1. Introduction

For any organization, securing and retaining their competitive advantage over other organizations are crucial for the success of the organization. One of the significant sources of such competitive advantage is the human capital (Bowen & Ostroff, 2004; Cunningham & Sagas, 2004). The quality of human capital is even more crucial in service organizations since it is closely related with customer satisfaction (Reichheld & Sasser, 1990). As Lings (2004) noted, “the attitudes and behaviors of customer contact employees influence customers’ perceptions of the service they receive” (p.405). In sum employees who are in contact with the customers play a key role in recruiting and retaining the customers and, thus, in the success of the organization.

The significance of the employee-customer interface was highlighted by Kotler (1994) who conceived of three types of marketing. First is the conventional marketing labeled external marketing whereby the organization tries to sell its products to external customers. The second type of marketing labeled interactional marketing is most germane to the present context as it focuses on service operations. The service employee is involved not only in the production of theservice in question but also in the marketing of that service through interpersonal interactions with the customers. This dual role of producer and marketer makes the service employee a critical resource that can create competitive advantage for an organization. Finally, internal marketing refers to the organization’s efforts in “attracting, developing, motivating, and retaining qualified employees through job-products that satisfy their needs” (Berry & Parasuraman, 1991; p. 151). A major thrust of internal marketing is creating a climate wherein the employees feel that the organization is concerned about their well being by facilitating their work performance as well as making their experiences pleasant and satisfying. When employees have a sense of well-being in the workplace, they are likely to be committed to the organization and would be willing to continue to work for the organization.

2. Review of Literature

The factors that contribute to employees’ positive attitudes toward the organization include the structural and process characteristics, job design, performance evaluation, and reward system (Chelladurai, 2006). An equally important factor and focal to the present study is the task and social support available in the work environment. In this study, we investigate the relationships among perceived support available at work, organizational commitment, motivation, work effort, and intention to leave among fitness club employees in the United States. In particular, environmental support which includes coworker support, supervisor support, and organizational support are examined as indicators of the latent variable of perceived support which is an antecedent of organizational commitment which, in turn, influences intention to leave and work effort. The moderating role of work motivation in the relationship between perceived support and organizational commitment is also examined. The proposed model is illustrated in Figure. 1. The following sections explicate the variables of the study and the relationships among them.

2.1. Organizational Commitment

Organizational commitment refers to the strength of personal attachment toward the organization (Arnold, Cooper, & Robertson, 1998). Recent approaches to the study of organizational commitment consider it to be composed of three dimensions—affective commitment, continuance commitment, and normative commitment (Meyer & Allen, 1991). Meyer and Allen (1991) define affective commitment as, “the employee’s emotional attachment to, identification with, and involvement in the organization. Employees with a strong affective commitment continue employment in the organization because they want to do so” (p.67). An employee is said to have continuance commitment to his/her organization when he/she stays at the organization due to the costs involved by leaving the organization (Burton, Lee, & Holtom, 2002; Meyer & Allen, 1991). In this case, employees continue their employment in the organization because they need to. Meyer and Allen (1991) define normative commitment as “a feeling of obligation to continue employment. Employees with a high level of normative commitment feel that they ought to remain with the organization” (p.67). These three dimensions of affective, continuance, and normative commitment will be included in the present study since these three forms of organizational commitment has been found to be associated with employees’ work behaviors (e.g. Chang & Chelladurai, 2003; Turner & Chelladurai, 2005).

2.2. Perceived Support as Antecedent of Organizational Commitment

One of the factors that help cultivate organizational commitment is employees’ perception of support available at work. That is, when employees feel they are supported by coworkers, supervisors, and/or organization, they are likely to develop a sense of commitment. Therefore, the types of support we included in this study are coworker support, supervisor support, and organizational support. Coworker support and supervisor support refer to emotional, instrumental, and/or informational support that comes from coworkers and supervisors respectively (Greenglass, Burke, & Konarski, 1997). Organizational support refers to individuals’ perceptions about how much the organization values the employees’ contributions and care about their well-being (Eisenberger, Huntington, Hutchison, & Sowa, 1986).

