Course Textbook Anderson, L. E., & Bolt, S. B. (2016). Professionalism: Skills for workplace success (4th ed.). Boston, MA: Pearson. PLEASE READ THE DIRECTIONS AND REQUIREMENTS IN DETAILUnit V PowerPo

ORIGINAL RESEARCH published: 17 January 2018 doi: 10.3389/fpsyg.2017.02365 Frontiers in Psychology | www.frontiersin.org 1January 2018 | Volume 8 | Article 2365 Edited by:

Riccardo Sartori, University of Verona, Italy Reviewed by:

Morteza Charkhabi, KU Leuven, Belgium Andrea Ceschi, University of Verona, Italy *Correspondence:Seungwoo Kwon [email protected] Specialty section:

This article was submitted to Organizational Psychology, a section of the journal Frontiers in Psychology Received: 31 August 2017 Accepted: 27 December 2017 Published: 17 January 2018 Citation:

Lee S, Kwon S, Shin SJ, Kim M and Park I-J (2018) How Team-Level and Individual-Level Con ict In uences Team Commitment: A Multilevel Investigation. Front. Psychol. 8:2365. doi: 10.3389/fpsyg.2017.02365 How Team-Level and Individual-Level Con ict In uences Team Commitment: A Multilevel Investigation Sanghyun Lee 1 , Seungwoo Kwon 2 * , Shung J. Shin 3 , MinSoo Kim 1 and In-Jo Park 4 1 School of Business, Hanyang University, Seoul, South Korea ,2 Korea University Business School, Seoul, South Korea, 3 The School of Business, Portland State University, Portlan d, OR, United States,4 School of Management, Wuhan University of Science and Technology, Wuhan, China We investigate how two different types of con ict (task con ict and relationship con ict) at two different levels (individual-level and team-level) in uence individual team commitment.

The analysis was conducted using data we collected from 193 e mployees in 31 branch of ces of a Korean commercial bank. The relationship s at multiple levels were tested using hierarchical linear modeling (HLM). The resul ts showed that individual-level relationship con ict was negatively related to team commit ment while individual-level task con ict was not. In addition, both team-level task and r elationship con ict were negatively associated with team commitment. Finally, only team-level relationship con ict signi cantly moderated the relationship between individu al-level relationship con ict and team commitment. We further derive theoretical implicatio ns of these ndings.

Keywords: individual-level con ict, team-level con ict, t ask con ict, relationship con ict, team commitment, multilevel analysis INTRODUCTION As modern organizations are increasingly adopting team-base d work structure, academics and practitioners have paid huge attention to team dynamics inclu ding team con ict ( Nesterkin and Porter eld, 2016 ). Because a team cannot perform well without managing team con ict e ectively ( de Wit et al., 2012; Bradley et al., 2015 ), it is critical to understand how team con ict in uences team dynamics and its success. Although team comm itment has been recognized as one of the critical determinants of team success ( Kukenberger et al., 2012; Mathieu and Gilson, 2012; Mahembe and Engelbrecht, 2013 ), there still has been a paucity of studies investigating the relationship between con ict and team commitment. To answer this call, this study examines the relationships between di erent types and levels of con ict and team commitment.

Team commitment refers to the relative strength of an indivi dual’s identi cation with and involvement in a particular team. It is likely to increase the team members’ (1) beliefs in, and acceptances of, the team’s goals and values; (2) willingness t o exert considerable e ort on behalf of the team; and (3) desire to maintain membership in the team ( Bishop and Scott, 2000 ). While team commitment plays critical roles for team success, intra group con ict, which is a natural phenomenon in any team setting, may have a signi cant relatio nship with team commitment. For example, intragroup con ict can incur interpersonal problems within the team, increasing negative emotions such as tension and anxiety and lowering positive em otions such as team satisfaction ( Jehn, 1995; De Dreu and Gelfand, 2008; de Wit et al., 2012; DeChu rch et al., 2013 ), which may Lee et al.Con ict and Team Commitment result in their weak team commitment. However, given the mixed ndings about the e ects of con ict on team success (e.g., Jehn et al., 1999; Pelled et al., 1999; de Wit et al., 2012 ) and some suggested boundary conditions for the relationship (e .g., Bradley et al., 2013 , 2015), we cannot simply expect a negative relationship between con ict and team commitment. In this study, we investigate the e ects of two di erent types of con ic t (i.e., relationship con ict and task con ict) on the two di e rent levels (i.e., individual- and team-level) on team commitme nt.

Prior studies have found that resolving relationship con i ct among team members is critical for team success as relations hip con ict often generates negative consequences on teams.

However, the ndings regarding the e ects of task con ict hav e been mixed; the results have included negative consequence s ( Jehn et al., 1999; Lovelace et al., 2001; De Dreu and Weingart, 2003; de Wit et al., 2012 ), positive consequences ( Jehn, 1995; Amason, 1996; Matsuo, 2006; Song et al., 2006; Olson et al., 2 007 ), and no signi cant consequences on teams ( Pelled et al., 1999; de Wit et al., 2012 ). In addition, while the bulk of con ict research (e.g., Baron, 1991; Amason, 1996; Jehn et al., 1997, 1999; Pelled et al., 1999; Bailey, 2000; Simons and Peterson, 2000; Jehn an d Mannix, 2001; Olson et al., 2007 ) has focused on the di erent con ict types in the last few decades, research on this issue h as largely proceeded with a single-level theory, and research h as not clearly speci ed the level of analysis in theory or measureme nt ( Korsgaard et al., 2008 ).

Given the lack of research on the relationship between con ict and team commitment, and given the fact that con ict research has neglected the multilevel nature of intragroup con ict, multilevel theorization and veri cation are nece ssary to enhance our knowledge of how con ict a ects team members.

To answer these calls, we adopt a multilevel perspective ( Klein and Kozlowski, 2000; Ceschi et al., 2014 ) by investigating the combined e ect of individual-level and team-level con ict o n individual attitudes toward the team (i.e., team commitmen t).

In the current study, we rst distinguish individual-level c on ict and team-level con ict. Then, at these two levels, we clarif y the e ects of task con ict and relationship con ict on team commitment. In addition, based on social information proces sing theory ( Salancik and Pfe er, 1978 ), we examine if the team- level con ict moderates the relation between the individua l- level con ict and team commitment.

Salancik and Pfe er (1978) emphasized that an individual’s attitude is in uenced by the social information about what others think, and thus indivi duals can acquire their own attitude or behavior by learning the information of the social environment. An important source o f information is the individual’s “immediate social environ ment” because it provides cues that individuals use to construct an d interpret events. We argue that team-level con ict can be an important source of social information. Because individual team members may collect and use social cues from their team in ord er to interpret the nature of their individual con ict, we propose that team-level con ict in uences how perceived individual con ict a ects the attitude toward their team such as team commitment. Our study contributes to con ict literature in several ways. First, we examine the relationship between con ict an dteam commitment, which has been rarely studied. It would be interesting to examine how within-team con ict a ects individual attitudes such as team commitment. Moreover, because team commitment is important for current and future team success ( Hackman, 1987 ), the investigation of the determinants would be helpful to better understand how to maximize team productivity. Secondly, we investigate the relationship at multiple levels analyzing the multilevel mo del with hierarchical linear modeling (HLM). The attitudes of te am members toward their team are in uenced by the team-level con ict as well as each team member’s personal experience with other members (i.e., individual-level con ict). Howe ver, past studies have mostly focused on the e ects of individual- level con ict perceptions on individual-level outcomes and very few studies have examined the e ects of team-level con ict on individual-level attitudes. We ll this void by investigating how team-level con ict relates to individua l team commitment. Furthermore, with the multilevel approach, we may be able to better understand the multilevel phenomenon of how team con ict in uences the individual-level attitud e of team commitment. Finally, we theorize and test cross-lev el interactions between individual-level and team-level con icts on team commitment. To date, the con ict literature says littl e about the consequences of the interactional e ect of individual- a nd team-level con ict. Because individual- and team-level co n icts may in uence each other, studying this interaction could prov ide a more robust understanding of how con ict a ects a team and its members.

