Mediation of Distributive Justice on Dyadic Relationship between Leaders and Followers with Personal Outcomes

Abstract

Purpose - This study examined the relationship between dyadic relationship between leaders and followers (DRLF), distributive justice (DISJ), job satisfaction (JSTC), and organizational commitment (OGCM). Research design, data, and methodology - 200 sets of survey questionnaires were distributed to the employees at a municipal office in East Malaysia using purposive sampling technique. Only 60 percent or 115 questionnaires were returned to the researchers. The survey data were analysed using the SmartPLS due to its ability to deliver latent construct scores, handle small sample size problems and estimate relationship between many constructs in the hypothesized model. Results - The findings indicated that there is a significant correlated direct relationship between DRLF and DISJ and mediating relationship between DRLF, DISJ and personal outcomes, which are JSTC and OGCM. Conclusions - This study confirms that DISJ does act as an important mediating variable in the relationship between DRLF with JSTC and DRLF with OGCM. Other dimensions of personal outcomes, such as extra-role behaviour, job motivation and service quality should be considered in future study because they are found to be the important outcomes of the relationship between DRLF and DISJ. The importance of these issues need to be further advanced in future research.

1. Introduction

In an era of competitive and demanding environments, dyadic relationship between leaders and followers (DRLF) is crucial in enhancing organizational effectiveness (Yang & Ju, 2011). It refers to formal and/or informal relationship between managers and subordinates that occur in units/departments to achieve organizations’strategies and objectives. A review of the recent literature highlights that a high quality relationship between managers and subordinates may have a significant impact on workplace justice especially distributive justice (DISJ) (Sindhu et al., 2017) which enhance the personal outcomes. These personal outcomes mentioned are job satisfaction (JSTC) (Zafar et al., 2017) and organizational commitment (OGCM) (Maslyn et al., 2017).

Even though the studies have prominent significance towards personal outcomes, little is discussed about the mediating effect of DISJ in the workplace leadership research literature. Thus, this situation inspires the researchers to fill in the gap of literature by quantifying the mediating effect of DISJ in the relationship between DRLF and personal outcomes.

2. Literature Review

Previous studies about workplace leadership had not produced sufficient findings to explain the whole meanings of DRLF and DISJ, as well as relationship between such variables. For example, a survey of 279 business-to-business salespeople in USA by Schwepker (2016) had only focused on the relationship between one element of DRLF, namely participation in quota setting and one element of DISJ, namely fairness in reward allocation. Even thoughthis study produced significant results, it did not sufficient to explain the effect of DRLF on DISJ in the organizations. Meanwhile, a survey of 306 professionals from 30 software organizations operating in different parts of India by Bhal (2006) had only emphasized on the relationship between one feature of DRLF, namely affective based relationship between leaders and followers, and DISJ. Outcomes of this study had not supported the relationship between variables in the organizations.

Recent studies about workplace leadership have used single constructs of DRLF and DISJ, as well asprovided strong theoretical and empirical findings to support the relationship between such constructs. The role of DRLF as an important predictor of personal outcomes has gained strong support from the notion of organizational behaviour theory. For example, Blau's (1964) social exchange theory emphasizes that interpersonal transaction between managers and subordinates is usually occur beyond economic gain, and this situation may strongly invoke followers’ perceptions of DISJ in organizations. Further, Graen's (1976) role making theory posits that leaders usually interact with followers based on loyalty, respect and trust in performing daily jobs may strongly invoke followers’ perceptions of DISJ in organizations. Application of these theories in workplace leadership shows that quality relationship between leaders and followers may act asan important determinant of personal outcomes(Sindhu et al., 2017; Yusniati et al., 2016).

Empirical studies supported the effect of DRLF on DISJ were summarized in [Table 1].

[Table 1] Research Findings Supported the Relationship between DRLF and DISJ

The findings of these surveys showed that the managers arehighly concerned about mutual trust, respect and obligation in relationship with their subordinates. The willingness of managers to practice such quality relationships in performing day-to-day job operations had strongly invoked subordinates’perceptions of distributive justice in the respective organizations. Thus, it was hypothesized that:

[H1] There is a positive relationship between DRLF and DISJ.

