Development and validation of the leader–member exchange scale in the Indian context

Abstract

Purpose

The primary objective of the present study was to develop an appropriate scale for evaluating LMX by investigating how individuals personally perceive and encounter distinct relationships (both high quality and low quality) with their supervisors, with a specific focus on the Indian context.

Design/methodology/approach

The scale was administered on a sample of 290 middle-level managers from two large manufacturing organizations located in North India.

Findings

The factors identified as important for the construct of leader-member exchange were affect, loyalty, and contribution.

Research limitations/implications

The internal consistency reliability of the LMX contribution dimension is very low. Future researchers should add a few additional items to increase the reliability of the contribution scale of LMX scale so that it fulfills adequate criteria of reliability. Further, the supervisor–subordinate relationship from both supervisor and subordinate perspectives should also be examined.

Practical implications

This study has made significant advancements in the field of LMX. The findings will also be utilized by the authorities of the organization in focusing future training for its managers.

Social implications

The findings of this research will help not only advancement in the field of LMX but will also help the manager using LMX to influence subordinates to have better knowledge on which factors to focus on to get better results.

Originality/value

Overall, the results of the current study provide evidence for the sound reliability and validity of the leader-member exchange scale with employees of Indian manufacturing organizations, supporting its use with these populations. Further, this scale is suitable not only in Indian culture but also in the Western cultural context, as the results corroborate the findings of Western scholars, indicating a fair level of cross-cultural validity. However, future research should also address the cross-validation of the factor structure of LMX on other samples and occupations.

Citation

Srivastava, U.R., Mohaley, S., Jaiswal, A. and Singh, M. (2024), "Development and validation of the leader–member exchange scale in the Indian context", IIMT Journal of Management, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/IIMTJM-11-2023-0047

Introduction

In today’s dynamic environment, both leaders and the members of their teams form an essential part of any organization. The success and growth of any organization markedly depend on competent leadership. The process of obligating a group of people to specific objectives in order to accomplish long-term goals is known as leadership. Sage Valmiki portrayed the analogy of an organization without a leader to a group of livestock without a shepherd, an army without a commander, or a night without the moon in his Sanskrit epic from ancient India: The Ramayana. After many years of leadership research and thousands of studies on the issue, the real meaning of leadership remains still elusive. Unlike many other theories of leadership which are either leader-centered or follower centered, leader-member exchange (LMX) is a theory that emphasizes relationships. It describes the two-fold relationship between leaders and followers encompassing elements of mutual faith, respect, and mutual obligation (Hoye, 2004; Schyns, 2006). Several academicians have empirically demonstrated and argued that the relationship between subordinates and managers determines subordinates’ behavior and attitudes (Manzoni and Barsoux, 2002) and this leads to several individual specific and administrative or organizational outcomes. LMX theories have very often studied the specific relationship between immediate officers/superiors (leaders) and followers (members).

One of the main features of leader-member exchange theory is that leaders find few of their subordinates as being more consistent regarding their behavior, interpersonal and communication skills, or job competencies. As a result, leaders do not interact with all of their subordinates in the same manner; rather, they develop unique relationships and exchanges with each of their subordinates. Several exchange studies have shown that the LMX theory has originated from three main theories: dyad linkage, social exchange theory, and role theory. The LMX theory, formerly known as the Vertical Dyad Linkage (VDL) theory, was acknowledged in the exemplary writings of Graen and his associates (Dansereau et al., 1973; Dansereau et al., 1975) around 45 years ago. In the VDL model, it has been argued that a leader's behavior can vary greatly depending on the members of their work group.

The VDL approach argues that a two-fold relationship is formed between a leader and his subordinate and that the leader has a unique and different relationship with each of his subordinates. This behavior of the leader, where he treats each of his subordinates differently, leads to the formation of two groups of subordinates, namely an in-group and an outgroup. The in-group is made up of a select group of dependable followers, with whom the leader typically develops a unique, more cohesive, trusting, and higher-quality exchange relationship. The leftover followers, with whom the leader still maintains a more formal relationship, make up the outgroup. The two-way relationship between a leader and subordinate can broadly be categorized as low-quality, also known as outgroup, and high-quality, also known as in-group, relationships (Graen and Uhl-Bien, 1995; Uhl-Bien et al., 2000). In-group members who have high exchange quality relationships with their leaders have several advantages over outgroup members. They receive formal and informal rewards, trustworthiness, increased financial, emotional, and instrumental support, and increased interpersonal relationships (Dienesch and Liden, 1986; Liden et al., 1997). On the other side, outgroup members with low exchange quality relationships with their leaders receive fewer advantages beyond the conditions of a formal employment contract. They are not considered to be trustworthy and do not receive any kind of support (Bang, 2008; Hogg et al., 2005).