THEORETICAL FRAMEWORKS AND HYPOTHESES The Multilevel Nature of Intragroup Con ict According to Klein and Kozlowski (2000) , organizations are multilevel systems, with individuals nested within groups, an d groups nested within organizations. As a result, the entities — individuals, dyads, teams, groups, organizations, and so on—ar e tied with each other ( Klein et al., 1994 ). Accordingly, certain levels inevitably interact with each other: characteristi cs of one level have e ects throughout levels above and below ( Rousseau, 1985 ). Con ict issues are no exception to this multilevel interaction. In particular, con ict may initially exist only a mong certain members of a team ( Jehn, 1995 ), but such isolated individual-level con ict can quickly escalate into a team-l evel con ict with a set of attributes speci c to that particular tea m.

In turn, this team-level con ict will in uence individual- level con ict, and the behaviors and attitudes of the team’s membe rs.

In the next section, following Rousseau’s (1985) suggestion, we begin with the explicit description of the properties that di erentiate individual-level con ict from team-level con ict.

While individual-level con ict refers to the individual’s recognition of con ict experienced from rsthand interperso nal interactions with a speci c team member(s), team-level con ict refers to the team members’ recognition of con ict existing i n the team as a whole regardless of the focal individual’s involvem ent.

Here, team-level con ict is a set of summary perceptions Frontiers in Psychology | www.frontiersin.org 2January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment re ecting an interaction between personal and organizational characteristics ( James and Jones, 1974 ). In addition, individual- level and team-level con ict are distinguished by the refer ence of the con ict. While individual-level con ict is a ected by interpersonal characteristics (i.e., team member referenc e), team- level con ict is in uenced by team characteristics (i.e., w ork team reference).

The Effect of Individual-Level Con ict on Team Commitment Individual-level con ict refers to an interpersonal incompatibility or di ering perceptions such as a di erence of opinions and/or an unmatched relationship in interacting with others. Individual-level con ict can have both negati ve and positive consequences on a team and its team members.

For example, some studies have reported that individual- level con ict is related to negative consequences because t he con ict causes tension and hostility and reduces team membe r satisfaction and team productivity ( March and Simon, 1958; Pondy, 1967; Hackman and Morris, 1975; Farh et al., 2010; de Wit et al., 2012 ). On the other hand, other studies have found positive consequences of individual-level con ict. For exa mple, it can stimulate innovative thinking and the creation of new ideas ( Coser, 1956; Walton, 1969; Deutsch, 1973 ).

As an attempt to reconcile such contrasting ndings, con ict researchers suggested distinguishing within-tea m con ict into task con ict and relationship con ict. Althou gh the de nitions vary ( Pinkley, 1990; Priem and Price, 1991; Jehn, 1995; Amason, 1996 ), they consistently emphasize distinguishing relationship/a ective/social-emotional c on ict from task/substantive/cognitive/goal-oriented con ict because they lead to di erent consequences. Here, relationship con ict refers to interpersonal incompatibility and includes tensio n, annoyance and animosity among team members; while task con ict refers to disagreements among team members regardi ng the content of their decisions and includes di erences in viewpoints, ideas, and opinions ( Jehn, 1995 ). In other words, task con ict is rooted in the substance of the task that a team is undertaking whereas relationship con ict derives from t he emotional, a ective aspects of the team’s interpersonal relat ions ( Guetzkow and Gyr, 1954 ).

Individual-level task con ict is expected to have signi can t in uence on satisfaction or team commitment. Individual-l evel task con ict can lead to positive consequences for certain ty pes of tasks such as creative performance and decision quality in t op management teams ( Schweiger et al., 1989; Jehn, 1995; Simons and Peterson, 2000 ), in particular via a self-regulated process (e.g., Ceschi et al., 2017 ). However, according to self-veri cation theory ( Swann et al., 2004 ), individuals become dissatis ed when their perspectives or solutions are challenged by group membe rs, since it can be interpreted as a negative assessment of their own abilities. Therefore, individual-level task con ict i s likely to negatively relate to satisfaction ( Jehn, 1995; de Wit et al., 2012 ).

Task con ict may also cause tension and resentment among team members leading to dissatisfaction with the interperso nal interaction ( Amason and Schweiger, 1994; Gamero et al., 2008 ) and decreasing the desire of team members to stay on the team ( Schweiger et al., 1986 ).

In interpersonal relationship con ict, team members are lik ely to experience interpersonal tension and negative emotions suc h as anxiety and fear, which leads to a less positive attitude tow ard the team ( Walton and Dutton, 1969; Jehn and Bendersky, 2003; Dijkstra et al., 2005; de Wit et al., 2012 ). Accordingly, relationship con ict, which is related to nervousness, hatred, and negati ve self-valuation, is likely to reduce team commitment and inc rease turnover intention ( Elron, 1997; Bayazit and Mannix, 2003; Raver and Gelfand, 2005; Rispens et al., 2007; Jehn et al., 2008; Rispens, 2012 ).

In sum, both task and relationship con icts at individual level are expected to be negatively related to team commitmen t.

Accordingly, we hypothesize as follows:

Hypothesis 1a. Individual-level task con ict will be negat ively related to team members’ commitment.

Hypothesis 1b. Individual-level relationship con ict wil l be negatively related to team members’ commitment.

The Effect of Team-Level Con ict on Team Commitment Team-level con ict is a shared perception amongst the team members ( Klein and Kozlowski, 2000 ). Team-level con ict is not objective and is not an observable concept, but it can be measu red by the degree of agreeableness among team members ( West et al., 2009 ). Accordingly, team-level con ict is expected to work as a contextual factor at the team level that can predict individu al- level attitudes beyond the e ects of the individual-level co n ict experience. For example, a team member may perceive high level of team-level con ict even when he or she does not have any interpersonal con ict with team members. In this case, he or she would not want to be involved in a con ict that has already occurred among other group members. That is, he or she may want to stay away from the con ict among other group members.

Individual team members are likely to adopt avoidance behavio r particularly when they perceive all types of con ict that arise from irreconcilable di erences among team members ( O’Neill et al., 2013 ). Avoidance behavior will aggravate di erences among group members since they cannot understand the reasons for others’ positions ( Tjosvold, 2008 ). Therefore, individuals are not likely to identify with the team or to be committed to the team .

In this case, the team member’s team commitment is likely to be negatively in uenced by the team-level con ict. Team-lev el con ict is a contextual variable that may globally in uence a n individual’s commitment toward the team. Therefore, regard less of how much con ict each team member has individually with other team members, team-level con ict will have a negative in uence on team commitment. With a perception of task or relationship con ict among other members, a team member is likely to have a negative attitude toward the team in the form of nervousness and worry, which will then lower his or her team commitment. Thus, we propose that team-level con ict has incremental e ects on team commitment over the e ects of individual-level con ict. Frontiers in Psychology | www.frontiersin.org 3January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment Hypothesis 2a. Holding individual-level con ict constant, team- level task con ict is negatively related to team commitment.

Hypothesis 2b. Holding individual-level con ict constant , team- level relationship con ict is negatively related to team commitment.

A Cross-Level Interaction between the Two Different Levels of Con ict In addition to examining the incremental e ect of team-level con ict on team commitment, our study explores the moderatin g role of team-level con ict on the relationship between indi vidual- level con ict and team commitment. In particular, we propose that team-level con ict may also a ect how a team member re ects his/her perception of an individual-level con ict (e. g., arising from his/her disagreement with other team members) on his/her team commitment. In other words, team-level con ict may a ect how team members perceive individual-level disagreements and emotional con icts with other team membe rs, in turn in uencing their general attitude toward their team .

