The influence of perceived leader engagement on employee engagement and the moderating role of their inter-personal relationship

(1)

The influence of perceived leader engagement on

employee engagement and the moderating role of their

inter-personal relationship.

(2)

Abstract

This study aims to investigate the relationship between the engagement of the leader, as perceived by the employee, and employee engagement. The direct relationship between perceived leader engagement and employee engagement is investigated, as well as the moderating role of the inter-personal relationship between the leader and the employee. This inter-personal relationship was characterized in this study by three subjects: leader-member exchange, need for leadership and the duration of the cooperation between the leader and the employee. Data was collected from an online survey with 130 respondents across a wide range of industries and hierarchical levels within The Netherlands. The basic condition to participate in the online survey was an employment in which the employee has a direct relationship with his/her supervisor. Results showed that there was a statistically significant positive relationship between perceived leader engagement and employee engagement. This relationship was significantly positively moderated by leader-member exchange. There was not enough evidence to prove the moderating role of need for leadership and cooperation tenure. However, taking two moderators (leader-member exchange and need for leadership) together increased the level of variance on employee engagement that was explained by the model. Practical and theoretical implications were discussed and implications for further research were given. Finally, the limitations and conclusions of this study were provided.

Keywords: Employee Engagement, Perceived Engagement, Leader-Member Exchange, Need For Leadership, Cooperation Tenure

(3)

List of Figures and Tables

Figure 1: Conceptual Model ... 6

Figure 2: Two-way interaction effect between leader-member exchange and perceived leader engagement on employee engagement ... 22

Table 1: Descriptive statistics, Correlations and Reliabilities ... 19

Table 2: Employee engagement and leader-member exchange ... 21

Table 3: Employee engagement and need for leadership ... 23

Table 4: Employee engagement and cooperation tenure ... 24

(5)

Introduction

Since the 1990s, employee engagement has been discussed thoroughly in academic literature. Many researchers investigated the factors that might influence engagement and what it may affect. It is stated that employee engagement is a significant predictor of desirable organizational outcomes, such as productivity and profitability (Buckingham & Coffman, 1999), but also to positive organizational behavior such as personal initiative and learning motivation (Sonnentag, 2003), extra-role behavior (Salanova, Agut & Pieró, 2003) and proactive behavior (Salanova, Llorens, Peiró & Schaufeli, 2003). In addition to the organizational aspect, engagement is also important for the individual. It seems obvious to state that work is an important factor in human well-being. Most employees spend more time at work than with their families and they even consider their work as a part of their being. Therefore, work engagement is important in everybody’s interest. You might then think “Well, let us strive to get the employees to the highest level of engagement!”, but what will happen if you encourage employees to be engaged while their supervisor is not?

Both past and recent literature mention the importance of leadership in the determination of employee engagement. Kahn (1990) suggests that employees experience dimensions of personal engagement during their daily task performances. Engagement occurs when one is cognitively vigilant and/or emotionally connected to others (i.e. their supervisors). More recent literature also argues that leadership can affect the meaningfulness of employees’ work as measured by work engagement (Strickland, Babcock, Gomes, Larson, Muh & Secare, 2007). The relationship between the employee and the leader, when it comes to engagement, is often discussed in current literature. Previous research has investigated several effects of leadership on engagement, such as different leadership styles or the individual traits of the leader. But to the best of my knowledge it has never been investigated whether the employee’s work engagement is influenced by the perceived work engagement of their supervisors. After all, the supervisor is actually an employee as well. This effect, in which a supervisor has to act in accordance with both their own interests and the interests of their teams, is for example examined in the change management literature (Kanter, Stein & Jick, 1992). Combining this insight with the literature of leadership and work engagement is one of the purposes of this research.

(6)

Prior research stated that there are three domains within leadership: the follower, the leader and their relationship (Graen & Uhl-Bien, 1995). Since the follower and the leader are represented in this study by employee engagement and perceived leader engagement, their inter-personal relationship is the remaining factor. One could expect that there is an interaction between the several variables, in which the effect of perceived leader engagement on employee engagement may differ when the quality of the inter-personal relationship differs. Thus, the inter-personal relationship between the leader and the employee is the interaction effect and will be studied as a moderating effect on the statistical relationship between perceived leader engagement and employee engagement.

Figure 1: Conceptual model

The leader-member exchange (LMX) theory (Dansereau, Graen & Haga, 1975) describes the relationship between supervisors and subordinates based on the quality of their exchanges (high to low). This score, on the quality of exchanges, is an important determinant for the inter-personal relationship. Need for leadership indicates an interaction between the leader and the employee, in which the employee can perceive interventions of the leader as (un)wanted or (un)necessary (De Vries, Roe & Taillieu, 2002). This score, the amount of need for leadership which the employee perceives, tells something about the personal preferences of the individual and can therefore influence the inter-personal relationship between the leader

(7)

and the employee. The cooperation tenure scores the amount of years the employee and the supervisor work together. One could reasonably assume that the perception of the employee about the supervisor is likely to change over time as a function of his/her experiences with the leader and the organizational processes (Dickson, 1997; Rush & Russel, 1988). Therefore, the amount of years of cooperation tenure might be a determinant factor in the inter-personal relationship between the employee and the supervisor.

The effect of the perceived leader engagement on employee engagement and the moderating role of their inter-personal relationship, will be investigated in this study. First, an overview of existing knowledge from prior research will be provided in order to ‘set the scene’ for this study and hypotheses. Second, the design of this study will be described, including the characteristics of the sample and the strategy of data collection and analyses. Subsequently, the results of the statistical analyses will be described, providing tests of the hypotheses. To conclude, the results and limitations will be discussed and implications for further research will be given.

(8)

Literature Review

Employee Engagement

In simple terms, being engaged is all about feeling responsible for and committed to superior job performances (Britt, 2003). More detailed, Kahn (1990) described engagement as a unique and important motivational concept: the harnessing of an employee’s full self in terms of physical, cognitive and emotional energies during work role performances. Even though the term ‘engagement’ appeals to the imagination of most people, there are many explanations of what this term implies. There are also many alternative terms that are related to engagement, such as commitment, job satisfaction and employee well-being. The definition of Schaufeli, Salanova, González-Romá & Bakker (2002, p.74) is the basic assumption of this study. They stated that “Engagement is a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication, and absorption. Rather than a momentary and specific state, engagement refers to a more persistent and pervasive affective-cognitive state that is not focused on any particular event, object, individual, or behavior. Vigor is characterized by high levels of energy and mental resilience while working, the willingness to invest effort in one’s work and persistence even in the face of difficulties. Dedication refers to being strongly involved in one's work and experiencing a sense of significance, enthusiasm, inspiration, pride and challenge. Absorption is characterized by being fully concentrated and happily engrossed in one’s work, whereby time passes quickly and one has difficulties with detaching oneself from work”. These characteristics of engagement, vigor, dedication and absorption, will be the underlying concepts of engagement in this study.

