capacities used to engage with others effectively, cooperation skills as capacities used to maintain positive interpersonal relationships, emotional resilience skills as capacities used to regulate emotions, and innovation skills as capacities used to generate creative ideas. Prior research has emphasized the conceptual resemblance between SEB skills and the Big Five personality traits (Kankaraš & Suárez-Álvarez, 2019; Napolitano et al., 2021; Primi et al., 2021). This resemblance is not surprising, considering that personality traits serve as foundational elements of SEB skills (Hoekstra & Van Sluijs, 2003). Specifically, the SEB skill domain of self-management captures the capabilities most relevant to task completion and achievement, thus resembling conscientiousness (Heckman & Kautz, 2012); social engagement skills capture the key dimensions of effective communication, thus resembling extraversion (DeYoung et al., 2013); cooperation skills captures the most prominent aspects of positive interpersonal relationship, thus resembling agreeableness (McCrae & Costa, 1989); emotional resilience skills captures the capacities of managing one’s emotional life, thus resembling emotional stability (vs. neuroticism; Diener et al., 2003); finally, innovation skills represent capabilities oriented toward learning and exploring, thus resembling openness to experience (Wilmot & Ones, 2019).

In conjunction with their integrative framework, Soto et al. (2022) further introduced the Behavioral, Emotional, and Social Skills Inventory (BESSI) as a novel assessment tool for SEB skills. Unlike trait inventories where respondents assess the accuracy of descriptive statements reflecting their typical behavioral patterns (Kankaraš & Suárez-Álvarez, 2019), each item in BESSI describes a specific skill, and respondents assess how well they can perform that skill. In five studies with multiple samples of students and adults, Soto et al. (2022) found that BESSI displayed excellent psychometric properties.

Overall, BESSI holds promise as a framework for integrating research on SEB skills. However, it is important to note that Soto et al. (2022) and subsequent studies that adopted the BESSI framework (e.g., Lechner et al., 2022; Sewell et al., 2023; Soto et al., 2023, 2024) primarily focused on adolescence and student samples within a Western context. Consequently, the utility of this new framework in other developmental stages, such as workforce professionals, remains unclear. Examining SEB skills in the workplace is crucial, since work-related activities often demand the utilization of higher levels of these skills. This is likely due to the complex nature of work tasks, which require a high level of collaboration, professionalism, and resilience, with a focus not only on personal growth but also on the growth of the organization (Kuron et al., 2015). For instance, students are typically expected to work on tasks independently and receive individual evaluations, while employees are often accustomed to cooperative work with colleagues and collective evaluations (Wendlandt & Rochlen, 2008).

In the present study, we focused on five key workplace outcomes, including OCB, counterproductive work behavior (CWB), objective measures of job performance, job satisfaction, and relationship quality with coworkers. They represent some of the most important and widely studied outcomes in organizational research (e.g., Berry et al., 2007; Judge et al., 2017; Ocampo et al., 2018; Settoon & Mossholder, 2002; Viswesvaran, 2002). After all, job performance, citizenship behaviors, and counterproductive work behaviors directly impact the overall performance of an organization; job satisfaction is closely linked to employees’ turnover intention and mental well-being; relationship quality with coworkers reflects the social dynamics at work and plays a crucial role in employees’ interpersonal connections.

Existing research has shown that trait-like social, emotional, and behavioral characteristics are related to a broad range of consequential work outcomes. For example, within the Big Five framework, conscientiousness positively predicts all forms of work role performance (Dudley et al., 2006; Hogan & Holland, 2003; O’Neill & Allen, 2011). Extraversion and agreeableness positively predict citizenship behaviors and quality of interpersonal relationship (Chiaburu et al., 2011; Connelly et al., 2022; Vater & Schröder-Abé, 2015). Finally, emotional stability (vs. neuroticism) and openness to experience are related to job satisfaction and counterproductive work behavior (Bowling et al., 2011; Judge et al., 2002; M. Mount et al., 2006; Swider & Zimmerman, 2010; Zhang et al., 2020). Within the BESSI framework, we speculate that SEB skills are also linked to similar work-related outcomes, given they share similar behavioral referents as the Big Five traits.

Research question 1 (RQ1): Are SEB skills related to (a) organizational citizenship behaviors; (b) counterproductive work behaviors; (c) job performance; (d) job satisfaction; and (e) relationship quality?

