Leader Power, Conflict Handling Styles, and Subordinate Compliance: A Study on Information Technology Professionals in Turkey
Article Category: Research Article
Pubblicato online: 29 mar 2018
Pagine: 18 - 35
Ricevuto: 25 lug 2017
Accettato: 10 feb 2018
DOI: https://doi.org/10.2478/ijme-2018-0003
Parole chiave
© 2018 Ada T. Cenkci, published by Sciendo
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
The changing environment of the business world has created new challenges in relation to maintaining peace at the workplace. Rapid pace of business, escalated competition, workplace diversity, and flattened organizational structures are among these new challenges [Muir, 2000]. Modern organizations need to continuously learn and make changes to remain competitive, but such a change can also bring conflict within organizations. Hence, today’s organizations and their leaders are in crucial need of structures and methods to manage conflict effectively. Former studies indicate that leaders’ styles of handling conflict can influence a number of important employee outcomes, such as satisfaction with supervision and work [Richmond et al., 1983], job performance [Rahim et al., 2001], turnover intention [Chan et al., 2008], and anxiety/depression [Way et al., 2016]. Moreover, because of their work roles, employees may have different goals and behaviors, which can also lead to conflict. In such an environment, power is a vital instrument for leaders to influence others in getting things done and accomplishing organizational goals.
Power is the capability to affect others’ actions, thoughts, and emotions [Borkowski, 2011]. Leaders utilize different power bases to influence followers by attracting one or more of their needs [Hellriegel and Slocum, 2007]. Previous research has shown that leaders’ power base choices affect a number of organizational outcomes, such as the degree of cooperative behavior and organizational commitment [Munduate and Dorado, 1998], employee job stress [Erkutlu and Chafra, 2006], and organizational citizenship behavior [Reiley and Jacobs, 2016]. Hence, it is important for leaders to determine their power bases and styles of handling conflict effectively in order to maximize their influence, manage organizational conflict well, and influence employee outcomes, such as subordinate compliance. Organizations that direct these issues well and effectively manage their IT human capital can acquire distinctive competencies.
IT professionals are chosen as the focus of this research, because they have characteristics that differ from those of other professionals [Armstrong et al., 2007]. The IT professional can be described as an individual who works directly with the development, implementation, support, and/or management of computer-based information systems, in particular, software applications and computer hardware [Marchewka, 2006; Messersmith, 2007; Rose, 2009]. An IT professional job functions include positions such as information system professional, programmer, developer, and software engineer [Maudgalya et al., 2006], all requiring highly skilled employees. In general, they have a high level of intelligence and education, consequently demanding very competitive wages [Maudgalya et al., 2006]. They exhibit certain distinguishing characteristics, such as youth, mobility, short tenure, adaptability to change, and sensitivity to their work (owing to their viewing their output as an extension of their personality) [Peterson, 1987; Thite, 2006]. IT careers, more than most other professions, necessitate long hours, travel, and frequent updating of skills Ahuja, 2002.
Moreover, an operational IT infrastructure is critical in most of today’s organizations as they continuously need to implement new IT initiatives to overcome the competition. The focus on IT initiatives usually causes its professionals to work longer hours and, hence experience more work–home life conflict, compared to employees in other functional units [McGee, 2003Messersmith, 2007]. In addition to long and unfixed work hours and lack of distinction between work and home life, the following are among the stress factors affecting today’s IT work environment: growing market pressures, tight deadlines, budgetary limitations that frequently lead to understaffing (therefore, overworking), and managerial policies that do not have an understanding of the IT operations [Maudgalya et al., 2006]. For the reasons listed here, the present study examines the influence of leaders’ power bases and styles of handling conflict on subordinates’ attitudinal and behavioral compliance. The study has been conducted on Turkish IT professionals.
Pasa et al. [2001] stated that for a significant amount of time, the dominance of American management theory created the assumption that a “good” manager in the US would also be a good manager in other countries. However, differences in national culture require different management practices [Pasa et al., 2001]. This study also provides a contribution to the existing literature by examining leadership practices of the IT industry in Turkey. Furthermore, this research contributes to the literature by conducting an investigation on various aspects affecting IT professionals employed in various industries. In the Turkish context, there has been very little research covering the various industries in which IT professionals are employed.
It should also be noted that the Turkish IT market is one of the fastest developing markets in Europe [Turkey Information Technology Report, 2011], and it is expected that IT spending in Turkey will grow faster than the world average. IT spending on hardware, software, IT services, and telecommunication services in Turkey are expected to reach 35 billion USD by 2018 [ Information and Communication Technologies (ICT), no date (n.d.)]. Employment in the Turkish IT industry can be expected to accumulate as the market size becomes bigger. The expected growth in the Turkish IT industry increases the importance of this study in terms of providing insights into how to manage IT human capital effectively.
As an important force for leader effectiveness [Barrett, 2010], power can be defined as the capacity of a person, team, or organization to influence other people [French and Raven, 1959]. Leaders achieve goals, and power is an instrument for facilitating this achievement [Robbins and Judge, 2009]. Without power, it would not be possible for leaders to use their influence to get things done [Barrett, 2010]. Individuals can have power over others due to a variety of reasons, such as gender, social class, and ethnicity. The emphasis here is on the nature of power as a social resource in organizations [Hewison, 2005]. Several classifications have been made of social power bases in organizational settings. However, French and Raven’s [1959] power taxonomy (coercive, reward, legitimate, expert, and referent) has been widely accepted by scholars, being also drawn upon for this work, the categories of which are defined below.
