Method of assessing political skill for personnel selection: Development and validation of a situational judgment test

1.   Introduction

Organizations are the arena of politics. As a significant social ability, political skill is pivotal to an organization’s core competitiveness and is a significant focus of research on organizational behavior^[1]. Political skill is defined as the ability to influence others by integrating existing knowledge and experience based on an understanding of self, others, and situations^[2]. Political skill predicts numerous work outcomes, including job performance and leadership effectiveness^[3–4]. Consequently, evaluating applicants’ political skill during personnel selection is vital. Especially when jobs involve more interpersonal interactions, the effect of political skill on performance is more significant^[5].

Although researchers have emphasized the significance of accessing political skill in personnel selection, it is seldom implemented in practice due to the lack of effective assessment methods. Currently, political skill is typically measured using self-reported scales. While these self-assessments are valuable in empirical research, they lack practical validity in real-world selection scenarios. Candidates tend to provide answers that align with sociocultural norms to achieve higher scores or to conceal their weaknesses, resulting in social desirable responses^[6]. These responses distort the results and diminish the authenticity of the scores^[7–8]. As a result, self-reported measures fail to provide effective suggestions for organizational personnel decisions.

Researchers have found that using the situational judgment test (SJT) to measure interpersonal skills can mitigate social desirability bias and improve measurement validity^[9]. This insight suggests a potential approach for assessing political skill in personnel selection. SJT is a low-fidelity simulation test that evaluates a subject’s relevant knowledge and skills based on their behavioral judgments and choices in work-related scenarios^[10]. SJT contains multiple behavioral choices that are more closely related to real life. Subjects need to use their actual abilities to make judgments, thus making it more difficult for them to fake^[9]. Therefore, the primary purpose of our study is to develop a situational judgment test of political skill (SJT-PS) and to evaluate its reliability and validity through a pilot test.

The second purpose of this study was to investigate the criterion-related validity of the SJT-PS. First, to assess whether the SJT-PS can mitigate the distortion of self-reported measurements caused by social desirability, we compared the correlations between the SJT-PS and self-reported political skill with social desirability. Additionally, previous research has indicated that individuals with high political skill are more likely to form strong and stable relationships with coworkers, thereby increasing their workplace popularity^{[11, 12]}. This, in turn, can positively influence workplace outcomes such as enhancing team members’ job performance and job satisfaction^[13–15]. Therefore, we choose TMX and workplace popularity as criteria to test the criterion-related validity and incremental validity of the SJT-PS.

2.   Study 1: SJT-PS development and pilot test

2.1   Definition and measurement of political skill

Pfeffer^[2] proposed analyzing organizational behavior from a political perspective and first introduced the concept of “political skill”. Mintzberg^{[16, 17]} also noted that organizations are political arenas in which members not only intend to engage in political behaviors but also should possess the necessary skills to perform political behaviors effectively. Ferris et al.^[18] further explored and expanded the concept of political skill, defining it as the ability to understand others at work effectively and use such knowledge to influence others in ways that enhance one’s personal or organizational objectives. This conceptualization of political skill encompasses four dimensions: apparent sincerity, social astuteness, interpersonal influence, and networking ability.

Apparent sincerity enables individuals to appear authentic, genuine, and honest, which modifies how others perceive behavioral intentions. Social astuteness involves the incisive observation of others, as well as the ability to understand people, the self, and social interactions and to accurately interpret the behavior of oneself and others. Interpersonal influence incorporates a flexible, adaptive nature that allows individuals to adjust and calibrate their behavior to different and changing situations, which tends to elicit targeted and desired attitudinal and behavioral responses from others. Finally, networking ability vests individuals with the capacity to identify and develop diverse networks of relationships^[19].

In addition to developing the theoretical foundation of political skill, Ferris and colleagues developed and validated several psychometric measures. The initial unidimensional measure^[20] was soon replaced by the multidimensional political skill inventory^[18]. This four-dimensional scale has been widely used in empirical studies^[21–23]. However, there has been limited research on developing a political skill assessment specifically applicable to personnel selection scenarios.