Today, this kind of relationship between superiors (leaders) and their subordinates (followers) forms a very crucial aspect in establishing long-lasting and positive workplace relations (Wang et al., 2005). This differing relationship is also formed due to limitations of time and energy. Due to limited time, leaders develop close working relationships with a few of their followers, sharing their personal resources only with them to perform their tasks. Therefore, this discrimination on the part of the leader eventually relegates the member to the outgroup (Harter and Evanecky, 2002). Moreover, the inequality becomes visible when a member of the group considers this division as biased and unjust. Members remain in their groups irrespective of their talents, efforts, or capabilities. It is found that this differentiation obviously affects productivity (Northouse, 2007), morale, and overall organizational effectiveness (Harter and Evanecky, 2002). Therefore, using the concept of leadership making as proposed by Graen and Uhl-Bien (1995), leaders should develop high-quality relations with all of their followers rather than just a few. In other words, working with each one of them as partners is required, especially in a country like India, where teams are preferred over individuals. Therefore, when leaders treat all their followers alike, the LMX process is perceived to be more equitable. But sometimes it is also not possible for a leader to treat everyone in the same way, as there are few individuals that have a great relationship with the leader, are more trustworthy, work much harder, and never let leaders down. Still, as a leader, they should make a reasonable effort to reestablish a relationship with outgroup members. Leaders should give a fair chance to each and every member of the group, so as to reduce the chances of conflict between members. Even if leaders do not trust a few of the members completely, they should not make it very evident in front of their followers, making them feel less worthy to be a part of the group. Substantial research has consistently highlighted that high-quality LMX relationships result in positive individual and organizational outcomes such as higher performance (Liden et al., 1993), fostering organizational commitment (Schriesheim et al., 2000), increment in job satisfaction (Liden and Graen, 1980; Hooper and Martin, 2008), promoting organizational citizenship behaviors (Hofmann et al., 2003; Bhal and Ansari, 2000), influencing perceptions of organizational justice (Rupp and Cropanzano, 2002; Wayne et al., 2002), increasing perceived organizational support (Wayne et al., 2002), stress reduction (Bernas and Major, 2000), and decrement in intentions to quit (Vecchio, 1985).

Uni-dimensionality vs multi-dimensionality of LMX construct

Most of the earlier research on LMX, while studying the impact of LMX on subordinate outcomes, treated it as a unidimensional construct confined to only job-related tasks, leaving no scope for assessing any kind of social exchange among individuals. As a result, unidimensional LMX measurements were used in the earlier research projects. These unidimensional measures were very narrow and simplified descriptions not adequately represent the interpersonal process between leaders and subordinates. The simplicity of using shorter, unidimensional scales, lack of awareness, interest, and curiosity in researching the various dimensions, efforts needed in explaining various dimensions of the LMX relationship, and psychometric issues with existing multidimensional measures were some likely causes of measuring LMX utilizing unidimensional measures as proposed by Graen and Uhl-Bien (1995) and Schriesheim et al. (1999).

This has sparked a significant debate about whether LMX is best understood and assessed as a unidimensional or multidimensional conception (Dienesch and Liden, 1986; Graen and Uhl-Bien, 1995; Liden and Maslyn, 1998). Later on, it was proposed by Dienesch and Liden (1986) that LMX was essentially a multidimensional construct. In this line, various researchers like Liden et al. (1997), and Bhal and Ansari (1996) have proposed that exchange relationships between leaders and subordinates in organizations seem to be multidimensional in nature. As a result, Dienesch and Liden (1986) and Liden and Maslyn (1998) proposed a multidimensional measure of LMX and the LMX relationship's quality actually has three components:

Affect

According to Dienesch and Liden (1986) affect can be defined as shared liking, mutual attraction, and affection between members of a dyadic relationship which is more directed by interpersonal members’ attraction rather than by professional or work-related ethics. This dimension holds significant theoretical importance within these relationships, particularly in high-quality LMX scenarios. However, it is important to understand that not all high-quality LMX relationships will essentially reveal high levels of affect, because differential relationships may emphasize various dimensions of LMX, with some being more prominent than others. Despite this proposition, several academicians have strongly supported affect as the most crucial dimension of LMX relationships (Liden et al., 1993).

Loyalty

In words of Dienesch and Liden (1986), loyalty is defined as the degree to which a leader (immediate officer/supervisor) and a member (subordinate) are supportive to each other’s behavior, attitudes, and ethical values. According to Scandura et al. (1986), loyalty denotes the feelings individuals have about their relationship, and how it affects how leaders encourage followers to work and take on various activities.

Contribution

Dienesch and Liden (1986) defined contribution as the ways through which each member of the dyad perceives and evaluates the extent, direction, and quality of the work-related efforts they contribute toward achieving the shared goals, whether those goals are explicit or implicit in nature. According to Graen (1976), contribution, though it may always be explicitly described as an LMX relationship quality dimension, theory shows that actions related to work affect the formation of quality relationships of the LMX. Despite the fact that researchers have consistently mentioned the contributions of both leader and the member, members’ behaviors have been the main focus of both theory and measurement.

