Workplace incivility, work engagement, and turnover intentions: Multi-group analysis

Abstract The purpose of this study is to examine workplace incivility in work engagement and employee turnover intention in the work place in Indonesia. In addition, this study also examines the differences in intensity of these relationships by gender.In total, 644 employees from various business industries in Indonesia completed a quantitative survey relating to their perceptions of workplace incivility, work engagement, and turnover intention. Data were analyzed with partial least square multi-group analysis technique (PLS-MGA).The results show that workplace incivility has a negative effect on work engagement, and a positive effect on turnover intention, and the intensity of these relationships differby gender. Work engagement has been shown to have a negative effect on turnover intentions and to play a mediating role in workplace incivility and turnover intention relationship.Relevant recommendations are presented to HR managers and practitioners to prevent unethical practices in organizations and encourage communication and interaction by promoting civil behavior in the workplace.

Abstract: The purpose of this study is to examine workplace incivility in work engagement and employee turnover intention in the work place in Indonesia. In addition, this study also examines the differences in intensity of these relationships by gender.In total, 644 employees from various business industries in Indonesia completed a quantitative survey relating to their perceptions of workplace incivility, work engagement, and turnover intention. Data were analyzed with partial least square multi-group analysis technique (PLS-MGA).The results show that workplace incivility has a negative effect on work engagement, and a positive effect on turnover intention, and the intensity of these relationships differby gender. Work engagement has been shown to have a negative effect on turnover intentions and to play a mediating role in workplace incivility and turnover intention relationship.
ABOUT THE AUTHORS Dr. Irsan Tricahyadinata is an Assistant Professor at the Faculty of Economics and Business Universitas Mulawarman, Samarinda, Indonesia. His current research interests include human resources management and organization behavior. Email: irsan.tricahyadinata@feb.unmul.ac. id Hendryadi is a lecturer at Sekolah Tinggi Ilmu Ekonomi Indonesia Jakarta, Jakarta, Indonesia. His current research interests are Islamic work ethics, leadership, workplace incivility, and quantitative research methods. Email: hendry. basrah@gmail.com Dr. Suryani is an Associate Professor at the Islamic State Institute of Lhokseumawe, Aceh, Indonesia. Her research interests include organizational culture, leadership, and ethics from the Islamic perspective. Email: suryani.uin@gmail.com Dr. Saida Zainurossalamia ZA is an Assistant Professor at the Faculty of Economics and Business Universitas Mulawarman, Samarinda, Indonesia. His research interests are marketing strategic and organizational behavior. Email: saida.zainurossalamia.za@feb.unmul.ac.id Prof. Sukisno Selamet Riadi is a Professor at the Universitas Mulawarman, Samarinda, Indonesia. He has published many articles in the area of human resources management and organizational behavior. Email: sukisno.selamet. riadi@feb.unmul.ac.id

PUBLIC INTEREST STATEMENT
In the last two decades, incivility in the workplace has become one of the issues that has attracted the attention of organization researchers and practitioners. Various empirical studies have been conducted to identify and explain the causes and consequences of incivility for individuals and organizations. However, previous studies conducted in Europe, the US, and Asian countries almost entirely found different forms of incivility in the workplace, due to differences in values or norms of politeness in different regions. Responding to this issue, the present research expands the study of workplace incivility in the context of Indonesian culture and clarifies the difference from previous researches on the role of gender on the impact of workplace incivility on work engagement and turnover intention.
Relevant recommendations are presented to HR managers and practitioners to prevent unethical practices in organizations and encourage communication and interaction by promoting civil behavior in the workplace.