2.2   Situational judgment test

SJT is a low-cost selection method based on situational simulation, which involves identifying a critical job-related situation in the stems, setting multiple possible behaviors as response options, and then asking subjects to choose the most effective option^{[24, 25]}. In the 1990s, Motowidlo et al.^[26] found that SJT could effectively measure individual competencies, which caused widespread concern about SJT.

The SJT-PS developed in this study is a construct-driven SJT. Traditionally, SJT has been associated with a variety of constructs and has been used to measure a wide range of job-related attributes^[27–29]. However, the heterogeneous nature of traditional SJT often results in exploratory factor analyses revealing few judgmental factors that accurately reflect specific trait dimensions and scores^[30]. In response to this issue, scholars have proposed developing construct-driven SJTs^{[31, 32]}. Construct-driven SJT is designed to assess “inherent beliefs about the general effectiveness of actions that express traits to varying degrees”, which partly reflects candidates’ trait levels^[33]. Constructed-driven SJT has been shown to predict job performance better and improve the effectiveness of personnel selection^[34]. Recent research has developed SJTs for constructs such as emotion regulation and social skills, demonstrating acceptable criterion-related validity in predicting work-related behaviors and life events^{[32, 35]}.

2.3   Process of developing the SJT-PS

The study followed the design steps proposed by Guenole and Lievens et al. for construct-driven SJT^{[33, 36]} and referenced previous construct-driven SJT development practices^{[32, 37]}.

2.3.1   Develop item stems of the SJT-PS

We drew upon Motowidlo’s suggestion of using the critical incident method to collect practical work scenarios related to political skill^[26]. We invited 28 participants (19 males, 9 females) to participate in an open-ended survey. They came from a variety of industries, were experienced, and all had managerial experience. The average age of the participants was 38 years, with an average tenure of 13 years in the workforce. Initially, we distributed materials about political skill to them and asked them to read the materials carefully. Then, we asked them to recall and write in detail about a situation they had encountered at work that required political skill. We collected 21 scenarios that belonged to the four dimensions of political skill (including 5 apparent sincerity, 6 social astuteness, 5 interpersonal influence, and 5 networking ability). We refined and standardized the descriptions of these scenarios by addressing any readability issues or inappropriate content. Through this process, we formulated a total of 21 item stems.

After identifying item stems, it is imperative to assess content validity to ensure alignment with the theoretical underpinnings of political skill. In this study, we used the item-level CVI (I-CVI)^[38] to evaluate content validity. We invited 15 EMBA students from a university in Anhui Province to be raters. Initially, we provided an overview of political skill and its four dimensions to ensure its familiarity with the theoretical framework. Subsequently, the raters were asked to assess the representativeness of each situational item in relation to its corresponding political skill dimension using a Likert-type scale ranging from 1 (strongly disagree) to 4 (strongly agree). A rating of 1 or 2 indicated that the item did not adequately represent the associated political skill dimension, while a rating of 3 or 4 signified strong alignment. Finally, we calculated the number of people with a rating of 3 or 4 divided by the total number of raters as the content validity coefficient. Considering that the consistency of evaluations among multiple experts may be influenced by random factors, it was necessary to adjust for random agreement using the K* statistic (where values greater than 0.74 indicate excellent agreement, 0.60−0.74 denote good agreement, and 0.40−0.59 indicate fair agreement)^[39]. Based on the K* values, we deleted four items with low content validity (K* values less than 0.6) and ultimately retained 17 items.

2.3.2   Develop response options for the SJT-PS

Response options for traditional SJT are derived from key event interviews. Conversely, in construct-driven SJT, psychologists are typically in charge of developing response options based on target construct theory^[31]. Response options of construct-driven SJT represent varying degrees of the construct being measured, yet all of them are valid in the given situation^[34].

In accordance with recommendations for construct-driven SJT development^{[31, 36]}, this study developed response options based on theoretical conceptualizations of political skill. We selected a group of eight graduate students of industrial and organizational psychology familiar with the construct of political skill as subject matter experts and then asked them to refine the key features of political skill’s four dimensions from the relevant literature^{[18, 40, 41]}. The response options for each item were constructed based on these features. Each item contains three response options, reflecting high, medium, and low levels of political skill.