Further, extant studies revealed that the mediating effect of DISJ in the hypothesized model has received strong support from the notion of distributive fairness theory. For example, Adams (1963) equity theory mentions that fair treatment in exchanging and distributing inputs (e.g., effort, loyalty and sacrifice) and outputs (e.g., return) may positively affect individuals’ actions. Thibaut and Walker's (1975) control theory state that employees who perceived justice on the outcomes they received may lead to induce favourable behaviour. Application of these fairnesstheories in a workplace leadership model shows that DISJ is an important link in between DRLF and personal outcomes. This notion has received strong support from DRLF research literature.

Several further studies were done using indirect effect model to assess distributive justice towards personal outcomes based on different organizational settings. The supported outcomes of these studies displayed that the ability of managers to practice high quality relationship with their subordinates in executing day-to-day job operations had invoked subordinates’perceptions of distributive justice. [Table 2] summarized the findings as follows:

[Table 2] Research Findings Supported the Mediating Effect of DISJ in the Relationship between DRLF and Personal Outcomes.

The literatures have been used as a foundation to establish a conceptual framework as exhibited in [Figure 1].

[Figure 1] Conceptual Framework

Based on the framework, the following hypotheses are established:

[H2] DISJ mediates the relationship between DRLF and JSTC

[H3] DISJ mediates the relationship between DRLF and OGCM

3. Research Method

This study applies quantitative methodology through leadership research literatures, in-depth interview and actual survey as the procedure of data collection. According to Creswell (2014) and Sekaran and Bougie (2015), application of this procedure may collect accurate, less bias dataand high quality data (Azman et al., 2009).

This study was conducted at a municipal office in East Malaysia. At the first step of data collection, survey questionnaire was drafted based on the DRLF literature. After that, a back translation technique was used to translate the survey questionnaire into English and Malay languages in order to increase the validity and reliability of research results (Sekaran & Bougie, 2015; Creswell, 2014; Peng, 2013).

3.1. Measures

The survey questionnaire was divided into four sections shown in [Table 3] as follows:

[Table 3] Number of Items According to Constructs

All items are to be rated based on 7-itemsscale ranging from "strongly disagree/dissatisfied" (1) to "strongly agree/ satisfied" (7). The demographic variables were used as controlling variables because this study focused on measuring employees' attitudes.

3.2. Sample

The unit of analysis for this study is employees at the organization. The name of this organizationremains anonymous for some confidentiality reasons. Purposive sampling plan was used to distribute 200 survey questionnaires to employees in the organizations. This sampling plan was chosen because the organization head had not provided a list of registered employees to the researchers and this conditiondid not allow the researchers to select participants using a random technique. From the number, 115 (60 percent) usable survey questionnaires were returned to the researchers. The participants answered the survey questionnaires based on their consents and a voluntarily basis.

The SmartPLS was utilized to analyse the instrument of this study because it could deliver latent construct scores, handle small sample size problems and estimate relationship between many constructs in the hypothesized model (Hair et al., 2017). The procedures of data analysis included few analysis processes and steps. First, the validity and reliability of instrument were determined using a confirmatory factor analysis. Second, the structural model was inspected based on the path coefficients for example standardized betas (β) and t statistics. For a direct effects model, t statistics is greater than 1.65 (one tail testing) for the relationship between variables will show a significant hypothesis. Conversely, for a mediating model, t statistics is greaterthan 1.96 (two tail testing) for the relationship between variables will show a significant hypothesis. Third, the value of R2is used as a criterion for determining the overall predictive strength of the model based on the rules: 0.02 (weak effect), 0.26 (substantial effect) and 0.13 (moderate effect) (Cohen, 1988). Fourth, the value of f2was used as a measure to determine the effect size of predicting variable in the model based on the criteria: 0.02 (weak), 0.15 (medium) and 0.35 (large) (Hair et al., 2017). Finally, the value of Q2for dependent variable higher than zero will show that the model has predictive relevance (Hair et al., 2017).