Introduction
In the last two decades, workplace incivility has received the attention of organization researchers and practitioners. Various empirical studies have been conducted to identify and explain the causes and consequences of incivility for individuals and organizations. Intensive researches involving large samples in Europe andAmerica throughout 1999-2009 show that workplace incivility has become a serious concern (e.g., Cortina et al., 2001;Lewis & Malecha, 2011;C. Pearson & Porath, 2009). Almost all studies conducted in America and Europe show an increasing trend of uncivil behaviors that are accepted by employees of various sectors/industries. This condition confirms that workplace incivility is a global problem that requires immediate attention from human resource and organization professionals (Ghosh et al., 2013).
Most of the preliminary researches on workplace incivility were conducted in the United States, and later developed in Australia (Birks et al., 2017;Loh & Loi, 2018;Warrner et al., 2016) and Canada (Smith et al., 2010). In Asia, there were several large-scale studies such as in China (Chen et al., 2018;Ma et al., 2018;Shi et al., 2018;Zhou et al., 2015), Korea (Hyun et al., 2018;Son & Jang, 2017); India (Sharma & Singh, 2016); Philippines (Bulloch, 2017), Singapore (Ho & Tan, 2018;Loh, 2015;Torres et al., 2017), and Malaysia (Arshad & Ismail, 2018;Dahri & Ab Hamid, 2018;Koon & Pun, 2018;Lim, 2016). Nearly all studies have found different forms of workplace incivility because there are differences in values or norms of politeness in different regions. This makes incivility not have a general form that can be accepted by all research, and thus, the issue of workplace incivility still leaves a gap to be explored in the context of cross-cultural research. In Asia, Yeung and Griffin (2008) found that 77% of respondents surveyed from various organizations in China, Hong Kong, India, Japan, Singapore, and Korea reported to have received disrespectful behavior from their coworkers or superiors at least once a year. The attention of researchers and practitioners on workplace incivility is reflected in various studies in numerous countries. They involve almost all types of work which show that workplace incivility continues to increase significantly and is increasingly worrying.
This study makes a number of contributions to the literature of workplace incivility, work engagement, and turnover intention. First, the proposed models test the interrelation of workplace incivility, work engagement, and turnover intention that were previously tested separately, such as the relationship between workplace incivility and work engagement (Beattie & Griffin, 2014;Reio & Sanders-Reio, 2011;Yeung & Griffin, 2008) and turnover intention (Cortina et al., 2013;Rahim & Cosby, 2016;Sharma & Singh, 2016). Second, work engagement in various studies has a mediating role in relation to various intention turnover antecedents (e.g., Agarwal & Gupta, 2018;Memon et al., 2016), but no one has tested the mediating role of work engagement in the workplace incivility-turnover intention relationship. Thus, this study is directed to expand previous researches by examining the role of work engagement as a mediator of workplace incivility-turnover intention relationship. Third, this research develops a multi-group model of analysis by considering the factor of gender (e.g., Cortina et al., 2013;Hendryadi & Zannati, 2018;Sliter et al., 2012). The gender was found to be inconsistently affecting perceptions of workplace incivility. Therefore, this study is present to close that gap.
The first objective of the present study is to investigate the relationship between workplace incivility, work engagement, and turnover intention. More specifically, this study examined the direct and indirect relationship of workplace incivility to turnover through work engagement. The second objective of the study is to expand the area of workplace incivility, work engagement, and turnover intention, as well as gender-diversity and gender equality research in the context of work environment in Indonesia. This information will provide better insights for managers and organizations on the more effective ways to manage communication patterns and interactions within organizations, especially in Indonesia. This is followed by the next section that focuses on the literature review and the research hypotheses, methodology,including research design, sampling procedures, measurement and analysis techniques. The next section deals with the results and discussion. An understanding of this form of incivility behavior can help companies to develop policies and rules related to communication patterns and interactions in order to create a more effective work environment. The paper concludes with the practical and theoretical implications of the findings and significance of the study about Indonesian organizations.