“Coercive power” refers to the capacity of the power holder to take something away from the target person or to punish her/him for not complying with a request [Spoelstra and Pienaar, 2008]. “Reward power”, being its opposite, pertains to the ability to provide things that others want or need in exchange for desired behaviors [O’Connell and Cuthbertson, 2009]. “Legitimate power” is the authority assigned to a person with a social position within a group [ibid], while “expert power” rests upon the influencee’s belief that the influencer possesses important knowledge, information, or skills in a desirable area [Busch and Wilson, 1976]. As jobs have become more specialized, dependence on experts has increased. Specialists such as computer specialists, tax accountants, and industrial psychologists are capable of gaining power because of their expertise [Robbins et al., 2009]. Finally, “referent power” is based on identification with another person [French and Raven, 1959], and popularity or charisma is often used to describe this form of power [O’Connell and Cuthbertson, 2009]. Next, the topic of conflict is discussed.
Conflict is inevitable and unavoidable among humans. Relationships among two or more social beings (humans, groups, organizations, or nations) can become incompatible or inconsistent for different reasons. For instance, conflict can arise when these entities desire a scarce resource or when they differ in attitudes, values, and/ or skills [Rahim, 2011]. There is a wide variety of definitions for conflict. For example, Putnam and Poole [1987] define it as the interaction of interdependent individuals, who see opposition of goals as well as values and who perceive the other side as potentially obstructing the reaching of these goals. Next, how organizational members attempt to deal with conflict is discussed.
Follett [1940] was among the first to propose that there might be patterns to response conflict. She proposed that individuals can manage conflict in three main ways: domination, compromise, and integration. Blake and Mouton [1964] offered a grid to classify conflict behaviors, according to concern for production or concern for people. The five styles of handling interpersonal conflict in this model are forcing, withdrawing, smoothing, compromising, and problem-solving. Based on Blake and Mouton’s work, Rahim and Bonoma [1979] classified the handling of interpersonal conflict under two main dimensions: concern for self and concern for others. The first dimension is the extent to which an individual tries to satisfy his/ her personal concerns, while the second refers to the degree to which a person attempts to satisfy others’ concerns. The combination of these two dimensions creates five styles: integrating, obliging, dominating, avoiding, and compromising. Based on this model, Rahim [1983] developed the Rahim Organizational Conflict Inventory (ROCI), which has been widely used by researchers.
“Integrating” (or collaborative) style is associated with “high concern for self and for others” [Rahim and Buntzman, 1989, p. 197; Gudykunst, 2005]. This style emphasizes problem-solving and seeks a result that provides both sides with what they want [Phillips and Gully, 2012]. “Obliging” (or accommodating) style is characterized by “low concern for self and high concern for others” [Rahim and Buntzman, 1989, p. 197; Gudykunst, 2005Collins, O’Rourke, 2009]. This involves trying to minimize differences and emphasizing common points so as to satisfy the other party’s concerns [Rahim et al., 2002]. “Dominating” (or competitive-controlling) style involves “high concern for self and low concern for others” [Rahim and Buntzman, 1989, p. 197; Gudykunst, 2005]. This is confrontational [Collins and O’Rourke, 2009], being described as having a win–lose orientation or a forcing behavior aimed at asserting one’s position [Rahim et al., 2002]. “Avoiding” style reflects “low concern for self and for others” [Rahim and Buntzman, 1989, p. 197], which is a passive CMS that involves ignoring the conflict or denying that it exists [Phillips and Gully, 2012]. “Compromising” style involves “intermediate concern for self and others” [Rahim and Buntzman, 1989, p. 197]. This has been associated with a give-and-take concession approach, with the aim to get to a midpoint agreement [Gudykunst, 2005], and by following this style, each party sacrifices something to end the conflict. This middle-ground style shows a moderate concern for one’s personal interests and a moderate concern for those of the other party [Phillips and Gully, 2012].
Conflict management styles (CMSs) that treat the other side with a moderate-to-high level of concern, i.e., integrating, obliging, and compromising, are named as “cooperative” CMSs. Similarly, styles in which little concern is given to the other party, i.e., avoiding and dominating, are called “uncooperative” CMSs [Rahim et al., 2000Song et al., 2000; Milic et al., 2011]. Conflict handling styles can differ from culture to culture. For instance, the study by Morris et al. [1998] indicated that Chinese managers tend to use the avoiding style more because they have relatively high conformity and tradition value-orientation. However, US managers tend to use the competing style more, since they have relatively high value orientation toward individual achievement [Morris et al., 1998]. In the next section, the topic of compliance with the supervisor’s wishes is examined.
How people organize and relate to each other to accomplish planned goals is a central issue in organizational and administrative theory. The overreaching problem in organizations is securing follower compliance [Porter et al., 2003]. The compliance variable is an ideal criterion to associate with leader power bases, because it is most directly related with the outcomes of power use [Rahim and Afza, 1993]. Compliance is about achieving the result aimed for from the use of power [Fairholm, 2009]. It indicates that people are following the direction of the person of power, even though they may not agree with the orders given. Resistance, on the other hand, means that employees intentionally attempt to avoid carrying out orders or they will try to disobey instructions [Daft, 2008].
Warren [1968] differentiated between attitudinal and behavioral compliance. The former is the degree to which a target person is willing by him/herself to fulfill an actor’s wishes (with or without acting accordingly), while the latter refers to the extent to which a target person adheres behaviorally to those wishes (with or without being willing to do so by him/herself) [Emans et al., 2003]. Rahim and Afza [1993] stated that a power base is effective to the degree to which it causes both attitudinal and behavioral compliance, where the latter results from a target’s wish to get favorable reactions or to avoid the actor’s unfavorable reactions. On the other hand, attitudinal compliance is the product of an actor’s influence on the target’s self-definition, which creates real persuasion and true internal change that remains in the absence of monitoring [Moscovici, 1976; Pérez and Mugny, 1990; Pérez, 1994; Emans et al., 2003].