Social information processing theory ( Salancik and Pfe er, 1978 ) helps explaining why this would be the case. Social informatio n processing theory ( Salancik and Pfe er, 1978 ) argues that an individual’s attitude is in uenced by social information a bout what people around them think. In the same vein, we argue that team-level con ict can be an important source of social information so that team-level con ict plays a moderating ro le in the relationship between individual-level con ict and t eam commitment. When individuals form attitudes toward their team, they use information gathered from other team members .

Via social information processing, the shared perception of team-level con ict is likely to in uence how individual tea m members process their perceptions of the con ict experienced at an individual level and in turn, how they form their attitu des toward their team. For example, when team-level con ict is hi gh, the negative relationship between individual team members ’ individual-level con icts and their team commitment is lik ely to be stronger than when team-level con ict is low. Under a high team-level con ict, members tend to respond to each other wit h more defensiveness and animosity even when the disagreemen t is constructive. This intensi es the negative e ect of indiv idual- level con ict on their attitude toward their team. On the con trary, when team-level con ict is low, the negative relationship b etween individual-level con ict and team commitment is likely to b e attenuated, because members may regard their individual-le vel con ict as just their own problems and not that of their team’s . In addition, they may take the interpersonal con ict less serio usly due to their perception of the team climate as positive. More speci cally, we argue that team-level task con ict is likely to aggravate the negative relationship between indi vidual- level task con ict and team commitment, and that team- level relationship con ict is likely to aggravate the negat ive relationship between individual-level relationship con ict and team commitment. Drawing on social information theory, our premise is that team-level con ict in uences how team members interpret and react to their own individual con ict with another team member(s). Therefore, we propose that team-level con ict is an important feature of a work team’s social context, and can enhance or weaken the previously hypothesized e ects of the individual level con ict on team commitment.

Hypothesis 3a. The relationship between individual-level task con ict and an individual’s team commitment will be moderated by team-level task con ict.

Hypothesis 3b. The relationship between individual-level relationship con ict and an individual’s team commitment will be moderated by team-level relationship con ict.

The model of this study is conceptualized in Figure 1.

METHOD We collected data from 452 employees in 57 branch o ces of a private commercial bank in Korea. Each of the branch o ces had a registered branch manager who was responsible for business operations. While the size of the branches, in terms of the number of employees and customers, varies a little depending on the location of the branch (e.g., urban or rural) , the structures, processes and contents of the work are identi cal across the branches. In addition, employees interacted with other employees within their own branch on a daily base with task interdependence. Thus, bank branches were conceptualized as de ned work teams of interdependent individuals. Typically, each of the bank branches is composed of a branch manager, one or two sub-branch manager (depending on branch size), and bank clerks who perform teller tasks. The members of each branch perform organizationally relevant tasks and share common goals and must coordinate with each other to carry out their tasks. The team-level data were collected from each branch and the individual-level data were collect ed from individuals within each branch. The response rate of the survey was 52.7 percent. We collected data from branch o ces with 3 or more employees (team size in branches ranged from 6 to 22 individuals) and excluded responses from the branches where <3 employees responded. As a result, the study’s nal sample was made up of 193 responses at 31 branches.

Among the respondents, 56.5% were male, and the average age was 31.8 years ( SD=5.94). The average working experience in banking was 7.4 years ( SD=6.71), and the average tenure in the current o ces was 1.47 years ( SD=1.02). Seventy-six percent of the respondents had obtained at least a 4-year college degr ee, 46.1% of which majored in business. This study was conducted in accordance with the Declaration of Helsinki and was approved by Korea University’s review board.

Measures To conduct a multi-level study, we measured both individual - level and team-level con ict separately with di erent levels of references. Frontiers in Psychology | www.frontiersin.org 4January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment FIGURE 1 |Hypothesized multilevel model of the intragroup con ict.

Team-Level Con ict Eight items adopted from Jehn (1995) were used to measure employees’ perceptions of the relationship con ict and task con ict at team level (4 items respectively). A sample item for team-level task con ict is “How often do people in your work team disagree about opinions regarding the work being done?” A sample item for team-level relationship con ict is “How muc h friction is there among members in your work unit?” (See Anne x for details.). Cronbach’s alpha was 0.93 for team-level task con ict and 0.95 for team-level relationship con ict. In using the individual responses to analyze the team- level con ict as a characteristic of the team, r wg (within-group interrater reliability) 1 was computed for the team-level con ict measures in order to assess the extent of consensus, agreeme nt, or within-unit variability, within a unit for the measures ( James et al., 1984 ). The inter-rater agreement estimates were computed for team-level con ict measures; mean r wg = 0.83 and median r wg = 0.90 for team-level task con ict, and mean r wg = 0.87, median r wg = 0.89 for team-level relationship con ict. James et al.

(1984) suggested that aggregation can be justi ed by a median agreement index of 0.70 or greater, and all of our estimates exceeded this criterion. Moreover, the ICC values for team-l evel task con ict [ICC (1) =0.07, ICC (2) =0.94] and team-level relationship con ict [ICC (1) =0.12, ICC (2) =0.96] were all statistically signi cant. Thus, these results support fo r the aggregation of the measures to the team level.

Individual-Level Con ict To provide individual-level measures of con ict in teams, employees responded to eight items modi ed to the individual level from Jehn’s (1995) team-level task con ict and relationship con ict scales (four items respectively). More speci cally, t o measure individual-level con ict, we asked team members to think about their individual con ict experiences with other 1 rwg = 1 (S 2 xj / 2 EU ). r wgis the within-group interrater reliability for a group of K judges on a single item X j, S 2 xj is the observed variance on X j, and 2 EU is the variance on X jthat would be expected if all judgments were due exclusively to random measurement error. team members rather than team unit as a whole. A sample item for individual-level task con ict is “How often do you disagree about opinions regarding the work being done with yo ur team members?” A sample item for individual-level relationsh ip con ict is “How much friction do you experience with your team members?” Cronbach’s alpha was 0.90 for individual- level task con ict and 0.93 for individual-level relations hip con ict. Because we used the same people in completing the same individual and team-based con icts, we assessed common method bias by allowing the error terms of individual and team con ict measured by same method to be correlated; stabl e correlations leading us to conclude that common method bias is not a signi cant issue in the study ( Podsako et al., 2003 ).

Team Commitment Team commitment was operationalized as the relative strengt h of an individual’s identi cation with, and involvement in, a particular team ( Mowday et al., 1982 ). Team commitment was measured with the 8-item scale developed by Bishop and Scott (2000) (Cronbach’s alpha =0.96). Sample items included the followings: “I talk up (brag about) this team to my friends as a great team to work on.” and “This team really inspires the very best in me in the way of job performance.” Individual-Level Control Variables We controlled for intragroup trust, in-house training period s, and education level. Intragroup trust is considered as a positi ve predictor of an individual’s attitude ( Driscoll, 1978 ).Simons and Peterson (2000) and Yang and Mossholder (2004) proposed that intragroup trust plays an important role in interpreting con i ct because task con ict can be easily misconstrued as being perso nal in nature or in motive, and thus, be erroneously categorized as relationship con ict. We measured intragroup trust using th e ve-item Likert-type scale used in Simons and Peterson’s (2000) and the coe cient alpha was 0.92. Sample items for intragroup trust included the followings: “We absolutely respect each oth er’s competence.” and “We expect the complete truth from each other.” Frontiers in Psychology | www.frontiersin.org 5January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment The in-house training program is a component of external compensation for employees and such compensation could be an important antecedent variable for an employee’s team commitment ( Porter et al., 1974 ). We controlled for all personal training, by asking each respondent for the total amount of in-house training they received (new employee orientation training was not included because every employee in the bank participated in this training).