Moreover, apart from these three dimensions, researchers have found a strong support for professional respect as the fourth major dimension. This dimension indicates the extent to which every member of the dyad has established prestige for his/her excellence in work, maintained in and out of the organization (Liden and Maslyn, 1998). The below Supplementary table (Table S1) provides summary of existing measures of LMX found in the research literature.

Relevance of LMX theory in the Indian context

Each culture is characterized by its own values, customs, traditions, and norms, which determine the attitudes and behaviors of a people. India is one of the most diverse lands found anywhere, all across the globe, with 29 states, 121 languages, and 1,369 dialects (Census of India, 2011), each with their own unique traditions and religions. Indian culture is very diverse in terms of multiculturalism, varied traditions and customs, languages, religions, food habits, regional practices, and collectivism (Budhwar and Varma, 2010). Indian society is strongly collectivistic, with a focus on relationship orientation in terms of higher unity and harmony, a stronger sense of empathy, increased feelings of support, an emphasis on fulfilling one's responsibilities to the group, and a belief in hierarchy. There is always “we” consciousness rather than ‘‘I” consciousness in people of collectivistic cultures.

Diversity, collectivism, and hierarchy have all been indicated as important characteristics of Indian work culture by a number of Indian researchers (Budhwar and Varma, 2010; Varma et al., 2005; Sinha, 1984). The concept of LMX as developed in Western individualistic culture is as valid in collectivistic culture or not emerges as an important research question for researchers all across the globe. Academicians are continually interested in exploring its antecedents and outcomes of LMX in Western individualistic cultures and collectivistic cultures.

In this streamline, Anand et al. (2011) conducted LMX research in Asia and other regions of the world and observed that LMX may function differently in cultures that are more collectivistic and have higher power distances. Supporting this finding, Rockstuhl et al. (2012) in a meta-analytic study of antecedents of LMX across 23 countries have also highlighted that the results of LMX are significantly connected to those in more collectivistic and higher power distance cultures.

In India, collectivist relationships are centered around the in-group, and a strong “in-group” attitude is developed and cultivated from an early age of life. The members of larger in-groups or collectives that take care of them in reciprocation for their loyalty. Right after birth individuals are integrated into strong, cohesive, and robust in-groups, most often extended families (which include maternal and paternal grandparents, uncle, aunt, etc.), that continuously defending them in reciprocation for their loyalty, and resist other in-groups. This goes parallel to the process of isolation and indifference toward the outgroup (the other community or residential location). To a large extent, in-group collectivism put more emphasis on the extent of cohesiveness, pride, and loyalty a member expresses toward their social group (Waldman et al., 2006). Likewise, the Indian business and work culture is also characterized by relationships, collectivism, and hierarchy. Researchers have asserted that in a collectivistic work culture, members of in-group are more likely to develop robust working relationships with their leaders, and in exchange they may derive an advantage in terms of more rewards than members of the outgroup (Liden and Maslyn, 1998). In India, power is a key factor in society, and with high power distance (Hofstede, 2001), older people are both respected and feared by the people.

In Indian society, an individual’s age is often linked with wisdom, knowledge, and insights, which are acquired through several years of professional and working experience. For example, the chairman, CEOs, or administrators at every level in the broader community and society are thought to have deeply rooted wisdom in terms of vision and mission, practical skills, experiential learning, recognition and respect for varied perspectives, and the necessity to balance customs or traditions with innovation. In Indian organizations, people have a natural disposition to seek guidance, information, and support from their senior employees. Their advice is often sought when required to take important decisions or navigate complex work challenges. As a result, respect for seniors with extraordinary working experience retains substantial importance and is profoundly ingrained in the work culture of many Indian organizations. This shows that Indian society accepts an unequal, hierarchical power distribution, and that individuals are aware of their position in the system (Juhasz, 2014).

This respect for seniority extends beyond the workplace to personal lives too and reflects the broader cultural values of respect and regard for older generations. In contrast to US organizational settings where employees are more accustomed to delegation and participative leadership, Indian people are more accustomed to respect for authority, and hierarchical structure is significant in most of the Indian organizations.

Employees are expected to show deference to their superiors and address them with professional, formal labeling, and honorifics. According to Sinha (1984), “subordinates readily accept the boss's higher status and respect and obey him provided he extends support to them and caters to their personal, and social needs” (p. 87). Sinha (1984) has given the model of Nurturant-Task Leadership in the Indian context. According to this concept, the ideal leader is both nurturing and task-oriented. Leaders who are reflective, contemplative, and considerate toward their subordinates demonstrate the nurturing dimension of leadership. According to Sinha (1984), nurturance is conditional upon the completion of the task.

