Highlight, copy & paste to cite:

Amah, O. E. (2009). Job Satisfaction and Turnover Intention Relationship: The Moderating Effect of Job Role Centrality and Life Satisfaction, Research and Practice in Human Resource Management, 17(1), 24-35.

Job Satisfaction and Turnover Intention Relationship: The Moderating Effect of Job Role Centrality and Life Satisfaction

Okechukwu E. Amah


Turnover affects the cost of operations and drains the organisation of inherent tacit knowledge. To manage these effects job satisfaction and turnover relationships have been studied. However, only moderate results have been obtained, a situation blamed on the exclusion of individual difference factors, and other relationships involving these factors. Using data from 400 employees of a new generation bank located in Lagos, Nigeria, the relationship among job satisfaction, life satisfaction, role centrality and turnover intention was evaluated. Job satisfaction was found to have a direct negative relationship with turnover intention. Life satisfaction and role centrality moderated this relationship, such that participants, who were low in life satisfaction and role centrality, had greater tendency to exit the organisation even at high levels of job satisfaction. These results indicate that the effect of job satisfaction on turnover can be enhanced in two ways; namely, when employees find congruence between their job and their self identity, and when involvement in such jobs enhances their overall life satisfaction. Consequently, human resources managers will be challenged to place a high priority on job design, and develop policies that help employees to balance their work and non work involvements so as to enhance their overall life satisfaction.


The exit of low performing employees is beneficial to an organisation. However, the exit of important and valuable human resources from the organisations leads to the loss of the tacit knowledge possessed by the leavers, and also has a huge effect on the cost of operation, as a result of the high cost of filling the positions that are vacant. Hence, turnover negates the benefit that should have been derived by organisations when they spend scarce resources attracting, selecting, socialising employees, and developing and retaining performing employees (Wright & Bonett 2007). Indeed, the cost of turnover has become an issue in recent times because of its drain on the operating profits of the organisations. For example, Hom and Griffeth (1995) stated that turnover cost is estimated at between 50 and 100 per cent of employee’s annual cost, and this cost does not include that of the tacit knowledge taken out of the corporation by the leavers. Moreover, a high rate of voluntary turnover is linked to low productivity (Huselid 1995), and poor future revenue growth (Baron, Hannan & Burton 2001). These negative effects justify the large number of studies on the pathways to employee turnover. Despite the existence of these studies, the understanding of the pathways to employees’ turnover is still ‘one sided’, emphasising the withdrawal model (Johns 2002).

The relationship between job satisfaction (JS) and employee turnover is one of the pathways to better understanding employee turnover - organisation linkages. This connection of employee responses has attracted the interest of researchers, because satisfied employees have a greater tendency to stay, and contribute to an organisation’s competitive advantage and productivity (Lee & Mowday 1987, Wright & Bonett 2007). In fact, Wright and Bonett (2007) stated that the negative relationship between JS and turnover is moderate, with substantial variation across studies. Two reasons have been given for this moderate relationship, namely, what Wright and Bonett (2007:143) referred to as “... temporal differences in when the turnover data were actually collected.”, and situational variables (Sweeney & McFarlin 2002). The former explanation is interpreted to be the inability to distinguish between voluntary and involuntary turnover based on how turnover data were acquired. One way of solving the turnover data collection problem is to use turnover intention (TI). Studies have shown that TI is a good measure of actual turnover behaviour, and that attitude is an immediate determinant of TI (Fishbein & Ajzen 1975, Ajzen & Fishbein 1980, Zimmerman 2008). Apart from this suggestion, there are two other ways to improve the magnitude of the job satisfaction turnover relationship. The first way is to study more variables including situational variables expected to be antecedents of turnover behaviour. The second strategy is to consider various forms of the relationship involving these antecedents. Wright and Bonett (2007) tested the interactive effect of one individual factor, psychological well being, and found a significant interactive effect on the JS turnover relationship. These authors also recognised the existence of other individual factors that are potential antecedents of employee turnover, and recommended that future studies might be designed to identify these variables.