Team-Level Control Variables Team size and age diversity were used in our study as team- level control variables. Team size a ects the severity of con icts between the members; when there are more employees in the group, the degree of con ict is higher ( Jehn, 1995 ). In this study, team size was operationalized as the number of registe red employees in a branch o ce. The in uence of age diversity on team members’ commitment was controlled, since Kunze et al.

(2011) suggested that age diversity had a negative e ect on the commitment of employees. To measure age diversity, we used the coe cient of variation ( Allison, 1978 ) and thus, divided each team’s standard deviation of age by the team’s mean age.

Analytical Approach Multi-Level Approach Because the individual-level data were nested within team- level data, we ran HLM analyses to test the hypotheses. HLM is a statistical technique that are used for analyzing data in a c lustered or “nested” structure, in which lower-level units of analysi s are nested within higher-level units of analysis. For example , employees are nested within teams, which are nested within company. In order to reduce multicollinearity, all the predictor vari ables were centered using the grand mean centering method. We took four steps to test the hypotheses. First, for the dependent variable (team commitment), we ran a set of null models with no predictors. Second, we then conducted level-1 analyse s to determine the signi cance of the hypothesized individual- level antecedents (individual-level con ict) in predictin g team commitment. A regression line was estimated for each of the 3 1 teams in this step for level-1 analysis. In the third step, lev el-2 analyses followed, to which we added team-level antecedent s to test for additional explained variance and assess the main e e cts of the team-level con ict climate. In the last step, we regres sed the slope estimates obtained from level-1 on the team-level predictors (team-level con ict) to detect cross-level inter action e ects.

RESULTS Table 1 presents the descriptive statistics and correlations for all the study variables at both the individual- and the team-leve l.

There was a signi cant positive correlation between intragr oup trust and team commitment ( r= 0.52). In addition, consistent with previous research ( Simons and Peterson, 2000; De Dreu and Weingart, 2003; de Wit et al., 2012 ), the perceptions of task and relationship con icts in the team had strong positive correl ation with each other ( r= 0.65). To avoid multicollinearity-related problems, we ran HLM analyses for task con ict and relationshi p con ict respectively.

Discriminant Validity Check We performed a con rmatory factor analysis on the six variables—individual-level task con ict, individual-le vel relationship con ict, team-level task con ict before aggr egation, team-level relationship con ict before aggregation, intr agroup trust, and team commitment—to establish their discriminan t validity (see Table 2). The con rmatory factor analysis with the six variables as distinct factors demonstrated good t to the data [ 2 (362, N= 193) = 822.86, 2 /df =2.27, IFI =0.92, CFI =0.92, RMSEA =0.081]. Further, the results of the chi-square di erence tests indicated that this four-factor model had a better t t o the data than the plausible alternative models (see Model 3 in Table 2 ). This six-factor model, compared with the alternative models, also provided superior point estimates for the t measures. These results supported the discriminant validity of the six variables as distinct constructs at the individual-l evel.

Hypotheses Testing Result of HLM Null Model We followed the HLM procedure recommended by Ho mann (1997) to test our hypotheses. Our hypotheses predict that both individual-level and team-level variables would have a signi cant negative relationship with the individual employees’ team commitment. In order for these hypotheses to be tested, there should be signi cant between-team variance in individual t eam commitment. A chi-square test showed that the between-team variance in team commitment was signi cant [ 2 (30) = 112.84, p < 0.001]. Based on this result from testing the null model, the next analysis was performed.

Individual-Level Con ict We estimated the level-1 model that did not include team- level con ict. We predicted that individual-level task con ict (Hypothesis 1a) and relationship con ict (Hypothesis 1b) wou ld be negatively associated with an employee’s team commitment.

The results indicate that individual-level relationship c on ict was negatively associated with team commitment ( = 0.24, p < 0.01), which supports Hypothesis 1b (see Tables 3,4).

However, contrary to the predictions of Hypothesis 1a, individual-level task con ict was negatively, but not signi cantly related to the team commitment ( = 0.16, p> 0.10).

Adding Team-Level Con ict To test Hypothesis 2a and 2b, we developed an HLM model in which the individual-level con ict variables were the lev el-1 predictors and then regressed the intercept coe cients obtai ned from level-1 on the measures of team-level con ict at level- 2. As reported in Tables 3,4, both team-level task con ict ( = 0.28, p< 0.05) and relationship con ict ( = 0.43, p < 0.01) demonstrated signi cant relationships with employee team commitment after we had accounted for individual-leve l con icts. Hence, Hypothesis 2a and 2b were supported. Frontiers in Psychology | www.frontiersin.org 6January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment TABLE 1 |Means, standard deviations, and correlations for variable s.

Individual-level Variable Means SD1 2 3 4 5 1. Task Con ict 4.20 1.24 2. Relationship Con ict 3.80 1.48 0.61** 3. Intragroup Trust 5.40 0.96 0.14 0.31** 4. Inhouse Training 0.49 1.18 0.13 0.15* 0.03 5. Education 2.63 0.76 0.00 0.01 0.14 0.01 6. Team Commitment 4.86 1.35 0.26** 0.43** 0.52** 0.07 0.08 Team-level Variable Means SD1 2 3 1. Team size 11.42 4.32 2. Age diversity 0.15 0.06 0.06 3. Task Con ict 3.81 0.80 0.28 0.13 4. Relationship Con ict 3.26 0.77 0.03 0.22 0.65* Individual-Level N =193; Team-Level N =31.

* p < 0.05, *p < 0.01.

TABLE 2 | Con rmatory factor analysis of level-1 variables.

Model Factor structure model χ2 (df) χ2 /df IFI CFI RMSEA RMSEA con dence interval 12 (1 df ) 1. One factor (alternative): All six scales together as one factor 3748.10(377) 9.94 0.42 0.42 0.216 0.210.222 2925.24(15) 2. Three factor (alternative): Task and relationship con ict constrained as one factor 1470.98 (371) 3.96 0.81 0.80 0.124 0.118,0.131 648.12 (9) 3. Three factor (alternative):

Individual-level con ict and Team-level con ict constrained as one factor 1369.87 (371) 3.69 0.82 0.82 0.118 0.112,0.125 547.0 (9) 4. Six factor (hypothesized): ITC, IRC, TTC, TRC, intragroup trust, and team commitment as distinct factors 822.86 (362) 2.27 0.92 0.92 0.081 0.074,0.089 – N = 193, All 2 values are signi cant at p <0.05. IFI, incremental t index; CFI, comparative t index; RMSEA, root-m ean-square error of approximation; ITC, individual-level task con ict; IRC, individual-level relationship con ict, TTC, team-level tas k con ict; TRC, team-level relationship con ict. All the nested chi squa re difference tests reported above are signi cant at p < 0.05. We decide to go on with the last model because it as the most parsimoniou s model.

Testing Cross-Level Interaction We predicted that the relationship between individual-level con ict and an employee’s team commitment would be moderated by team-level con ict such that the negative relationship would be stronger when there is a high level of team-level con ict (Hypothesis 3a and 3b). A prerequisite for testing these cross-level interactions i s that there should be signi cant random variance for the individu al- level con ict variables in the intercepts-as-outcomes mode ls estimated in the previous step ( Liao and Chuang, 2004 ). There was signi cant random variance in the slopes for individual-l evel task con ict ( 22 = 0.21, p< 0.01) and relationship con ict ( 22 = 0.14, p< 0.01), indicating the possible presence of team- level moderators. With the prerequisites ful lled, we intro duced team-level con ict as a level-2 moderator. Hypothesis 3a was not supported because there was no signi cant interaction between individual-level and team- level task con ict ( = 0.09, n.s.). However, the results indicate that the interaction of individual-level and team-level relationshipcon ict was signi cant ( = 0.15, p< 0.05), supporting Hypothesis 3b. We then tested whether this interaction e ect w as aligned with the hypothesized trend, utilizing a simple slope t est at cross-level interaction ( Preacher et al., 2006 ). The interaction plot ( Figure 2 ) graphically represents cross-level moderation, showing the relationship between individual-level relati onship con ict and employee team commitment in teams with (1) high and (2) low relationship con ict climates. As expected, individual-level relationship con ict had a stronger assoc iation with team commitment when team-level relationship con ict was high ( +1 SD) rather than low ( 1 SD). The results indicated that the simple slope at +1 SD of team-level relationship con ict was negatively signi cant ( B= 0.59, t= 2.28, p< 0.05), whereas the simple slope at 1 SD of team-level relationship con ict was not signi cant ( B= 0.25, t= 1.12, p> 0.05).