Next, several studies from the literature relating to “the relationship between bases of leader power and subordinate compliance” are examined. Meng et al. [2014] investigated the relationship between subordinates’ perceptions of leader power bases and the influence of these perceptions on subordinates’ attitudinal compliance, behavioral compliance, and satisfaction with supervision. This research had 86 respondents, who were postdoctoral and PhD students at a science institution in the UK. The findings revealed that subordinates perceived that the expert power base was used by their leaders most, followed by legitimate power. In addition, the outcomes indicated that a leader’s legitimate power and expert power were positively related to attitudinal compliance. Furthermore, legitimate power, coercive power, and expert power were positively related to subordinates’ behavioral compliance. Rahim and Afza’s [1993] research on 308 American accountants showed that expert and referent power bases were positively related to attitudinal compliance, while the referent and legitimate bases of power were positively related to behavioral compliance. Based on a literature review, Rahim and Buntzman [1989] concluded that expert, referent, and to some degree, legitimate power bases, in general, cause compliance from subordinates. The authors added that subordinates perceive coercive and reward power bases as weak compliance reasons. Rahim [1989] contended that the lack of consistent associations between power bases and compliance might be partially related to measurement and sampling inadequacies.
Considering the literature review provided, in this current study, the following hypotheses are formulated according to the bases of leader power and subordinate compliance:
In terms of the literature review on the “relationship between leader styles of handling conflict and subordinate compliance”, there are a limited number of studies investigating this relationship. As stated earlier, one piece of research is that by Rahim and Buntzman [1989]. This study was conducted on 301 American students of business administration, and the following findings were reported: integrating CMS was positively associated with attitudinal and behavioral compliance; obliging CMS was positively associated with attitudinal compliance; and a compromising style was negatively associated with behavioral compliance. Furthermore, as pointed out earlier in this paper, former studies have indicated that cooperative CMSs (integrating, obliging, and compromising), which focus on satisfying others’ concerns, generally produced positive outcomes for subordinates.
Hence, the following research hypotheses are formulated:
According to Rahim and Buntzman [1989, p. 197], an avoiding style involves “low concern for self and for others”, while a dominating one pertains to “high concern for self and low concern for others” [Rahim and Buntzman, 1989, p. 197]. The latter can help a person achieve individual goals, but as with the avoiding style, it is likely to result in an unfavorable evaluation by others [Singh, 2012]. In addition, former studies have indicated that an uncooperative CMS (dominating and avoiding) that ignores the needs of others is associated with negative job outcomes [Chan et al., 2008] and, thus, the following hypotheses are constructed:
Quantitative research is used in this study and it is a cross-sectional study. A questionnaire was deemed the appropriate data collection method. The model was tested using the SPSS (Statistical Package for the Social Sciences) software.
As mentioned earlier, the target population of this study is Turkey’s information technology (IT) professionals. Nonprobability techniques of convenience and snowball sampling were used to collect the data. According to Cooper and Schindler [2003], factors such as nonavailability of the population, or high costs, might lead researchers to use nonprobability sampling techniques. In this study, convenience and snowball sampling techniques were applied to increase the number of participants and to reach a variety of IT professionals employed in different industries. In total, 353 Turkish IT professionals participated in the survey over a period of 2 months. In the end, 72.5% of the people who started the questionnaire finished it; partially completed surveys were not used in the study.
A questionnaire, in Turkish, was distributed through a survey website to collect data from IT professionals. It was put online to reach IT professionals from a variety of industries and because of its convenience to the participants. It should be noted that IT professionals are considered Internet users, because it is related to their profession. The online distribution of the questionnaire might also have had benefits in terms of ensuring the confidentiality of responses, compared to distributing paper-based questionnaires in organizations. There were questions in the survey asking the participants to evaluate their leaders’ behavior and attitude and, therefore, the respondents might not have felt comfortable about filling in paper-based questionnaires at work.
The survey link was sent to the participants along with an invitation text, which included a briefing about the purpose of the study and assurances of the confidentiality of the responses. The survey invitation was sent to contacts working in various IT organizations in Turkey, and these people distributed the invitation within their organizations. The invitation to join the survey was also posted to a number of email groups related to Turkish IT professionals and groups of Turkish IT industry professional associations on an online networking site. Before distributing the questionnaire, a pilot study on 153 IT professionals was conducted. After the pilot study, the results from the data analysis were scrutinized, and the Turkish translation of the scales was reviewed. Then, as is explained in the following section, modifications were made to Rahim’s (1988) Behavioral and Attitudinal Compliance with Superior’s Wishes Scale (CSWS).
To measure IT professionals’ perceptions about their supervisors’ sources of power, Rahim’s (1988) Leader Power Inventory (RLPI) was chosen, which has 29 items. Rahim and Afza’s [1993] research provided evidence for the RLPI scale’s construct and criterion validities. Representative items include the following: “My superior has a pleasing personality”, “I approach my superior for advice on work-related problems because she (he) is usually right”, and “My superior’s position entitles her (him) to expect support of her (his) policies from me”. The instrument has five subscales: coercive, reward, legitimate, expert, and referent. The translation of the questionnaire was undertaken by the researcher, and Acar’s [2009] Turkish translation of the instrument was used in this translation as a resource. In the current study, Cronbach’s alpha coefficient was found to be 0.891 for the instrument.
To measure subordinates’ perceptions of supervisors’ CMSs with them, Rahim’s (1983) scale was used, namely, the ROCI-II, which has 28 items, but it was altered to measure subordinates’ perceptions. For instance, the item “I try to investigate an issue with my subordinates to find a solution acceptable to us” from the original questionnaire was modified as “My supervisor tries to investigate an issue with us to find a solution acceptable to us”. This alteration in wording to measure subordinates’ perspective has been used by several other studies, such as Rahim and Buntzman [1989] and Chan et al. [2008]. Sample items from the instrument are “My supervisor generally tries to satisfy our needs”, “My supervisor uses his/ her expertise to make a decision in his/ her favor”, and “My supervisor tries to avoid unpleasant exchanges with us”.