Workplace incivility
Empirical studies of workplace incivility are undeniably most influenced by Andersson and Pearson (1999, p. 457) who provide the definition of workplace incivility as "low-intensity behavior with ambiguous intent to harm the target, which violates workplace norms for mutual respect; rude, showing a lack of respect towards others." Spiral theory was then developed by Andersson and Pearson (1999) to explain how the chain effect of workplace incivility. Spiral phenomenon starts at the starting point where incivility behavior is expressed as uncivilized behavior by individuals, violates norms, or as a form of behavior that is unacceptable to victims (Andersson & Pearson, 1999). This situation then raises a desire to get a revenge. This desire will likely result in incivility in response to the impoliteness experienced. As the spiral continues, one or both parties tend to reach a tipping point because of anger, embarrassment, and humiliation, which can trigger intense intentional behavior such as violence or aggression. The spiral of incivility can continue until there is an agreement between the two parties to forgive each other, or one of the parties resigns. Furthermore, the primary spiral can trigger a secondary spiral. Secondary spirals are triggered by immodesty (Andersson & Pearson, 1999). For example, someone who witnesses a spiral of impoliteness tends to take similar actions; and therefore, workplace incivility then becomes increasingly widespread within the organization. Based on the spiral theory of incivility, it can be concluded that workplace incivility is a cycle that can be triggered from a small problem and then increases in intensity into rude behavior or more severe aggression in the workplace. C.M. Pearson and Porath (2005) propose that efforts to stop this spiral phenomenon can begin with building an organizational culture and climate based on mutual respect, and especially a culture that does not tolerate the behavior of impoliteness that occurs within the organization (C.M. Pearson & Porath, 2005).
Workplace incivility is a deviant behavior, verbally and non-verbally, such as a look of condescension, harsh words, impatience or a lack of respect for the dignity of others. As a result, employees who are victims of incivility have a tendency to decrease their commitment over time (Montgomery et al., 2004). Other examples of unethical behavior include not saying thank you, heeding co-workers' suggestions, texting or sending emails during meetings, making derogatory comments, showing hostility, invasion of privacy, exclusive behavior, gossiping and ignoring or insulting coworkers (C. Pearson & Porath, 2009).
Incivil behavior in the workplace is a part of employees' daily behaviors in interacting with each other in an organization. The perpetrators sometimes do not realize they have conducted this behavior, such as undervaluing coworkers (for example, not saying thank you for the simple assistance provided), or asking for help in polite words to subordinates or fellow coworkers. Other examples of this behavior include actions such as using condescending language, making verbal and non-verbal threats, gossiping, ignoring coworkers' requests, and showing disrespect for others at work (Holm et al., 2015). The hallmark of incivility is that sometimes the purpose is unclear, and is not intended to be detrimental to others. Although people sometimes behave rudely with a clear purpose to demean or insult others, at other times their disrespectful behavior can be caused by fatigue, carelessness, or indifference to local social norms. Thus, the existence of intention is unclear either from the perpetrator, the victim, or the bystander (Andersson & Pearson, 1999).

Relation of incivility to work engagement
Work engagement is relatively rarely studied in relation to workplace incivility (Beattie & Griffin, 2014). The concept of engagement refers to Kahn (1990) as psychologically present to exert physical, emotional, and cognitive energies into one's role. in engagement, people use and express themselves physically, cognitively, and emotionally while carrying out their work (Kahn, 1990). Schaufeli et al. (2002) provides a definition of work engagement as "a positive, fulfilling, workrelated state of mind that is characterized by vigor, dedication, and absorption". Vigor refers to high energy and mental endurance at work, a willingness to invest efforts to get the job done well, and struggle and perseverance when facing difficulties. Dedication refers to a sense of significance, enthusiasm, inspiration, pride, and challenges at work. Absorption is characterized as a person who is fully concentrated and really enjoys work where time passes quickly, and has a difficulty escaping from work (Bakker & Demerouti, 2007).
Incivility at work represents normative behavior that is contrary to the norms of civility; behaving rudely and disrespectfully, and demonstrating a lack of respect for others (Andersson & Pearson, 1999;Cortina et al., 2001). Beattie and Griffin (2014) conducted an important research among security personnel working in Australia. Their results found that security personnel have higher levels of stress when they experience incivility, but high support from superiors reduced this effect. In addition, the negative effect of incivility on work engagement is only significant for those who have low self-evaluation. Empirical support for workplace incivility relationships is also evidenced in the study of Reio and Sanders-Reio (2011) whose study was on computer company employees in America, and the study of Yeung and Griffin (2008) in Asia. Thus, if work engagement can be increased through interactions in the workplace such as support from colleagues and leaders, then workplace incivility as a form of behavior that tends to be demeaning, abusive treatment from the superior or colleagues will have a negative impact on employees' perceptions of their environment, and have implications for the low work engagement. Therefore, based on the theoretical description and empirical evidence above, the hypothesis proposed is: H1: workplace incivility is negatively related to employee work engagement