Indeed, SEB skills are related to Big Five personality traits, but they are by no means equivalent. Personality traits refer to enduring patterns of thinking, feeling, and behavior that remain relatively stable over time and across situation (Fleeson et al., 2015). As such, they represent characteristic tendencies, reflecting what an individual typically thinks, feels, and does across different contexts. By comparison, SEB skills within the BESSI framework are defined as functional capacities that represent what an individual is capable of doing in particular situations (Napolitano et al., 2021). In other words, instead of reflecting characteristic tendencies across situations, SEB skills act as individuals’ psychological toolbox that can be brought out or put away as needed by different situations (Soto et al., 2022). Thus, we further speculate that the SEB skill-outcomes relations noted in RQ1 would remain robust even after controlling for the effect of personality traits.

Research question 2 (RQ2): Do SEB skills provide incremental validity over the Big Five personality traits in predicting (a) organizational citizenship behaviors; (b) counterproductive work behaviors; (c) job performance; (d) job satisfaction; and (e) relationship quality?

Given that one’s skill levels do not always correspond to their trait levels, another important question arises: Do individuals with high skills and low traits, or those with low skills and high traits, perform similarly to individuals possessing both high skills and high traits? In other words, how do varying levels of skills and traits interact to affect behavior and outcomes? To answer this question, we draw from the motives as primary perspective of personality, which reconceptualizes personality traits as need-like constructs that motivate individuals to react in a certain way to circumscribed situational cues (Hudson & Roberts, 2014; McCabe & Fleeson, 2012). As an example, an individual’s level of conscientiousness essentially can be regarded as their desire to be self-disciplined or to pursue goals.

There are two models regarding how skills and motivational factors such as personality traits can together affect behaviors and outcomes. The first is an additive model, in which the effects of skills and motivational factors on outcomes are independent and compensatory (M. K. Mount et al., 1999; Sackett et al., 1998). In other words, individuals’ levels of motivation will not affect the relationship between skills and outcomes, and those with a low level of motivation can compensate for this deficit by possessing high levels of skills. Soto and colleagues’ descriptions of SEB skills and personality traits resemble this additive model. In particular, the vignette they included to differentiate SEB skills from traits describes a student who is typically shy but can take on the role of a leader when needed (Soto et al., 2021, 2024), suggesting that high levels of social engagement skills possessed by the student can compensate for a low level of motivation for engaging with others and may lead to positive outcomes.

The second model is related to a longstanding belief that skills and motivational factors interact to influence behaviors and outcomes (Vroom, 1964). This multiplicative model predicts that individuals will demonstrate the highest levels of positive outcomes when they have high skills and high motivation. Yet, when individuals possess low levels of motivation, they will demonstrate similarly low levels of positive outcomes regardless of their skill levels. In other words, skills and motivation are non-compensatory in this model. Several empirical studies on the relationship between cognitive abilities, motivation, and performance supported the multiplicative model (Atit et al., 2020; Harris-Watson et al., 2022; Perry et al., 2010).

As the two competing models regarding the effects of SEB skills and personality traits in a workplace setting have not been tested before, we examined the following research question.

Research Question 3 (RQ3): Do SEB skills interact with the Big Five traits in predicting (a) organizational citizenship behaviors; (b) counterproductive work behaviors; (c) job performance; (d) job satisfaction; and (e) relationship quality, such that these skill-outcomes relationship are stronger when traits are higher?

Overall, the main purpose of this study is to examine the role of SEB skills under the BESSI framework in a work setting by investigating (1) whether the five domains of BESSI predict important work-related outcomes, (2) whether they provide unique information beyond personality traits, and (3) whether they interact with personality traits in predicting the outcomes. We attempted to answer these questions by analyzing data from a sample of real estate agents in China.

We recruited full-time agents working in a real estate company in mainland China to participate in this study. This company was considered a suitable setting for the current research, as it involves field sales tasks that require high levels of persuasive effort, perseverance, and composure under pressure, all of which have a strong tie to personal qualities like traits and skills (Wihler et al., 2017). Given that there is no prior data on the effect size of SEB skills on workplace outcomes, we were unable to perform an informative power analysis to determine the appropriate sample size. Therefore, we choose to err on the side of large sample size by reaching out to all agents within this organization. Specifically, we circulated the study information with help from the Human Resource department of the company, and all participants all received a separate form, in which we explained the research purpose of the study. Also, in the instruction we assured the confidentiality of their survey responses to reduce the social desirability effect. After removing inattentive respondents who finished the questionnaires too quickly (under 5 minutes) or failed to answer 90% of the attention check questions correctly, there are 2,992 valid responses left (29% female, 71% male; $Mage$= 30.25, SD = 6.24; 75.8% holding a professional or bachelor’s degree). They all resided in China and had an average of 3.40 years of work experience (SD = 3.08) in the real estate industry. A post-hoc power analysis revealed that our valid sample size can afford us a power of 80% to detect a correlation of .051, and thus we considered our study well-powered.

