“Dark personality traits and job performance of employees: The mediating role of perfectionism, stress, and social media addiction”

The purpose of this study was to explore the indirect and direct relationships of Big- 5 and dark personality traits (i.e., extroversion, neuroticism, agreeableness, openness, conscientiousness, narcissism, Machiavellianism, psychopathy, sadism, and spiteful- ness) with job performance via perfectionism, stress, and social media addiction. A total of 514 private sector employees filled out a query sheet that included the assess- ment tools for the variables. Path analysis using a multiple mediation model indicated that neuroticism was negatively directly and indirectly related to job performance via stress and social media addiction. Machiavellianism and spitefulness were directly positively associated with job performance, and Machiavellianism-related higher so- cial media addiction diminished the direct positive effect of Machiavellianism on job performance, indicating complex relationships. Furthermore, stress, social media ad- diction, and perfectionism were related to different personality traits positively and negatively. Findings of the present study suggest that an anti-social personality may promote higher job performance. However, job performance may be adversely affected by the adverse consequences relating to these traits. Professionals and firms that at- tempt to increase job performance should take anti-social personality traits and their complex effects on job performance into account.


INTRODUCTION
The last century is characterized by a large number of studies of psychology on the relationship between work performance and personality. Personality was seen as an important factor for productivity analysis. It is a behavior that distinguishes one person from another (Beer & Brooks, 2011; Tuteja & Sharma, 2017; Cherry, 2019). Personality makes it possible to understand whether a person will be able to perform a certain job or not. Performance is the result of skills, knowledge, effort, and role abilities. It is customary to different individuals based on their personal traits. These features make them suitable for a certain type of work. It is therefore critical to know how different factors contribute to work (Yohannes, 2020), especially through stress and social media.
Digitalization affects substantial transformation of job nature as well as labor markets. Traditional HR visions on how each trait of potential employee related to performance on job-place may not be relevant in the new era. At the same time, a large fraction of traditional econom-ic and social interactions may still require more common traits that rely on rational (for each culture) expectations about performance on the job placement. While labor economics and HR practices are enriched by behavioral studies, that tie economics and psychology, further understanding of relations between traits and performance are needed. This paper uses the empirical approach grounded on data from a questionnaire that helps to avoid some theoretical biases in modeling the relations between traits and employee performance.
The paper is structured into six sections. Following the introduction, Section 1 presents a brief literature review of Big-5 and dark personality traits. Section 2 is devoted to applying a methodological approach and research methods. In Section 3, results from correlation analysis are presented. In Sections 4 and 5, data models are interpreted and commented on.

LITERATURE REVIEW
The first person to investigate the hypothesis that a complex taxonomy of human personality traits could be obtained by selecting a language was F. Galton (Galton, 1884; Shrout & Fiske, 1995). Later, Allport and Odbert (1936) applied Galton's hypothesis in practice, extracting more than 4,000 adjectives, which described the observed and relatively constant features in dictionaries at the time. In 1949, a pioneering multivariate study of personality was committed by J. Guilford (Guilford, 1949;Musek, 2017), who analyzed ten personality factors. Later, Mischel (1968) stressed that attitudes and behavior change depend on the situation and are not stable.
The paradigm returns to the adoption of the Big-5 model that emerged in the early 1980s (Goldberg, 1980;Joubert & Venter, 2013). Until 1983, experiments showed that the forecast of personal models is best correlated with actual behavior in stressful emotional postures, in contrast to the typical interview in neutral emotional postures (Boyle, 1983;Boyle et al., 1995 Gender studies of the Big-5 inventory showed that women are mostly slightly superior to men in neuroticism, extraversion, pleasure, and integrity (Costa et al., 2001). However, the gender difference in personality traits is greater in the more developed, wealthy, and gender-egalitarian countries. Men in highly developed regions of the world are less neurotic, extroverted, honest, and pleasant compared to men living in less affluent regions of the world. Women, as a rule, do not differ in personality traits in different regions of the world (Schmitt et al., 2008). Yet most literature sources do not consider the moderating effects of gender on the impact of personality traits on performance (Cubel et al., 2016).
Cultural differences also affect the Big-5. The degree of assessment of individualism in the country correlates (Kozlovskyi et al., 2021) with the ordinary extraversion, while individuals belonging to cultures that perceive significant imbalances in their power structures tend to be slightly higher than honesty Some companies and organizations around the world evaluate people based on the personality traits of the Big-5. Studies have confirmed that individuals who are considered leaders tend to show fewer neurotic traits, sustain a higher level of openness, a balanced level of integrity, and a balanced level of extraversion (Judge et al., 2002). Further systematic studies link burnout at work and home with neuroticism, and extraversionwith permanent positive work experience (Mehta, 2012). Some studies show that the results of professional activities are correlated with the personality traits of Big-5. Integrity is the highest level of overall performance (Sackett & Walmsley, 2014). Important predictors of promotion goals are extraversion, integrity, and pleasure (Spurk & Abele, 2011). FFM-based inventories have a long shelf life.
It is well known that the role of the individual in the performance of work is crucial because certain individual traits contribute to career growth at work. The results of various studies, in particular Salgado (1997)