The ROCI-II instrument is widely used, and it has five independent dimensions that represent interpersonal conflict: integrating, obliging, dominating, avoiding, and compromising. In Rahim and Buntzman’s [1989] study, the values of Cronbach’s alpha for the subscales ranged between 0.64 and 0.87. Rahim and Magner’s [1995] research with five different samples supported the convergent and discriminant validities of the scale. In this current research, a Cronbach’s alpha coefficient of 0.893 was obtained for the ROCI-II scale. The translation of the instrument was undertaken by the researcher, and Sirin’s [2008] Turkish translation of the instrument was used as a resource.
Rahim’s (1988) Behavioral and Attitudinal Compliance with Superior’s Wishes Scale (CSWS) was used in this study, which has satisfactory construct and criterion validities [Rahim and Buntzman, 1989]. Representative items include the following: “I like to do what my superior suggests” and “I follow the work procedures set up by my superior”. The translation was undertaken by the researcher. After the pilot study, it was seen that Turkish translations of some items in the CSWS have very similar meanings. Hence, scale Item 1 (I follow my superior’s orders) and item 3 (I prefer not to comply with my superior’s instructions) were removed for the subsequent main study. The final instrument has eight items. In this current research, the Cronbach’s alpha coefficient was found to be 0.870 for the modified CSWS.
All scales were translated from English to Turkish. The translations of the scales were examined by two bilingual academicians, and the items were reverse-translated to compare the English and Turkish versions. Corrections were then carried out to make the questionnaire more understandable, and the responses were anchored on a six-item Likert scale (completely disagree = 1, completely agree = 6).
The respondents were also asked to provide information about themselves and the organization they worked for. Demographic questions that were asked included the following: gender, age, highest level of education attained, marital status, organization industry category, level in the organizational hierarchy (top management, middle management, or nonsupervisory employees), tenure in the company (in years), and job experience (in years).
The descriptive statistics of the sample are given in Table 1. From the table, it can be seen that the participants’ mean age was 32.1 years, and 34.8% of the respondents were female, with 49.9% being single. Most of the participants had a university degree (66.3%). The mean of the tenure at work was 4.6 years, while the mean of total job experience was 9.9 years; most of the participants were nonsupervisory employees (56.7%).
Descriptive Statistics of the Sample
| Variable | N | Percentage | Mean | Standard Deviation | Range |
|---|---|---|---|---|---|
| Gender | |||||
| Male | 230 | 65.2 | – | – | – |
| Female | 123 | 34.8 | – | – | – |
| Age | – | – | 32.1 | 7.3 | 21–61 years |
| Marital Status | |||||
| Married | 177 | 50.1 | – | – | – |
| Single | 176 | 49.9 | – | – | – |
| Education Level | |||||
| High School | 17 | 4.8 | – | – | – |
| University | 234 | 66.3 | – | – | – |
| Masters’ Degree | 99 | 28.0 | – | – | – |
| PhD | 3 | 0.8 | – | – | – |
| Tenure | – | – | 4.6 | 5.5 | 1–35 years |
| Total Experience | – | – | 9.9 | 7.7 | 1–40 years |
| Position | |||||
| Top Management | 40 | 11.3 | – | – | – |
| Middle Management | 113 | 32.0 | – | – | – |
| Nonsupervisory Employee | 200 | 56.7 | – | – | – |
The industry category of the organizations in which the participants worked was classified. It emerged that most of the participants worked in the information and communications industry (44.1%), with the financial and insurance activities industry following this with 19.0%. Next came the 12.2% of the respondents who were employed in the manufacturing industry.
Factor analysis using the principal components solution with varimax rotation was used to find the factor structure of the leader power bases scale. Four factors were found, and these explained 62.486% of the total variance. This was despite the Rahim Leader Power Inventory (RLPI) [Rahim, 1988] having five factors, as the items belonging to expert and referent power loaded on a single one and this factor was thus named “expert and referent power”. This commonality could be related to the fact that the participants see expert and referent personal power bases as being closely associated. As indicated earlier, expert power is based on the influencee’s belief that the influencer has important knowledge, information, or skills in a desirable area [Busch and Wilson, 1976], whereas referent power rests upon identification with another person [French and Raven, 1959]. IT careers, more than most other professions, demand constant updating of skills [Ahuja, 2002]. In addition, IT employees possess a strong need for growth and personal development in comparison to professionals in other occupations [Lee, 2000]. Hence, for IT professionals, having expert power might be associated with having referent power, thus causing these factors to load on one factor, “Expert and referent power”. Table 2 shows the results of the factor analysis for the bases of the leader power scale and the Cronbach’s alpha values.
Results of the Factor Analysis of the Bases of the Leader Power Scale
| Bases of Leader Power (Overall) Cronbach’s Alpha: 0.891 | Factor Loading |
|---|---|
| When a tough job comes up, my superior has the technical “know-how” to get it done. | 0.815 |
| My superior has considerable professional experience to draw from in helping me to do my work. | 0.795 |
| My superior does | 0.790 |
| I prefer to do what my superior suggests because he (she) has high professional expertise. | 0.752 |
| I approach my superior for advice on work-related problems because she (he) is usually right. | 0.733 |
| My superior has specialized training in his (her) field. | 0.713 |
| My superior has a pleasing personality. | 0.682 |
| My superior is | 0.679 |
| I like the personal qualities of my superior. | 0.677 |
| My superior can recommend a promotion for me if my performance is consistently above average. | 0.827 |
| My superior can get me a bonus for earning a good performance rating. | 0.801 |
| My superior can recommend me for merit recognition if my performance is especially good. | 0.757 |
| If I put forth extra effort, my superior can take it into consideration to determine my pay raise. | 0.721 |
| My superior can provide opportunities for my advancement if my work is outstanding. | 0.694 |
| My superior | 0.620 |
| My superior can fire me if I neglect my duties. | 0.842 |
| My superior can fire me if my performance is consistently below standards. | 0.781 |
| My superior can see to it that I get no pay raise if my work is unsatisfactory. | 0.694 |
| My superior can suspend me if I am habitually late in coming to work. | 0.632 |
| My superior can take disciplinary action against me for insubordination. | 0.569 |
| I should do what my superior wants because she (he) is my superior. | 0.804 |
| My superior’s position entitles her (him) to expect support of her (his) policies from me. | 0.702 |
| It is reasonable for my superior to decide what he (she) wants me to do. | 0.627 |
| My superior has the right to expect me to carry out her (his) instructions. | 0.581 |
Note: *Reverse-scored items. Kaiser–Meyer–Olkin value: 0.918; Bartlett significance value: 0.000;
Factor analysis was used to find the factor structure of the “styles of handling conflict with subordinates” scale. Three factors were found, and these explained 67.139% of the total variance. The factor analysis results are given in Table 3.