When combining the definitions of Kahn (1990) and Schaufeli et al. (2002), one can conclude that engagement is a complex concept, that a) influences the individual through his/her physical, cognitive and emotional energies during work performances and b) refers to an overall situation, rather than a specific event, object, individual or behavior. This conclusion implies the importance of engagement, since it is a state of mind that influences the whole being of an individual at their work performances. Adding that conclusion to the arguments of Buckingham & Coffman (1999), who stated that employee engagement is a significant predictor of desirable organizational outcomes, such as productivity and

(9)

profitability, one can conclude that work engagement is important in everybody’s interest. Due to this crucial role, engagement is the main research subject in this study.

Even though Schaufeli et al. (2002) suggest that engagement refers to an overall situation, rather than a specific event, object, individual or behavior, Kahn (1990) suggests that employees can experience different dimensions of personal engagement during daily task performances. Engagement occurs when one is cognitively vigilant and/or emotionally connected to others (i.e. their supervisors). More recent literature also argues that leadership can affect the meaningfulness of employees’ work as measured by work engagement (Strickland et al., 2007). Other research showed that engagement is positively associated with job characteristics that might be labeled as resources, motivators or energizers, such as social support form co-workers and one's superior (Demerouti et al., 2001; Salanova et al., 2001, 2003; Schaufeli, Taris & Van Rhenen,2003). These authors were all describing the important role of leadership in determining employee engagement. It is therefore interesting to investigate the influence of the engagement of the leader, as perceived by the employee, on employee engagement.

Perceived Leader Engagement

The assumption that the level of engagement of the leader is relevant in the determination of the engagement of the follower is, among others, based on the fact that the supervisor is in most cases actually an employee as well. This effect, in which a supervisor has to cope with both their own interests and the interests of their teams at the same time, is for example examined in change management literature (Kanter, Stein & Jick, 1992). The authors stated that absorbing and coping with change themselves and passing it on, also involved the interpretation and translation of the change intent into implications for themselves and their teams. In other words, the leaders have two interests at the same time and the translation of the change for their teams, involves their personal interpretation. Therefore, the message towards the team is probably biased by the interpretation or personal preferences of the leader. This theory (Kaner, Stein & Jick, 1992), originated from change management literature, will be extended to the research field of engagement in this study.

(10)

As mentioned before, engagement is a state of mind that influences the whole being of the individual. A study among working couples showed that wives' levels of vigor and dedication uniquely contributed to husbands' levels of vigor and dedication, even when controlled for several work and home demands. The same effect applies to husband's levels of engagement that are likewise influenced by their wives' levels of engagement (Bakker, Demerouti & Schaufeli, 2003). These results on working couples imply that engagement is ‘contagious’, which means that it may cross over from one person to the other and vice versa. This process of transference or crossover by which one person ''catches'' the high level of engagement of the other person, may also be applicable on leaders and their subordinates. This assumption results in the following hypothesis:

Hypothesis 1: The perceived engagement of the leader has a positive relation with employee engagement

When looking at leadership and the engagement of leaders, it is important to keep in mind that being perceived as a leader is a prerequisite for being able to go beyond the formal role of influencing others (Lord & Mahler, 1991).This theory implies that one should keep the perception of the subordinates in mind while researching leadership. The (implicit) ideas that people have about their leader, called Implicit Leadership Theories (ILTs), are ideas about how leaders should behave and what is expected of them (Eden & Leviatan, 1975). Even though it sounds quite obvious that all employees have their own ILTs, it is important to keep these individual ideas in mind because ILTs have shown to be a possible bias in the measurement of actual leader behavior (Gioia & Sims, 1985). The authors argued that performance cues about the leader could be rated significantly different due to the fact that individuals respond in a way that reflects their own preconceptions about what leaders do. In order to take these biases of the individual preconceptions in account, this research will investigate the engagement of both the leader and the employee from the same source: the perception of the employee. In order to test at least part of the individual preferences of the employees, the inter-personal relationship between the leader and the follower was added as a moderating effect.

(11)

Influence of the inter-personal relationship

Even though most approaches to leadership are leader-centered, there are also theories that incorporate the follower, and the relationship between the leader and the follower, as second and third domains (Graen & Uhl-Bien, 1995). Issues of concern in the relationship-domain are reciprocal influence and the development and maintenance of effective relationships. Both concerns are applicable when it comes to leader engagement and employee engagement. In the short term, the reciprocal influence implies that the leader’s behavior and activities can have an influence on the follower and vice versa. They can therefore influence each other’s well-being, i.e. on engagement. In the long term, the level of engagement of both the leader and the follower can have an influence on the development and/or maintenance of the quality of the relationship between the leader and the follower. For example, when the employee is barely engaged, the leader is likely to give a lower rating of the quality of their relationship. The employee, in turn, is likely to do the same when the leader is barely engaged.

It is made clear that the relationship between the leader and the follower, the third domain according to Graen and Uhl-Bien (1995), is the linking factor between the engagement of the leader and the engagement of the employee. This inter-personal relationship will in this research be studied as a moderating effect on the relationship between perceived leader engagement and employee engagement. The inter-personal relationship was divided into three aspects: leader-member exchange, need for leadership and cooperation tenure. These aspects will be discussed in the following chapters.