2.3.3   Decide score procedure for the SJT-PS

The scoring method for respondents’ answers adheres to the process advocated by Bledow and Frese^[37]. The respondents were asked to indicate which of the three response options presented for each item they would perform most likely and least likely. For the most likely ratings, respondents receive +1 if they selected the response options high in political skill, 0 if they selected a response option with a medium level of political skill, and −1 if they selected a response option low in political skill. Responses for the question asking about what respondents would do least likely were scored inversely: If a response option low in political skill was selected as what would be done least likely, the score was +1; for the response option with medium levels of political skill, the score was 0; and for the response option with high levels of political skill, the score was −1. The most and least likely ratings were combined for each item, resulting in a score that could vary on a 5-point scale between −2 (low political skill) and +2 (high political skill). The respondents’ overall scale score on the SJT-PS was derived from the mean of the item scores.

2.3.4   Pilot test

Participants and procedure. A sample cohort of 150 EMBA students from a university in Anhui Province composed the sample population for testing the SJT-PS. We finally obtained 134 valid questionnaires and gave CNY 30 rewards to those participants. The average age of the participants was 41 years, with women comprising 53% of the sample. A total of 26.9%, 60.4%, and 12.7% of the participants held junior degrees or lower, bachelor’s degrees, and master’s degrees or higher, respectively. In terms of hierarchical level, 9.7% indicated that they held hierarchical levels, 21.6% held lower-level management positions, 54.5% held middle management positions, and 14.2% worked in upper management positions.

Confirmatory factor analysis. To assess the alignment of SJT-PS items with the dimensions of the political skill theoretical model, we conducted confirmatory factor analysis (CFA) using Amos 28.0. The 17 items of the SJT-PS were designated observational variables, while the four dimensions of political skill served as latent variables, forming the basis of the factor analysis model. The CFA results (as shown in Fig. 1) demonstrated that the standardized factor loading for each item on the corresponding latent variable ranged from 0.5 to 1.0, and all of them passed t tests (p < 0.001). Errors for all items were also significantly less than 0.70, indicating that the SJT-PS had adequate convergent validity.

Figure  1.  Path diagram of the confirmatory factor analysis.

DownLoad: Full-Size Img PowerPoint

We then compared the indices derived from the factor analysis with the goodness of fit. The results showed that χ²/df was 1.13, which was less than 3, the root mean square error of approximation (RMSEA) was less than 0.10, and the goodness-of-fit index (GFI) and the adjusted goodness-of-fit index (AGFI) basically reached the desired level of 0.9. In addition, the correlation indices (CFI = 0.99, NFI = 0.94) were both greater than 0.90. Finally, the parsimony goodness-of-fit index (PGFI) was 0.67, which was greater than 0.50, indicating that our model is concise. In summary, the goodness-of-fit indices were all within the acceptable range, meaning that the structure of the model was reasonable and that the SJT-PS we developed had good construct validity.

Reliability. Motowidlo et al.^[26] argued that Cronbach’s α was not suitable for evaluating the reliability of SJT and recommended the use of test-retest reliability instead. Similarly, Sackett and Lievens^[42] posited that test-retest reliability is more appropriate for SJT because the subjects who retake the test exhibit a greater degree of consistency. Therefore, this study employed test-retest reliability to assess the reliability of the SJT-PS. The SJT-PS was administered to the same subjects on two occasions, yielding 112 valid responses. The interval between the first and second waves of data collection was approximately two weeks. The correlation coefficient between the two scores was 0.925 (p < 0.01), which reached significance, indicating that the test-retest reliability of the SJT-PS was good.

In summary, the analyses indicate that the SJT-PS developed in this study has high reliability. It effectively reflects the structural dimensions of political skill, exhibiting both good reliability and validity. Figs. 2–5 provide sample SJT items for each of the four political skill dimensions.

Figure  2.  A sample SJT item of apparent sincerity.

DownLoad: Full-Size Img PowerPoint

Figure  3.  A sample SJT item of social astuteness.