Summarizing the result, the interaction plot revealed that t eam- level relationship con ict intensi ed the negative relati onship between individual-level relationship con ict and the team commitment of the team members. Frontiers in Psychology | www.frontiersin.org 7January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment TABLE 3 |HLM results for the effects of task con ict on team commitment .

Variables Control variable Individual-level predictors T eam-level predictors Cross level predictors γSE γ SE γSE γSE INDIVIDUAL- LEVEL PREDICTORS Intercept 4.82*** 0.16 4.88*** 0.11 4.88*** 0.11 4.90*** 0.11 Intragroup Trust 0.67*** 0.09 0.67*** 0.10 0.66*** 0.10 0.65 *** 0.10 In-house Training 0.08 0.08 0.05 0.06 0.03 0.08 0.02 0.08 Education 0.08 0.10 0.01 0.09 0.03 0.09 0.03 0.09 Task Con ict (Hypothesis 1a) 0.16 0.11 0.09 0.12 0.10 0.12 TEAM-LEVEL PREDICTORS Team size 0.04 0.03 0.03 0.02 0.02 0.02 0.02 0.02 Age Diversity 0.72 2.19 0.07 1.25 0.12 1.21 0.24 1.22 Team-level Task Con ict (Hypothesis 2a) 0.28* 0.12 0.33* 0.13 CROSS-LEVEL INTERACTIONS Team-level X Individual-level Task Con ict (Hypothesis 3a) 0.09 0.09 Model Deviance 583.81 536.07 532.04 535.9641.38 Individual-Level N =193; Team-Level N =31.

* p < 0.05, ***p< 0.001.

TABLE 4 | HLM results for the effects of relationship con ict on team co mmitment.

VariablesControl variable Individual-level predictors Team-level p redictors Cross level predictors SE SE SE SE INDIVIDUAL- LEVEL PREDICTORS Intercept 4.82*** 0.16 4.84*** 0.11 4.85*** 0.10 4.90*** 0.09 Intragroup Trust 0.67*** 0.09 0.56*** 0.10 0.55*** 0.10 0.54*** 0.10 In-house Training 0.08 0.08 0.05 0.07 0.05 0.07 0.02 0.06 Education 0.08 0.10 0.11 0.11 0.09 0.11 0.10 0.11 Relationship Con ict (Hypothesis 1b) 0.24** 0.08 0.18* 0.08 0.20* 0.08 TEAM-LEVEL PREDICTORS Team size 0.04 0.03 0.01 0.02 0.02 0.02 0.01 0.02 Age Diversity 0.72 2.19 0.65 1.34 0.14 1.30 0.40 1.26 Team-level Relationship Con ict (Hypothesis 2b) 0.43** 0.12 0.43*** 0.12 CROSS-LEVEL INTERACTIONS Team-level X Individual-level Relationship Con ict (Hypothesis 3b) 0.15* 0.07 Model Deviance 538.81 539.62 531.86 533.33 Individual-Level N=193; Team-Level N =31.

* p < 0.05, **p < 0.01, ***p< 0.001.

DISCUSSION In this study, we aimed to distinguish and verify two levels o f con ict (i.e., individual- and team-level). Then we invest igated how two levels of con ict and two types of con ict (relationsh ip con ict and task con ict) in uence team commitment. More speci cally, we examined the combined e ect of individual-lev el and team-level con ict on individual attitudes toward the t eam (i.e., team commitment) using a multi-level perspective ( Klein and Kozlowski, 2000; Ceschi et al., 2014 ).

First, we found that individual-level relationship con ic t was signi cantly and negatively related to team members’ team commitment. Even though team commitment is an important factor for team success, there has been a paucity of studies empirically testing these relationships. The nega tive association between individual-level relationship con i ct and team members’ team commitment is consistent with the ndings of previous studies that conducted individual-level analys is of intragroup con ict ( Jehn, 1995; Jehn et al., 1999; de Wit et al., 2012 ).

However, we found that individual-level task con ict did not signi cantly relate to team commitment. One possible explanation for this insigni cant relationship is that the t wo opposing mechanisms were operative at the same time. On one Frontiers in Psychology | www.frontiersin.org 8January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment FIGURE 2 |Plot of cross-level moderating effects of team-level relat ionship con ict on the relationship between individual-lev el relationship con ict and team commitment.

hand, self-veri cation theory ( Swann et al., 2004 ) posits that individuals are less likely to be committed to the team when t heir opinions are challenged by other group members, since it can b e interpreted as his or her inability or incompetence. On the oth er hand, according to self-perception theory ( Bem, 1967, 1972 ), individuals tend to infer their attitudes from their own beh avior in the same way that an outside observer might. Therefore, individuals who experience task con ict or disagreement wit h other members may judge themselves to be highly involved in the work and committed to the team. In other words, employees can infer that they are involved in the task con ict with othe r members because they have high motivation toward team tasks .

Therefore, experiencing task con ict can be considered to be a social cue or signal that an individual is committed to the te am. In addition, they can be satis ed because they had a chance to vo ice their opinions on the task at hand ( Simons and Peterson, 2000 ).

Thus, the insigni cant result might re ect the trade-o bet ween the negative and positive e ects of individual-level task con ict on team commitment. Second, we found that team-level task and relationship con icts had negative relationships with team commitment even after controlling for individual-level task and relat ionship con icts. By di erentiating the two levels of con ict, we wer e able to show the incremental e ects of team-level con ict on team commitment. Previous studies have not paid attention to the two di erent levels of intragroup con ict: individual- level and team-level. Using a multi-level approach, our study implies that team-level con ict may have incremental e ects on team commitment beyond the e ects of individual-level con ict. The negative relationship between team-level con ict (both task and relationship con ict) and the team commitmen t of employees is consistent with previous empirical studies on team con ict focusing on team level analysis. For example, Bailey (2000) found that con ict at the team-level decreased employee satisfaction with team. Interestingly, although we found t hat individual-level task con ict was not signi cantly related to team commitment, team-level task con ict had a negative impact on team commitment. These results suggest that once an individ ual is already involved in an individual-level task con ict, it can be a signal that he or she is committed to the team. However, when other group members are in a team-level task con ict (regardless the focal individual may or may not be involved in the task con ict), individuals will not be committed to th e team in order to avoid con ict. In addition to the CFA results for the discriminant validity, this nding implies that team - level task con ict is a distinct construct that is di erent fro m individual-level task con ict. Finally, by examining the cross-level interactions betwee n the con icts at the two levels, we found that a high degree of team-level relationship con ict aggravated the negativ e relationship between individual-level relationship con ict and team commitment. We theorized and tested the cross-level interactions between team-level and individual-level con icts on team commitment. While previous studies have investigate d interactions between the two types of con ict (e.g., task and relationship con ict) ( Eisenhardt et al., 1997; Mooney et al., 2007 ), the con ict literature has not paid attention to these cross-level interactions partly because it has not paid speci c attention to the level issue in con ict. Based on social info rmation theory, we proposed that team-level con ict aggravated the negative relationship between con ict and team commitment .