Another individual factor that has been identified as important in the attitude and work outcome relationship is work centrality (WC). This construct is the relative importance of work or family in the life of an individual. For instance, Carr, Boyar and Gregory (2008) stated that individuals may have strong ties to work that will have the potential of affecting their behaviour towards the work system. These authors found a significant interactive effect in the work family conflict and the turnover relationship. Earlier, Carlson and Kacmar (2000) called for studies to test the role of WC in the relationship involving attitudes and employee turnover as there is a viewpoint that WC focuses on the centrality of the entire work system without reference to the particular role of the individual. However, in the study reported in this article, the focus is on the centrality of the current role an individual is playing in the work place. Role centrality (RC) is thus, defined as the importance of the current role of an individual to the individual’s identity. This micro definition of centrality is not new as Paullay, Allinger and Stone- Romero (1994) stated that job involvement should be defined in relation to the current job of the individual, and not confounded with general work involvement. Similarly Mossholder, Settoon and Henagan (2005:609) defined a micro centrality, network centrality as “… relative numbers of direct and indirect links an individual has with others comprising a social network”.

In spite of the role of individual difference factors being acknowledged a considerable time ago by March and Simon (1958) this notion has been consistently overlooked. Only recently Trevor (2001) redefined these components to include economic and individual factors, and found support for this redefinition. The need to study more individual factors, particularly those unique to the individual in employee turnover research is further highlighted by the assertion made by Zimmerman (2008:310) that “…researchers need to consider whether some individuals have a propensity to quit regardless of having a work environment designed to increase their job satisfaction, whereas other employees may be more likely to stay even under less than ideal circumstance.”. Zimmerman (2008) recognised that something different from the entire work environment must be operating in the employee turnover pathways. This assertion agrees with the thinking in this study (reported in this paper) that RC is likely to be an important individual factor, which may play a major role in pathways to employee turnover. Zimmerman (2008) actually tested and formulated support for a turnover model that included individual factors and JS, as direct and indirect antecedents of turnover, but corroboration will be addressed in future studies.

The first aim of this study is to add to the literature by testing the effect of the interaction of RC, an individual factor, and JS in a turnover model. The second aim of this study is to test the direct and interactive role of life satisfaction (LS), a form of well being, in the turnover model. The outcome of this study will highlight the role of individual factors, and provide an alternative pathway in the turnover model.

The manuscript is organised in five parts. The next section describes the theory, explanation of the variables, and their relationships, leading to the various hypotheses that were tested. This is followed by the methodology, which describes the participants in the study, the measures used, and the statistical methods applied. The concluding sections contain the results obtained, discussion of the findings of this study, and conclusion drawn from the results in terms of implication and consequences for human resource management policies and practices in contemporary organisations.

Theory and Hypotheses

Predicting Turnover Intention: JS and LS

In practice JS is evaluated by an accumulation of the facets of the job or as a global affective response termed general satisfaction. Irrespective of how the concept has been assessed the construct of JS has often been found to be negatively related to turnover (Griffeth, Hom & Gaertner 2000, Wright & Bonett 2007). Other lines of enquiry and substantive theories of work behaviour have provided convincing theoretical explanations for employee - organisation linkages. For instance, the balance theory (Heider 1958), states that individuals tend to choose balance state in their behaviour in social settings by avoiding conflicting attitudes toward the same object. And by the theory of cognitive dissonance (Festinger & Carlsmith 1959), if an individual’s attitude and behaviour to an object are inconsistent, and the person experiences distasteful pressure, an attempt is likely to be made to reduce this pressure by making adjustment to achieve more acceptable consistency. For example, an employee who has high JS has positive evaluation of the job, but if such an employee intents to quit, the person’s attitude and behaviour about the job are inconsistent, and would create an imbalanced state. Since individuals prefer balance to imbalance in their social behaviour, and this situation is resolved by changing either the attitude or the behaviour (e.g., alter the state of the employee - organisation linkage). Consequently, it is likely that the relationship between JS and turnover intention is negative in terms of cognitive dissonant theory. Moreover, LS is a measure of subjective well being, and is the evaluation of the quality of the various aspects of life in which the individual is involved. Wright and Bonett (1992, 2007) found that the perception of well being and JS affect turnover independently. Hence, it is expected that LS, also a measure of subjective well being, may also be negatively related to turnover intention. These theoretical and empirical imperatives provide foundation for hypotheses one and two.

H1: JS is negatively related to turnover intention.

H2: LS is negatively related to turnover intention.