RESEARCH FRAMEWORK
This study combines the impact of Big-5 personality traits ( Figure 1) and dark personality traits ( Figure 2) in terms of job performance.

Participants and procedure
A total of 514 Turkish private sector adult employees (41% female; older than 21 years old) were recruited for the study via promoting a paper-and-pencil questionnaire in different factories in a Turkish city. The respondents gave their informed consent before participating in the study acknowledging that partaking in the study is anonymous and voluntary. Respondents did not receive any remuneration for participating in the survey. Ethical approval for the study was received from the university's ethical board before the participants' selection and complied with the Helsinki declaration (World Medical Association, 2013).

Measures
Performance Scale (PS): The Turkish form (Çöl, 2008) of the PS (Sigler & Pearson, 2000) was used to assess employees' job performance. The scale consists of four items (e.g., "I complete my tasks on time", "I make sure that products meet or exceed quality standards") on a 5-point Likert scale from "never" to "always". The Turkish scale form reported sound psychometric properties (Çöl, 2008). The coefficient of internal consistency in this study showed high values (Cronbach's α = .76).

Figure 1. Big-5 personality traits
Personality traits Extraversionindividual differences in social inclusion, assertiveness, and energy level Agreeablenessdifferences in compassion, respect, and perception of others Conscientiousnessdifferences in organization, productivity, and responsibility Neuroticism (emotional stability)differences in the frequency and intensity of negative emotions Openness to experience (intellect) -differences in intellectual curiosity, aesthetic sensitivity, and imagination Narcissismgrandiosity, pride, selfishness, and a lack of empathy The dark triad Psychopathyantisocial behavior, impulsivity, selfishness, callous and unemotional traits, and remorselessness Machiavellianismmanipulation and exploitation of others, absence of morality, unemotional callousness, and a higher level of self-interest

Positive and Negative Perfectionism Scale (PANPS):
The Turkish form (blinded for peer review) of the PANPS (Haase & Prapavessis, 2004) was used to assess perfectionism. The positive perfectionism subscale was used in the present study, which comprises seven items (e.g., "When I am competing against others, I am motivated by wanting to be the best") on a 5-point Likert scale from "absolutely disagree" to "absolutely agree". The Turkish adaptation study reported mostly good fit to the data in the confirmatory factor analysis (χ 2 = 30.32, df = 9, p < .001, RMSEA = .07 CI 90% [.05, .10], SRMR = .03, CFI = .99, GFI = .98). The internal consistency was neither high nor low in the present study (α = .60).