Leader-Member Exchange

The Leader-Member Exchange Theory (LMX) describes the relationship between supervisors and subordinates on a continuum of low- to high-quality exchanges. The LMX theory proposes that relationships develop over time based on differential degrees of relational qualities (Graen & Cashman, 1975). High-quality relationships are characterized by mutual trust, respect, liking and reciprocal influence (Liden & Maslyn, 1998). Low-quality

(12)

relationships are characterized by low levels of trust and obligation, in which followers only do what is defined as part of their job description (Graen & Uhl-Bien, 1995). In relationships with high-quality LMX, individuals trust their managers, and interpret meaning and intent of management initiatives in a positive way (Furst & Cable, 2008). Due to this positive attitude, the levels of employee engagement may increase. Furthermore, due to the positive attitude resulting from a high-quality inter-personal relationship, the effect of the relationship between perceived leader engagement and employee engagement gets even stronger. These assumptions result in the following hypotheses:

Hypothesis 2: The positive relationship between perceived leader engagement and employee engagement is moderated by the extent to which the employee has positive exchange with the leader (LMX), such that the relationship between perceived leader engagement and employee engagement is stronger for higher values of LMX.

Need for leadership

This study focused on need for leadership as a characteristic of the employees. Need for leadership seems to be of immediate relevance for what happens in the interaction between the leader and the subordinate. It indicates that there is an interaction between the leader and the employee, in which the employee can perceive interventions of the leader as (un)wanted or (un)necessary. The subordinate will welcome the leader’s intervention when he/she considers it as instrumental to the achievement of a work goal. However, an unwanted intervention, or one considered to be unnecessary, is likely to result in opposition or neglect at the side or the subordinate (De Vries et al., 2002). This theory is clearly focused on the individual as one of the three dimensions of leadership (Graen & Uhl-Bien, 1995). Just as with the ILTs, need for leadership is an individual preference that can influence the relationship between the leader and the follower. De Vries et al. (2002) argued that with low need for leadership, the relation between leadership and outcomes is weaker compared with high need for leadership. Extending this knowledge to the research aim of this study, one could assume that the relationship between perceived leader engagement and employee engagement gets weaker with low values of need for leadership. On the other hand, the

(13)

relationship gets stronger with high levels of need for leadership, since the employee is less self-reliant and more dependent on the interventions of the leader instead. One could assume that the employee is therefore more susceptible for the level of engagement of the leader. These assumptions result in the following hypotheses:

Hypothesis 3: The positive relationship between perceived leader engagement and employee engagement is moderated by the extent to which the employee experiences a need for leadership (NFL), such that the relationship between perceived leader engagement and employee engagement is stronger for higher values of NFL.

Cooperation Tenure

Another aspect that could influence the perceived leader behavior is the effect of changes over time. Given the fact that Implicit Leadership Theories (ILTs) are personal assumptions about leadership, that are based on socialization and prior experiences, one could reasonably assume that a person’s ILTs are likely to change as a function of his or her experiences with leaders and organizational processes (Dickson, 1997; Rush & Russel, 1988) and might therefore change over time. One could assume that the relationship between the employee and the leader will change due to the employee’s new experiences with the leader and the organizational processes. This enhanced relationship might have a positive influence on the relationship between perceived leader engagement and employee engagement. These assumptions result in the following hypotheses:

Hypothesis 4: The positive relationship between perceived leader engagement and employee engagement is moderated by the duration of the cooperation, such that the relationship between perceived leader engagement and employee engagement is stronger for higher values of cooperation tenure.

(14)

Method

Procedure

In order to collect the data, an online questionnaire was conducted among a heterogeneous group of respondents, which consists of employees at several levels from operational to managerial staff. The data were obtained from respondents within a wide range of industries in The Netherlands. The basic condition to participate in the survey was an employment in which the employee had a direct relationship with his/her supervisor. Respondents were requested to answer a number of questions about their personal situation and about their personal opinion and preferences. They were also asked to answer a number of questions about the personal situation of their supervisor and about their perceptions of the engagement of the supervisor. Two versions of the questionnaire were conducted, with a difference in the sequence of the questions. Version A (Appendix 1) consisted of questions about the supervisor first, followed by questions about the individual. Version B (Appendix 2) consisted of questions about the individual first and questions about the supervisor subsequently. The purpose of this subdivision is to examine whether respondents would answer significantly different if they had questions about their supervisor first, or questions about their personal situation first. Thus, whether they were biased about the preferences of the supervisor when they just answered the questions about their personal preferences or vice versa. Both versions consisted of a total of 64 questions.

Measures

All items in this study were extracted from validated scales. The questionnaires were published in Dutch, since it was the most common language in the target group. Employee

engagement was measured on the complete 17-term version of the Utrecht Work Engagement Scale (UWES; Schaufeli, Bakker & Salanova, 2006). This scale included items for each of the three underlying dimensions of work engagement: Vigor (6 items), Dedication (5 items) and Absorption (6 items), with Cronbach’s alpha of .95. A 7-point scale was used on these items

(15)

ranging from 1: “Never” to 7: “Always/Daily”. Example items were “At my work, I feel bursting with energy” and “My job inspires me”. To measure the perceived leader

engagement, the employees answered the same 17-term UWES scale for their supervisor. Cronbach’s alpha of .96. The employees were asked to picture themselves an average day and make an estimation of the behavior or preferences of their supervisor.

The inter-personal relationship was divided into three aspects: leader-member exchange (LMX), need for leadership (NFL) and cooperation tenure. The respondents were asked to answer these questions according to their own situation or preferences.

Leader-member exchange (LMX) was measured on a 7-item scale (Graen & Uhl-Bien, 1995) with Cronbach’s alpha of .86, focusing on the quality of the relationship between the employee and the supervisor. An example was “How well does your supervisor understand your job problems and needs?”. Six out of seven items were measured on a 5-point scale (1: “Not at all” to 5: “Fully”). The seventh item “How would you characterize your working relationship with your leader?” was measured on a 5-point scale with answer categories ranging from 1: “Very ineffective” to 5: “Very effective”.

Need for leadership (NFL) was measured on a 17-item scale (De Vries et al., 2002), with Cronbach’s alpha of .86, which indicates whether the employee perceives interventions of the leader as (un)wanted or (un)necessary. Example items were “I need my supervisor to set goals” and “I need my supervisor to motivate me”. All 17 items were measured on a 5-point scale, ranging from 1: “Not at all” to 5: “A lot”.

Cooperation tenure was measured with one question: “How long do you work together with your current supervisor?”. Respondents were asked to provide an answer in a number of years, ranging from 0 to 50, with one decimal point allowed.