We translated the measures from English to Chinese following Brislin’s (1980) translation-back translation procedure, except for the BFI-2, whose Chinese version was already validated by Zhang et al. (2022). Other scales were first translated to Chinese by a senior doctoral student in psychology who is proficient in Chinese and English and has the knowledge of personality and assessment. Another doctoral student with similar backgrounds then translated the measures back to Chinese, after which the two students reviewed both English versions and identified and adjusted the translated Chinese version accordingly. All other authors on the paper thoughtfully evaluated the back translated items and agreed on the accuracy of translation, which were then administered on the recruited samples. All items were responded on a 5-point scale unless noted otherwise.

We decided to measure participants’ self-concepts of SEB skills with the 45-item short form of the BESSI (BESSI-45; Soto et al., 2022) as opposed to its complete version comprising 192 items. The primary reason behind this decision is the efficiency of BESSI-45, which exhibits comparable psychometric properties to the full form (Soto et al., 2024), while requires significantly less time to complete. Considering that participants were constrained by time commitment in their work, such a short version emerges as the preferred option. Each SEB skills were measured by 9 items on the survey, and participants were asked to indicate the level of expertise with which they could perform each item on a scale from 1 = not at all well to 5 = extremely well. An example item for social engagement skills is “communicate my thoughts and feelings to other people”. In the present sample, McDonald’s omegas for the 9-item scales are .93 for self-management, .94 for social engagement, .95 for cooperation, .96 for emotional resilience, and .93 for innovation. All reliability estimates were above .70 and thus were deemed acceptable for research purposes.

The Big Five personality traits were measured using the Chinese version of the 60-item Big Five Inventory-2 (Zhang et al., 2022). Participants were asked to rate the items on an agreement scale ranging from 1 = strongly disagree to 5 = strongly agree. McDonald’s omegas for the 12-item scales are .86 for extraversion, .84 for agreeableness, .87 for conscientiousness, .88 for neuroticism/emotional stability, and .86 for openness to experience.

To measure OCB, we adapted the 10-item version of the Organizational Citizenship Behavior Checklist (OCB-C) created to avoid overlapping items (Spector et al., 2010). An example item is “I volunteered to attend meetings or work on committees on own time”, and participants were asked to indicate how often they engaged in such OCBs (1 = never to 5 = every day). McDonald’s omega is .82.

The CWB measure was adapted from the 10-item Counterproductive Work Behavior Checklist (CWB-C) by Spector et al. (2010). Its items are phrases describing specific counterproductive behaviors, such as “purposely wasted the employer’s materials/supplies”. Similar to OCB, participants rated how often they perform such behaviors on a scale from 1 = never to 5 = every day. McDonald’s omega is .73.

To objectively assess job performance, we used two measures from the company’s end-of-year sales performance data: one that recorded the total number of properties each agent had shown to customers, and the other that recorded the total number of properties each agent had successfully sold across 2021. We labeled them show performance and sales performance, respectively. We then divided each agent’s show and sales performance by the number of months they were employed in the company to find the average monthly performance, because in our sample some agents joined at different times within 2021. Given that the demand and pricing of housing tend to vary greatly across different cities in mainland China, we further performed within-city centering on both performance measures to remove the city effect.

Job satisfaction was measured with one item, “In general, how satisfied were you with your current job”, ranging from 1 = very dissatisfied to 5 = very satisfied. One-item measure for job satisfaction is considered acceptable, especially when space on our questionnaires is limited and the focus of this study is not on different facets of job satisfaction (Matthews et al., 2022).

We measured participants’ relationship quality with coworkers with an adapted four-item scale by Faraj and Sproull (2000). A sample item is “My colleagues and I share special knowledge and expertise with each other”, and participants were asked to indicate the extent to which they agreed with the items (1 = strongly disagree to 5 = strongly agree). McDonald’s omega is .86.

In the survey battery, we also included measures for transformational leadership of supervisors and participants’ goal orientation. They were primarily of interest to the real estate company from which we collected data, and we opted not to utilize them in the analysis as we did not find them relevant to our focal research questions. Besides this, we confirm that all conditions, data exclusions, and procedures for determining sample sizes have been reported in the paper. This study was not pre-registered.