DownLoad: Full-Size Img PowerPoint

Figure  4.  A sample SJT item of interpersonal influence.

DownLoad: Full-Size Img PowerPoint

Figure  5.  A sample SJT item of networking ability.

DownLoad: Full-Size Img PowerPoint

3.   Study 2: Criterion validity and incremental validity of the SJT-PS

When utilizing self-assessment to measure individual competencies for personnel selection, applicants can easily provide fake answers to meet social expectations. This practice can yield distorted results, posing challenges for organizations in selecting suitable employees^[43]. In response to this concern, some scholars have proposed that SJT may be more applicable for personnel selection. SJT necessitates candidates to make behavioral decisions by leveraging their genuine abilities within simulated scenarios, thereby mitigating the influence of socially desirable responses^[44]. To examine whether the SJT-PS is less influenced by social desirability and more applicable for personnel selection, Study 2 initially contrasted the correlations between the SJT-PS and self-reported political skill with social desirability.

Furthermore, from the point of view of practical value, one of the most important properties of an assessment is criterion-related validity, which is the extent to which the assessment can accurately predict a particular criterion. High criterion-related validity reflects the effectiveness of the measurement^[45]. In Study 2, we selected the TMX and workplace popularity as criteria to test the criterion-related validity of the SJT-PS. In addition, we also examined the incremental validity of the SJT-PS over self-reported political skill scale.

3.1   SJT-PS, self-reported political skill, and social desirability

Political skill is an important noncognitive competency. However, like many personality scales, political skill is predominantly measured using self-reported Likert scales. A major concern of many organizational stakeholders in using assessment devices based on self-assessment by applicants in applied personnel selection settings has been the potential for subjects to fake their responses influenced by social desirability^[46].

Social desirability (SD) refers to the tendency of subjects to respond to test items in a way that conforms to socially expected responses^[6] and demonstrates behaviors that conform to the norms of the dominant social culture or societal values. It can lead individuals to receive greater appreciation and acceptance from others^[47].

To mitigate the impact of social desirability on individuals and enhance the validity of measurement results, some researchers have advocated for the evaluation of noncognitive abilities through SJT^[43]. SJT centers on assessing candidates’ perceptions of response options, thereby elucidating the cognitive processes underpinning their judgment. Specifically, by revealing the thought process leading up to response judgment, SJT allows for a more thorough assessment of candidates’ characteristics, which can increase their criterion-related validity^[48]. The SJT-PS in this study is precisely an assessment developed based on the situational judgment test method. Building upon prior analyses, we believe that compared with the self-reported political skill scale, the SJT-PS can reduce social desirability response bias and is more suitable for personnel selection. In addition, if an assessment is free from social desirability bias, scores on the instrument should not be related to scores on a measure of social desirability^[49]. Based on the above analysis, we propose:

Hypothesis 1a: The self-reported political skill is positively related to social desirability.

Hypothesis 1b: The relationship between self-reported political skill and social desirability is stronger than the relationship between the SJT-PS and social desirability.

3.2   SJT-PS, self-reported political skill, and team-member exchange

With the widespread adoption of the teamwork model in contemporary enterprises, communication and interaction among team members are becoming more frequent. Interpersonal relationships among team members are crucial for organizational performance^[50]. Seers^[51] first proposed TMX based on the concept of leader-member exchange (LMX) to describe mutually beneficial interactions among team members. TMX embodies a high-quality reciprocal mechanism wherein team members can readily exchange knowledge, experience, and assistance within the team^[52].

According to social exchange theory, both participants in a social interaction engage with the expectation of receiving reciprocal benefits. By offering assistance and support to one another, a participant creates an obligation for the other party to reciprocate^[53]. This fostered the emergence of TMX. Consequently, team members tend to classify their colleagues based on their social performance within the workplace. They cultivate high-quality exchange relationships characterized by trust, respect, and support with certain coworkers while maintaining only average-level relationships with others. Individuals possessing high political skill can influence or reshape others’ perceptions by proactively building quality exchange relationships with them^[54].