Relation of incivility to turnover intention
Employee turnover refers to when an employee decides to leave an organization voluntarily (Shaw et al., 2005). The employee's decision to leave the organization is very expensive for both individuals and organizations (Lee et al., 2004). Three basic components are generally considered when calculating employee turnover costs, including recruitment costs, replacement costs, and training costs (Cascio, 2000). Researchers have found workplace incivility is negatively related to productivity and job satisfaction, and positively related to absenteeism, tardiness, and desire to resign (Lim et al., 2008;Rahim & Cosby, 2016). Various studies provide important notes about the effects of incivility in the workplace on turnover intentions (Cortina et al., 2013;Rahim & Cosby, 2016;Sharma & Singh, 2016). From their study on business administration students in the United States, Rahim and Cosby (2016) found that workplace incivility is positively related to the intention to resign. In various sectors, Sharma and Singh (2016) and Cortina et al. (2013) found that workplace incivility can increase the level of intention to resign. The studies of Hendryadi and Zannati (2018)provide preliminary empirical evidence regarding the link between workplace incivility and employee turnover intention in Indonesia. Both studies found a positive effect of workplace incivility on employee turnover intention. Disrespectful behavior occurs in general and in many organizations, and organizations often do not understand their harmful effects, and most managers are not prepared to deal with it. Due to their experience as victim of workplace incivility, employees tend to reduce work effort, time on the job, and job performance. (C. M. Pearson & Porath, 2005;Rahim & Cosby, 2016). That is, the higher the employee's perception of incivility in their workplace, the intention to leave or find a new job is higher. Therefore, based on the abovementioned researches, the second hypothesis is proposed: H2: workplace incivility is positively related to turnover intention

Relation of work engagement to turnover intention
In the view of the Job Demands-Resources model (J D-R Model), the main contributors to work engagement come from two factors: job demand and job and personal resources. Job resources such as social support (both from colleagues and superiors), performance feedback, skills variation, autonomy, and learning opportunities have positive implications for work engagement (Bakker & Demerouti, 2007). That is, employees who receive more work resources (for example, the support of colleagues and superiors) then the tendency to have a sense of attachment is higher. Second, personal resources are positive self-evaluations related to resilience and refer to individual feelings about their ability to control the environment (Hobfoll, 2002). Personal resources can be in the form of self-efficacy, optimism, self-esteem, endurance, and so on (Bakker & Demerouti, 2007). Job demands refer to aspects of work that require ongoing physical and/or psychological effort and are therefore associated with certain physiological and/or psychological costs (Hakanen et al., 2008). In the J D-R Model, job demands act to moderate the relationship between job resources and personal resources to work engagement. Job demands can be in the form of work, mental, emotional, and physical demands. The main effect of work engagement in the J D-R model is employee performance, both in the form of role performance, extra role performance, creativity, and in terms of the company's financial performance (Bakker & Demerouti, 2007).
Several recent studies have found that work engagement has a negative effect on turnover intention (Agarwal & Gupta, 2018;Babakus et al., 2017;Caesens et al., 2016;Lu et al., 2016;Memon et al., 2015Memon et al., , 2016De Simone et al., 2018). Using data from employees and supervisors in 29 hotels in North America, Rahim and Cosby (2016) found that supervisor level employees had significantly higher work engagement and lower turnover intentions than line level employees. Caesens et al. (2016) conducted a study on 647 employees in Belgium and found that the relationship between work engagement and employee turnover intention was curvilinear. Curvilinear relationship is a form of relationship between two variables where one variable increases followed by another variable, but this increase only occurs at a certain point. (like an inverted U curve). Another form of curvilinear relationship is when one variable increases, and the other decreases to a certain point, and after that the two variables increase together (forming a U curve). Work engagement also has direct and indirect effects with turnover intention (De Simone et al., 2018).
Besides having a direct effect on turnover intention, work engagement in various studies has a mediating role (e.g., Agarwal et al., 2012;Agarwal & Gupta, 2018;Gupta & Shaheen, 2017a, 2017bMemon et al., 2016). Agarwal et al. (2012) prove that work engagement is negatively correlated with turnover intention, and mediate the relationship between LMX and turnover intention. In another research, Agarwal and Gupta (2018) proved work engagement as a mediator for the relationship between job characteristics and turnover intention. The mediating role of work engagement on the relationshipbetwen psychological capital and turnover intention was supported by Gupta and Shaheen (2017a) and the moderating role of personal resources in the relationship between work engagement and turnover intention (Gupta & Shaheen, 2017b). Memon et al. (2016) found that work engagement has a negative relationship with turnover intention, and work engagement also mediates the relationship of job satisfaction with turnover intention. Based on the empirical evidence, this study argues that the relationship model between workplace incivility and turnover intention can be mediated by work engagement in line with support for workplace incivility-work engagement relationship (Gupta & Shaheen, 2017a;Reio & Sanders-Reio, 2011;Yeung & Griffin, 2008) and the relationship work engagement-turnover intention (Agarwal & Gupta, 2018;Babakus et al., 2017;Caesens et al., 2016;De Simone et al., 2018;Lu et al., 2016;Memon, 2016). Thus, the hypothesis proposed is: H3: work engagement is negatively related to turnover intention H4: work engagement mediates the relationship between workplace incivility and turnover intention Demographics are the characteristics of employees that distinguish them from other employees. Demographic elements can be in the form of age, gender, education, years of service, and other social factors. A number of researchers have theorized that categorizing by gender can be offensive. Women are more likely to be targets of disrespectful behavior at work than men (Cortina et al., 2013;Welbourne et al., 2016). According to this idea, women are considered more sensitive to social behavior than men. The consequence is that they are more likely to face interpersonal problems, such as workplace incivility, than men. Cortina (2008) suggests that disrespectful behavior may be targeted more often at women and ethnic minorities (in the American context). This theory is supported by research findings that assert that women experience more incivility in the form of harassment at work than men (C.M. Pearson & Porath, 2005;Cortina et al., 2001;Lim et al., 2008). Reio and Sanders-Reio (2011) found that women experience more incivility from their coworkers and men experience more incivility from their superiors. However, the majority of studies comparing men and women show that incivility generally has the same negative effect on women and men (Cortina et al., 2001;Lim et al., 2008). More recent research found that 65 percent of women compared to 47 percent of men experienced "general incivility" in their workplaces (Cortina et al., 2013).
Based on the various empirical evidence, there are two important things to note, first, immoral behavior in the form of harassment is more likely to be accepted by women from their coworkers (C.M. Pearson & Porath, 2005;Reio & Sanders-Reio, 2011); and younger in age (Lim & Lee, 2011). Second, the difference in results can be explained based on the assumptions put forward by Rousseau et al. (2008) that national culture tends to influence social values, and how individuals perceive and respond to workplace incivility. For example, it is possible for employees who work in countries with high power distance cultural values (such as Indonesia and Japan) to tend to regard neglect by their superiors as a reasonable behavior, rather than employees who work in countries with lower power distance (like America and Europe). Thus, this study considers that gender is likely to only have an impact and strength in the relationship between workplace incivility and work engagement and turnover intention.