Interaction of LS and JS to Predict Turnover Intention

According to Wright and Bonett (2007), well being is a primary resource, which results from a combination of secondary resources. Thus, LS, being a measure of subjective well being results from the combination of the secondary resources, JS and family satisfaction (Frone, Russell & Cooper 1992, Akerele, Osamwonyi & Amah 2007). These secondary resources (JS and family satisfaction) are held in bundles or what (Hobfoll 1989) referred to as caravans. Individuals do not assess the optimum level of each resource in a bundle, which will be independent of the levels of the other resources that will be included in the same bundle. Thus, when making decisions involving the level of each secondary resource in any bundle, there is a holistic view of the entire range of resources in the bundle (both available and potential levels).

This decision process can be understood by considering the consumer behaviour analyses in the theory of indifference curves (Koutsoyiannis 1979). An indifference curve shows the combination of goods that yield the same level of satisfaction, and when indifference curves are shown in a graph, they form an indifference map. Individuals will always prefer the bundle of goods that provide higher satisfaction, and are on a higher indifference curve in the individual’s indifference map. From the theory of the indifference curve, an individual can accept a lower level of a good in the bundle, if a higher level will negatively affect the level of satisfaction derived from the other goods in the same bundle. This decision becomes very critical if this higher level results in the individual moving to a lower indifference curve in the indifference map. Similarly, LS can be compared to the level of satisfaction in the indifference curve, while job and family satisfaction are in the bundle, and their combination will result in a level of LS. It is likely that an employee will remain in a job with low JS, if one with higher satisfaction will negatively affect family satisfaction, and cause lower overall LS. For example, an employee in a current job with low satisfaction, and low commuting, may not exit the job, even when offered a new job with higher JS, if the new job involves greater commuting that will negatively affect family satisfaction, and thus, put the individual on a low overall life satisfaction. Hence, the individual will prefer to remain in the current job to optimise LS, and this relationship is expressed in hypotheses three.

H3: LS will interact with JS such that the relationship between JS and turnover intention will be weaker for high life satisfaction participants.

RC as Moderator of the Relationship Between JS and Turnover Intention

The work of Kanungo (1982) and Paullay, et al. (1994) established the distinction between job and WC. The former is a micro definition involving the current job, while the latter is a general measure relating to an entire work system, and reflects the centrality of participating in the work domain to individual’s self identity. When comparing individual differences on the centrality of the work and family domains, as was done by Carr, et al. (2008), the WC is the construct to be used. The concept of RC, like the job involvement concept, adopts a micro view of the definition of centrality by considering the importance of an individual’s present job in the definition of self. Thus, RC captures an individual’s perception that the current job defines his/her self worth and self esteem. Like identity RC is an individual value measure (Rothbard & Edwards 2003), the state is enduring, it is not situational, and is resistant to change.

There is no consensus yet as to the definition of value. For the purpose of this study, value is defined from the work of Kluckhohn (1951:395-398) as “…a conception, explicit or implicit, distinctive of an individual or characteristic of a group, of the desirable which influences the selection from available modes, means and ends of action.”. Thus, value can be an individual or a social phenomenon, and is accepted as a component of super ego, which when violated is likely to result in feelings of guilt, shame, and ego deflation. Contribution to the framework has been given by Smith (1963), who stated that in order to maintain self esteem, individuals become thoroughly practiced in evoking values to justify their actions and behaviours. In summary, RC (as an individual value), is unique, it varies in importance, and is important in determining the attitudes and behaviour of individuals in the work place (Schwartz 1994).

Turnover decision, like any other decision in the social sciences, involves cost benefit analyses. Some employees may consider leaving their job very costly, due to their great commitment to the workplace, or when they define their role as central to the definition of their personality and self identity (Kluckhohn 1951, Smith 1963, Judge 1993, Carr, et al. 2008). Borrowing from the assertion of Carr, et al. (2008) on WC, RC may likely determine how decisions are made and actions that are taken. Based on these assumptions it is likely that even in situations of low JS, employees may perceive that the cost incurred when they leave a particular job may be higher than the benefit they will get by leaving, because they consider the current work role as central to their self identity. Thus, the negative relationship between JS and turnover may be weaker. Individuals’ work family conflict levels have been found to depend on their WC level, and thus, WC moderated the relationship between work family conflict and work attitude (Carr, et al. 2008).