Thus, the leader is concerned if the subordinate is task-oriented and works hard. Similar to many non-Western cultures, a paternalistic attitude is part of the leadership process under Nurturant-Task Leadership. In one study, Thakur (2010) asserted that Indian employees are highly sensitive to the rank/position of people in organizations, and such understanding and cognizance mold their behavior toward it. Hierarchy is crucial in maintaining the organizational structure. At work, they are accustomed to a system of hierarchy where superiors are compiled and respected even knowing that they may not always make wise decisions. Thus, it is clear that the hierarchical system in organizations indicates appreciation for the seniority and expertise of people in higher positions. Thus, the need to develop high quality LMX is more crucial in the Indian context. In this streamline, Indian researchers have emphasized that the effectiveness of leadership depends on the nature of the leader–subordinate relationship (Sinha, 1984). In a very recent study, Shah and Patki (2020) asserted that LMX has been regarded as a very significant theory to understand the overall context of Indian leadership. The LMX construct is particularly important in the collectivistic culture of India because of the person-oriented nature of Indian society. Hence, studying the leader–subordinate relationship in the Indian context is very relevant.

Objective of the study

The inclusive or overall purpose of this study is to advance the research on LMX by exploring how individuals themselves perceive and experience differential relationships (high quality and low quality) with their supervisors and by developing a suitable scale for the assessment of LMX with special reference to the Indian context that could enhance future LMX research. The dimensions proposed by Liden and Maslyn (1998) were taken into consideration.

Development of the LMX scale

Overview and planning phase

The four dimensions proposed by Liden and Maslyn (1998) were followed in the present study. While examining the multidimensionality of the LMX construct in the Indian context through qualitative investigation, Srivastava (2012) have found that the kinds of relationships and the kinds of work-related resources exchanged were the two major themes that emerged from the analysis. The findings further indicated that the participants’ descriptions of their relationship with their supervisor/immediate officer included affect, loyalty, perceived contribution, and professional respect dimensions as proposed by Liden and Maslyn (1998). Taking this into consideration the findings of a qualitative study conducted by Srivastava (2012), theoretical and empirical extensive review of LMX literature, a review of existing measures of LMX a pool of items was constructed. A total of 54 items were written to reflect the four dimensions (affect, loyalty, perceived contribution, and professional respect). The items were reviewed carefully for confusing grammar and terminology. The selection and wording of items are based on clear theoretical formulations that is consistent with the literature on the LMX construct.

A panel of five experts (specializing in OB and leadership research) judged the items. Experts were recommended to drop, change, or mark unclear items. Using the data from the experts’ ratings, decisions were made as to whether items should be retained, revised, or discarded. Thus, the initial item pool was reduced to 27 items after discarding items containing difficult words or phrases the items judged by the experts as not relevant to the content areas thought to represent the theoretical domain. The twenty-seven items that were included in the final data collection were based on the agreement of all the experts that the item fitted well to a precise dimension of LMX. This task was done to enhance surety that the items included in the final analysis were precisely reflecting the underlying dimensions and theory of LMX. The resulting 27 items, dimension-wise, were distributed as follows: affect (ten items), loyalty (six), professional respect (six), and contribution (five). Every item was rated on a five-point rating scale from strongly disagree (1) to strongly agree (5). The reverse scoring pattern was used for false-keyed items. Higher scores on the scale were indicative of high-quality LMX relationships.

Participants

The participants of the present study comprised of 290 middle-level managers from two large manufacturing organizations located in North India (Bokaro Steel Plant, Bokaro, and Hindustan Aeronautics Limited, Lucknow). The participants of both of these organizations consistently worked in teams and interacted on a regular basis. Many times they were taking support of their supervisors in order to complete many elements of their jobs. Thus, it is apparent that the LMX relationship was an important aspect for participants of both of these organizations to get their jobs done. Initially, approximately 365 questionnaires were administered to the target sample, out of which 290 (79.45%) were regarded as complete.

Table 1 represents the demographic characteristics of the sample.

Control variables

The demographic factors that were taken into account in the present study are as follows: gender, age, marital status, income, job designation, educational qualification, coworkers, experience, and size of the team of the organization.

In recent years, researchers have emphasized that the above-mentioned sociodemographic influence the attitudinal and behavioral responses of the employees and organizational characteristics variables (Bernerth et al., 2007). In the context of assessing LMX, Liden and Graen (1980) addressed the issue of collecting data on structural and demographic data, indicating such information is used to analyze, measure, and detect differences between manager and subordinate.

Procedure

A set of structured questions along with a letter requesting participation were distributed to every employee. To obtain formal permission for the study, the researchers contacted the organization's human resource (HR) managers. To secure collaboration, the HR managers briefly explained the objectives of the study to the possible respondents. The respondents completed the survey during their work hours and were assured about confidentiality of information.

Data analysis

The data analysis was conducted in four stages:

• (1)

  In the first phase, item total correlation was performed to determine how well each item contributed to the measurement of its respective dimension as operationally defined.