First, individuals with high political skill attach importance to interpersonal interactions. They perceive social interactions as opportunities rather than burdens and believe that they can attain desired outcomes through interactions with others^[11]. Therefore, they demonstrate a proactive approach to social interactions and believe that they can excel in such situations and win the favor of others. In addition, individuals with high political skill demonstrate social astuteness. They can always positively recognize the conditions and use appropriate strategies to achieve their objectives with different individuals and circumstances^[55], which makes them highly favored by the team. Finally, individuals with high political skill possess networking ability. They are good at building and utilizing relationships that are advantageous to their careers^[18]. Individuals who have networking ability produce a profound understanding of social relationships conducive to goal achievement^[19]. They can quickly establish friendships and alliances with others who share their interests^[56]. Therefore, as teams play an increasingly important role in modern enterprises, individuals with high political skill will strive to build reciprocal relationships with other team members to gain more resources to help themselves. Given the hypothesis presented in the previous section that SJT can overcome the limitations of self-assessment, it is expected that the SJT-PS will enhance criterion-related validity. We propose:

Hypothesis 2a: The SJT-PS is positively related to TMX.

Hypothesis 2b: The correlation between the SJT-PS and TMX is stronger than the correlation between self-reported political skill and TMX.

3.3   SJT-PS, self-reported political skill, and workplace popularity

Attaining acceptance by one’s social group is a primary goal for most individuals^[57]. In the workplace, popularity is defined as “generally being accepted by one’s peers”. In other words, workplace popularity is the consensus of an individual’s peers regarding the individual’s level of social acceptance within the office, company, or work team. According to social influence theory^[58], individuals with a good network of relationships are able to receive more benefits and preferential treatment from others, which translates into their social capital. Workplace popularity can be considered a form of capital, as it can lead to more trust, respect, and obedience from others, which helps individuals achieve personal goals^[59].

This study posits that the four dimensions of political skill exert direct and significant impacts on increasing workplace popularity. First, individuals exhibiting high political skill demonstrate elevated social astuteness, enabling them to keenly observe their environment, perceive the preferences and intentions of others, and adapt accordingly, thereby fostering acceptance^[60]. Second, individuals with high political skill who exhibit sincerity are more prone to engendering the trust of others^[60], a crucial element for garnering acceptance. Furthermore, individuals with high political skill can depend on their high levels of sociability to establish dependable alliances with others^[61]. Finally, individuals with high political skill possess interpersonal influence ability, which can be utilized to cultivate a positive reputation^[62]. In summary, individuals with high political skill are more likely to gain acceptance and affection from others, making them more popular at work. Given the hypothesis presented in the previous section that SJT can overcome the limitations of self-reported scales, it is expected that the SJT-PS will enhance the criterion-related validity. We propose the following:

Hypothesis 3a: The SJT-PS is positively related to coworkers’ ratings of popularity.

Hypothesis 3b: The correlation between the SJT-PS and workplace popularity is stronger than the correlation between self-reported political skill and workplace popularity.

3.4   Method

3.4.1   Validation study sample

In Study 2, we conducted a paired study to gather data. We designed and distributed questionnaires through the Credamo platform and recruited subjects using the sample matching function. Each subject was matched with one of their coworkers. After that, we sent explanatory statements and two types of questionnaires, one for the subjects and the other for their paired coworkers. Specifically, each coworker was in charge of evaluating the subjects’ exchange relationships with other team members and their popularity at work. Subjects were required to indicate their political skill and social desirability and subsequently completed the SJT-PS questionnaire. To simulate scenarios wherein participants might falsify responses to attain higher scores, we told the subjects that we would give rewards based on their scores to increase their motivation to obtain higher scores. Finally, a total of 251 useable individual-coworker dyadic matched responses were collected. The final sample consisted of 48% males and 52% females. The subjects had an average age of 26–30 years and an average working experience of 1–5 years. In terms of job level, 38.4% were general employees, 33.2% were junior managers, 21.2% were middle managers, and 7.2% were senior managers. On average, the subjects had worked with their coworkers for three years.