The findings in the Rothbard and Edwards (2003) study of the identity, and utilitarian motives of individual time involvement in roles has potential to shed more light on the interactive effect of RC. These two authors discovered that the identity motive operates on the premise that people invest more in a role with which they identify as it enhances their self esteem and self actualisation. But whether people can maintain their organisational linkage, and invest more in a role when they have low JS was somewhat answered by the finding of Rothbard and Edwards (2003) that people can accommodate the cost of displeasure and not withdraw, if they perceive that the expected long term benefit is more than the cost. These findings indicate that the centrality of a role to an individual may likely moderate their JS-turnover relationship. Hence, hypothesis four is tendered.

H4: RC will moderate the relationship between JS and turnover intention such that for employees who see their role as central to the definition of their identity, the relationship between JS and turnover intention will be weak.


Site and Participants

The site of the survey is a new generation bank with its head office in Lagos, Nigeria, and branches through out the states in the country. New generation banks unlike their old generation counterparts are highly automated, adopt very aggressive marketing strategies, and employ individuals with varying academic backgrounds. In order to retain their employees, the new generation banks also have very generous salary structures, and grant autonomy to their employees. The bank has a total staff strength of over 5,000. However, the participants for this survey were drawn from the 1,200 staff in the head office of the company located in Lagos. The participants average age was 38 years (s.d. = 0.9); the average tenure was 10 years (s.d. = 1.08); a total of 60 per cent were men; 45 per cent are married, and 65 per cent are either senior or junior employees.


This survey utilised cross sectional data acquired through self report instruments. The Human Resources Manager of the bank granted permission to the researcher to send questionnaires to employees of the bank. A contact person was appointed to work with the researcher. A total of 800 questionnaires were distributed to a random sample of the employees selected from the 1,200 employees in the bank’s head office by using the phone list of the company. Every member of the organisation is listed in the phone list, and thus, has equal chance of being selected. Some 400 usable questionnaires were returned (50 per cent return rate). Participants were assured of the confidentiality of information provided.


The questionnaire had two parts. The first part contains five questions that captured the demographic data. These data were job status, gender, age, tenure and marital status. The second part of the questionnaire contains the study variables 1. Job satisfaction, 2. Life satisfaction, 3. Role centrality, and 4. Turnover intention. This part contains 17 items that collectively provided responses for the four study variables. All the study variables, excluding the demographic data, were measured with six point Likert scales that ranged from strongly disagree to strongly agree.


Most studies on voluntary turnover control for some demographic variables (Carr, et al. 2008, Wright & Bonett 2007). The following were controlled in this study: Job status was measured as (1. Junior, 2. Senior, 3. Supervisor, and 4. Manager); Gender as (1. Male and 2. Female); Age as (1. under 30 years, 2. 31-40 years, 3. 41-50 years, 4. 51-60 years, and 6. above 60 years); Tenure measures as (1. less than 5 years, 2. 5-10 years, 3. 11-15 years, 4. 16-20 years, and 5. above 20 years); and Marital status as (1. Married, 2. Single, and 3. Others).

Job Satisfaction

Job satisfaction can be measured as a summation of the perception of satisfaction individuals derive from the facets of a job, or as an overall evaluation of the entire job. The latter definition is adopted in this study. Thus, job satisfaction is defined as an affective evaluation of the overall job situation without reference to the various facets of the job. The scale contains five items taken from the work of Anderson, Coffer and Byerly (2002). The Cronbach (1951) alpha obtained in the Anderson, et al. (2002) study was 0.74, while the value for the current study is 0.73

Life Satisfaction

The scale used in this study contains four items taken from the work of Diener, Emmons, Larsen and Griffins (1985). The scale measures an individual’s perception of the quality of his/her overall life, without specific reference to the aspects of life contributing to the evaluation of the overall LS. The Cronbach alpha obtained by Diener, et al. (1985) was 0.70, while the value in the current study is 0.76.

Role Centrality

This is taken from the five item WC scale used by Carr, et al. (2008). The WC scale measured the relative importance of work and family to the definition of an individual’s self identity. The measure was rephrased to the perception of the importance of the individual’s current role to the definition of the individual’s identify. This was done to conform to the concept of role identify used in the current study. Since family variables were not involved the aspect of the items relating to relative centrality between work and family was removed. The Cronbach alpha obtained by Carr, et al. (2008) was 0.93, while the value in the current study is 0.79.

Turnover Intention

The three item scale from the work of Aryee, Budhwar and Chen (2002) was used. This scale measures the intention of the individuals to voluntarily quit their current job. The Cronbach alpha obtained by Aryee, et al. (2002) was 0.79, while the value in the current study is 0.77.