• (2)

  In the second phase, exploratory factor analysis (EFA) using varimax rotation was performed on 20 items.

• (3)

  In the third phase, confirmatory factor analysis (CFA) was used to analyze whether the data is fitting to a hypothesized measurement model based on prior research or theory and to examine the goodness of fit of the finally selected items of the scale with the proposed dimension of LMX and competing models.

• (4)

  Finally, Cronbach’s alpha coefficients and descriptive statistics, including mean, SD, and range of scores, were computed for each dimension of LMX as well as the overall LMX scale.

Results

The primary goal of LMX scale was to establish sound psychometric properties. The areas of psychometric analysis of the scale included the descriptive statistics (mean and SD), item analysis, EFA, CFA, and internal consistency reliability.

Item analysis

Item analysis of the items of LMX scale has been done by the point biserial correlation method (r _pbi). Point biserial correlation provides a measure of relationship between variables (total scores of a scale) and two categorized or dichotomous variable (positive and negative scores). This statistic is interpreted as a measure of the degree to which the continuous variables differentiate or discriminates between the two categories of the dichotomous variables. The results indicated that 20 items (see Appendix) (out of 27) were found to be significant at 0.01 level. Seven items were deleted from the final scale as their item-total correlations were found to be nonsignificant. The range of r _pbi for overall LMX Scale were 0.145–0.634 (Table 2).

Exploratory factor analysis

EFA using principal components with varimax rotation were conducted on the 19 remaining items, using best practices of participant-to-item ratios above 1:10 (Comrey and Lee, 1992; Everitt, 1975). Kaiser–Meyer–Olkin (KMO) (Kaiser, 1974) measure was used to check the sampling adequacy of data for conducting factor analysis. The value of KMO ranges from 0 to 1. Researchers have suggested that the KMO value of 0.50 and more is considered suitable for EFA (Tabacnick and Fidell, 2001). In the present study, the KMO measure of sampling adequacy was 0.887, sufficiently high for conducting EFA. Bartlett's test of sphericity (Bartlett, 1950) tests the hypothesis that your correlation matrix is an identity matrix, which would indicate that your variables are unrelated and therefore unsuitable for structure detection. Accordingly, it should be significant (p < 0.05) for factor analysis to be suitable (Tabacnick and Fidell, 2001). For the present study, the results of Bartlett’s test of sphericity were highly significant (p < 0.00001), indicating that the intercorrelation matrix was suitable for factor analysis. The findings of EFA of 19 items yielded three factors: LMX-affect, LMX-loyalty, and LMX-perceived contribution, which altogether accounted for 56.49% of variance after rotation. Varimax has generally been regarded as the best and most widely used rotation in psychological research (Fabrigar et al., 1999). Items were considered part of a factor if they had loadings of 0.50 or greater on one factor and loadings below 0.35 on other factors (Bryant and Yarnold, 1995). The factor loadings of all the items on their respective factors were fairly high (>0.50), thus fulfilling the criteria of Bryant and Yarnold (1995).

We considered it desirable to exclude those items that load highly on two or more factors because it might be an indication that they have an ambiguous meaning. We formulated an additional criterion that the loadings on the other factors should be 0.35 or lower. The items were broken down as follows: LMX-affect (12 items; eigenvalue = 6.286, accounting for 31.429% of the variance in the items), LMX-loyalty (5 items; eigenvalue = 3.603, accounting for 18.013% of the variance in the items), and LMX-contribution (2 items; eigenvalue = 1.410, accounting for 7.050% of the variance in the items). The dimension of professional respect (Liden and Maslyn, 1998) did not emerge in this study. The items of professional respect were loaded on the affect dimension (Table 3).

Confirmatory factor analysis (CFA)

CFA were conducted on selected 19 items, and the results substantiated the findings of EFA. The CFA was performed on a raw data base with no missing values using the maximum likelihood (ML) procedure to test if the three-factor confirmatory model supported the basic scale structure identified in exploratory factor analyses. i.e. the three identified discrete factors comprise one underlying construct. The goodness of fit of a three-factor model (for the chosen 19-items) was tested in comparison to compelling models through the use of the AMOS 20 version. Results showed that all items loaded onto their respective latent constructs. Loadings ranged from 0.58 to 0.81 for affect, 0.73 to 0.87 for loyalty, and 0.42 to 0.54 for contribution, all ps < 0.001 (Figure 1). The factor loadings of each item are provided in Table 3.

Models fit to data was assessed through multiple goodness-of-fit indices and respective reference values. A nonsignificant χ² test indicates a good absolute fit of the model to the data, but this test is sensitive to sample size. In this way, other indices were employed in order to minimize the occurrence of errors of Type 1 and Type 2. The results were interpreted following recommendations of Schermelleh-Engel et al. (2003) holding that an acceptable model fit occurs when the chi-square likelihood ratio (χ²/df) has values lower than three, the Bentler Comparative Fit Index (CFI) has values greater than 0.95, and the Root Mean Square Error of Approximation (RMSEA) and the Standardized Root Mean Square Residual (SRMR) has values lower than 0.08. The model fit indices are provided in Table 4. The fit of the three-factor model to the data was satisfactory, goodness of fit (GFI) = 0.89; root-mean-square error of approximation (RMSEA) = 0.06; Standardized (RMR) = 0.05, comparative fit index (CFI) = 0.94.