Sample procedure
The research sample was taken from 12 companies from various business sectors in Jakarta, Indonesia. A total of 644 respondents were involved in completing the questionnaire in 2019. A total of 52.48% of the respondents were male, and the majority (69%) had a senior high school education. The majority of respondents (64.13%) are still single. Most of the respondents are aged 21-25 (45.50%).

Measurements
Workplace incivility was adapted from a seven-item scale developed by Cortina et al. (2001) to measure the extent to which employees experienced incivility at work in the past year. Items are rated on a five-point Likert scale (never = 1, rarely = 2, sometimes = 3, often = 4, most of the time = 5). Example of the item: "How often in the past year have you received … … … .". This scale was tested and has a Cronbach α internal consistency of 0.87 (Chen et al., 2013); and 0.89 (Rahim & Cosby, 2016). In this study, internal consistency was obtained at 0.87.

Work engagement
Work engagement is measured on a short Utrecht Work Engagement scale (UWES-9) developed by Schaufeli et al. (2006). UWES includes three subscales that reflect the basic dimensions of attachment, including: vigor (three items; for example, "I feel enthusiastic about going to work"), dedication (three items; for example, "I want to know about the results of my work"), and absorption (three items; for example, "time goes by so fast while working"). Each item is rated on a five-point Likert scale from 0 ("never") to 4 ("always"). The Cronbach Alpha coefficient of 0.76 in this study that met internal consistency requirements.

Turnover intention
Three items of the turnover intention subscale from the Michigan Organizational Assessment Scale (Seashore et al., 1982) were used to measure respondents' turnover intentions. Example items for this scale are, "I am actively looking for a new job," and "I often think of quitting". Responses ranged from 1 (strongly disagree) to 5 (strongly agree). The possible range of scores is 3 to 15. A higher score on the scale indicates a higher level of intention to stop. This scale has been shown to have adequate reliability and validity (α = 0.91, Karim et al., 2015). In this study, the value of Cronbach Alpha was 0.80.