As this is the first time the BESSI has been applied in a Chinese population, we first examined its reliability, factorial validity, and correlations with the Big Five personality factors to establish its basic psychometric properties. Specifically, we adopted exploratory structural equation modeling (ESEM; Asparouhov & Muthén, 2009) with target rotation to examine the structure of the Chinese version of BESSI to take into consideration cross-loadings common to most measures (Hopwood & Donnellan, 2010; Zhang et al., 2022). Each indicator was modeled as categorical. Chi-square, the Comparative Fit Index (CFI), the Tucker-Lewis Index (TLI), the root mean square error of approximation (RMSEA), and the standardized root mean squared residual (SRMR) were used to evaluate model fit.

To examine the first sets of research questions, we regressed each proposed outcome on all SEB skills. To examine the second set of research questions, we performed hierarchical regression, which allows us to find whether SEB skills explain a significant amount of variance in each of the proposed outcomes after accounting for the effect of the Big Five traits. To examine the third set of research questions, we mean centered SEB skills and the Big Five traits and created their product terms, on which we then regressed each proposed outcome. In light of the increasing methodological skepticism on significance testing (Funder & Ozer, 2019) and our large sample size, we also considered change in $R2$ aside from significance tests. Specifically, we examined our RQ3 by computing the amount of change in $R2$ between the additive models and multiplicative models. Only change in $R2$ greater than 1.96 % ($f2=.02$) is considered practically meaningful (Cohen, 1988). ESEM was conducted in Mplus 8.0 (Muthén & Muthén, 2017) and all other analyses were conducted in R (version 4.1.1; R Core Team, 2021), using the lavaan package (Rosseel, 2012). The data and R codes can be accessed through this link.

The five-factor ESEM model showed an adequate fit in our sample: χ² (775) = 22792.790, p < .001, CFI = .928, TLI = .908, RMSEA = .097, and SRMR = .033. In general, the cutoff score for poor fitting models is .90 for CFI/TLI and .10 for RMSEA/SRMR (Kline, 2010; Yuan et al., 2015). As shown in Table S1, inspection of factor loading patterns shows that the majority of the items have moderate to large loadings on their intended target factor, supporting the factorial validity of the BESSI. There were seven items that had substantial cross-loading with absolute value greater than .30. As Table 1 shows, scores for the five skills also demonstrated good reliability (McDonald’s omega = .93 – .96), and moderate correlations with their corresponding personality domains (rs = .57 – .69). These findings align well with the English version of BESSI (Soto et al., 2022), where the correlations between the SEB skills and their corresponding personality domains ranged from .67 to .79, due to the conceptual resemblance between the Big Five domains and BESSI. In sum, the BESSI demonstrated good psychometric properties in the Chinese sample.

The first set of research questions examined whether the five domains of BESSI predict important work-related outcomes. We first computed Pearson correlations of the BESSI skill domains with the five proposed outcomes, as shown in Table 2,¹. We then regressed each outcome on the set of the five SEB skill domains to further test these relationships while controlling for overlap between the skill domains. Table 3 presents the standardized coefficients and confidence intervals. Most expected relationships were statistically significant²: OCB was positively related to all five SEB skill domains ($R2$ = .42). Its strongest skill predictor was cooperation skills (β = .22, CI = [.18, .26]), followed by self-management skills (β = .18, CI = [.14, .22]) and social engagement skills (β = .17, CI = [.12, .21]) that had similar effects, and the weakest predictors were emotional resilience skills (β = .10, CI = [.05, .15]) and innovation skills (β = .09, CI = [.05, .12]). CWB was best predicted ($R2$ = .10) by self-management skills (β = −.24, CI = [−.30, −.19]) and emotional resilience skills (β = −.20. CI = [−.26, −.14]), followed by social engagement skills (β = .10, CI = [.04, .15]), which had a positive relationship. Show performance was unrelated to all SEB skill domains ($R2$ = .00). Sales performance was positively related to social engagement skills (β = .10, CI = [.03, .16]) but negatively related to emotional resilience skills (β = −.08, CI = [−.15, −.00]), but the variance explained by SEB skills was tiny ($R2$ = .01). Job satisfaction was positively related to all SEB skills ($R2$ = .16) except for innovation; its strongest skill predictor was self-management skills (β = .18, CI = [.13, .23]), followed by emotional resilience skills (β = .10, CI = [.05, .16]), cooperation skills (β = .09, CI = [.03, .14]), and social engagement skills (β = .06, CI = [.01, .11]) that shared similar effects. Finally, relationship quality was positively related to all five SEB skill domains ($R2$ = .27). Its strongest skill predictor was cooperation skills (β = .28, CI = [.23, .33]), while the effects of social engagement skills (β = .09, CI = [.04, .14]), emotional resilience skills (β = .08, CI = [.03, .14]), innovation skills (β = .07, CI = [.03, .12]), and self-management skills (β = .07, CI = [.02, .12]) were relatively smaller. Note that except for show and sales performance, all other work-related outcomes were self-reported by participants themselves. Taken together, these findings suggest that participants’ self-concepts of SEB skills are mostly robustly related to their self-reported outcomes at work but not so much with objective performance.