Principal component analyses with varimax rotation were performed to ensure that the RC items located on their factor, and also that a four factor structure adequately represented the study variables. Since a four factor structure was established by past studies, the analyses stipulated that four factors should be extracted. The hypotheses involving direct relationships, and the interaction effects were tested with hierarchical multiple regression analyses. The control variables were entered in step 1, the main effects of the study variables in step 2, the interaction variables in step 3, while the interaction terms were entered in step 4. The interpretation of the regression analyses followed the approach by Cohen and Cohen (1983), in which the coefficients for the demographic variables were obtained in step 1, those of the study variables in step 2, those for the interaction variables in step 3, and those of the interaction terms in step 4. Also, following the recommendation of Aiken and West (1991), the variables in the interaction term were centred prior to calculating the terms to avoid multicollinearity.

To assess the form of the significant interaction terms, the method recommended by Aiken and West (1991) was followed. To test the form of LS*JS, the participants were divided into two groups. The first group contains participants with values of LS less than the mean LS minus 1 standard deviation, while the second group has LS values more than the mean LS plus 1 standard deviation. The testing of RC*JS terms followed similar procedure, but the categorisation is based on values of RC. Thereafter, regression analyses were carried for the various groups using JS as the independent variable and turnover intention as the dependent variable.


Preliminary Analyses

Eigenvalues and screenplot justified the stipulated four factor model. The four factors extracted 68 per cent of the variance in the items. The factor loadings for the RC scale are between 0.65 and 0.80. An unrotated factor analyses was also conducted. The first factor extracted only 16 per cent of the variance, compared to 52 per cent extracted by the other three factors, which is an indication that common method variance is not an issue (Podsakoff & Organ 1986). The square of the correlation between any two factors is lower that the variance extracted by each factor. This result shows that the factors discriminated very well (Koufteros, Vonderembse & Doll 2002).

The Cronbach alpha for the study variables are above the 0.7 limit recommended by Cronbach (1951). Correlation analyses in Table 1 indicate that job satisfaction is negatively correlated with turnover intention (-.17, p < 0.01), life satisfaction is negatively correlated with turnover intention (-.11, p < 0.05), and role centrality is also negatively correlated with turnover (-.19, p < 0.01). These relationships support the hypotheses 1 and 2.

Table 1
Descriptive statistics and zero-order correlation
Variables Mean SD C V% 1 2 3 4 5 6 7 8
1. Job satisfaction 4.46 1.06 0.73 63
2. Role centrality 2.90 1.08 0.79 54 .04
3. Life satisfaction 3.24 1.04 0.76 64 .20** .27**
4. Turnover intention 3.37 1.47 0.77 66 -.17** -.19** -.11*
5. Job status -.10* .02 .05 -.13**
6. Gender -.25** -.15** .08 -.01 .12*
7. Age .10* -.03 .03 -.14** .25** .01
8. Organisational tenure -.07 -.01 -.06 -.18** .28** .12* .50**
9. Marital status -.07 -.06 .01 .10* -.19** .03 .09 -.13*

Notes: a. N = 400.
b. C = Cronbach Alpha, and V% = Variance explained per cent.
c. SD = Standard deviation, *p < 0.05 and ** p < 0.01.

Hierarchical Multiple Regression

After controlling for the control variables, job satisfaction was negatively related to turnover intention (-.11, p < 0.05), showing support for hypothesis 1. Life satisfaction is non significantly related to turnover intention (-.08, p > 0.05), indicating a lack of support for hypothesis 2. Two of the interaction terms namely, LS*JS (-.24, p < 0.01), and RC*JS (-.36, p < 0.01) were significant. These findings give support to hypotheses 3 and 4.

Table 2
Moderated regression analyses for the tests of turnover intention
Variables Step 1 Step 2 Step 3 Step 4
Job status -.08 -.08 -.10 -.09
Gender .04 .02 -.06 -.06
Age -.08 -.08 -.07 .09
Organisational tenure -.13* -.12* -.13* -.11*
Marital status .07 .07 .08 .09
Job satisfaction (JS) -.11* -.09 -.09
Life satisfaction (LS) -.07 -.07
Role centrality (RC) -.18** -.20**
JS*LS -.24**
JS*RC -.36**
ΔR2 .10 .02 .04 .08
R2 .10 .12 .16 .24
F 7.351** 4.322* 9.041** 9.845**

Notes: a. N = 400.
b. JS = Job satisfaction, LS = Life satisfaction and RC = Role centrality.
c. *p < 0.05 and ** p < 0.01.