3.4.2   Measures

All the scales in the current study were adapted according to China’s national conditions and modified according to expert opinions.

SJT-PS. The scale was constructed as described above. The SJT-PS contains 17 items, each with three response options. Overall, respondents’ overall scores on the SJT-PS were derived from the mean of the 17 item scores.

Self-reported political skill. The 18-point Likert-type scale^[18] was used to measure participants’ self-reported political skill. The coefficient α was 0.87, and an example item was “I spend a lot of time and effort at work networking with others”.

Social desirability. Social desirability was measured with the Chinese short version of the Marlowe–Crowne Social Desirability Scale (MCSD), which has good reliability and validity. It was modified by Wei et al.^[63] The revised scale consisted of 13 items, such as “Sometimes it is difficult for me to keep things going if I do not get encouragement from others”. The subjects were asked to rate whether the descriptions of the items corresponded to their actual situation, with “no” scoring 0 and “yes” scoring 1. The higher the score, the stronger the tendency to conform to social expectations. Three of the items were positive statements, and 10 were negative statements. The coefficient α was 0.89.

Team-member exchange. The scale for coworkers’ ratings of team-member exchange was adopted from the ten-item questionnaire developed by Seers et al.^[64] The coefficient α was 0.84. Sample items include “Team members understand his/her problems and requirements well”.

Workplace popularity. Coworker ratings of workplace popularity were measured with eight items developed by Scott and Judge^[57]. An example item was “The person for whom I am completing this survey is well known”. Participants were instructed to solicit responses from one of their coworkers. The coefficient α was 0.81.

Control variables. We controlled for employee demographics, including gender (coded 1 = male and 2 = female), age (coded 1= < 20 years; 2 = 21–25 years; 3 = 26–30 years; 4 = 31–35 years; 5 = 36–40 years; 6 = > 41 years), education (coded 1 = junior college degree or lower; 2 = bachelor’s degree; 3 = master’s degree or higher), tenure (coded 1 = < 1 year; 2 = 1–5 years; 3 = 6–10 years; 4 = > 10 years), and coworking years (coded 1 = < 1 year; 2 = 1–3 years; 3 = 4–6 years; 4 = > 6 years).

3.5   Results

3.5.1   Correlation analysis

The mean value, standard deviation, and correlation coefficient of the main variables in this study are shown in Table 1.

Table  1.  Means, standard deviations, and correlations of each variable.

Variables	1	2	3	4	5	6	7	8	9	10	11
1. Gender
2. Age	0.09
3. Education	−0.01	−0.09
4. Years of working	0.06	0.41**	−0.12*
5. Hierarchical level	0.07	0.34**	−0.03	0.23***
6. Coworking years	0.09	0.33**	−0.09	0.38***	0.25**
7. SJT-PS	−0.05	0.01	−0.04	0.14*	0.18**	0.04
8. Self-reported political skill	0.09	−0.01	0.09	0.04	0.15*	−0.03	0.17**
9. Social desirability	0.04	0.12*	0.03	0.04	0.21**	0.07	−0.03	0.31**
10. TMX	−0.01	0.05	−0.08	0.14*	0.18**	0.23**	0.37***	0.18**	0.05
11. Workplace popularity	0.05	0.08	−0.05	0.14*	0.15*	0.19**	0.38***	0.19**	0.04	0.71**
M	0.48	3.41	2.10	2.70	1.97	2.26	1.11	4.58	0.61	4.61	4.21
SD	0.50	1.19	0.51	0.90	0.94	0.83	0.36	0.51	0.31	0.73	0.56
*** p < 0.001; ** p < 0.01; * p < 0.1.

 | Show Table

DownLoad: CSV

Table 1 shows that self-reported political skill was significantly related to social desirability (r = 0.31, p < 0.01), while SJT-PS had no significant relationship with social desirability. Furthermore, the scaling ranges of the SJT-PS and the self-reported political skill scale were different in this study, so we used the coefficient of variation (CV) to compare the dispersion degree of the data from the two measurements^[64]. The data were more centralized as the CV decreased and became more decentralized as the CV increased. The CV of the SJT-PS (32%) was greater than that of the self-reported political skill score (11%). This suggested that the SJT-PS score distribution was more uniform, while the self-rating score distribution was more concentrated in higher score ranges. Based on the above analysis, we believe that compared to self-reported measurements, the SJT-PS will reduce the impact of social desirability on subjects and minimize measurement bias. This finding verifies Hypothesis 1a and 1b.