Control Variables

Several control variables were added to the questionnaire, including some demographics. Age, gender, job tenure, education level and industry were asked. The control variable position was added in order to investigate the hierarchical position of the respondent. For the supervisors, employees were asked to answer about the demographics of their supervisors or to estimate as closely as possible. The control variables were tested whether

(16)

they had a relationship with employee engagement. These results (Appendix 3) indicate that none of the control variables had a significant correlation with employee engagement. Therefore, the control variables were not included in the statistical analyses of this study.

Sample

A total of 186 employees participated in the questionnaire. 92 employees conducted version A and 94 employees conducted version B. A total of 130 respondents completed the whole questionnaire. 56 respondents completed version A and 74 respondents completed version B. Of all respondents, 40.8% were male and 59.2% were female. Of all supervisors, 73.8% were male and 26.2% were female.

The average age of all respondents was 38.5 years, with a standard deviation of 11.8 years. The minimum age was 19 years and the maximum age was 63 years. From all supervisors, the average age was 37.2 years with a standard deviation of 7.8 years. The minimum age was 25-34 years and the maximum age was 65-74 years.

The average job tenure among the employees was 9.5 years, with a standard deviation of 9.7 years. Of all employees, 41 years was the longest job tenure. Among the supervisors, the average job tenure was 5-10 years, and the maximum job tenure was 40 years.

The results showed that the highest educational level among all respondents was a WO diploma (24.6%) and the lowest educational level was secondary school (3.8%). Other respondents had an MBO diploma (9.2%) or an HBO diploma (59,2%).

The education level for the supervisors was as follows: an HBO diploma was most common (45.4%) and secondary school was least common (0.8%). Other supervisor had an MBO diploma (5.4%) or an WO diploma (30.0%). For the remaining 18.5%, the respondents did not know the educational level of their supervisor.

23.1% of the employees and supervisor were working in financial services. 20.0% of the respondents were working in a teaching institute and 10.0 % of the respondents were working in commercial services. 23.1% of the respondents were in a governmental employment and 4.6 % were working in the retail industry. The remaining respondents were working in the tourism and recreation industry (3.1%), the transportation industry (3.1%) or other industries (6.9%).

(17)

When it comes to the hierarchical position, 59.6% of the respondents were operational staff. The managerial staff was divided into team management (8.5%), middle management (13.8%), and senior management (7.7%). Only 3.1% of the respondents operated on a board level and the remaining 10.0% of the respondents indicated to operate otherwise.

Strategy of Analysis

The SPSS software package was used to perform the statistical analysis of this study. Data was gathered from an online survey with two versions: version A and version B. The sequence of the questions was the single difference between the two versions. Prior to the statistical analysis, an independent samples t-test was conducted to compare employee engagement in version A and version B. The results (Appendix 4) showed that there was no significant difference (t (128) = -.680; p = .498) between version A (M = 5.27; SD = .98) and version B (M = 5.39; SD = 1.04). Therefore, the respondents of both versions were merged into one database.

The next step was to compute scale means according to the validated scales. The reliability of the scales was reported, as well as the possible improvement if one item was deleted (Appendix 5). Results showed that all scales had an excellent reliability, which would not significantly improve if any items were deleted.

Prior to the statistical analyses to test the hypotheses, a check for normality was conducted. The skewness levels of variables should be around zero for a normal distribution and are significant (p <.05) when their z-scores are greater than 1.96 (Field, 2009). Significant skewness had been detected at three of the five variables in this study. Employee engagement had moderate negative skewness (s = -.69, p <.05) , perceived leader engagement had substantial negative skewness (s = -1.02, p <.05) and cooperation tenure had moderate positive skewness (s = .78, p <.05). The skewness levels of these variables indicated that the dataset was not normally distributed. However, a transformation of the data did not result in improved scores in the regression analyses. Therefore, the dataset was treated as a normal distribution and was not transformed.

After the check for the assumption of normality, descriptive statistics of all variables were conducted. Subsequently, a correlation matrix was conducted to show the correlations

(18)

between the main variables in this study. Both the correlation matrix and the scores on the variance inflation factor (VIF) were checked for multicollinearity problems. The correlation matrix reported that all correlations were equal to or less than r = .30 and the VIF scores were well below the cut-off value of 10 (Netter, Wasserman & Kutner, 1990), which indicated that there were no issues on multicollinearity.

To test the hypothesis with a direct effect, a linear regression model was used. To test the interaction effects of the moderating hypotheses, multiple regression models were used. Prior to the moderated regression analyses, the variables were standardized by computing their Z-scores. A final regression model was conducted with all moderating variables included, in order to investigate which moderator has the greatest influence on the relationship between perceived leader engagement and employee engagement.

This study assumed an alpha-level of α .01 or α.05 in all statistical analyses. All results were rounded to two decimal places.

(19)

Results

Descriptive statistics and correlations

The descriptive statistics, correlations and scale reliabilities for the main variables in this study are provided in Table 1. It shows that the means of employee engagement (M = 5.43) and perceived leader engagement (M = 5.57), both measured on a 7-point scale, are quite high. Leader-member exchange (M = 3.61) and need for leadership (M = 2.53) are both measured on a 5-point scale and the table shows a quite high mean on leader-member exchange. Cooperation tenure was indicated in a number of years, with a mean of 2.81 years. The standard deviations of the variables employee engagement (SD = .89), perceived leader engagement (SD = .98), leader-member exchange (SD = .64) and need for leadership (SD= .60) are quite small. This means that a majority of the respondents answered these questions quite similarly.

As mentioned before, the internal reliabilities (Cronbach’s alpha coefficients) of all scales used in this study were analyzed. This analysis indicated very high coefficients, which are shown in parentheses on the diagonal in Table 1.

Table 1: Descriptive statistics, Correlations and Reliabilities

M SD 1. 2. 3. 4. 5.

1. Employee engagement 5.43 .89 (.95)

2. Perceived leader engagement 5.57 .98 .35** (.96)

3. Leader-member exchange 3.61 .64 .29** .30** (.86)

4. Need for leadership 2.53 .60 .19* .28** .17* (.86)

5. Cooperation tenure 2.81 1.79 -.21* -.09 -.05 -.14 -

N= 130 for Employee engagement; Perceived leader engagement, Leader-member exchange, Need for leadership; N= 78 for Cooperation tenure ** Correlation is significant at the 0.01 level (1-tailed).

* Correlation is significant at the 0.05 level (1-tailed).

Descriptive statistics, correlations and reliabilities are reported for non-transformed data. Cronbach’s alpha coefficients are provided in parentheses on the diagonal.