Considering the empirical overlap between SEB skills and the Big Five personality traits, our second set of research questions investigated whether SEB skills, as measured by BESSI, provided incremental validity over the Big Five traits when predicting work outcomes. To accomplish this, we regressed each of the work-related outcomes on (a) the Big Five traits, (b) the five SEB skill domains, and (c) the combination of traits and skill domains. Table 4 shows the proportion of variance in each outcome explained by each set of predictors. The results reveal that adding the set of SEB skill domains as predictors provided a substantial increment in explained variance over the Big Five traits for OCB ($ΔR2$ = .11) and relationship quality ($ΔR2$ = .09), and a statistically significant but much smaller incremental effect for CWB ($ΔR2$ = .01) and job satisfaction ($ΔR2$ = .02). By comparison, adding the Big Five traits as predictors provided a substantial increase over the set of SEB skill domains for CWB ($ΔR2$ = .11) and job satisfaction ($ΔR2$ = .04), and a statistically significant but much smaller incremental effect for OCB ($ΔR2$ = .02) and relationship quality ($ΔR2$ = .01). As for both measures for job performance, the increments are almost negligible for both SEB skills and Big Five traits ($ΔR2$ = .00). Yet, it is noteworthy to point out here that the combination of SEB skills and the Big Five traits together only accounted for 1.0% variance in show performance and 1.3% variance in sales performance, suggesting that there may be important external factors that were not captured in this study, such as the economic recession and customers’ reluctance to go out due to quarantine policy during the pandemic. Above all, the skills do provide unique information over traits in predicting some work-related outcomes.

Our last set of research questions pertains to possible additive and multiplicative effects of SEB skill domains and their personality referents on the work outcomes. Table 5 presents the standardized regression estimates and $R2$ from the additive models and multiplicative models³. The interaction term between self-management skills and conscientiousness positively predicts OCB (β = .09, CI = [.06, .12]), job satisfaction (β = .06, CI = [.03, .10]), and relationship quality with coworkers (β = .08, CI = [.04, .11]). The interaction term between social engagement skills and extraversion positively predicts OCB (β = .04, CI = [.01, .07]), show performance (β = .04, CI = [.00, .08]), and relationship quality (β = .04, CI = [.01, .07]) and negatively predicts C (β = −.05, CI = [-.08, -.02]). The interaction term between cooperation skills and agreeableness positively predicts OCB (β = .08, CI = [.05, .10]) and job satisfaction (β = .05, CI = [.02, .09]). The interaction term between emotional resilience skills and emotional stability positively predicts OCB (β = .08, CI = [.05, .11]), show performance (β = .04, CI = [.00, .08]), job satisfaction (β = .04, CI = [.01, .08]), and relationship quality (β = .07, CI = [.04, .10]). Finally, the interaction between innovation skills and openness to experience positively predicts OCB (β = .06, CI = [.02, .09]), sales performance (β = .06, CI = [.01, .10]), and relationship quality (β = .05, CI = [.01, .08]).

However, given that our study had a large sample size, to which null hypothesis significance testing is sensitive in general, we further examined change in $R2$ between models. As shown in Table 5, the change in $R2$ from the additive models to the multiplicative models was too little to matter, with most of the interactions accounting for merely 1% of variance in the outcomes or less. Thus, despite the statistical significance, the inclusion of the skill-trait interactions resulted in almost negligible incremental validity in the predictions of work outcomes beyond the additive effects of skills and traits. Together, the current study did not find strong support for the multiplicative model and instead showed evidence for the additive model.