4. Results

[Table 4] shows that most respondents were aged between 20 and 29 years old (44.3%) females (54.8%), with 5 years and above working experience (45.2%) and medium secondary school (SPM/MCE) holders (43.5 %).

[Table 4] Respondents Profile

[Table 5] depicts the results of convergent validity and reliability assessment. Factor loading for all items that represent DRLF were from 0.708 to 0.823, DISJ were from 0.713 to 0.829, JSTC were from 0.721 to0.809 and OGCM were from 0.735 to 0.816. Thus, all items that represent the research constructs had factor loadings greater than 0.70 (Fornell & Larcker, 1981; Hair et al., 2017), indicating that they meet acceptable standard of convergent validity. Conversely, average variance extracted (AVE)11 AVE-Average Variance Extracted values for DRLF were 0.565, DISJ was 0.611, JSTC was 0.583, and OGCM was 0.590. These AVE values were above the threshold value of 0.50 (Fornell & Larcker, 1981; Hair et al., 2017; Henseler et al., 2009) all constructs meet the acceptable standard of convergent validity analysis. While, the Cronbach's Alpha values and composite reliability for all constructs were above 0.80 (Hair et al., 2017; Nunnally & Bernstein, 1994), indicating that the constructs had high internal consistency.

[Table 5] The Convergent Validity and Composite Reliability

[Table 6] shows the results of discriminant validity. The values of AVE square root in diagonal for DRLF (0.752), DISJ (0.782), JSTC (0.763), and OGCM (0.768) were greater than the squared correlation with other constructs in off-diagonal, showing that these constructs fulfil the requirements of discriminant validity (Hair et al., 2017).

[Table 6] Discriminant Validity Assessment

[Table 7] exhibits the outcomes of discriminant analysis for all constructs. The values of all indicators (items) for their own constructs are more than 0.70, and loweron other constructs, indicating that all constructs meet the standards of discriminant analysis (Hair et al., 2017).

[Table 7] Factor Loadings and Cross Loading for Different Constructs

[Table 8] presents the outcomes of descriptive statistics and variance inflation factor (VIF) VIF - Variance Inflation Factor for all constructs. The mean values for all constructs are from 5.4493 to 5.5797 indicating that the levels of DRLF, DISJ, JSTC and OGCM range from high (4) to the highest level (7). Conversely, the values of VIFfor all constructs are less than 5.00, indicating that the constructs are free from the serious collinearity problems (Hair et al., 2017).

[Table 8] Descriptive Statistics and Variance Inflation Factor

[Table 9] denotes the outcomes of testing H1. The inclusion of DRLF in the analysis had explained 40 percent of variance in DISJ. In terms of explanatory power, this model has large effect. The outcomes of testing the hypothesis showed that DRLF was positively and significantly related to DISJ (β=0.630, t=6.735), therefore H1 was supported. This result confirms that DRLF is an important determinant of DISJ.

[Table 9] Outcomes of Testing the H1

Note: Significant at t >1.96; p<0.05(one tail testing)

LLCI = Lower Level Confidence Interval, ULCI = Lower Level Confidence Interval

[Table 10] shows that the outcomes of testing the indirect effects model. The inclusion of DRLF and DISJ in the analysis had explained 30 percent of JSTC. While, the inclusion of DRLF and DISJ in the analysis had explained 68 percent of OGCM. In terms of explanatory power, this model has substantial effect. Further, the outcomes of testing the hypotheses showed two important findings: first, relationship between DRLF and DISJ was significantly related to JSTC (β=0.212, t=2.528), therefore H2 was supported. Also, the 95% Bootstrap Confidence Interval (CI) CI- Confidence Intervaldoes not straddle a 0 in between [LLCI=0.069, UL=0.396], indicating that DISJ mediates the relationship between DRLF and JSTC (Preacher & Hayes, 2008). Second, relationship between DRLF and DISJ was significantly related to OGCM (β =0.302, t=4.445), therefore H3 was supported. Also, the 95% Bootstrap Confidence Interval (CI) does not straddle a 0 in between [LLCI=0.160, ULCI=0.421], indicating that DISJ mediates the relationship between DRLF and OGCM (Preacher & Hayes, 2008). In overall, the above results have supported the mediating effect of DISJ in the hypothesized model.