(20)

Table 1 provides the correlation coefficients and shows that employee engagement is significantly positively related to perceived leader engagement (r = .35, p < .01), to leader-member exchange (r = .29, p < .01) and to need for leadership (r = .19, p < .05). Leader-member exchange and need for leadership are also significantly positively related to perceived leader engagement (respectively r = .30, p < .01 and r = .28, p. < .01). These correlations show a small to medium effect, but the effect of the correlation between perceived leader engagement and employee engagement is slightly bigger. The results in Table 1 show that cooperation tenure is only significantly correlated to employee engagement (r = -.21, p < .01). The correlation results in Table 1 imply some direct relationships, especially between perceived leader engagement and employee engagement, which is mentioned in Hypothesis 1.

Direct effect

Hypothesis 1 stated that perceived leader engagement has a positive influence on employee engagement. This direct relationship was examined by the use of linear regression. The model was statistically significant with F (1,128) = 21.89; p < .01 and explained 14.6 % of the variance in employee engagement. Perceived leader engagement showed to have a significant effect on employee engagement (β = .38, p < .01) which means that employees score higher on their own engagement when they score their supervisor higher on his/her perceived leader engagement. Therefore, Hypothesis 1 is supported.

Moderating effects

To test the moderating effects, multiple regression models were used. Table 2, 3 and 4 are visualizing the results of these regression analyses, with employee engagement as the dependent variable. Two models were specified in this analysis. Model 1 includes the independent variable (perceived leader engagement) and the moderating variables a) leader-member exchange, b) need for leadership and c) cooperation tenure. Subsequently, the interaction effect was included (model 2) by adding the respective product terms of the independent variables with the moderating variables. Table 2 reports the interaction term of

(21)

leader-member exchange, Table 3 reports the interaction term of need for leadership and Table 4 reports the interaction term of cooperation tenure.

Hypothesis 2 stated that leader-member exchange has a moderating effect on the relationship between perceived leader engagement and employee engagement, such that the positive relationship between perceived leader engagement and employee engagement gets stronger for higher values of leader-member exchange. The results of the regression analysis are presented in Table 2.

Table 2: Employee engagement and leader-member exchange

Employee engagement

Model 1 (a) Model 2 (b)

Perceived leader engagement .29** .37**

Leader-member exchange .20* .17*

Perceived leader engagement x Leader-member exchange .20* R .40 .44 R² .16 .19 R² change .16 .04 F 11.80 9.98

N=130. Standardized regression coefficients are reported.

Results are reported for non-transformed data. **. Coefficient is significant at the 0.01 level *. Coefficient is significant at the 0.05 level

a. Predictors: (Contstant), Perceived leader engagement, Leader-member exchange

b. Predictors: (Contstant), Perceived leader engagement, member exchange, Perceived leader engagement x Leader-member exchange

The results indicate that the moderating effect of leader-member exchange was supported, since there was a significant positive moderating effect (β = .20, p < .05) of leader-member exchange. This implies that the relationship between perceived leader engagement and employee engagement gets stronger when the employee scores higher on leader-member exchange. The moderated regression model (model 2) was statistically significant with F (1,126) = 9.98 and explained 19.2% of the variance in employee engagement. The entry of leader-member exchange as a moderator to the model, explained an additional 3.5% of the variance in employee engagement (R² change = .04, p < .05). Therefore, Hypothesis 2 is supported.

(22)

To plot this interaction, high and low levels of the moderator terms were represented by one standard deviation above and below the means of perceived leader engagement and leader-member exchange. As visualized in Figure 2, there is a stronger effect of perceived leader engagement on employee engagement in inter-personal relationships with high leader-member exchange. In inter-personal relationships with low leader-leader-member exchange, the effect of perceived leader engagement on employee engagement is smaller.

Figure 2: Two-way interaction effect between leader-member exchange and perceived leader engagement on

(23)

Hypothesis 3 stated that need for leadership has a moderating effect on the relationship between perceived leader engagement and employee engagement, such that the positive relationship between perceived leader engagement and employee engagement gets stronger for higher values of need for leadership.

Table 3: Employee engagement and need for leadership

Model 1 (a) Model 2 (b)

Need for leadership .10 .10

Perceived leader engagement x Need for leadership

-.05

R .36 .36

R² .13 .13

F 9.42 6.33

a. Predictors: (Contstant), Perceived leader engagement, Need for leadership

b. Predictors: (Contstant), Perceived leader engagement, Need for leadership, Perceived leader engagement x Need for leadership

The correlation matrix (Table 1) indicates that there is a significant positive correlation between need for leadership and employee engagement (r = .19, p < .05) and between need for leadership and perceived leader engagement (r = .28, p < .01). On the other hand, the regression analysis (Table 3) indicates a very small negative moderating effect of need for leadership, which is in contrast to the hypothesized positive interaction. However, both the regression model (F (1,126) = 6.33; p = .60) and the moderating effect of need for leadership were not significant (β = -.05, p < .60). Therefore, Hypothesis 3 is not supported.

(24)

Hypothesis 4 stated that cooperation tenure has a moderating effect on the relationship between perceived leader engagement and employee engagement, such that the positive relationship between perceived leader engagement and employee engagement gets stronger for higher values of cooperation tenure.

Table 4: Employee engagement and cooperation tenure

Model 1 (a) Model 2 (b)

Cooperation tenure -.19 -.17

Perceived leader engagement x Cooperation tenure

.03

R .39 .39

R² .15 .15

F 6.67 4.40

a. Predictors: (Contstant), Perceived leader engagement, Cooperation tenure

b. Predictors: (Contstant), Perceived leader engagement, Cooperation tenure, Perceived leader engagement x Cooperation tenure

The correlation matrix (Table 1) showed that cooperation tenure was only significantly correlated to employee engagement (r = -.21, p < .05). The moderated regression model (Table 4) was not significant (F (1,74) = 4.40; p = .83) and showed no significant effect (β = .03, p = .83) of cooperation tenure as a moderator in the relationship between perceived leader engagement and employee engagement. Therefore, Hypothesis 4 is not supported.

As mentioned before, a final regression model was conducted in order to test which moderating variable in this study had the strongest effect. Due to the fact that cooperation tenure had considerably less respondents than the other moderating variables (N=78 for cooperation tenure and N=130 for other variables, as reported in the subscripts of Table 2, 3 and 4), cooperation tenure was not included in the final regression model. The results are presented in Table 5.