The relationship between perceived support at work and employee commitment is explained by social exchange theory (Thibaut & Kelley, 1967; Setton, Bennett, & Liden, 1996). According to Blau (1964), social exchange is based on a quid pro quo, which means ‘this for that.’ Thus, it is more likely that a person will provide support to the others (i.e., the supervisor and coworkers in our context) who support the person (Bowling, Beehr, Johnson, Semmer, Hendricks, & Webster, 2004). In other words, if employees feel their supervisor is providing support for them, they would work for the supervisor’s benefit, and this way, the supervisor would provide more support to the employees. Similarly, employees form their perceptions on how much the organization values their contributions and care about their well-being and respond to it with job attitudes and/or organizational behaviors that facilitate organizational success (Eisenberger, Armeli, Rexwinkel, Lynch, & Rhoads, 2001; Eisenberger et al., 1986).

Figure 1. Hypothesized model of the relationship among perceived support, organizational commitment, motivation, and work outcomes

From another perspective, social identity theory (Tajfel & Turner, 1979; Turner, 1982) also helps to explain the relationship between organizational support and organizational commitment (Fuller, Barnett, Hester, & Relyea, 2003). Social identity theory proposes that employees remain loyal when they feel their organization values and appreciate them (Tyler, 1999). According to Fuller et al. (2003), “when people feel that their organization values and appreciates them, it is a sign of organizational respect for them or of their high status within the organization” (p.789). This feeling of high status increases employees’ social identity which, in turn, strengthens their commitment to the organization (Tyler, 1999).

Previous studies have shown the relationship between perceived support and organizational commitment (e.g. Eisenberger et al., 2001; Pack, Jordan, Turner, & Haines, 2007). For example, Nelson and Quick (1991) found that supervisor support and coworker support had a positive impact on organizational commitment among newcomers to the organization while it had a negative impact on their intent to leave the organization. Bartlett’s (2001) study showed that employees were more affectively committed to the organization when they received supervisor and coworker support. In Ko et al.’s (1997) study, perceived supervisor support and coworker support were antecedents of affective commitment while supervisory support was highly related with continuance commitment. Eisenberger et al. (2001) found that organizational support increased the employees’ feeling for and caring about the organization’s welfare and goals, and this feeling led to the increase of their commitment to the organization. In a similar vein, Eisenberger et al. (1986) and Eisenberger, Fasolo, and Davis-LaMastro (1990) showed that employees’ emotional attachment towards the organization is affected by the employees’ perception that the organization cares for the employees’ welfare. In sport setting, Pack et al. (2007) showed that organizational support explained 46.2% of the variance in affective commitment and 39% of the variance in normative commitment among student employees in a recreational sport department in a large university.

Based on previous research, this study investigates the influence of perceived support from coworkers, supervisors, and organization on the employees’ level of commitment affective commitment, continuance commitment, and normative commitment in fitness industry. Accordingly, we propose the following hypothesis.

Hypothesis 1. Perceived support available at work will be positively related to employee commitment to the organization.

In testing the above hypothesis, coworker support, supervisor support, and organizational support will indicate the second order latent variable of perceived support, and affective commitment, continuance commitment, and normative commitment will indicate the second order latent variable of organizational commitment.

2.3. Consequences of Organizational Commitment

Traditionally, intention to leave the organization has been considered as an important outcome of organizational commitment. More recently, the importance of other work related behaviors influenced by organizational commitment has been highlighted (Chang & Chelladurai, 2003; Turner & Chelladurai, 2005; Wasti, 2005). More specifically, organizational commitment has been shown to be positively related to attendance, work effort, job involvement, job satisfaction and employee retention (Randall, 1990).

In the current study, however, work effort and intention to leave are included as outcome variables of affective commitment due to their importance in the performance of service organizations. As the employees directly interact with customers in performance of the service, the amount and nature of the effort employees put into their work would influence the quality of the service provided. And the quality of the service is a strong predictor of satisfaction with service and repeat purchase of that service. In addition, predicting employees’ intention to leave and reducing it is crucial because losing valuable employees adversely affects an organization’s immediate performance. Further, replacing current employees and training new employees are costly. Thus, understanding the factors that are closely related to employees’ intention to leave the organization is especially important in fitness organizations where the turnover rate is known to be relatively high (Pack et al., 2007).