This paper focused on the BESSI framework of SEB skills and provided new information on its criterion space, incremental validity, and joint effects of SEB skills and traits, obtained from a large sample of 2,992 workforce professionals in China. Our findings suggest that similar to their personality referents, these skills represent another group of personal qualities that are related to positive work-related outcomes such as citizenship behaviors and job satisfaction. Moreover, such relationships remain robust after the effects of personality traits are accounted for. Finally, we found that the joint effects of SEB skills and traits are more additive than multiplicative.

The present study provides three key implications about SEB skills. First, individuals’ self-concepts of SEB skills as measured by BESSI predict self-reported work-related outcomes. Previous studies have highlighted many specific skill-outcome relationships that are robust across multiple life domains, including adolescents’ and young adults’ happiness, health, social life, and academic achievements (Soto et al., 2022, 2024). Yet, application of BESSI in work settings, where adults invest a majority of their time and energy, are rarely investigated. Using a non-western sample of workforce professionals, the present study extended BESSI’s nomological network to work settings and found that SEB skills were positively related to several crucial work-related outcomes, including job satisfaction, OCB, and relationship quality with coworkers. However, show and sales performance, the two most important performance measure for real estate agents, were largely unrelated to their SEB skills. This is surprising, as previous research linked SEB skills with objective outcomes such as school attendance and grades (Soto et al., 2023, 2024). One post hoc explanation for such null finding on objective outcomes is that our data was collected in 2021, during which the massive outbreak of COVID-19 had led the Chinese government to implement a strict quarantine policy that discouraged many unnecessary outdoor activities. In other words, we suspect that whether agents could successfully show and sell houses to customers may depend largely on the customers’ willingness to visit under the quarantine policy, which could not be controlled by the agents themselves. On the other hand, other outcomes that assess job attitudes and voluntary behaviors, such as agents’ job satisfaction and OCB, were less likely to be directly affected by clients’ behaviors. Nevertheless, these findings address calls for research to investigate SEB skills among workforce professionals (e.g., Soto et al., 2024; Yoder et al., 2020) and provide evidence that their self-concepts of SEB skills are meaningfully linked to self-reported work-related outcomes.

Second, despite the conceptual resemblance between SEB skills and traits, skills provide unique information beyond traits that matters in predicting work-related outcomes. In our study, we found that SEB skills, as measured by BESSI, demonstrated incremental validity over and above personality traits for OCB, relationship quality, and job satisfaction. Therefore, while an individual’s tendencies in thinking, feeling, and behaving certainly play a role in achieving positive outcomes (Beck & Jackson, 2022), such individuals’ capacity to selectively regulate their thoughts, emotions, and behaviors should not be overlooked because they also contribute to success and thriving at the workplace. These findings reaffirm the theoretical and empirical distinction between skills and traits, emphasizing the advantage of BESSI as a novel assessment tool that explicitly evaluates SEB skills as distinct skills rather than traits or preferences, setting it apart from previous taxonomies and inventories in the field.

Furthermore, supporting Soto et al. (2024), our findings suggest that some work outcomes may be strongly affected by consistent behavioral tendencies over time, while others may be largely determined by the psychological toolbox that can be brought out when needed. For example, the inclusion of traits accounted for an additional 11% of variance in CWB over skills, while the inclusion of skills merely explained 1% additional variance, suggesting CWB was main related to traits than skills. Conversely, the inclusion of traits accounted for merely 1% additional variance in OCB over skills, while that of skills explained an additional 11% of variance, indicating skills as the more powerful predictors of OCB. These findings resonate with the hybrid approach proposed by Primi et al. (2021) in understanding SEB skills. They argued that certain SEB dimensions such as ‘responsibility’ may be more effectively conceptualized and measured through the trait approach. On the other hand, dimensions like ‘cooperation skills’ may be better understood through a functional capacity approach.

Similarly, work-related outcomes may be categorized in such a way that some outcomes may be better conceptualized through a trait approach, thus best predicted by traits, while others may be better understood through a skill approach, thus best predicted by skills. For example, because traits reflect average tendencies, they might be especially important for such outcomes as CWB, which represent undesirable behaviors driven by aggregate hostile feelings and negative attitudes towards organizations and their members over time (Spector & Fox, 2005). On the contrary, because SEB skills are functional capacities that people enact when needed, they might be especially important for such outcomes as OCB, which involves voluntary actions that contribute positively to the work environment, such as resolving conflicts between colleagues and volunteering for additional tasks (N. P. Podsakoff et al., 2009). Effectively engaging in such OCB may require a set of SEB skills, including conflict resolution skills and self-management skills. Therefore, different degrees of incremental validity we observed in our study may be related to the nature of the work-related outcomes.