Table  3.  Regression analysis of self-reported political skill, SJT-PS, and workplace popularity.

Variables	First step	Second step		Third step
		Self-reported political skill		Self-reported political skill + SJT-PS
Gender	0.04	0.02		0.05
Age	−0.04	−0.03		0.01
Education	−0.03	−0.05		−0.03
Years of working	0.08	0.07		0.04
Hierarchical level	0.11	0.07		0.03
Coworking years	0.16*	0.18**		0.13*
Self-reported political skill		0.19**		0.13*
SJT-PS				0.32***
R²	0.06	0.10		0.19
ΔR²		0.04		0.09
F	2.63	3.64***		7.02***
ΔF		9.14***		27.89***
*** p < 0.001; ** p < 0.01; * p < 0.1.

 | Show Table

DownLoad: CSV

The correlation analysis results show that the SJT-PS were positively related to TMX (r = 0.37, p < 0.001), and their relationship was stronger than the relationship between self-reported political skill and TMX. The SJT-PS was positively related to workplace popularity (r = 0.38, p < 0.001), and this relationship was stronger than that between self-reported political skill and workplace popularity. These findings suggested that the SJT-PS had good validity and could effectively predict the relationships between individuals and team members, as well as their popularity at work. The predictive effects are all better than those of the self-reported scale, providing preliminary evidence for Hypothesis 2 and Hypothesis 3.

Finally, an individual’s gender, age, and educational background had no significant correlation with their TMX or workplace popularity. Those with more working years, higher hierarchical levels, and longer spending time with coworkers had better exchange relationships and were more popular in the workplace.

3.5.2   Regression analysis

To further explore the relationships between the SJT-PS and TMX and workplace popularity and whether the SJT-PS has incremental validity over the self-reported political skill scale, we conducted a multilayer regression analysis on the predictor variables. The results of the specific regression analysis are shown in Tables 2 and 3.

Table  2.  Regression analysis of self-reported political skill, SJT-PS, and TMX.

Variables	First step	Second step		Third step
		Self-reported political skill		Self-reported political skill + SJT-PS
Gender	−0.04	−0.065		−0.03
Age	−0.09	−0.08		−0.05
Education	−0.07	−0.09		−0.07
Years of working	0.06	0.04		0.02
Hierarchical level	0.15*	0.12		0.08
Coworking years	0.26***	0.29***		0.24***
Self-reported political skill		0.20**		0.15*
SJT-PS				0.27***
R²	0.11	0.15		0.22
ΔR²	0.11	0.04		0.07
F	5.15***	6.10***		8.35***
ΔF	5.15***	10.61***		20.56***
*** p < 0.001; ** p < 0.01; * p < 0.1.

 | Show Table

DownLoad: CSV

In the first step, we examined the impact of political skill on TMX. First, we put the control variables of gender, grade, educational background, hierarchical level, and time spent with coworkers into the first layer of the regression analysis. Second, we put self-reported political skill into the second layer of regression analysis. Self-reported political skill (β = 0.20, p < 0.01) has a significant positive impact on TMX. In the case of considering control variables, self-reported political skill explained 15% of the variance in the TMX. Third, by placing the SJT-PS on the third level of the regression analysis, it can be seen that the SJT-PS (β = 0.27, p < 0.001) has a significant positive impact on the TMX. Under the premise of controlling the variable of self-reported political skill, the SJT-PS increased the interpretation of TMX by 7%. The SJT-PS is effective and unique in predicting individuals’ TMX and has incremental validity over the self-rating method. Therefore, Hypothesis 2a and Hypothesis 2b are verified.