[Table 10] Outcomes of Testing H2 and H3 Relationships β

Note: LLCI = Lower Level Confidence Interval, ULCI = Upper Level Confidence Interval

[Table 11] exhibits the type of mediation analysis based on Hair et al. (2017) and Zhao et al. (2010) procedures. The results of this test showed that DISJ has partially mediated: a) the relationship between DRLF and JSTC; and b) the relationship between DRLF and OGCM.

[Table 11] Type of Mediation Analysis Assessment

Note: Significant at t > 1.645, p <0.05(two tail testing)

LLCI = Lower Level Confidence Interval, ULCI = Lower Level Confidence Interval

5. Discussion

The findings of this study exposed that DISJ acts as an important mediating variable in the relationship between DRLF with JSTC and OGCM. In the context of this organization, management has high awareness to practice good relationship with subordinates as a mean to accomplish their organizations’ strategy and objectives. The majority of respondents perceived that the levels of DRLF, DISP, JSTC, and OGCM are high. This situation explains that the ability of managementto appropriately practice high quality interaction with subordinates will strongly invoke subordinates’perceptions of DISJ. Consequently, this perception may lead to greater JSTC and OGCM in the organization.

This study provides three major contributions: theoretical, robustness of research methodology and practical. According to theoretical contribution, the results of this study expose that DISJ has mediated the effect of DRLF on JSTC and OGCM. This finding is consistent with extended studies by Lee (2000), Hassan and Chandaran (2005), Bhal and Ansari (2007), Gichira et al. (2016) and Zafar et al. (2017). Concerning the robustness of research methodology, the survey questionnaire employed in this study has satisfied the acceptable standards of the validity and reliability analyses. This condition could lead to accurate and reliable research results.

From a practical contribution, the findings of this study may useas important guidelines by management to improve leadership behaviour in the organizations. In order to achieve this aim, management needs to focus on the following aspects: first, leadership training program curriculum should updated to upgrade managers’ knowledge and skills in communicating, involving and engaging with diverse employee backgrounds and expectations. These abilities may assist managers to influence employees in achieving their job targets. Second, pay rises and levels to higher performing employees should revisited according to current organizational strategy and job challenges. If higher performing employees feel that they receive the type, level and/ amount of reward equal with their contributions, this may motivate them toimproved customer service and productivity. Third, participative work culture should be promoted in order to enhance cooperation and collaboration between management and employees in performing yearly key performance indicators. If the above suggestions are given more attention, this may encourage employees to support their organizations’strategic business mission and vision.

6. Conclusions

The results of this research confirm that DISJ acts as an important mediating variable in the relationship between DRLF and personal outcomes, which areJSTC and OGCM. This result also has supported and broadened the previous studies mostly done in Western countries.

This research acknowledges some limitations: first, data gathered from the cross-sectional research design have neglected the development issue and detail causal correlation between the variables of interest. Second, this study has not measured specific dimensions for DRLF, DISP, JSTC, and OGCM. Third, thisstudy is done at a municipal council office. Finally, purposivesampling plan cannot control response bias. These limitations may reduce the ability of generalizing the results of this study to other organizational settings.

This study provides few suggestions to strengthenfuture research: first, a longitudinal research design should be used in future study because it able to detect patterns of change and the direction and magnitude of causal relationships amongst variables of interest. Second, a larger sample size should be taken because it may represent the population. Third, more than one organizations should be used in future study because their results may show similarities and differences within different organizational settings. Fourth, other components of DRLF such as trust, honesty, and decision-making should be utilized in future research because they are widely recognized as important determinants of DISJ. Other elements of DISJ such as adequacy of outcome and award basis should be utilized in future study because they are widely acknowledged as important mediating variables in between DRLF and personal outcomes. Finally, other dimensions of personal outcomes, such as extra-role behaviour, job motivation and service quality should be considered in future study because they are found to be important outcomes of the relationship between DRLF and DISJ. The importance of these issues needs to be further advanced in future research.