Finally, the present study addressed a foundational question concerning the joint effects of skills and traits on work-related outcomes. Our findings did not provide strong support for the multiplicative models that skills and traits interact to affect outcomes and suggested clearly that the effects of skills and traits on work-related outcomes were additive rather than multiplicative. Moderated multiple regression analyses showed that change in $R2$ from the additive models to the multiplicative models was around .01 or less. Therefore, the skill-trait interaction tends to provide very little incremental validity beyond the additive effects of skills and traits, which renders it relatively unimportant to outcomes. These findings support Soto and colleagues’ idea that skills may not be a resource that only individuals with high traits can allocate towards tasks; instead, they can compensate for low levels of traits when needed for specific situations (Soto et al., 2024). Moreover, given personality traits can be reconceptualized as motivational factors (Hudson & Roberts, 2014), our findings align with prior research that challenges multiplicative models of skills and motivation. For example, a meta-analysis by Van Iddekinge and colleagues (2018) showed that additive models of skills and motivation accounts for about 90% of explained variance while the inclusion of interaction accounts for only 9%, suggesting that skills and motivation exert independent effect on such outcomes as job performance rather than interactive effects.

The present study points to several actionable plans that organizations can consider to improve their talent management practices. First, while our findings suggested the importance of both SEB skills and personality traits in predicting crucial work outcomes as those that emphasize both. However, incorporating an additional measure into existing assessment practices could incur additional expenses. Depending on the outcomes practitioners aim to investigate, they may prioritize one set of predictors over another if time is of great importance. For certain outcomes such as job satisfaction, we found that skills and traits provided small incremental validity over each other, suggesting both were equally powerful predictors. In this case, it may be more efficient to assess either SEB skills or traits alone than to assess both. For outcomes such as CWB that reflect how a person feels about their jobs across situations, traits might emerge as stronger predictors than SEB skills, which makes it more efficient to use trait measures. On the contrary, for outcomes such as OCB and relationship quality, which requires a diverse set of skills, it may be more appropriate to assess SEB skills than traits.

Second, the general support for the additive model of skills and traits suggests that organizations should allow job applicants to compensate for lower scores on personality traits with higher scores on skills (Van Iddekinge et al., 2018). Rather than relying on a multiple cutoffs system where one must pass a personality test to be eligible for a skill test, it may be more effective to set up a cutoff based upon the combination of scores on both skills and traits. Third, if skills and personality traits interact to influence work outcomes, this would imply that workplace interventions designed to improve skills should particularly target employees with high levels of traits (Leiß & Rausch, 2022; Vignoli & Depolo, 2019). Yet, the present results challenge this idea and suggest that employees of all trait levels will benefit from skill improvement interventions.

Finally, what should a workplace intervention on SEB skills improvement look like? In our opinion, the first step in an SEB skill intervention is to assess employees’ current level of skills. Specifically, BESSI-45 can be used to measure employees’ self-concepts of each of the five SEB skill dimensions so that areas that need attention and improvement can be identified. Second, once those areas are identified, targeted training interventions can be implemented and should be tailored to the specific skills that need to be trained. For example, employees who need training on social engagement skills can work on Interpersonal Perception Task (IPT; Costanzo & Archer, 1989). This task may involve watching videos of live scenes of real-life social interaction (e.g., a man is talking to a woman) and answering questions regarding the scenes in terms of social, emotional, and nonverbal cues (e.g., what is the relationship between the man and the woman? Who is enjoying the conversation and who is not?). Following completion, participants receive feedback on their performance, along with guidance on areas for improvement. Additionally, they are provided with instructions on which social and emotional cues to attend to, as well as common biases and mistakes that can hinder accurate analysis of social situations (Riggio & Reichard, 2008). The third step is reflection, in which employees are asked to reflect on what they’ve learned in the training sessions through, for example, keeping a diary or reviewing the sessions with their managers.