Technique of data analysis
The first part of this analysis is designed to examine the psychometric measures of incivility, work engagement, and turnover intentions. The second part of the analysis is designed to test the hypothesis. Partial least square structural equation modeling (PLS-SEM) with a multi-group analysis (MGA) approach is used to test the model. Data analysis was performed using SmartPLS 3.0 software (Hair et al., 2012)

Common method biases
Before further statistical analysis is performed, the common method variance (CMV) is examined first. CMV identifies false correlations that usually occur in cross-sectional data collection methods (e.g., surveys conducted at the same time) are used to measure variables (Tehseen et al., 2017). In this study, CMV was assessed using the Harman single factor test with the principal component analysis (PCA) approach (Podsakoff et al., 2003). The CMV test results show that there is no single dominant factor in the three constructs which explains more than 50 percent of the total variance. Thus, it can be stated that CMV is not a serious problem in this study.

Descriptive analysis
The first analysis begins with descriptive statistics that show the means and standard deviations for the study variables (Table 1) Table 1 shows that the average score of workplace incivility, work engagement, and turnover intentions for the male group are slightly higher than for women. Unexpectedly, men have a higher level of perception than women in workplace incivility. This data differs from previous research which states that women are more likely to be targets of incivility (Cortina, 2008;Welbourne et al., 2016). Next is that men have lower levels of work engagement and turnover intention than women. An independent-samples t-test was performed to examine whether the workplace incivility, work engagement, and turnover intention differ significantly between the gender. As indicated in Table 1, no statistically significant differences were found in the three constructs by gender (all sig. in the independent t-test > 0.05).

Structural equations model
This study uses a one-stage measurement model and the explanation of the results adopts the recommendations of Henseler et al. (2016), Hair et al. (2012Hair et al. ( , 2014. The first stage of testing is to evaluate the measurement model (outer model) to test the validity and reliability of the extract, and then to evaluate the significance of the parameters (inner model) to prove the relationship between constructs.

Measurement model evaluation (outer model)
Evaluation of the measurement model is carried out to check internal consistency and construct validity (convergent validity and discriminatory validity) as determined by Hair et al. (2014). In the first stage, internal consistency is evaluated using the Cronbach α coefficient and construct reliability (CR). The results of the analysis show that all latent constructs meet the requirements for internal consistency, namely CA> 0.70 (WI = 0.88; WE = 0.75; TI = 0.80). Next is evaluating the reliability of the composite (CR) and giving results in line with CA, where the CR is entirely above 0.70 (WI = 0.91; WE = 0.86; TI = 0.88). It can be concluded that the measurement model has met all the reliability requirements as recommended by Hair et al. (2014) and Chin (2010).
Construct validity is to check convergent validity and discriminant validity. Convergent validity is assessed through loading factor (λ) and average variance extracted (AVE). Fornell and Larcker (1981) recommend that the cut-off value for AVE is 0.50, and the loading factor is above 0.70. As shown in Table 2, all loading factors> 0.70 and all AVE greater than 0.50, and the results can be considered satisfactory (Hair et al., 2012).  Note: WIC = workplace incivility, WE = work engagement; TI = turnover intention. Tricahyadinata et al., Cogent Psychology (2020), 7: 1743627 https://doi.org/10. 1080/23311908.2020.1743627 between latent variables. The measurement model is stated to have good discriminant validity if the correlation between latent variables is lower than the square of AVE. As shown in Table 3, the AVE squared for all latent variables is greater than the correlation between variables. Thus, it can be concluded that the scale used in this study has sufficient construct validity.

Structural model testing (Inner model)
After evaluating the measurement model, the next step is to evaluate the structural model. Criteria for evaluating structural models as suggested by Hair et al. (2014). They consist of R 2 , f 2 and Q 2 . Next is to examine the path coefficient, and the significance of the path. R 2 dependent variable explains the amount of variance explained by the model that represents the relevant predictive value with the cut-off value as follows: R 2 values of 0.75, 0.50, or 0.25 can be categorized as strong, moderate and weak (Hair et al., 2011). The value of R 2 obtained for this model shows that both variables (work engagement and turnover intention) have weak predictive power (R 2 = 0.06 and R 2 = 0.18 for each latent endogenous variable).
In addition to evaluating the R 2 values of all endogenous constructs, changes in the R 2 value when certain exogenous constructs are eliminated from the model can be used to evaluate whether if any constructs are removed it can have a substantive impact on endogenous constructs. This measurement is called ƒ 2 effect size (Hair et al., 2012). The guidelines for assessing ƒ 2 are the values of 0.02 (small effect), 0.15 (moderate), and 0.35 (large) (Cohen, 1988;   The next evaluation model is to use blindfolding to validate the crossing of each construct. Stone-Geisser's Q 2 value which is greater than zero indicates that the exogenous constructs have predictive relevance for endogenous constructs (Hair et al., 2011). The results showed that the Q 2 value for work engagement was 0.07, and the turnover intention was 0.18. These results indicate that workplace incivility has weak predictive relevance for work engagement, but is good enough for turnover intention.
Finally, to test the suitability of the theoretical model, a conservative parameter of the SRMR value is used, i.e. if the SRMR value is less than 0.08, it indicates good fit (Hair et al., 2011). SRMR value based on the analysis results obtained is 0.06 (<0.08). It can be stated that for theoretical testing, the results of this analysis are good enough.