(25)

Table 5: Employee engagement, leader-member exchange and need for leadership

Model 1 (a) Model 2 (b)

Leader-member exchange .19* .18*

Need for leadership .09 .08

Perceived leader engagement x Leader-member exchange

.21* Perceived leader engagement

x Need for leadership

-.10

R .40 .45

R² .16 .21

R² change .16 .04

F 8.20 6.40

a. Predictors: (Contstant), Perceived leader engagement, Leader-member exchange, Need for leadership,

b. Predictors: (Contstant), Perceived leader engagement, Leader-member exchange, Need for leadership, Perceived leader engagement x Leader-member Exchange, Perceived leader engagement x Need for leadership.

The results in Table 5 indicate that the regression model was significant (F(2,124) = 6.39; p < .05) and explained 20.5 % of the variance in employee engagement. As presented in Table 3, the regression model of the moderating effect of need for leadership was not significant with F (1,126) = 6.33; p = .60. Thus, there was not enough evidence to prove the moderating effect of need for leadership. However, when the moderating effects of need for leadership and leader-member exchange are combined, the model explains a significant proportion of variance on employee engagement (R² = .20 with F(2,124) = 6.39; p < .05). Model 2 (Table 5) shows that the combined moderating effects of leader-member exchange and need for leadership, improved the level of variance on employee engagement that was explained by this model (R² change = .04, p < .05).

Adding cooperation tenure as a third moderator to the model did not improve the moderating effects and significance levels (Appendix 6), since the model was not significant (F(3,70) = 3.42; p = .55).

To conclude the overall analytical results of this study, there was a statistically significant positive relationship between perceived leader engagement and employee engagement. This relationship was significantly positively moderated by leader-member

(26)

exchange. There was not enough evidence to support the moderating role of need for leadership and cooperation tenure. However, taking two moderators (leader-member exchange and need for leadership) together increased the level of variance on employee engagement that was explained by this model.

(27)

Discussion

This chapter elaborates on the results of the statistical analyses. Theoretical and practical implications will be discussed by comparing the results with prior research and by discussing the contribution of this study in the practice of employees, supervisors and organizations.

Perceived leader engagement and employee engagement

This study investigated whether the engagement of the leader, as perceived by the employees, had an influence on employee engagement. Prior research argued that leadership can affect the meaningfulness of employees’ work as measured by work engagement (Strickland et al., 2007). In addition to the findings of Strickland et al. (2007), the results of the present study showed that the leader can influence the work engagement of the employee, more specifically, by the leader’s own engagement since there was a statistically significant positive relation (β = .38, p < .01) between perceived leader engagement on employee engagement (Table 1). Bakker et al. (2003) found that engagement has ‘contagious’ characteristics in their research among working couples. Even though the relationship in the present study was based on a working relationship rather than a romantic relationship, the findings are similar in that they showed a significant relationship between two persons that are related to each other. The present study adds to the results of Bakker et al. (2003) since the working relationship between employees and supervisors is more relevant to organizations.

In practice, the results of this study indicate that supervisors should be aware of the fact that their engagement may influence the engagement of their employees. As stated by Buckingham & Coffman (1999), employee engagement is a significant predictor of desirable organizational outcomes, such as productivity and profitability. Therefore, organizations are likely to spend much time, effort and money on fancy programmes to improve the engagement of their employees. However, the results of this study show organizations that their investments could be neutralized when they do not pay attention to the engagement of the supervisors. Getting back to the aforementioned statement that supervisors are actually also employees themselves, one could conclude that organizations should pay attentions to the

(28)

engagement of all their employees, in its broadest sense: from operational staff to management and board members.

Leader-member exchange as a moderator

The hypotheses of this study stated that the positive relationship between perceived leader engagement and employee engagement was moderated by their inter-personal relationship. This relationship incorporated leader-member exchange, need for leadership and cooperation tenure as separate moderators. The results (Table 2) indicated that leader-member exchange has a positive moderating effect (β = .20, p < .05). This is consistent with prior research which found that in relationships with a high quality of leader-member exchange, individuals trust their managers and interpret the meaning and intention of management initiatives in a positive way (Furst & Cable, 2008). Due to this positive attitude, the levels of employee engagement may increase. In contribution to the direct relationship in the research of Furst and Cable (2008), this study found a significant influence of leader-member exchange as a moderator on the relationship between the perceived engagement of the leader and the engagement of the employee. The results of this study showed that the relationship between perceived leader engagement and employee engagement gets stronger when the employees score higher on leader-member exchange. In other words, when the employees scored the relationship with their supervisor as ’high quality’, the influence of the engagement of the supervisor on the engagement of the employee gets even bigger. Figure 2 clearly visualized this moderating effect of leader-member exchange.

In practice, this additional insight can be used as a resource in increasing employee engagement. If organizations are willing to invest in the engagement of their employees, they should keep the engagement of the supervisor in mind, but also the quality of the relationship between both persons. The fact that the leader-member exchange moderates the relation between perceived leader engagement and employee engagement, gives organizations multiple ways to increase employee engagement. For example, when a certain supervisor is not very engaged to his/her work and is also not a very change-oriented, organizations can instead work on both aspects (quality of the relationship and engagement of the leader) at the same time on a slower pace. The moderating effect of leader-member exchange will

(29)

accelerate the effects of the time, money and effort that was spent on both aspects. This will prevent the supervisor from getting the feeling of being rushed or forced, but will in the meantime still affect, and hopefully increase, the engagement of the employees.

Need for leadership as a moderator

In addition to De Vries et al. (2002), who argued that with low need for leadership, the relation between leadership and outcomes is weaker compared with high need for leadership, this study hypothesized a moderating effect of need for leadership, such that the relationship between perceived leader engagement and employee engagement is stronger for higher values of need for leadership. However, the results (Table 3) did not support this hypothesis. Even though need for leadership was significantly positively correlated to employee engagement (r = .19, p < .05) and to perceived leader engagement (r = .28, p < .01), there was no significant moderating effect of need for leadership (β = -.05, p < .60) on the relation between perceived leader engagement and employee engagement. An explanation for these results can be found in a discussion of cause and effect. Even though need for leadership is significantly related to both perceived leader engagement and employee engagement, the nature of this relation is not clear. It could be that leaders show more support and structure when the employees need them to do so, or the leaders could invoke the amount of need for leadership of the employee with this kind of leadership (De Vries et al., 2002).