Mulaik et al. (1989) have suggested that even models which are good-fitting can be subjected to misspecification, therefore alternative models should be taken into consideration while using structural equation modeling. The competing model in the present study was a single-factor model, in which all items represent a single dimension. The results demonstrated in Table 4 indicated the superiority of the three-factor model over the one-factor model, as all the model indices, namely affect, loyalty, and contribution, do not fall under the acceptable range. Thus, analyses of the data yielded good evidence of construct validity. In conclusion, the results of EFA and CFA both confirmed the three dimensions of affect, loyalty, and contribution as proposed by Dienesch and Liden (1986).

Internal consistency reliability

After going through EFA and CFA, the final measure of LMX consisted of 19 items reflecting affect, loyalty, and perceived contribution dimensions of LMX. For the first factor Affect (No. of items = 12) Cronbach’s alpha was found to be 0.91. A reliability analysis of the five items of the second factor loyalty showed a Cronbach’s alpha of 0.89. Finally, the third factor perceived contribution had only two items in common, although from the inspection of item content it can be concluded that the meaning of the scale is comparable. Hence, we could form a scale for LMX-perceived contribution. But this scale showed a lowest alpha of 0.54. Thus, it is clear that internal consistency reliability results generated by Cronbach’s alpha (α) were above 0.7 (Hair et al., 1998) for LMX dimension except for the contribution dimension. A reliability analysis of the 19-item scale showed a Cronbach’s alpha of 0.86 (Table 2). Thus, the predicted reliability values were deemed favorable for the entire LMX scale.

Discussion

The present study aimed to advance the research on LMX by exploring how individuals themselves perceive and experience differential relationships (high quality and low quality) with their supervisors by developing a multidimensional measure of LMX that will be suitable in the Indian context taking into consideration the dimensions proposed by Dienesch and Liden (1986) and Liden and Maslyn (1998). In the present study, both EFA and CFA were used to examine the underlying multiple dimensions of LMX. Researchers have illustrated that CFA permits better integration of theory and measurement in comparison to EFA (Hughes et al., 1986; Schermelleh-Engel et al., 2003). In this regard, Church and Burke (1994) have emphasized, “Whereas exploratory methods are advantageous when no a priori structure can be hypothesized, confirmatory methods provide significance tests and goodness of fit indices for hypothesized models” (p. 93). The logic of CFA is to compare a reproduced matrix of covariance of the a priori hypothesized structural model with the “real” matrix observed with empirical indicators. The degree of overlap between the matrices is finally evaluated by using goodness-of-fit indexes (Fiorilli et al., 2015).

The major general findings of the study indicated that, in particular, the three-factor structural model appears remarkably stable both in EFA and CFA and is consistent with the studies of Liden and Maslyn (1998), Dienesch and Liden (1986). As we have already mentioned while studying the impact of LMX on subordinate outcomes, most of the earlier research conceptualization treated LMX as a unidimensional construct. Supporting this unidimensional model over all other models would validate the theoretical premise of Graen and his associates that LMX is a one-dimensional concept (Graen and Scandura, 1987; Graen and Uhl-Bien, 1995).

Thus, in this line of arguments, simultaneously, a unifactor model was also tested. According to the findings, the three-factor structure was supported by a CFA and was a better fit to the data than the alternatively tested one-factor model. The factors that emerged in this study were not only based on statistical criteria, but they were also consistent with the theoretical and methodological assumptions of Dienesch and Liden (1986). The findings of the present study are also in line with the results of Phillips et al. (1993) who have verified Dienesch’s items on a sample of middle-level managers of retail organizations. The results of CFA also provided support for affect, loyalty, and contribution as distinct dimensions of LMX relationships.

Thus, it can be apparent from the above-mentioned discussion that the results of both EFA and CFA confirmed the three dimensions of affect, loyalty, and contribution as proposed by Dienesch and Liden (1986). The dimension of professional respect (Liden and Maslyn, 1998) did not emerge in this study. The items of professional respect were loaded on the affect dimension. Evidence for not getting enough evidence for professional respect might be that the scale of professional respect has much in common with affect. Also, the selection of the study sample or the cultural differences between the USA and India may have an influence on LMX. The member’s perceptions of the professional respect of their leaders are grounded in the affective dimension. For example, a subordinate who shows strong affection toward his or her supervisor is more likely to perceive their supervisor as outstanding at his or her profile of work. Thus, it seems quite possible that subordinates who like their superiors recognize their professional excellence in due course of the relationship. Thus, in the Indian context, subordinates’ affection toward their superiors is also manifested in acknowledging their professional recognition inside and outside of the organization.