Also, we found that team-level relationship con ict modera ted the relationship between individual-level relationship c on ict and team commitment such that the negative relationship was stronger when team-level relationship con ict was prevalen t.

This result suggests that immediate social context (i.e., t eam-level Frontiers in Psychology | www.frontiersin.org 9January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment relationship con ict) serves as an important factor that in uences how individual-level relationship con ict re ect on employ ee attitudes toward the team. In contrast, with regard to task con ict, there were no cross-level interactions between the two levels. The di ere nt results between the two types of con ict can be explained by the social information processing perspective of Salancik and Pfe er (1978) . In particular, Salancik and Pfe er (1978) suggest that as information becomes increasingly ambiguous , individuals rely more on social comparisons to assess it.

Compared to the relationship con ict information, task con ict information is more certain and less ambiguous because task con ict is dependent on task contents and perceived more objectively. Thus, when experiencing individual-level task con ict, individuals may rely less on cues from their surrounding environments in interpreting con ict informat ion.

Conversely, relationship con ict information is more unce rtain and ambiguous because it tends to be more interpersonal and emotional. Therefore, the social information processing is more likely to occur in relationship con ict information process ing, and as a result individuals are more a ected by the social cont ext in determining their attitude toward their team. Future stu dies should focus on investigating the mechanisms that drive thi s interesting nding.

Implications for Practice This paper provides interesting implications for practice. In or der to make employees be more committed to the organization, managers should try to minimize relationship con ict regar dless of its level (individual or team). However, this paper provides di erent prescriptions for dealing with di erent levels of task con ict (individual and team level). First, team-level tas k con ict should be minimized since individuals do not want to become embroiled in a dispute with their team members. That is, employees want to avoid factional activities. In contrast, t o some extent, individual-level task con ict can be recommen ded since it does not reduce team commitment of employees. It can be a signal that team members are committed to the team. Furthermore, the satisfaction of employees could incr ease because they have opportunities to express their opinions regarding the task. This suggests that team leaders who are concerned with improving the commitment of their team members should not only di erentiate between task con ict and relationship con ict but also distinguish between individual-level and team-le vel con ict. With regard to task con ict, it is important for team leaders to resolve con ict at team-level such as task coordi nation in teams. However, with regard to relationship con ict, dea ling with con ict at team-level is not su cient. It is required fo r team leaders to know who have the relationship con ict and to deal with con ict personally.

LIMITATIONS Our study has several limitations. First, there can be a generalizability issue because this study used Korean sample s only. In Korea, con ict is less accepted than in Western count riesbecause the traditional corporate culture of South Korea is based on Confucianism that emphasizes harmonious relations hip with others. However, the result of our study would be even stronger in Western societies where con ict is generally mo re accepted.

Second, we measured individual-level and team-level con i ct variables from same person, a practice that may introduce common method bias ( Podsako et al., 2003 ). In the future, more objective data that are focused on team-level con ict shoul d be examined. Third, we used only employee team commitment as a dependent variable and measured team commitment using a self- report. In order to generalize our ndings to team literature , future studies need to use objective performance measures in order to test the distinct e ects of team-level and individua l-level con icts on team performance. Another limitation of our study is that the sampled branches were not identical in terms of team size. However, we believe that this limitation does not a ect the interpretation of the result not only because the branches belonged to the same organization, but also because we controlled for the team si ze e ect and any other team-dependent e ect with the HLM analysis. Finally, this study showed that there was no signi cant relationship between the individual-level task con ict an d team commitment. However, task con ict can be transformed into emotional relationship con ict as time passes ( Simons and Peterson, 2000; De Dreu and Weingart, 2003; Yang and Mossholder, 2004 ). Because this study used only cross-sectional data, the con icts were measured at a certain point of time.

Therefore, future studies may need to include longitudinal data to trace how a certain type of con ict changes into another type of con ict across time. We can then investigate the dynamic relationship between the two di erent types of con ict at both individual and team levels.

CONCLUSION We found that team-level con ict, either relationship or ta sk, had an incremental e ect on individual members’ team commitment .

In addition, while individual-level task con ict had a non- signi cant negative in uence on team commitment overall, team-level task con ict had a negative in uence on team members’ team commitment. Further, we found team-level relationship con ict aggravated the negative relation bet ween individual-level con ict and team commitment. These nding s expand our understanding of team con ict by revealing that team-level con ict is distinct from individual-level con ict.

This implication encourages future studies to investigate t he respective roles of individual- and team-level con icts sepa rately in team dynamics, and their interplay for team success. As a practical implication, managers should pay attention not only to individual-level relationship con ict, but also to team-l evel relationship con ict because a high level of team-level con ict may aggravate the negative in uence of individual-level co n ict on team commitment and performance. Frontiers in Psychology | www.frontiersin.org 10January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment AUTHOR CONTRIBUTIONS SL wrote the rst draft and edited by SK. MK analyzed data with SL. SL, SK, MK, SS, and I-JP generated idea for this study.FUNDING The research for this article was supported by the grant from Korea University Business School.

REFERENCES Allison, P. D. (1978). Measures of inequality. Am. Sociol. Rev.43, 865–880.

doi: 10.2307/2094626 Amason, A. C. (1996). Distinguishing the e ects of functiona l and dysfunctional con ict on strategic decision making: resolving a paradox for top ma nagement group. Acad. Manage. J. 39, 123–148. doi: 10.2307/256633 Amason, A. C., and Schweiger, D. (1994). Resolving the paradox of con ict, strategic decision making, and organizational performance. Int. J. Con .

Manage. 5, 239–253. doi: 10.1108/eb022745 Bailey, D. E. (2000). Modeling work group e ectiveness in high- technology manufacturing environments. IIE Transact.32, 361–368.

doi: 10.1080/07408170008963913 Baron, R. A. (1991). Positive e ects of con ict: a cognitive pe rspective.Empl.

Respons. Rights J. 4, 25–36. doi: 10.1007/BF01390436 Bayazit, M., and Mannix, E. A. (2003). Should I stay or should I g o? Predicting team members’ intent to remain in the team. Small Gr. Res.34, 290–321.

doi: 10.1177/1046496403034003002 Bem, D. J. (1967). Self-perception: an alternative interpretation of cognitive dissonance phenomena. Psychol. Rev.74, 183–200. doi: 10.1037/h0024835 Bem, D. J. (1972). “Self-perception theory,” in Advances in Experimental Social psychology ed L. Berkowitz (New York, NY: Academic Press), 1–16.

Bishop, J. W., and Scott, K. D. (2000). Organizational commitmen t and team commitment in a team environment. J. Appl. Psychol.85, 439–450.

doi: 10.1037/0021-9010.85.3.439 Bradley, B. H., Anderson, H. J., Baur, J. E., and Klotz, A. C. (2015) . When con ict helps: integrating evidence for bene cial con ict in g roups and teams under three perspectives. Group Dyn. Theory Res. Pract.19, 243–272.

doi: 10.1037/gdn0000033 Bradley, B. H., Klotz, A. C., Postlethwaite, B. E., and Brown, K. G. ( 2013). Ready to rumble: how team personality composition and task con ict interac t to improve performance. J. Appl. Psychol.98, 385–392. doi: 10.1037/a0029845 Ceschi, A., Demerouti, E., Sartori, R., and Weller, J. (2017). Deci sion- making processes in the workplace: how exhaustion, lack of resources and job demands impair them and a ect performance. Front. Psychol.8:313.

doi: 10.3389/fpsyg.2017.00313 Ceschi, A., Dorofeeva, K., and Sartori, R. (2014). Studying tea mwork and team climate by using a business simulation. Eur. J. Train. Dev.38, 211–230.

doi: 10.1108/EJTD-01-2013-0004 Coser, L. A. (1956). The Functions of Social Con ict . New York, NY: Free Press.