Assessing the Form of Interaction

The results of the regression analyses were plotted in Figures 1 and 2. As shown in Figure 1, at low level of JS, participants with high RC have a lower departure rate than those with low RC. The slope of the JS turnover relationship for low RC participants is positive and significant (.11, p < 0.05), while that of high RC participants is negative, but non significant (-.047, p > 0.05). These results imply that participants with low RC are likely to quit their job even at high JS, while high RC participants are insensitive to levels of JS. Figure 2 indicates that at low JS, participants with low LS had a higher turnover rate than those with high LS. The slope for low LS participants is positive and significant (0.15, p < 0.05), while that for participants who experienced high LS is negative and significant (-.09, p < 0.05).

Figure 1
Interaction of role centrality and job satisfaction in predicting turnover intention
Interaction of role centrality and job satisfaction in predicting turnover intention

Figure 2
Interaction of life satisfaction and job satisfaction in predicting turnover intention
Interaction of life satisfaction and job satisfaction in predicting turnover intention


There are three key important aspects of this study, namely, to test the direct effect of LS on turnover intention, to test the effect of the interaction of LS and JS in a turnover model, and lastly to test the effect of the interaction of RC and JS in a turnover model. The testing of these interactions was to fill the gap identified by Wright and Bonett (2007), Zimmerman (2008), and Carlson and Kacmar (2000) in the current studies involving the relationship between JS and turnover intention. A major contribution of this study is the testing of the interaction effect of RC instead of the general WC adopted by past studies.

The results of the regression analyses indicate that JS has negative significant relationships with turnover intention, while LS does not. The negative relationship between JS and turnover intention agrees with the balance theory (Heider 1958) that individuals prefer balance to an imbalanced state, and the cognitive dissonance theory of social attitude, that there must be consistency between an individual’s attitude and behaviour towards an object, so as to eliminate pressure associated with inconsistency. The non significant relationship between LS and turnover is contrary to the results obtained by Wright and Bonett (2007), that psychological well being is negatively related to turnover behaviour. The additive model (Frone, et al. 1992, Akerele, et al. 2007) posits that family and JS jointly determine an individual’s LS. Consequently, it is possible that employees do not make turnover decisions based on absolute values of LS alone. They may consider the relative contributions of the various domains of life they are involved in, and make decision based on these contributions, and their life preferences. For example, if the major contributor to a perceived LS is in a domain of life not considered by the individual as central to the definition of his/her identity, the level of LS alone may not be critical in making turnover decisions.

It was found LS moderated the relationship between JS and turnover such that individuals with high LS had lower turnover rate at all levels of JS. The Wright and Bonett (2007) results indicated a significant interaction between psychological well being and JS in predicting turnover. Analogous to the consumer behaviour analyses using indifference curve theory, the obtained study result agrees with the fact that individuals view the quality of their lives as contributions from the various domains in which they are involved, and thus, make decisions based on the absolute levels of LS, and satisfactions from these domains. For example, an individual could have a job with high satisfaction, but the satisfaction in other areas of life may be so affected that the overall LS will be low. In such a situation the individual may elect to stay in a low satisfaction job that does not have a substantial negative effect on satisfaction in other areas of life. In this way the individual will operate at a higher level of LS.

A salient observation of the study was RC moderated the relationship between JS and turnover intention. Indeed, turnover level was lower for high RC employees at all levels JS. For low RC individuals, a high JS is insufficient to make the individual stay when the job is perceived not to be central to the definition of individual’s identity. This study result agrees with the postulation of Rothbard and Edwards (2003), that an individual may accommodate some displeasure, and not quit when the individual perceives that the role being played enhances self definition. This study finding also agrees with the observation of Maertz and Campion (1998), that work relationship can be a reason for individuals staying in an organisation even when they do not like some features of the work system. In the alternative, when the role is not central to self definition, the individual may quit no matter the level of pleasure provided.

The relationship between RC and turnover intention was not hypothesised. However, the regression analyses show that RC has significant negative relationship with turnover intention (-.19, p < 0.05). This agrees with the result of Mossholder, et al. (2005:608) that network centrality is negatively related to turnover, and that employees consider their current role important for self definition are more embedded within the organisation and will be “… less susceptible to forces that could dislodge them from their organisation”.