Multi-group analysis
In the next step, bootstrap analysis is used to assess the significance of the path coefficient. The minimum number of bootstrap samples is 5,000 as recommended by Hair et al. (2014). To conduct group comparisons, the structural model uses a multi-group approach (PLS MGA). Table 4 displays the p-value and confidence intervals obtained for each path coefficient. Table 4 and Figure 1, work incivility explains a significant relationship in work engagement (β = -.24, p-value < 0.01) for all groups; β = -.19, p-value < 0.01 for male group and β = -.29, p-value < 0.01 for female group. There was also a significant relationship between work incivility and turnover intention: β = .34, p-value < 0.01 (for all groups). Work engagement is also proven to have a significant relationship with turnover intention turnover intention: β = −.20, p-value < 0.01 (for all groups); β = −.25, p-value < 0.01 (for male group); and β = −.15, p-value < 0.05 (for female group). Lastly, the relationship between workplace incivility-turnover intention is proven to be mediated by work engagamenet (β = -.05, p-value < 0.01).  Notes: * p < 0.05; ** p < 0.01, The entire group. Direct Effect. Indirect Effect.

Discussion
This article empirically examines the relationship between workplace incivility (WI), work engagement (WE), and turnover intention (IT) in Indonesian business sector organizations. Table 4 presents the results of the structural model analysis. Based on the results of the analysis, it can be stated that WI has a negative relationship with WE (coefficient = −0.24, p-value 0.00), so H1 is supported. This shows that when employees feel a high level of incivility, it can lower their level of work engagement. This finding reinforces the results of a previous study conducted by Beattie and Griffin (2014), Sanders-Reio (2011), andYeung andGriffin (2008). Workplace incivility is a contradictory form of good social relations within an organization, and can have implications for the low level of work engagement. The results of the analysis further showed that the relationship between workplace incivility and work engagement was significant for the two groups, both for men and women. Path coefficient values −0.19 (men) and −0.29 (women) show that women have a greater tendency to lessen their work engagement than men if exposed to disrespectful behavior in the workplace.
The next hypothesis shows that workplace incivility plays an important role in predicting employee turnover intention. The path coefficient value is significant and positive between WI and TI (coefficient = 0.34, p-value 0.00) so that H2 is supported. These results reinforce previous empirical evidence (Hendryadi & Zannati, 2018;Rahim & Cosby, 2016;Sharma & Singh, 2016). That is, the higher the employee's perception of incivility in their workplace, the intention to leave or find a new job is higher. Therefore, based on the abovementioned researches, the second hypothesis is proposed: Multi-group analysis shows that both men and women have an equal response related to turnover intention (coefficient of 0.34 for the whole group). The results of this study differ from that of Hendryadi and Zannati (2018) that found the effect of workplace incivility on turnover intention was greater in the male group than in the female; and does not support differences in receiving incivility between men and women (Cortina, 2008;Welbourne et al., 2016). It can be said that workplace incivility can increase the intention to find another job among employees. When this intention increases, employees will have the intention to consider quitting their jobs. This effect generally applies to men and women, and there is no difference between men and women in responding to workplace incivility for turnover. Thus, both men and women who have experience workplace incivility have the same tendency to leave their organizations. This result shows the positive effects of working in a social work environment where employees treat one another with respect and refrain from disrespectful behavior in their daily work.
Finally, the analysis found that WE negatively predicted turnover intention (coefficient = −0.20, p-value 0.00), so that H3 was supported. These results are consistent with previous empirical evidence (e.g., Agarwal & Gupta, 2018;Babakus et al., 2017;Caesens et al., 2016;De Simone et al., 2018;Lu et al., 2016;Memon, 2016). Based on the results of the multi-group analysis obtained, interesting results are shown where the relationship of work engagement with turnover intention is greater in the male group (coefficient of −0.25) compared to the group of women (−0.15). It means that men have more tendency to consider leaving the job than women. Besides having a direct effect on turnover intention, work engagement in this study was proven to mediate the relationship of workplace incivility and turnover intention (coefficient 0.05, p-value 0.00). This finding is the first test on the indirect relationship of workplace incivility to turnover intention by placing work engagement as a mediator, so that it has a theoretical contribution to the impact of workplace incivility on turnover intention.