Results of the Factor Analysis of the Styles of Handling Conflict with Subordinates Scale
| Handling Conflict With Subordinates (Overall) Cronbach’s Alpha: 0.893 | Factor Loading |
|---|---|
| Usually proposes a middle ground for breaking deadlocks. | 0.858 |
| Collaborates with us to come up with decisions acceptable to us. | 0.856 |
| Negotiates with us so that a compromise can be reached. | 0.852 |
| Tries to work with us to find solutions to a problem that satisfies our expectations. | 0.845 |
| Generally tries to satisfy our needs. | 0.828 |
| Tries to integrate his/her ideas with our ideas to come up with a decision jointly. | 0.827 |
| Tries to investigate an issue with us to find a solution acceptable to us. | 0.826 |
| Tries to work with us for a proper understanding of a problem. | 0.825 |
| Tries to bring all our concerns out in the open so that the issues can be resolved in the best possible way. | 0.822 |
| Exchanges accurate information with us to solve a problem together. | 0.818 |
| Tries to satisfy our expectations. | 0.795 |
| Tries to find a middle course to resolve an impasse. | 0.788 |
| Accommodates our wishes. | 0.761 |
| Uses a “give-and-take” approach so that a compromise can be made. | 0.682 |
| Often goes along with our suggestions. | 0.631 |
| Gives in to our wishes. | 0.553 |
| Tries to stay away from disagreement with us. | 0.835 |
| Avoids an encounter with us. | 0.790 |
| Tries to keep his/her disagreement with us to himself/herself in order to avoid hard feelings. | 0.648 |
| Tries to avoid unpleasant exchanges with us. | 0.567 |
| Attempts to avoid being “put on the spot” and tries to keep his/her conflict with us to himself/herself. | 0.529 |
| Usually allows concessions to us. | 0.526 |
| Sometimes uses his/her power to win a competitive situation. | 0.769 |
| Uses his/her authority to make a decision in his/her favor. | 0.726 |
| Uses his/her expertise to make a decision in his/her favor. | 0.706 |
| Uses his/her influence to get his/her ideas accepted. | 0.701 |
| Is generally firm in pursuing his/her side of the issue. | 0.699 |
Note: Kaiser–Meyer–Olkin value: 0.958; Bartlett significance value: 0.000;
The first factor was composed of items from the integrating, compromising, and obliging styles. As stated earlier, CMSs that are treating the other side with a moderate-to-high level of concern, i.e., integrating, obliging, and compromising, are known as “cooperative” CMSs. On the other hand, styles in which little concern is shown for the other party, i.e., avoiding and dominating, are defined as being “uncooperative” CMSs [Rahim et al., 2000; Song et al., 2000; Milic et al., 2011]. In this current study, since the first factor was composed of items from the integrating, compromising, and obliging styles, it was named the “cooperative style”.
The ROCI-II [1983] offered five dimensions that represent interpersonal conflict: integrating, obliging, dominating, avoiding, and compromising. However, in this study, as a result of factor analysis, only three dimensions were found. This result might be due to the fact that the study was conducted with Turkish IT professionals. In other words, cultural differences or the participants’ occupational differences might have caused such a result. It should also be noted that this research involved measuring participants’ perceptions about their supervisors. There have been other studies on styles of handling conflict conducted in Turkey using ROCI-II. However, in many of these, either the participants evaluated their own CMSs (such as the studies by Ozgan [2006] and Kilic [2006]) or when the researchers did ask the participants’ perception regarding their supervisors, factorial analysis was not conducted (such as the research by Polat [2008] and Gunes [2008]). Consequently, a comparison between these listed former studies conducted in the Turkish context on the one hand and the current research on the other regarding the factor structure of the styles of handling conflict cannot be made.
Factor analysis was also used to find the factor structure of CSWS. Two factors were found, which explained 73.091% of the total variance. The first factor was composed of items mostly from behavioral compliance, just a few attitudinal compliance items. Table 4 shows the outcomes of the factor analysis scale and the Cronbach’s alpha values.
Results of the Factor Analysis of Compliance with Supervisor’s Wishes Scale
| Compliance with Supervisor’s Wishes (Overall) Cronbach’s Alpha: 0.870 | Factor Loading |
|---|---|
| I comply with the instructions of my superior. | 0.882 |
| I do what my superior suggests. | 0.859 |
| I follow the work procedures set up by my superior. | 0.850 |
| I like to do what my superior suggests. | 0.836 |
| I comply with the directives of my superior. | 0.814 |
| I prefer to follow the work procedures set up by my superior. | 0.812 |
| I prefer not to comply with the directives of my superior. Reverse-scored items. Kaiser–Meyer–Olkin value: 0.880; Bartlett significance value: 0.000; | 0.915 |
| I don’t like to follow my superior’s orders. Reverse-scored items. Kaiser–Meyer–Olkin value: 0.880; Bartlett significance value: 0.000; | 0.649 |
The overall reliability of the scale was checked and found to be 0.870. The Cronbach’s alpha value for Factor 2 (Attitudinal Compliance) was found to be 0.507. Clark and Watson [1995] recommended a minimum value of 0.15–0.20 for the mean inter-item correlation for broad higher-order constructs (e.g., extraversion) and a range of 0.40–0.50 for those tapping narrower constructs (e.g., talkativeness). In addition, the items were quite relevant and created a meaningful factor. Hence, none of the Factor 2 items were removed for the subsequent analysis.