The results of the present study provide valuable insights on the scope of the multidimensionality of LMX theory by examining how effectively it can be applied in a collectivistic work culture. In-group and outgroup membership are distinct features of a collectivistic society. Researchers have asserted that in a collectivistic work culture, members of the in-group value interpersonal relationships with their superiors, receive personal favors and benefits in exchange for hard work and devotion, and are more likely to develop robust working relationships with their leaders (Graen et al., 1982; Liden and Maslyn, 1998). Similar to the previous studies, the dimensions of LMX found in this study can be utilized to cultivate high-quality LMX in the conducive work environment. Further, high-quality LMX can be facilitated by collectivistic Indian cultural values of interpersonal harmony within in-groups (for protecting the group’s interest), respect for seniority, and good social relationships. These cultural values create positive work setting for developing high-quality LMX relationships. Thus, it is clear that the LMX construct is particularly important in the collectivistic culture of India because of the person-oriented nature of Indian society.

Implications, limitations, and future research suggestions

This study was designed with two objectives in mind. One is to extend studies of LMX to the Indian context, and the second is to examine LMX from the perspective of the subordinate. Further, the findings of the present study have fairly validated the multidimensionality of LMX processes and offer many insights into the workings of LMX. This study has made significant advancements in the field of LMX because it has taken a multidimensional approach to gathering information, which is fairly new. The importance of multiple dimensions of LMX is implicated in sustaining and developing LMX relationships. Thus, it is recommended that future researchers should examine how these dimensions differentially influence work-related outcomes. The findings will also be utilized by the authorities of the organization in focusing future training for its managers.

Future research should assess the social exchange between superiors and subordinates from the leader’s perspective and examine if the leaders perceive the same degree of exchange as their subordinates. This means that the supervisor using LMX to influence subordinates will have a better understanding of what factors to emphasize to get better results in terms of individual and organizational outcomes. The present study included LMX quality from the perspective of subordinates only. It does not include the perceptions of superiors to analyze how they perceive and react to similar exchange relationships with their individual subordinates at work. Hence, it is recommended that future researchers examine the exchange relationships between supervisor and subordinate from both perspectives. An examination of both perspectives is required to compare how leaders and subordinates view their relationship, both in qualitative and in quantitative terms. Substantial research has shown that higher-quality LMX has been associated with a number of beneficial outcomes for both individuals and organizations (Liden et al., 1993, 2000; Terpstra-Tong et al., 2020; Aggarwal et al., 2020; Wu, 2018).

Thus, within organizations, both supervisors and employees, are required to be educated regarding the advantages of developing and sustaining high-quality LMX relationships, especially in times of organizational change. The current results will be of great advantage to organizations if they translate these findings into actions such that the supervisor’s effectiveness in encouraging all of their subordinates to be “in-group” members is enhanced. Consequently, supervisors are required to be aware of how to enhance and maintain high-quality LMX relationships with their employees for prolonged periods of time by utilizing the multiple dimensions of LMX (affect, loyalty, and contribution) found in the study.

Similar to the findings of Liden and Maslyn (1998), the internal consistency reliability of the LMX–contribution dimension is very low. However, from the examination of the item content of the contribution, it is clear that the meaning of the scale is comparable. Hence, a scale for contribution has been formed. In view of this, it is recommended that future researchers should add a few additional items to increase the reliability of the contribution scale of LMX scale so that it fulfills adequate criteria of reliability. In summary, the results regarding the reliability and validity of the scales derived from the factor structure of LMX are satisfactory and deserve further research.

Conclusion

In general, the results of the present study provided support for the hypothesized research model and indicated that this model is theoretically and practically meaningful and provides an adequate fit to the data. This scale is suitable not only in the Indian culture but also in the Western cultural context, as the results corroborate the findings of Liden and Maslyn (1998), Dienesch and Liden (1986), and Phillips et al. (1993), indicating a fair level of cross-cultural validity. However, future research should also address the cross-validation of the factor structure of LMX on other samples and occupations. Further, they still have a good opportunity to further develop the potentiality of the LMX approach by providing greater clarity in the theoretical conceptualization of the construct and by enhancing measurement for examining LMX both from the leader’s and subordinate’s perspectives.