The present study has several strengths, including its non-western sample of workforce professionals, the inclusion of both personality traits and SEB skills, and inclusion of self-reported and objectively measured outcomes at the workplace. However, we acknowledge several limitations that may warrant attention. First, we assessed workers’ SEB skills with a brief, self-report measure (i.e., BESSI-45). Such measures can be prone to bias. For example, because of blind spots in self-knowledge, some workers may overestimate or underestimate the extent to which they enact certain skills (Vazire & Carlson, 2011). Furthermore, using this short-version measure only allows us to conduct domain-level analysis and thus sacrifices for efficiency some valuable information or nuances that could be possibly captured by facet-level analysis instead (Soto & John, 2019). After all, there are a total of 32 skill facets organized across the five skill domains. Examining the SEB skills at the facet levels will likely allow researchers to uncover more nuances in their relations with personality traits and consequential outcomes. Therefore, we encourage future work to (1) use peer or supervisor report of participants’ SEB skills, (2) conduct facet-level analysis, and (3) further investigate whether the skill-outcome relationships would differ depending on the assessment methods, such as the comparison between the full-length measures and the brief measures or that between self-reported measures and other-reported measures.

Second, there may be concern about the social desirability effect, as the questionnaires were distributed to agents by the Human Resources department of the company. Indeed, agents might want to present favorably to their employers and thus exaggerated their levels of SEB skills and helping behaviors. However, there are several reasons why we believe our results were not likely to be substantially biased by the social desirability effect. First, at the beginning of the survey, we made it clear that this study was independently conducted by researchers from a university, and that their managers or employers have no access to their responses. Besides, the HR department was only in charge of distributing the survey link through group email and WeChat groups. No physical copies of the survey were handed out and collected. Therefore, we believe the threat of socially desirable responses should have been minimized by this procedure. Second, even if socially desirable responding were substantial, they would most likely inflate bivariate correlations. In multiple regressions where multiple predictors (similarly impacted by social desirability bias) are used to predict the outcome variable, the regression coefficients reflect the unique contribution of each predictor after controlling for all other predictors, including social desirability effect. Therefore, the multiple-regression-based results (e.g., incremental validity) were more likely to show a more conservative estimate instead of inflated results.

Relatedly, the third limitation concerns the cross-sectional design where all workers’ skills, traits, and work-related outcomes were measured at a single measurement occasion, raising concern about common method variance (CMV; P. M. Podsakoff et al., 2012). Given that a majority of our significant findings (i.e., results in Table 3) were related to self-reported predictors and outcomes, it is likely that such findings might be explained by CMV. However, as previously mentioned, CMV were controlled for in multiple regression, so the findings we observed in Table 3 – 5 were unlikely to be artefacts of CMV. That said, we encourage future research to explore ways to minimize the potential impact of CMV, such as incorporating one or more temporal lags in the study design.

The fourth limitation of this study is related to the applicability of its findings to other industries or contexts. Specifically, this study focused exclusively on real estate agents during the unprecedented challenges posed by the COVID-19 pandemic. Undoubtedly, such settings may present a unique opportunity to investigate skills and work-related outcomes. For example, agents tend to feel satisfied about their job when they possess high SEB skills, even amid times of crisis, yet their objective performance may be affected by national policies relevant to this specific period. As such, while our findings may offer valuable insights within this specific population and timeframe, caution is warranted when extrapolating conclusions to different industries or periods. Future research may develop theoretical explanations on why the relationships between SEB skills and work-related outcomes may be different depending on the nature of the work performed and the period people live in. For example, how is the effect of social engagement skills on job performance different in customer-facing work (in which interaction with customers is the key) compared to technical work?

Finally, the null findings on objective performance suggest that focusing solely on behavioral consequences (e.g., number of houses shown) to gauge performance may not be appropriate, particularly during times of crisis. This is because such consequences may to some extent depend upon factors outside of agents’ control, including the market conditions and national policy during the pandemic. Therefore, we highly encourage future researchers to incorporate metrics such as supervisors’ evaluations of agents’ communication efforts with prospective clients, which would provide a more accurate assessment of their performance.

The present study provides three key conclusions that advance our understanding of SEB skills within the BESSI framework. First, it successfully extends their criterion space by showing that self-concepts of SEB skills predict self-reported work-related outcomes, including OCB, CWB, job satisfaction, and relationship quality with coworkers. Second, it replicates previous findings that SEB skills provide unique information beyond traits in predicting these outcomes. Third, it provides evidence that the joint effects of skills and traits are additive instead of multiplicative.

Contributed to conception and design: LC, BZ, JL

Contributed to acquisition of data: JL

Contributed to analysis and interpretation of data: LC, BZ

Drafted and/or revised the article: LC, BZ

Approved the submitted version for publication: LC, BZ, JL

We declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

This research was partially supported by National Key R&D Program of China (2020YFC2003000) awarded to Dr. Jian Li.

All data and R scripts can be accessed through this link.