This study is based on cross sectional data acquired through self report instruments. This strategy prevents inference of causality, and also the results obtained may contain some level of common method variance. However, the fact that the first factor in an unrotated factor analyses extracted only 16 per cent variance compared to 52 per cent extracted by the other three factors, is indication that common method variance may not be severe (Podsakoff & Organ 1986). Generalisation can be enhanced in future studies by testing the micro definition of centrality in other contexts.


A major contribution of this study is the adoption of a micro definition of centrality, which considers the centrality of the individual’s role in self definition. The study argues that the micro level definition is well suited when within domain study is involved. This approach followed the work of Paullay, et al. (1994), who disaggregated the macro definition of Kanungo (1982) into general WC and micro job involvement. Job involvement like RC is always used in within work studies, and provides better result than the macro definition confounded with WC (Paullay, et al. 1994).

The current study has made contribution in the turnover model, by identifying three additional pathways to employee turnover, namely, through the interactive effects of LS and RC and the direct effect of RC. The significant interactive roles of LS and RC have implications for organisational participants. Firstly, it implies that managers must be aware of the level of LS; the contributions made by the various domains to this level of LS, and individual preference as to the importance of each domain in defining self identity. Thus, it is necessary to consider the centrality of the job role of an individual in self definition, while analysing the turnover pathway involving employee work attitude. The fact that the employee is generally satisfied with the overall job system does not guarantee low turnover, if the employee does not see the current role as central to self definition. Consequently, organisations are encouraged to design work environments that enhance employee JS, and also ensure that the employees are happy with the jobs assigned them, and see such jobs as part of their self definition. When they perceive leaving the organisation will have high cost to them, and thus, discourage turnover. Secondly, personality tests are frequently used in the selection processes by organisations. Organisations should go further by identifying the role preferences of employees also, and by appropriate use of high performance human resources policies, they can reinforce the individual’s perception of the centrality of the role assigned them, thereby managing turnover. Life satisfaction is affected by job and family satisfaction (Akerele, et al. 2007). Thus, managing life satisfaction involves robust management of job and family satisfaction, which attracts attention for the establishment of human resources policies that will help employees balance their work and family involvements.


Okechukwu Amah is an employee of Chevron Nigeria Limited and has worked for the organisation for 25 years. He obtained his PhD from the University of Benin, Benin City in Nigeria. He is a part time lecturer in Business Administration, at Lagos State University in Lagos, Nigeria. His research interests include work family conflict, organisational behaviour and training.

Email :


Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting interaction. Newbury Park, CA: Sage.

Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and predicting social behaviour. NJ: Prentice-Hall.

Akerele, A., Osamwonyi, I., & Amah, O. E. (2007). Work-family conflict model: Application of Aryee’s model. African Journal of the Psychological Study of Social Issues, 10(1), 3-16

Anderson, S. E., Coffer, B. S., & Byerly, R. T. (2002). Formal organisational initiatives and informal workplace practices: Links to work-family conflict and job related outcomes. Journal of Management, 28(6), 787-810.

Aryee, S., Budhwar, P. S., & Chen, Z. X. (2002). Trust as a mediator of the relationship between organizational justice and work outcomes: Test of social exchange model. Journal of Organizational Behavior, 23(3), 267-285.

Baron, J. N., Hannan, M. T., & Burton, M. D. (2001). Labour pains: Change in organizational models and employee turnover in your high-tech firms. American Journal of Sociology, 106(4), 960-1012.

Carlson, D. S., & Kacmar, K. M. (2000). Work-family conflict in the organization: Do life role values make a difference? Journal of Management, 26(5), 1031-1054.

Carr, J. C., Boyar, S. L., & Gregory, B. T. (2008). The moderating effect of work family centrality on work-family conflict, organisational attitudes, and turnover behaviour. Journal of Management, 34(2), 244-262.

Cohen, J., & P. Cohen. (1983). Applied multiple regression/correlation analysis for the behavioural sciences. (2nd ed.). Hilsdale, NJ: Erlbaum.

Cronbach, L. (1951). Coefficient of alpha and internal structure tests. Psychometrica, XVI, 297-334.

Diener, E., Emmons, R. L., Larsen, R. J., & Griffin, S. (1985). The satisfaction with life scale. Journal of Personality Assessment, 49(1), 71-75.

Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. Journal of Abnormal and Social Psychology, 58(1), 203-210.

Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behaviour: An introduction to theory and research. Reading, MA: Addison-Wesley.