In practice, it is clear that multiple factors are influencing a certain aspect as the same time. This is also the case when it comes to employee engagement. In order to combine some of the moderators in this study, a combination of leader-member exchange and need for leadership was made in one model. Cooperation tenure was not added due to the considerably lower number of respondents (Table 4). The results (Table 5) showed that the combination of leader-member exchange and need for leadership as moderators explained 20.5% of the variance (p < .05) on employee engagement, which was more than the variances explained by leader-member exchange as a moderator (19.3%, p < .05) and need for leadership as a moderator (not significant with R² = .13, p = .60).

This study provided insights on the (moderating) effects of perceived leader engagement, leader-member exchange and need for leadership. These insights are useful for

(30)

organizations who are willing to invest in their employee engagement. Employee engagement could be increased when the leader’s engagement is increased, since the results indicated a significant positive relationship between perceived leader engagement and employee engagement. However, this effect on employee engagement could be stronger when the quality of the relationship (leader-member exchange) is incorporated, and even stronger when leader-member exchange and need for leadership are incorporated together.

Cooperation tenure as a moderator

As mentioned by Eden and Leviatan (1975), implicit leadership theories (ILTs) are ideas that employees have about their leaders, about how leaders should behave and what is expected of them. These ILTs are likely to change over time (Dickson, 1997; Rush & Russel, 1988). Based on these findings one could assume that these changing ILTs might influence the relationship between the employee and the leader when they are working together for more years. However, the results of this study did not support this hypothesis. There was no significant effect (β = .03, p = .83) of cooperation tenure as a moderator. This could be explained by the findings of research on schema change, since ILTs are seen as cognitive schemas. Jelinek, Schmircich and Hirsch (1983) and Labianca, Gray and Brass (2000) proposed that once schemas are established, they tend to endure and are resistant to change, even when disconfirming information is presented. Poole, Gioia and Gray (1989) and Labianca et al. (2002) found that schemas can change when employees have undergone experiences or have received information that made them question the validity of their existing schemas. That implies that the ideas that employees have about their supervisors do not causelessly change over time, which seems to be consistent with the findings in the present study, in which there was no correlation between cooperation tenure and perceived leader engagement (r = .09, p = .21). In practice, the results of the present study imply that cooperation tenure is not an important moderator on the relationship between perceived leader engagement and employee engagement.

(31)

Limitations and suggestions for further research

This study had some limitations in the design and execution. A first limitation was the way of selecting the sample of the population. Since the sample was derived from a large heterogeneous population (Dutch speaking employees with a direct relationship with their supervisor), a non-probability sampling technique called self-selection sampling was used. This means that respondents were invited through different types of social media. A pitfall of this technique is that respondents often participate in the survey because they have strong feelings or opinions about the research topic (Saunders & Lewis, 2012). To improve the quality of the data, further research could use a more precise sampling techniques like quota sampling, based on quota that represent the population.

Another limitation of this study is the nature of the research sample. The research sample consisted of employees within The Netherlands (or more precisely: Dutch speaking employees). This affected the generalizability of the findings in this study, which were only representative for The Netherlands in the current time frame. Further research could investigate the effect of perceived leader engagement on employee engagement and the moderating effects of leader-member exchange, need for leadership and cooperation tenure in other countries, in order to find results that are more generalizable. Furthermore, the sample was fairly homogeneous when it comes to the demographics of the supervisors. Most of the supervisors were male (73.8 %) and most were between 45 and 54 years old (52.3%). The small differences between the supervisors could have resulted in smaller statistical effects. Further research could prevent this small differences between the supervisors in the sample, by using the aforementioned quota sampling technique.

The third limitation is related to the control variables in this study. Some of them might have decreased the validity of the overall research, which is caused by the fact that respondents had to estimate some of the control variables for the supervisor. For example, the answers for age and education level of the supervisor were provided with answer categories (see Appendix 1 and 2). The reason that these answer categories were added to the questionnaire design, was to lower the threshold for respondents to answer. If one does not know the age of his/her supervisor, an answer category like ’45 to 54 years’ is easier to choose, which results in higher response scores. However, the answer categories could have

(32)

resulted in a lower validity of the control variables because they might not have been specific enough. Further research could improve the validity of the control variables by making them as clear, concise and specific as possible. In that way, the control variables can provide a better context for the research.

The fourth limitation of this study is the fact that the measures of the variables in this study shared common methods, which threatens the validity of the conclusions drawn from the results of the study. Since the data of this study was gathered by a self-report questionnaire and all variables were measured at the same time in one survey, it is susceptible of common method bias. The influence of common method bias is a major concern in survey-based research, especially for self-report measures (Spector, 2006), since these are inherently biased by the respondent's feelings at the time they answered the questionnaire. Additionally, the answers in this study were given by a common rater and are likely to have a tendency of social desirability of the respondent. This is a potential source for common methods bias (Podsakoff, MacKenzie, Lee & Podsakoff, 2003). Furthermore, Likert scales were used in the questionnaire, which are characterized as itemized rating scales. Although the use of these scale formats are argued to make is easier for respondents to complete the questionnaire, it was also found in prior research that scale formats and anchors systematically influence responses (Tourangeau, Rips & Resinski, 2000). For example, with Likert scales one could assume that respondents are likely to respond towards the middle in order to avoid the extremes. It was stated that Likert scales are also a potential source for common method bias (Podsakoff et al., 2003). Thus, the present study has multiple characteristics that could potentially lead to common method bias. This might have influenced the correlations between the variables in this study and are therefore a thread to the validity of the conclusions drawn from this research. Further research should try to identify what the measures of the predictor and criterion variables have in common and eliminate or minimize it through the design of the study.

The fifth limitation of this study was caused by a practical issue. The basic criterion to participate in the questionnaire was an employment in which there was a direct relationship with the supervisor. However, it was not made clear what the quality of the relationship should be, in order to avoid influencing the respondents. Some of the respondents indicated that they did have a direct relationship with their supervisor, but the frequency of meeting each other was very low (i.e. three times a year). Therefore, they could not make a decent

(33)

estimation of the preferences of their supervisor. This might have been one of the causes that from a total of 186 respondents, only 130 completed the whole questionnaire. Further research should make a more clear overview of the basic criterion to participate in the survey, in which also a minimum number of meeting times per year is indicated.