Figures

Figure 1

Confirmatory factor analysis of multidimensional leader member exchange scale

Table 1

Demographic details of the participants

Demographic characteristics		Number	Percentage (%)
Variable	Levels
Age (in years)	24–35 years	79	27.2
	36–45 years	40	14.1
	46–55 years	95	32.8
	56–60 years	76	25.9
		Mean (SD)
		45.24 (11.71)
Gender	Male	264	91
	Female	26	9
Marital status	Married	250	86.2
	Unmarried	40	13.8
Salary in rupees (per month)	Less than 20,000	20	6.9
	20,000–40,000	148	51
	40,000–above	122	42.1
Educational qualification	Postgraduate and above	86	29.65
	Graduate	180	62.07
	Undergraduate	24	8.28
Designation	Upper Middle	120	29.66
	Middle	96	61.03
	Lower	74	9.31
Coworkers (same level employees)	Worker without coworker	53	18.28
	1–5	129	44.48
	6–10	50	17.24
	11–209	58	20.00
Length of service in years (experience)	Up to 10 years	72	24.83
	11–20 years	35	12.07
	21–30 years	98	33.79
	31–40 years	85	29.31
Size of team	2–20	124	42.76
	21–60	79	27.24
	61–300	52	17.93
	301–800	35	12.07

Source(s): Table by authors

Table 2

Mean (M), standard deviation (SD), range item total correlation, and Cronbach’s alpha (α) for leader-member exchange (LMX) scale and its dimensions (N = 290)

	No. of items	Mean	SD	Range item total correlation	Alpha
Overall LMX	19	66.96	9.99	0.145–0.634	0.86
LMX-affect	12	47.48	7.62	0.473–0.634	0.91
LMX-loyalty	5	15.41	5.32	0.632–0.471	0.89
LMX-contribution	2	8.22	1.17	0.145–0.231	0.54

Source(s): Table by authors

Table 3

Exploratory factor analysis of multidimensional leader–member exchange scale

Factor (F) name and items	Loadings
	1	2	3
F1: Affect
Item 19	0.82
Item 2	0.79
Item 7	0.79
Item 20	0.76
Item 13	0.73
Item 5	0.72
Item 17	0.71
Item 1	0.70
Item 16	0.64
Item 10	0.62
Item 6	0.61
Item 3	0.59
F2: Loyalty
Item 12		0.87
Item 15		0.87
Item 9		0.84
Item 8		0.81
Item 11		0.76
F3: Contribution
Item 4			0.82
Item 18			0.57

Note(s): All loadings are significant

Source(s): Table by authors

Table 4

Summary of goodness-of-fit indices for one-factor and three-factor models of measurement

Model fit indices	χ2/df (Should be < 3)	GFI (>0.90 for acceptable fit)	AGFI (>0.85 for acceptable fit)	CFI (>0.90 for acceptable fit)	RMSEA ((0.05–0.8) for acceptable fit)	RMR (Close to zero)	SRMR ((<0.10) for acceptable fit)
One-factor	5.767	0.757	0.693	0.716	0.134	0.227	0.1737
Three-factor (hypothesized)	2.095	0.891	0.858	0.937	0.064	0.047	0.0513

Note(s): One-Factor Model: It represents Unidimensional of global LMX

Three-Factor Model: It represents Affect Factor, Loyalty factor and Contribution factor

Source(s): Table by authors

Declaration of conflicting interests: The authors declare that there are no conflicts of interest regarding the publication of this paper.

Appendix

Table A1

Leader–member exchange scale (LMX)

S.No.	Statement	Strongly disagree	Disagree	Neutral	Agree	Strongly agree
1.	I appreciate the personal qualities of my supervisor/immediate officer
2.	My supervisor/immediate officer truly cares about my well-being
3.	My supervisor\ immediate officer truly feels that I am responsible and trustworthy for the organization
4.	I am willing to handle responsibility and complete task that extend beyond my job description and/or employment contract
5.	I have enough trust and confidence in my supervisor\ immediate officer
6.	I am influenced by my supervisor\ immediate officer ‘s knowledge, expertise and competence on the job
7.	My manager gives me enough support whenever I am in trouble. (R)
8.	My manager lacks the ability to motivate employees. (R)
9.	My manager does not provide substantial resources and opportunities for the accomplishment of goals. (R)
10.	My supervisor\ immediate officer trusts me a lot in sharing personal and work-related information
11.	My manager is unwilling to help me if I need some special favor. (R)
12.	My manager does not recognize my work-related potential. (R)
13.	I have great respect for my supervisor\ immediate officer
15.	My supervisor\ immediate officer comes forward to defend my working pattern whenever required
16.	I have faithful/trustworthy relations with my supervisor/immediate officer
17.	I appreciate my supervisor/immediate officer’s professional and communication skills
18.	I put hard efforts for my supervisor/immediate officer (which is not normally required) to accomplish my supervisor/immediate officer goals
19.	I really enjoy working with my supervisor\ immediate officer
20.	If my supervisor\ immediate officer comes to know that I am in some problem, he/she will come forward help even if I have not approached him/her

Note(s): Item 14 * Deleted from the Final Scale

Items 1,2,3,5,6,7,10,13,16,17,19,20 represents affect dimension; Items 8,9,11,12,15 represents the loyalty dimension; Items 4,18 represents Contribution dimension

R-reverse coded item

Source(s): Table made by authors

Supplementary material

The supplementary material for this article can be found online.