In the second step, we examined the impact of political skill on workplace popularity. First, we put the control variables of gender, grade, educational background, hierarchical level, and time spent with coworkers into the first layer of the regression analysis. Second, we put self-reported political skill into the second layer of regression analysis. Self-reported political skill (β = 0.19, p < 0.01) has a significant positive impact on workplace popularity. In the case of considering control variables, self-reported political skill explains 10% of workplace popularity. Third, by placing the SJT-PS on the third level of the regression analysis, it can be seen that the SJT-PS (β = 0.32, p < 0.001) has a significant positive impact on workplace popularity. Under the premise of controlling the variable of self-reported political skill, the SJT-PS increases the interpretation of workplace popularity by 9%. Therefore, Hypothesis 3a and Hypothesis 3b are verified.

4.   Conclusions

In this study, we developed a situational judgment test of political skill for personnel selection and validated that it had good reliability and validity. The results of the confirmatory factor analysis revealed that the SJT-PS closely aligned with the theoretical framework of political skill and exhibited favorable structural validity. Regression analyses confirmed that the SJT-PS had a weak correlation with social desirability and a significant positive correlation with both TMX and workplace popularity. Additionally, this study revealed that the SJT-PS had significant incremental validity when self-reported political skills were controlled for. This suggested that the SJT-PS would be less influenced by social desirability than would the self-reported political skill scale. The SJT-PS provides an effective way to assess individuals’ political skill for personnel selection.

The theoretical contribution and practical insights of this study are mainly manifested in the following three aspects. First, this study developed a political skill SJT for personnel selection, which improves the measurement of political skill. Prevailing empirical studies mainly rely on self-reported scales to assess political skill, but evaluating it in the context of selection questionnaires is subject to faking. To address this issue, the present study developed the SJT-PS and compared the correlations of the SJT-PS and self-reported political skill with social desirability. This study showed the feasibility and validity of using SJT to assess an individual’s political skill for personnel selection. In addition, using self-reported measures to gather data may lead to common method bias^[65]. Our SJT-PS could also be used in future empirical studies to minimize the problem of homophily bias.

Second, this study explored the relationship between political skill and TMX, as well as the relationship between political skill and workplace popularity. Previous studies have focused predominantly on the relationship between political skill and leader-member exchange. This emphasis stems from the pivotal role of leaders, who acting as agents of the organization, discharge managerial duties formally and informally on their behalf. They possess the authority to evaluate subordinates’ performance, provide supervision and mentorship, and adjudicate promotions. Establishing a robust LMX can help an individual obtain more resources and opportunities within the organization^[66]. However, modern enterprises have transitioned away from the traditional top-down hierarchical organizational model toward a small and flexible team-based project-oriented corporate structure^[67], where the reciprocal relationships established with team members are more important than those with leaders. However, there is little research on political skill, TMX, and workplace popularity. This study revealed that political skill had significant positive impacts on TMX and workplace popularity. The SJT-PS can better predict the two criteria than self-assessment. Our study adds the relevant literature.

The findings of this study have significant implications for enterprises engaging in personnel selection. Enterprises should not only directly use self-reported measurements from empirical research but also select more appropriate assessment methods, such as SJT, to improve the effectiveness of personnel selection. Talent is the cornerstone of enterprise development. As a simulation test, SJT offers the ability to assess candidates’ actual abilities needed for work and evaluate their future performance through behavioral judgment. Therefore, integrating SJT into personnel selection practices can ameliorate the deficiencies inherent in extant self-reported methods, thus serving as an efficacious complement to contemporary selection techniques for modern enterprises.

While this study adhered to stringent procedures and criteria, some limitations can be improved and expanded upon in the future. First, this study only explored the effects of SJT-PS on TMX and workplace popularity without examining its effects on other important outcomes. Future research can investigate the impact of the SJT-PS on various organizational outcomes using diverse samples. Second, in the development of the SJT-PS, this study did not strictly distinguish different industries. Future investigations of political skill assessment methods could refine this aspect by tailoring assessments to specific industries and organizational contexts, thereby augmenting measurement validity. Last, we only gathered one coworker’s rating of the two criteria, which is not rigorous enough. Future research should consider employing the method of multiperson evaluation when gathering criterion data.