(34)

Conclusion

This study investigated the relationship between perceived leader engagement and employee engagement and the moderating effect of the inter-personal relationship between the employee and the supervisor. Graen and Uhl-Bien (1995) indicate that there are three domains of leadership approaches. The domains of leader-centered and follower-centered theories were represented in this study by perceived leader engagement and employee engagement. The third domain, the relationship between the leader and the follower, was represented in this study by the three moderating variables: leader-member exchange (which scores the quality of the relationship between the employee and the supervisor), need for leadership (which scores whether the employee perceives interventions of the leader as (un)wanted or (un)necessary) and cooperation tenure (which scores the amount of years the employee and the supervisor work together).

The results showed that there was a significant positive relationship between perceived leader engagement and employee engagement. That implies that employees score higher on their own engagement when they score their supervisor with higher values on his/her engagement. In practice, this means that supervisors have to take their own engagement into account when they want to improve the engagement of their employees. Positive moderating roles of need for leadership and cooperation tenure were hypothesized, but the results were not significant and did not support these hypotheses. On the other hand, the results did show significant positive moderating role of member exchange. Higher scores on leader-member exchange (implying a good quality of the relationship between the employee and the leader) resulted in a stronger relationship between perceived leader engagement and employee engagement. In practice, it is clear that multiple factors are influencing a certain aspect as the same time. For organizations who are willing to invest time, money and effort in increasing their employee engagement, this study provided insights on the (moderating) effects of perceived leader engagement, leader-member exchange and need for leadership.

(35)

References

Bakker, A.B. Demerouti, E. & Schaufeli, W.B. (2003). Crossover of burnout and engagement among working couples. Manuscript in preparation.

Britt, T.W. (2003). Aspects of identity predict engagement in work under adverse conditions.

Self & Identity, 2, 31-45.

Buckingham, M. & Coffman, C. (1999). First, Break all the Rules. Simon and Schuster, New York, NY.

Dansereau, F., Graen, G.B., & Haga, W. (1975). A vertical dyad linkage approach to leadership in formal organizations. Organizational Behavior and Human Performance, 13,

46-78.

Demerouti, E., Bakker, A.B., Janssen, P.P.M. & Schaufeli, W.B. (2001). Burnout and engagement at work as a function of demands and control. Scandinavian Journal of Work,

Environment & Health, 27, 279-286.

Dickson, M. W. (1997). Universality and variation in organizationally shared cognitive prototypes of effective leadership. Unpublished doctoral dissertation, University of Maryland,

College Park.

De Vries, R. E., Roe, R. A., & Taillieu, T. C. (2002). Need for leadership as a moderator of the relationships between leadership and individual outcomes. The Leadership Quarterly, 13(2), 121-137.

Eden, D., & Leviatan, U. (1975). Implicit leadership theory as a determinant of the factor structure underlying supervisory behavior scales. Journal of Applied Psychology, 60, 736–

741.

Field, A. (2009). Discovering statistics using SPSS. Third edition. London, UK: SAGE

publications Ltd.

Furst, S.A. & Cable, D.M. (2008). Employee Resistance to Organizational Change: Managerial Influence Tactics and Leader–Member Exchange. Journal of Applied Psychology,

93: 453–462.

Gioia, D.A., & Sims, H.P. (1985). On avoiding the influence of implicit leadership theories in leader behavior descriptions. Educational and Psychological Measurement, 45, 217–243. Graen, G.B. and Cashman, J.F. (1975). ‘A role making model in formal organizations: A developmental approach’, in J.G. Hunt and L.L. Larson (eds), Leadership Frontiers, Kent,

(36)

Graen, G.B., Uhl-Bien, M. (1995). Relationship-based approach to leadership: Development of a Leader-Member Exchange (LMX) Theory of leadership over 25 years: Applying a multi-level multi-domain perspective. Leadership Quarterly, 6(2), 219-247.

Jelinek, M., Smircich, L., & Hirsch, P. (1983). Introduction: A code of many colors.

Administrative Science Quarterly, 28, 331–338.

Kahn, W. A. (1990). Psychological conditions of personal engagement and disengagement at work. Academy of Management Journal, 33: 692–724.

Kanter, R.M., Stein, B.A. & Jick, T.D. (1992). The Challenge of Organizational Change: How companies experience it and leaders guide it. New York.

Labianca, G., Gray, B., & Brass, D. J. (2000). A grounded model of organizational schema change during empowerment. Organization Science, 11, 235–257.

Liden, R.C. & Maslyn, J.M. (1998). Multidimensionality of Leader–Member Exchange: An Empirical Assessment through Scale Development. Journal of Management, 24: 43–72. Lord, R.G., & Maher, K.J. (1991). Leadership and information processing. London:

Routledge.

Netter, J., Wasserman, W. & Kutner, M. H. (1990). Applied Linear Regression Models: Regression, Analysis of Variance, and Experimental Designs. Holmwood, IL: Irwin

Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y., & Podsakoff, N. P. (2003). Common method biases in behavioral research: a critical review of the literature and recommended remedies. Journal of applied psychology, 88(5), 879.

Poole, P. P., Gioia, D. A., & Gray, B. (1989). Influence modes, schema change, and organization transformation. The Journal of Applied Behavioral Science, 25, 271–289.

Rush, M. C., & Russell, J. E. A. (1988). Leader prototypes and prototypecontingent consensus in leader behavior descriptions. Journal of Experimental Social Psychology, 24, 88–104. Salanova, M., Agut, S., & Pieró, J.M. (2003). Linking organizational facilitators and work engagement to extrarole performance and customer loyalty : The mediating role of service climate. Submitted for publication.

Salanova, M. Grau, R., Llorens, S., & Schaufeli, W.B. (2001). Exposición a las tecnologías de la información, burnout y engagement: el rol modulador de la autoeficacia profesional [Exposure to information technology, burnout and engagement: about the role of professional self-efficacy]. Psicología Social Aplicada, 11, 69-89.

Salanova, M., Llorens, S., Peiró, J.M., & Schaufeli, W.B. (2003). How emotions predict self-efficacy: The mediating role of job engagement. Submitted for publication.

The influence of perceived leader engagement on employee engagement and the moderating role of their inter-personal relationship