DeChurch, L. A., Mesmer-Magnus, J. R., and Doty, D. (2013). Mo ving beyond relationship and task con ict: toward a process-state perspective .J. Appl.

Psychol. 98, 559–578. doi: 10.1037/a0032896 De Dreu, C. K. W., and Gelfand, M. J. (2008). “Con ict in the workpla ce: sources, functions, and dynamics across multiple levels of analysis,” in The Psychology of Con ict and Con ict Management in Organizations eds C. K. W. De Dreu and M. J. Gelfand (New York, NY: Erlbaum), 3–54.

De Dreu, C. K., and Weingart, L. R. (2003). Task versus relationsh ip con ict, team performance, and team member satisfaction: a meta-analysis. J. Appl. Psychol.

88, 741–749. doi: 10.1037/0021-9010.88.4.741 Deutsch, M. (1973). Con ict Resolution: Constructive and Destructive Process es.

New Haven, CT: Yale University Press.

de Wit, F. R., Greer, L. L., and Jehn, K. A. (2012). The paradox of i ntragroup con ict: a meta-analysis. J. Appl. Psychol.97, 360–390. doi: 10.1037/a0024844 Dijkstra, M. T. M., Van Dierendonck, D., and Evers, A. (2005). R esponding to con ict at work and individual well-being: the mediating role of ight behavior and feelings of helplessness. Eur. J. Work Organ. Psychol.14, 119–135.

doi: 10.1080/13594320444000254 Driscoll, J. W. (1978). Trust and participation in organizational d ecision making as predictors of satisfaction. Acad. Manage. J.21, 44–56. doi: 10.2307/255661 Eisenhardt, K. M., Kahwajy, J. L., and Bourgeois, L. J. (1997) . How management teams can have a good ght. Harv. Bus. Rev.75, 77–85.

Elron, E. (1997). Top management teams within multinational corporations: e ects of cultural heterogeneity . Leadership Q.8, 393–412.

doi: 10.1016/S1048-9843(97)90021-7 Farh, J. L., Lee, C., and Farh, C. I. (2010). Task con ict and t eam creativity:

a question of how much and when. J. Appl. Psychol.95, 1173–1180.

doi: 10.1037/a0020015 Gamero, N., Gonzalez-Roma, V., and Peiro, J. M. (2008). The in uen ce of intra- team con ict on work teams’ a ective climate: a longitudinal stud y.J. Occup.

Organ. Psychol. 81, 47–69. doi: 10.1348/096317907X180441 Guetzkow, H., and Gyr, J. (1954). An analysis of con ict in deci sion-making groups. Hum. Relat. 7, 367–381. doi: 10.1177/001872675400700307 Hackman, J. R. (1987). “The design of work teams,” in Handbook of Organizational Behavior ed J. W. Lorsch (Englewood Cli s, NJ: Prentice Hall), 315–342.

Hackman, J. R., and Morris, C. G. (1975). Group task, group interact ion process, and group performance e ectiveness: a review and proposed integratio n.Adv.

Exp. Soc. Psychol. 8, 45–99. doi: 10.1016/S0065-2601(08)60248-8 Ho mann, D. A. (1997). An overview of the logic and rational of hie rarchical linear models. J. Manage. 23, 723–744. doi: 10.1177/014920639702300602 James, L. R., Demaree, R. G., and Wolf, G. (1984). Estimating withi n group interrator relate with and without response bias. J. Appl. Psychol.69, 85–98.

doi: 10.1037/0021-9010.69.1.85 James, L. R., and Jones, A. P. (1974). Organizational climate: a review of theory and research. Psychol. Bull. 81, 1096–1112. doi: 10.1037/h0037511 Jehn, K. A. (1995). A multi-method examination of the bene ts and detriments of intragroup con ict. Adm. Sci. Q.40, 256–282. doi: 10.2307/2393638 Jehn, K. A., and Bendersky, C. (2003). “Intragroup con ict in o rganizations: a contingency perspective on the con ict-outcome relationship,” i nResearch in Organizational Behavior: An Annual Series of Analytical Ess ays and Critical Reviews eds R. M. Kramer and B. M. Staw (Oxford: Elsevier Science Ltd), 187–242.

Jehn, K. A., Chatwick, C., and Thatcher, S. M. B. (1997). To ag ree or not to agree: the e ects of value congruence, individual demographic dis similarity, and con ict on workgroup outcomes. Int. J. Con . Manage.8, 287–305.

doi: 10.1108/eb022799 Jehn, K. A., Greer, L. L., Levine, S., and Szulanski, G. (2008). The e ects of con ict types, dimensions, and emergent states on group outcomes .Group Dec.

Negotiat. 17, 465–495. doi: 10.1007/s10726-008-9107-0 Jehn, K. A., and Mannix, E. (2001). The dynamic nature of con ic t: a longitudinal study of intragroup con ict and group performance. Acad. Manage. J.44, 238–251. doi: 10.2307/3069453 Jehn, K. A., Northcraft, G. B., and Neale, M. A. (1999). Why di ere nces make a di erence: a eld study of diversity, con ict, and performance in work group.

Adm. Sci. Q. 44, 741–763. doi: 10.2307/2667054 Klein, K. J., Dansereau, F., and Hall, R. J. (1994). Levels issues i n theory development, data collection and analysis. Acad. Manage. Rev.19, 195–225.

Klein, K. J., and Kozlowski, S. W. J. (2000). From micro to meso: criti cal steps in conceptualizing and conducting multilevel research. Organ. Res. Method3, 211–236. doi: 10.1177/109442810033001 Korsgaard, M. A., Jeong, S. S., Mahony, D. M., and Pitariu, A. H. ( 2008).

A multilevel view of intragroup con ict. J. Manage.34, 1222–1252.

doi: 10.1177/0149206308325124 Kukenberger, M. R., Mathieu, J. E., and Ruddy, T. (2012). A cro ss-level test of empowerment and process in uences on members’ informal learning and team commitment . J. Manage.41, 987–1016. doi: 10.1177/01492063124 43559 Kunze, F., Boehm, S., and Bruch, H. (2011). Age diversity, age di scrimination climate and performance consequences: a cross organizational stu dy.J. Organ.

Behav. 32, 264–290. doi: 10.1002/job.698 Frontiers in Psychology | www.frontiersin.org 11January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment Liao, H., and Chuang, A. (2004). A multilevel investigation of factors in uencing employee service performance and customer outcomes. Acad. Manage. J.47, 41–58. doi: 10.2307/20159559 Lovelace, K., Shapiro, L. D., and Weingart, L. R. (2001). Maximizi ng cross- functional new product teams’ innovativeness and constraint ad herence:

a con ict communications perspective. Acad. Manage. J.44, 779–783.

doi: 10.2307/3069415 Mahembe, B., and Engelbrecht, A. S. (2013). The relationship betw een servant leadership, a ective team commitment and team e ectiveness. J. Hum. Resour.

Manage. 11, 1–10. doi: 10.4102/sajhrm.v11i1.495 March, J. G., and Simon, H. (1958). Organizations. New York, NY: Wiley.

Mathieu, J. E., and Gilson, L. L. (2012). “Critical issues and t eam e ectiveness,” in Oxford Handbook of Organizational Psychology , ed S. J. W. Kozlowski (New York, NY: Oxford University Press), 910–930.

Matsuo, M. (2006). Customer orientation, con ict, and innova tiveness in Japanese sales departments. J. Bus. Res.59, 242–250. doi: 10.1016/j.jbusres.2005.06.002 Mooney, A. C., Holahan, P. J., and Amason, A. C. (2007). Don’t tak e it personally:

exploring cognitive con ict as a mediator of a ective con ict. J. Manage. Stud.