Instagram Addiction Scale (IAS):
The unidimensional IAS (Yurdagül et al., 2019) was used to assess problematic social media use by replacing the word "Instagram" with "social media". The IAS consists of six items (e.g., "How often in the past year have you spent a lot of time thinking about social media or planned using social media?") on a 5-point Likert scale from "very rarely" to "very often". It evaluates the components of the dependence (i.e., salience, conflict, withdrawal, mood modification, tolerance, and relapse) is described in the biopsychosocial context of addiction (Griffiths, 2005). The internal consistency coefficient was neither high nor low in this study (Cronbach's α = .62).
Perceived Stress Scale (PSS): The Turkish form (Küçükusta, 2007) of the PSS (Cohen et al., 1983) was used to measure the stress levels of participants. Even though the original form comprised 14 items, the four-item Turkish form reported good psychometric properties (Küçükusta, 2007). The items (e.g., "I have felt that I was unable to control the important things in my life lately", "I have found that I could not cope with all the things that I had to do") were scored on a 5-point Likert scale from "never" to "always". The internal consistency was very high in the present study (Cronbach's α = .88). . Participants were asked to analyze the extent to which individual traits are consistent with their nature, from 1 ="absolutely disagree" to 7 = "absolutely agree" (e.g., I am a narcissist = selfish, self-centered; I am Machiavellian = manipulate others and exploit them, lie to achieve own goals at the expense of others; I am psychopath = insensitive to the requests of others, devoid of repentance, indifferent to the morality of actions; I am sadist = enjoy causing intentional harm and pain to others; I am spiteful = ready to harm even myself to harm others).
Neuroticism (negatively) was directly (β = -.15, p < .01) and indirectly (β = -.06, p < .01) related to job performance via stress and social media addiction. Machiavellianism was directly positively (β = .20, p < .001) and indirectly negatively (β = -.05, p < .01) associated with job performance via social media addiction. Spitefulness was directly positively related to job performance (β = .12, p < .01). Furthermore, total effects of extroversion (positively directly and negatively indirectly related to job performance), openness (negatively directly and positively indirectly related to job perfor-mance), and psychopathy (negatively directly and positively indirectly related to job performance) on job performance (Fried et al., 2008) were nonsignificant despite their significant direct and indirect effects (Table 1).
relationships are not illustrated as a table or figure for clarity. The tested model explained 16%, 11%, 13%, and 17% of the variance in stress, social media addiction, perfectionism, and job performance respectively.

DISCUSSION
The empirical results mostly confirm the intuitive idea that psychologically unstable traits are negatively correlated with performance. At the same time, this finding requires additional conceptualization. For job places, that require psychological resilience and emotional stability, more "flat" personalities are looked preferable. This is relevant mostly for traditional economic and social interactions. The situation may differ in the case of socalled creative industries where talent and personnel admissions may play a more important role. When exceptional creativity coincides with some emotional instability, empirical results of the paper and relevant paradigm may be challenged. It is doubtful that neurotic traits are quite common to hesitate general business performance, but more flexible labor organization and horizontal organizational structures may assimilate destructive traits retaining benefits from creativity.
Also, important to note that the most important drivers of performance (the highest correlation ratio (Appendix A,

CONCLUSION
Transformation of the nature of the job requires a deeper understanding of which traits are more consistent with employee performance. The shift from hierarchies to horizontal structures as well as industrial economy-mode to freelance-mode may affect the rethinking of each trait's contribution to the success of the business. The paper demonstrates that traditional traits stamped as emotional instability mostly negatively correlate with employee performance while more traditional "labor ethic" traits mostly positively. In addition, it was found that in some cases similar patterns like Perfectionism and Narcissism differently contribute to performance. Perfectionism demonstrates the highest level of correlation with performance. At the same time, Machiavellianism traits demonstrate mixed results. From one point of view, such a trait is a less important camper to more traditional hierarchical structures like Agreeableness and Conscientiousness. From another point of view, Machiavellianism demonstrates a strong contribution to performance during the effect decomposition procedure. Such "disappointing" patterns show that Machiavellianism as a trait is inherited by some dualism that may reproduce strong and weak sides in different social environments and different incentive structures. It is assumed that the modality of the economy is more important in that regards camper to, for example, cultural attitudes, but more robust statements in that direction require further research.   Note: SMA = social media addiction; * means p < .05, ** means p < .01, *** means p < .001.