Table 5 shows the means and standard deviations of all the scales used, as well as for their subscales. According to the findings, subordinates perceive that their supervisors used expert and referent power more than other power bases and that coercive power was the power base least used by the latter. In addition, supervisors used a cooperative style more than other styles when dealing with conflict with subordinates. Finally, the avoiding style was the least used CMS by supervisors.
Means and Standard Deviations of Scales and Subscales
| Scale | Mean | Standard Deviation |
|---|---|---|
| Expert and Referent Power (Factor 1) | 3.877 | 1.232 |
| Reward Power (Factor 2) | 3.547 | 1.239 |
| Coercive Power (Factor 3) | 3.510 | 1.056 |
| Legitimate Power (Factor 4) | 3.611 | 0.962 |
| Cooperative Style (Factor 1) | 3.713 | 1.120 |
| Avoiding Style (Factor 2) | 3.171 | 0.921 |
| Dominating Style (Factor 3) | 3.331 | 1.169 |
| Behavioral Compliance (Factor 1) | 4.184 | 0.913 |
| Attitudinal Compliance (Factor 2) | 4.212 | 1.114 |
Table 6 shows the correlations among the study variables.
Correlations Among Study Variables
| Variables | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 |
|---|---|---|---|---|---|---|---|---|---|
| – | 0.662 Correlation is significant at the 0.01 level (two-tailed) | –0.062 | 0.375 Correlation is significant at the 0.01 level (two-tailed) | 0.767 Correlation is significant at the 0.01 level (two-tailed) | 0.401 Correlation is significant at the 0.01 level (two-tailed) | –0.512 Correlation is significant at the 0.01 level (two-tailed) Correlation is significant at the 0.01 level (two-tailed) | 0.425 Correlation is significant at the 0.01 level (two-tailed) | ||
| – | –0.013 | 0.255 Correlation is significant at the 0.01 level (two-tailed) | 0.668 Correlation is significant at the 0.01 level (two-tailed) | 0.315 Correlation is significant at the 0.01 level (two-tailed) | –0.435 Correlation is significant at the 0.01 level (two-tailed) Correlation is significant at the 0.01 level (two-tailed) | 0.275 Correlation is significant at the 0.01 level (two-tailed) | |||
| – | 0.322 Correlation is significant at the 0.01 level (two-tailed) | –0.104 | –0.144 Correlation is significant at the 0.01 level (two-tailed) | 0.279 Correlation is significant at the 0.01 level (two-tailed) Correlation is significant at the 0.01 level (two-tailed) | 0.069 | ||||
| – | 0.294 Correlation is significant at the 0.01 level (two-tailed) | 0.248 Correlation is significant at the 0.01 level (two-tailed) | –0.018 0.668 Correlation is significant at the 0.01 level (two-tailed) | 0.273 Correlation is significant at the 0.01 level (two-tailed) | |||||
| – | 0.576 Correlation is significant at the 0.01 level (two-tailed) | –0.615 Correlation is significant at the 0.01 level (two-tailed) Correlation is significant at the 0.01 level (two-tailed) | 0.304 Correlation is significant at the 0.01 level (two-tailed) | ||||||
| – | –0.292 Correlation is significant at the 0.01 level (two-tailed) Correlation is significant at the 0.01 level (two-tailed) | 0.033 | |||||||
| – –0.127 correlation is significant at the 0.05 level (two-tailed). | –0.340 Correlation is significant at the 0.01 level (two-tailed) | ||||||||
| - | 0.384 Correlation is significant at the 0.01 level (two-tailed) | ||||||||
| - |
Before the regression analysis, the linearity of the model was checked by looking at the scatterplot matrices. The data were also examined to see whether the errors were normally distributed and the variances of the residuals were constant. The residual scatterplots were also checked and, overall, it was seen that these assumptions were met. In addition, multicollinearity was tested for by examining the variance inflation factor (VIF) values, eigenvalues, and the collinearity diagnostics table. Table 6 indicates that there are significant correlations between some of the independent study variables (such as “expert and referent power” and cooperative CMS). However, the additional multicollinearity tests listed indicate that multicollinearity was not a problem for this research.
Regression analyses were conducted between the independent and dependent variables, the results of which are shown in Table 7. In this table, only the significant results are listed. For the first analyses, the independent variables are CMSs and bases of leader power, and the dependent variable is compliance with the supervisor’s wishes. The findings indicate that cooperative (Factor 1) and dominating (Factor 3) CMSs, “expert and referent power” (Factor 1), and legitimate power (Factor 4) explain the variance of subordinate behavioral compliance. Legitimate power has the greatest explanatory power on behavioral compliance, compared to the other independent variables (β= 0.497).
Regression Analyses for Compliance with the Supervisor’s Wishes
| Dependent Variable: Behavioral Compliance (Factor 1) | |||
|---|---|---|---|
| Independent Variables: | Beta | t | p One-tailed |
| Cooperative Style (Factor 1) | 0.274 | 4.629 | |
| Dominating Style (Factor 3) | 0.178 | 3.922 | |
| Expert and Referent Power (Factor 1) | 0.248 | 4.430 | |
| Legitimate Power (Factor 4) | 0.497 | 12.946 | |
| R | |||
| t | p One-tailed | ||
| Avoiding Style (Factor 2) | –0.212 | –4.151 | |
| Dominating Style (Factor 3) | –0.24 | –4.137 | |
| Expert and Referent Power (Factor 1) | 0.310 | 5.097 | |
| Legitimate Power (Factor 4) | 0.205 | 3.987 | |
| R |
Moreover, the outcomes pointed to avoiding (Factor 2) and dominating (Factor 3) CMSs, “expert and referent power” (Factor 1), and legitimate power (Factor 4) explaining the variance in subordinate attitudinal compliance. Avoiding and dominating CMS factors are negatively associated with attitudinal compliance, while the “expert and referent power” factor has the greatest explanatory power regarding attitudinal compliance compared to the other independent variables (β= 0.310).