44, 733–758. doi: 10.1111/j.1467-6486.2006.00674.x Mowday, R. T., Porter, L. W., and Steers, R. M. (1982). Employee-Organizational Linkages: The Psychology of Commitment, Absenteeism, and Tu rnover. New York, NY: Academic Press.

Nesterkin, D., and Porter eld, T. (2016). Con ict management and performance of information technology development teams. Team Perf. Manage.22, 242–256.

doi: 10.1108/TPM-05-2016-0018 O’Neill, T. A., Allen, N. J., and Hastings, S. E. (2013). Examining the “pros” and “cons” of team con ict: a team-level meta-analysis of task, re lationship, and process con ict. Hum. Perf.26, 236–260. doi: 10.1080/08959285.2013.

795573 Olson, B. J., Parayitam, A., and Bao, Y. (2007). Strategic decisi on making: the e ects of cognitive diversity, con ict, and trust on decision outco mes.J. Manage. 33, 196–222. doi: 10.1177/0149206306298657 Pelled, L. H., Eisenhardt, M. K., and Xin, R. K. (1999). Exploring the b lack box:

an analysis of work group diversity, con ict, and performance. Adm. Sci. Q.44, 1–28. doi: 10.2307/2667029 Pinkley, R. L. (1990). Dimensions of con ict frame: disputant int erpretations of con ict. J. Appl. Psychol. 75, 117–131. doi: 10.1037/0021-9010.75.2.117 Podsako , P. M., MacKenzie, S. B., Lee, Y. L., and Podsako , N . P. (2003).

Common method biases in behavioral research: a critical review of th e literature and recommended remedies. J. Appl. Psychol.88, 879–903.

doi: 10.1037/0021-9010.88.5.879 Pondy, L. R. (1967). Organizational con ict: concepts and mod els.Adm. Sci. Q. 12, 296–320. doi: 10.2307/2391553 Porter, L. E., Steers, R. M., Mowday, R. T., and Boulian, P. V. (1 974). Organizational commitment, job satisfaction, and turnover among psychiatric tech nicians.J.

Appl. Psychol. 59, 603–609. doi: 10.1037/h0037335 Preacher, K. J., Curran, P. J., and Bauer, D. J. (2006). Computatio nal tools for probing interaction e ects in multiple linear regression, multilevel modeling, and latent curve analysis. J. Educ. Behav. Stat.31, 437–448.

doi: 10.3102/10769986031004437 Priem, R., and Price, K. (1991). Process and outcome expectations f or the dialectical inquiry, devil’s advocacy, and consensus tech niques of strategic decision making. Group Organ. Manage.16, 206–225.

doi: 10.1177/105960119101600207 Raver, J. L., and Gelfand, M. J. (2005). Beyond the individual victim: linking sexual harassment, team processes, and team performance. Acad. Manage. J.

48, 387–400. doi: 10.5465/AMJ.2005.17407904 Rispens, S. (2012). The in uence of con ict issue importance on t he co-occurrence of task and relationship con ict in teams. Appl. Psychol. Int. Rev.61, 349–367.

doi: 10.1111/j.1464-0597.2011.00473.x Rispens, S., Greer, L. L., and Jehn, K. A. (2007). It could be wors e: a study on the alleviating roles of trust and connectedness in intragroup con ict s.Int. J. Conf.

Manage. 18, 325–344. doi: 10.1108/10444060710833450 Rousseau, D. M. (1985). Issues of level in organizational rese arch: multi-level and cross-level perspectives. Res. Organ. Behav.7, 1–37.

Salancik, G. R., and Pfe er, J. (1978). A social information proce ssing approach to job attitudes and task design. Adm. Sci. Q.23, 224–253. doi: 10.2307/2392563 Schweiger, D., Sandberg, W., and Ragan, J. (1986). Group approac hes for improving strategic decision making: a comparative analysis of dialec tical inquiry, devil’s advocacy, and consensus approaches to strategic decision making. Acad. Manage. J. 29, 51–71.

Schweiger, D., Sandberg, W., and Rechner, P. (1989). Experien tial e ects of dialectical inquiry, devil’s advocacy, and consensus approaches to strategic decision making. Acad. Manage. J.32, 745–772. doi: 10.2307/256567 Simons, T., and Peterson, R. (2000). Task con ict and relations hip con ict in top management team: the pivotal role of intragroup trust. J. Appl. Psychol.85, 102–111. doi: 10.1037/0021-9010.85.1.102 Song, M., Dver, B., and Thieme, R. J. (2006). Con ict management and innovation performance: an integrated contingency perspective. J. Acad. Mark. Sci.34, 341–356. doi: 10.1177/0092070306286705 Swann, W. B. Jr., Polzer, J. T., Seyle, D. C., and Ko, S. J. (2004). Finding value in diversity: veri cation of personal and social self-views in di verse groups.Acad.

Manage. Rev. 29, 9–27. doi: 10.5465/AMR.2004.11851702 Tjosvold, D. (2008). The con ict-positive organization: it d epends on us.J. Organ.

Behav. 29, 19–28. doi: 10.1002/job.473 Walton, R. E. (1969). Interpersonal Peacemaking: Confrontations and Third Party Consultation . Reading: Addison Wesley.

Walton, R. E., and Dutton, J. M. (1969). The management of interde partment con ict: a model and review. Adm. Sci. Q.14, 73–84. doi: 10.2307/2391364 West, B. J., Patera, J. L., and Carsten, M. K. (2009). Team level positivity:

investigating positive psychological capacities and team level outcomes.

J. Organ. Behav. 30, 249–267. doi: 10.1002/job.593 Yang, J., and Mossholder, K. W. (2004). Decoupling task and relati onship con ict:

the role of intragroup emotional processing. J. Organ. Behav.25, 589–605.

doi: 10.1002/job.258 Con ict of Interest Statement: The authors declare that the research was conducted in the absence of any commercial or nancial relati onships that could be construed as a potential con ict of interest.

The reviewer, AC and handling Editor declared their shared a liati on.

Copyright © 2018 Lee, Kwon, Shin, Kim and Park. This is an open-acc ess article distributed under the terms of the Creative Commons Attribu tion License (CC BY).

The use, distribution or reproduction in other forums is perm itted, provided the original author(s) or licensor are credited and that the ori ginal publication in this journal is cited, in accordance with accepted academic prac tice. No use, distribution or reproduction is permitted which does not comply with thes e terms.

Frontiers in Psychology | www.frontiersin.org 12January 2018 | Volume 8 | Article 2365 Lee et al.Con ict and Team Commitment ANNEX Measurement of individual-level con ict and team-level con ict (task con ict).

Item 1 Individual-level How often do you disagree about opinions regarding the work being done with your team members?

Team-level How often do people in your work team disagree abo ut opinions regarding the work being done?

Item 2 Individual-level How frequently do you experience con icts about ideas with your team members?

Team-level How frequently are there con icts about ideas in y our work unit?

Item 3 Individual-level How much con ict about work do you exper ience with your team members?

Team-level How much con ict about work is there in your work un it?

Item 4 Individual-level To what extent do you experience differences of opinion with your team members?

Team-level To what extent are there differences of opinion i n your work unit?

(Relationship Con ict) Item 1 Individual-level How much friction do you experience wi th your team members?

Team-level How much friction is there among members in your w ork unit?

Item 2 Individual-level How much personality con icts do you experience with your team members?

Team-level How much are personality con icts evident in your work unit?

Item 3 Individual-level How much tension do you experience wit h your team members?

Team-level How much tension is there among members in your wo rk unit?

Item 4 Individual-level How much emotional con ict do you experience with your team members?

Team-level How much emotional con ict is there among members in your work unit?

Frontiers in Psychology | www.frontiersin.org 13January 2018 | Volume 8 | Article 2365