The summary of the results
The relationship between workplace incivility, work engagement, and turnover intention has been proven by various previous studies in cross-cultural contexts. This study not only focuses on investigating the direct effect of workplace incivility on turnover intention, but also examines the role of work engagement as a mediator using multi-group design by placing gender as a distinguishing factor between variables. The results show that workplace incivility has a negative effect on work engagement, and a positive effect on turnover intention, and the intensity of these relationships differs by gender. Work engagement has been shown to have a negative effect on turnover intentions and to play a mediating role in workplace incivility and turnover intention relationship. The subsequent sections discuss theoretical contributions, managerial implications, limitations and direction for future studies.

Theoretical contributions
This study contributes to the existing knowledge about the impact of workplace incivility on work engagement and turnover intention through several ways. First, this study has proven that workplace incivility has a negative impact on work engagement, and positively on turnover intention. In addition, the relationship of workplace incivility with turnover intention is proven through work engagement, thereby expanding previous research studies that only focus on the direct effects of workplace incivility to work engagement (Reio & Sanders-Reio, 2011;Yeung & Griffin, 2008) and the relationship between work engagement and turnover intention (Agarwal & Gupta, 2018;Babakus et al., 2017;Caesens et al., 2016;De Simone et al., 2018;Lu et al., 2016;Memon, 2016). Second, the relationship of workplace incivility to work engagement and turnover intention was found to vary by gender.Specifically, this study highlights how male and female employees react when they see or experience incivility in their workplace. There are even fewer studies examining gender differences on relationship between workplace incivility and turnover intention (Cortina et al., 2001;Lim et al., 2008;C.M. Pearson & Porath, 2005;Reio & Sanders-Reio, 2011). This study supports previous researches comparing men and women showing that general disability has the same negative effect on women and men (Cortina et al., 2001;Lim et al., 2008) and found that gender played a moderating role in the relationship between workplace incivility-turnover intention, and and confirm previous empirical studies (e.g., Cortina et al., 2013;Hendryadi & Zannati, 2018;Sliter et al., 2012).

Practical implications
Based on the empirical evidence previously explained, there are several key implications for organizations to consider when creating a policy. First, managers must pay attention to workplace incivility situations to reduce employee turnover intentions, especially workplace incivility which can lead to the decrease of work engagement and the increase of turnover intentions. Our findings show that workplace incivility has a key role in reducing work engagement and increasing turnover intention. It needs to be followed up through various strategic steps from the human resource management to prevent the more worrisome effects of workplace incivility. Therefore, it is important for top management to recognize the existence of incivility in the workplace and to stop it as early as possible. For example, to foster a work environment that promotes mutual respect, organizations must encourage cultural values that explicitly prohibit all forms of disrespectful behavior among employees, and between supervisors and subordinates. Second, so that managers or supervisors have the ability to identify and improve communication patterns between employees, it is advisable to regularly attend communication and leadership training so that communication between superviors and subordinates, and among employees can be effectively carried out.

Limitations and future research directions
There are still many limitations in this study. First, the data was collected from business sector organizations in Indonesia using convenience sampling, which can limit research generalizations. Therefore, it is recommended that future researchers replicate this research in various sectors using random sampling. Second, the majority of respondents in this study consisted of unmarried employees aged between 21-30 years, with high school education. These respondent characteristics may not represent the values and perceptions of senior employees (over 35 years). Future research is recommended to use a more varied sample by adding the percentage of employees over 35 years. Finally, the cross-sectional nature of this study has limited claims for causality. Therefore, subsequent research needs to use a longitudinal design to test changes over time which will be very valuable to test causality.