Frone, M. R., Russell, M., & Cooper, M. L. (1992). Antecedents and outcomes of work-family conflict: Testing a model of the work-family interface. Journal of Applied Psychology, 77(1), 65-78.

Griffeth, R. W., Hom, P. W., & Gaertner, S. (2000). A meta-analysis of antecedents and correlates of employee turnover: Update, moderator tests, and research implications for the millennium. Journal of Management, 26(3), 463-488.

Heider, F. (1958). The psychology of interpersonal relations. New York: Wiley

Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist, 44(3), 513-528.

Hom, P. W., & Griffeth, R. W. (1995). Employee turnover. Cincinnati, OH: South-Western.

Huselid, M. A. (1995). The impact of human resources management practices on turnover, productivity and corporate financial performance. Academy of Management Journal, 38(3), 635-672.

Johns, G. (2002). The psychology of lateness, absenteeism, and turnover. In N. Anderson, D. Ones, H. K. Sinangil, & C. Viswesvaran (Eds.), Handbook of industrial, work and organizational psychology (Vol. 2, 232-252). Thousand Oaks, CA: Sage.

Judge, T. A. (1993). Does affective disposition moderate the relationship between job satisfaction and voluntary turnover? Journal of Applied Psychology, 78(3), 395-401.

Kanungo, R. N. (1982). Measurement of Job and Work Involvement. Journal of Applied Psychology, 67(3), 341-349.

Kluckhohn, C. K. (1951). Values and value orientations in the theory of action. In T. Parson & E. Shils (Eds.), Toward a general Theory of action (388-433). Cambridge: Harvard University Press.

Koufteros, X. A., Vonderembse, M. A., & W. J. Doll. (2002). Examining the competitive capabilities of manufacturing firms. Structural Equation Modeling, 9(2), 256-282.

Koutsoyiannis, A. (1979). Modern microeconomics. (2 ed.). Hong Kong: Macmillan

Lee, T. W., & Mowday, R. T. (1987). Voluntarily leaving an organisation: An empirical investigation of Steers and Mowday’s model of turnover. Academy of Management Journal, 31(5), 721-743.

Maertz, C. P., & Campion, M. A. (1998). 25 years of voluntary turnover research: A review and critique. In C. L. Cooper, & I. T. Robertson (Eds.), International review of industrial and organisational psychology (Vol. 13, 49-81). Chichester, England: Wiley.

March, J. G., & Simon, H. A. (1958). Organisations. New York: Wiley.

Mossholder, K. W., Settoon, R. P., & Henagan, S. C. (2005). The relational perspective on turnover: Examining structural, attitudinal and behavioral predictors. Academy of Management Journal, 48(4), 607-618.

Podsakoff, P. M., & Organ, D. W. (1986). Self-report in organizational research: Problems and prospects. Journal of Management, 12(4), 545-560.

Paullay, I. M., Allinger, G. M., & Stone-Romero, E. F. (1994). Construct validation of two instruments designed to measure job involvement and work centrality. Journal of Applied Psychology, 79(2), 224-228.

Rothbard, N. P., & Edwards, J. R. (2003). Investment in work and family roles: A test of identity and utilitarian motives. Personnel Psychology, 56(3), 699-730.

Schwartz, S. H. (1994). Are there universal aspects in the structure and contents of human values? Journal of Social Issues, 50(1), 19-45.

Smith, M. B. (1963). Personal values in the study of lives. In R. W. White (Ed.), The study of lives (324-347). New York: Atherton Press.

Sweeney, P. D., & McFarlin, D. B. (2002). Organizational behaviour: Solution for management. Irwin: McGraw-Hill.

Trevor, C. O. (2001). Interactions among actual ease-of-movement determinants and job satisfaction in the prediction of voluntary turnover. Academy of Management Journal, 44(4), 621-638.

Wright, T. A., & Bonett, D. G. (1992). The effect of turnover on work satisfaction and mental health: Support for a situational perspective. Journal of Organisational Behaviour, 13(3), 603-615.

Wright, T. A., & Bonett, D. G. (2007). Job satisfaction and psychological well-being as nonadditive predictors of workplace turnover. Journal of Management, 33(2), 141-160.

Zimmerman, R. D. (2008). Understanding the impact of personality traits on individual’s turnover decisions: A meta-analytical path model. Personnel Psychology, 61(2), 309-348.