The first research hypothesis is “supervisors’ legitimate power positively influences subordinate behavioral compliance (1a) and attitudinal compliance (1b)”. This is accepted, and this underlines the importance of legitimate power in getting subordinate attitudinal compliance and behavioral compliance. The outcomes indicate that leaders’ legitimate power (meaning power based on their respective positions) is an effective tool to get IT professionals’ attitudinal compliance and behavioral compliance, simultaneously. The finding is in line with former studies. For instance, Meng et al.’s [2014] study on postdoctoral and PhD students at a science institution in the UK pointed to leader’s legitimate power being positively related to subordinates’ attitudinal and behavioral compliance.
The second hypothesis is “expert power positively influences subordinate behavioral compliance (2a) and attitudinal compliance (2b)”, while the third is “supervisors’ referent power positively influences subordinate behavioral compliance (3a) and attitudinal compliance (3b)”. During the factor analysis stage, the items belonging to expert and referent power loaded onto one factor, named as “expert and referent power”. Thus, the second and third hypotheses were not tested. The outcomes indicate that supervisors’ “expert and referent power” (that is, their expertise, special skill, knowledge, follower attention, respect, and/ or admiration) positively influences employee behavioral and attitudinal compliance. Such results are in line with hypotheses (2a), (2b), (3a), (3b), and former literature. For instance, based on a literature review, Rahim and Buntzman [1989] concluded that, in general, expert, referent, and to some degree, legitimate power bases receive subordinate compliance. Moreover, Palenzuela [2001] investigated prekindergarten teachers’ perceptions of their supervisory relationship with their educational specialists. The outcomes show that the expert, legitimate, referent, and informational power bases of these specialists have the most influence on the attitudinal and behavioral compliance of teachers.
As a result of the factor analysis, it was seen that the integrating, compromising, and obliging CMS items loaded on one factor, named as “cooperative style”. Hence, hypotheses 4, 5, and 6 were not tested. The findings indicate that a cooperative style positively influences employee behavioral compliance, which is in line with the results of former studies. Chan et al. [2008] point out that previous research has found that managers’ cooperative CMS (integrating, obliging, and compromising) generally produces positive subordinate outcomes. The findings imply that when supervisors use more cooperative styles, IT professionals comply with their wishes more.
Hypothesis 7 is “supervisors’ avoiding CMS negatively influences subordinate attitudinal compliance” and hypothesis 8 is “supervisors’ dominating CMS negatively influences subordinate attitudinal compliance”. These hypotheses are both accepted. An avoiding style is associated with “low concern for self and for others” [Rahim and Buntzman, 1989, p. 197], while a dominating one involves “high concern for self and low concern for others” [Rahim and Buntzman, 1989, p. 197]. Dominating and avoiding styles are likely to result in an unfavorable evaluation by others [Singh, 2012]. In addition, as stated earlier, former studies indicate that managers’ demonstration of dominating and avoiding CMSs is related to negative employee outcomes [Chan et al., 2008]. Thus, it can be said that the hypothesis testing results are in line with results from former studies.
Moreover, even though it was not hypothesized, the outcomes indicate that dominating CMS positively influences employee behavioral compliance, while it negatively influences attitudinal compliance. Furthermore, dominating CMS may increase conflict, and the losing party could try to retaliate [Phillips and Gully, 2012]. Hence, supervisors should use dominating CMS with caution. In addition, according to the research outcomes, subordinates perceive that their supervisors use a cooperative style more than other styles in dealing with conflict with their subordinates. The cooperative style is composed of items from the integrating, compromising, and obliging styles. Furthermore, the IT professionals indicated that their supervisors exerted “expert and referent power” more frequently than the other power bases. Moreover, it was reported that coercive power is the power basis least used by supervisors. Another research work conducted in the Turkish context was by Kozan et al. [2014], which involved the examination of the power bases and managerial intervention strategies vis-à-vis subordinates’ conflicts on a sample of 39 supervisors and 165 subordinates. This study was conducted with employees from different organizations in a variety of industries, including the textile, banking, mining, and government ministries. The contrasting findings indicated that supervisors use legitimate and expert power most, followed by referent power. The higher use of legitimate power seen in Kozan et al.’s [2014] study, compared to this current research, may be due to the differences in the research samples. In other words, their research did not involve targeting a specific group of professionals in Turkey, while this study was singly aimed at IT professionals. The results were expected because, in general, IT professionals are highly skilled and educated [Maudgalya et al., 2006], with a high need for learning and they possess a strong desire to be challenged [Lee, 2000]. Hence, IT supervisors may well prefer to use “expert and referent power” and the cooperative CMS more frequently to manage such a group of workers than their counterparts in other sectors.
While leadership, social power, and conflict management are widely researched concepts, these aspects still need further attention by scholars and practitioners alike, because they are highly critical issues for organizations. These matters are especially important for IT professionals, because, as stated earlier, they possess characteristics (such as high skill levels, education, and intelligence) that differ from those in many other professions [Maudgalya et al., 2006; Armstrong et al., 2007]. In addition, an operational IT infrastructure is highly important for competing in today’s business world.
The results of the study partially support the influence of leader power bases and styles of handling conflict on subordinate compliance. Furthermore, the outcomes underline that leaders’ expert and referent power, as well as legitimate power, need special attention. Leaders of IT professionals can use these findings in different ways. For instance, IT managers could use a combination of their expert and referent power bases to accomplish behavioral and attitudinal compliance of the employee. To utilize their expert power base, they could make their expertise easily accessible by subordinates when needed and recommend solutions to problems that employees are facing. In addition, it would be beneficial for IT managers to develop a good reputation for their work-related expertise. Moreover, to make use of a referent power base at work, organizations can hire IT supervisors who have charisma and who can get admiration, attention, and respect from IT employees. In addition, supervisors can aim to be role models or mentors to their subordinates. Legitimate power, which is the ability to influence because of a formal position [Spoelstra and Pienaar, 2008], can also be used to influence employee attitudinal and behavioral compliance.
In terms of legitimate power, it should be noted that today’s IT work environments require flexible work schedules [Richard, 2009] and have a less-distinct legitimate power structure. IT work environments vary from industry to industry, and in some branches, such as the software industry, there tends to be a laid-back, flexible workspace [Careers in Information Technology, 2009]. The lines of legitimate power are blurred in more-organic type of organizations such as these, and an employee may work for more than one boss at the same time. In addition, the leaders and subordinates may have almost equal organizational standing [Griffin and Moorhead, 2012]. Such issues may create challenges in terms of application of legitimate power to get IT employee compliance. Hence, these organizations could use different practices to make the authority of IT managers’ clearer, such as writing more detailed job descriptions for them and their subordinates. It would also be beneficial to define clearly the job responsibilities of IT employees who work on different projects at the same time, under the supervision of different managers.
Moreover, the research outcomes show that special attention should be given to the leader’s cooperative CMS. Cooperative CMSs (integrating, obliging, and compromising) focus on satisfying others’ concerns [Chan et al., 2008]. IT supervisors could apply cooperative CMS to achieve employee behavioral compliance. IT leaders might utilize cooperative CMS with subordinates by applying techniques such as openness, exchange of information, minimizing differences, and emphasizing common points, or even making sacrifices, depending on the situation. Furthermore, leaders’ avoiding and dominating CMSs should be used with caution, because both of these styles negatively influence subordinate attitudinal compliance. However, such CMSs could be used when the situation requires them. For instance, a downsizing situation may require a manager to utilize dominating CMS, and the supervisor might be obligated to make decisions that will not be well received [Collins and O’Rourke, 2009]. In contrast, an avoiding CMS may be beneficial when issues are not important or when the costs related with challenging someone outweigh the benefits [Tosi and Plati, 2011]. Moreover, IT leaders are encouraged to assess their preferred power bases and CMSs with subordinates, as well as the influence of these factors on subordinate compliance, in order to increase their leadership effectiveness. For such assessment, organizations can offer leaders development opportunities, such as executive coaching or management development training.
With these findings, this research has contributed to the literature in several different ways. As stated earlier, different cultures necessitate different managerial practices [Pasa et al., 2001]. This research has provided contributions to the existing literature by examining leadership practices in the Turkish IT industry. In addition, as previously discussed, there is scarce empirical research that examines the relationship between leader styles in relation to handling conflict and subordinate compliance. This study has added to the literature by investigating the association between these variables. Moreover, in the Turkish context, there has been limited research conducted on different industries in which IT professionals work. This study questionnaire has been administered to IT professionals employed in various industries. Furthermore, the Turkish IT market is one of the fastest-developing ones in Europe [Turkey Information Technology Report, 2011] and, hence, this situation increases the importance of this study.
Some of the limitations of this study also need to be noted. First of all, the data were collected through responses to questionnaires from IT professionals with self-reported measures. Hence, the answers represented the perceptions of employees, such as those about supervisors’ use of power bases or conflict handling styles with subordinates. This is a limitation, because both the independent and dependent variables were collected from the same source, which could have led to common method variance due to single-source bias. Second, convenience and snowball sampling were used to collect data, which might limit the generalizability of the survey results. Moreover, the data collection was conducted through an online survey website. Even though using the Internet ensured the confidentiality of responses and helped reaching a wide range of participants, it also limited tracking the number of leaders who were evaluated and the names of the organizations in which the participants work. However, asking questions that could have identified the respondents’ leader or organization name would have restricted their willingness to complete the questionnaire. Furthermore, it is common for IT professionals to work on more than one project at the same time [Schwalbe, 2011]. Thus, if a respondent has more than one supervisor, the IT professional may hesitate about which one of the supervisors he/ she should consider when answering the questionnaire. These issues should be taken into account in future studies deploying the research model.
Some aspects of this study need to be investigated with further research. Future studies could test alternative models with additional variables, such as organizational climate, different leadership styles (e.g., transactional or transformational), employee role ambiguity, role conflict, or intention to quit. In addition, the sample size could be increased to test the validity of the hypothesized model. Moreover, in this research, participants employed in different industries completed the survey instrument. This study could thus be replicated to conduct industry-specific analyses (such as for the financial activities, manufacturing, or education industries).
In this research, to measure leaders’ styles of handling conflict with subordinates, the ROCI-II (1983) was used. After conducting the data analysis, it was seen that some of the items in the scale did not aggregate under the proposed subscales. As indicated earlier in this study, such a result might be due to cultural differences or participants’ occupational differences. In future studies, additional modifications could be made to the scale to account for the Turkish context and the IT profession. Furthermore, for this research, questionnaires were used to collect data. Future studies could also use qualitative research methods (such as interviews or focus groups) to generate more insights and expand our understanding on the subject. In addition, future studies on this topic, especially in non-Western settings, would prove fruitful.
Overall, the outcomes of this study can benefit organizational leaders and human resource practitioners in terms of understanding the influence of leaders’ bases of power and styles of handling conflict on IT professionals’ attitudinal and behavioral compliance. The results may also be useful for organizations in increasing leadership effectiveness and delivering better management of IT human capital.