The relationships among perceived support, organizational commitment, work effort, and intention to leave are explained by social exchange theory. As discussed earlier, social exchange theory emphasizes reciprocity (Blau, 1964). Therefore, when employees perceive that they are receiving support, they would build commitment and the commitment will show as positive work behaviors. Shanock and Eisenberger (2006) supports this notion by stating “workers trade effort and dedication to their organizations for such tangible incentives as pay and fringe benefits and such socioemotional benefits as esteem, approval, and caring” (p.689).

Previous studies have shown how one’s commitment to the organization translated into positive behaviors. For example, Turner and Chelladurai (2005) found that normative and affective commitments were negatively related to the intercollegiate coaches’ intent to leave the organization normative commitment having a stronger relationship while continuance commitment was unrelated. Similarly, Turner, Jordan, and DuBord (2005) examined the impact of four forms of organizational commitment (affective, continuance high sacrifices, continuance low alternatives, and normative commitment) on desire to stay in the organization among student employees in a university recreational sport department and found that all four forms of commitment were significantly related with desire to stay explaining 13.4% of the variance.

Regarding work effort, Chelte and Tausky (1986) found that organizational commitment was significantly and positively related with employees’ perceived work effort although the level of significance differed in three different groups of university employees. Lee and Gao’s (2005) study on Korean retail business showed that affective commitment increased employees’ work effort whereas it decreased their intention to leave the organization. Sager and Johnston (1989) also found that sales employees with higher organizational commitment perceived themselves as working harder and their job searching behavior decreased dramatically. Based on such literature, we propose the following hypotheses.

Hypothesis 2. Organizational commitment will be positively related to employee’s work effort.

Hypothesis 3. Organizational commitment will be negatively related with employees’ intention to leave.

2.4. Motivation as a Moderator of Support-Commitment Relationship

Motivation is defined as “the willingness to exert a high level of effort towards organizational goals, conditioned by the employee’s ability to satisfy some individual need” (Rai, 2004; p.43). Motivation is divided into two categories—intrinsic motivation and extrinsic motivation. Basically, some people are motivated because they find the work itself interesting (intrinsic motivation), and not because of the consequences it would bring to the individual. On the other hand, some people are motivated by the extrinsic consequences such as money and promotion (extrinsic motivation) (Gagne & Deci, 2005). In other words, intrinsic motivation includes feelings of interest in and enjoyment of the job while extrinsic motivation includes perceptions that their work will be rewarded with something that they value (Pinder, 1997).

Although there have not been many studies that examined the influence of motivation on organizational commitment, the possibility has been discussed by many researchers (e.g. Bartlett, 2001; Buchanan, 1974; Eby, Freeman, Rush, & Lance, 1999; Kuvaas, 2003; Steers, 1977; Whitener & Walz, 1993). Those studies that empirically examined the relationship have supported a positive relationship between work motivation and organizational commitment. For example, Eby et al. (1999) indicated in their model that intrinsic motivation influences affective commitment, which has an impact on outcome variables. Their study showed that intrinsic motivation explained 23% of the variance in affective commitment. Similarly, Bartlett (2001) found that intrinsic motivation was positively related to affective and normative commitment while it was slightly negatively related to continuance commitment. In addition, Kuvaas (2003) found that both intrinsic motivation (sense of ownership) and extrinsic motivation (company shares) led to an increase of the employees’ affective commitment.

While many studies have investigated the relationship between perceived support and organizational commitment and the relationship between work motivation and organizational commitment, the study of the moderating role of motivation in the relationship between perceived support and organizational commitment has largely been ignored. Given the relationship between support and organizational commitment and the relationship between motivation and organizational commitment, the authors propose that employees’ intrinsic and extrinsic motivation will moderate the relationship between support and organizational commitment. That is, if someone has high level of work motivation, the effect of perceived support on organizational commitment may not be significant since his/her organizational commitment can solely stem from his/her high work motivation. On the other hand, for an employee with low work motivation, an effect of perceived support will be stronger on his/her level of organizational commitment.

As there is not enough literature on the moderating role of motivation in the relationship between perceived support and organizational commitment, we present the following non- directional hypotheses.

Hypothesis 4. Intrinsic motivation will moderate the relationship between perceived support and organizational commitment.

Hypothesis 5. Extrinsic motivation will moderate the relationship between perceived support and organizational commitment.

We tested the above hypotheses in the context of fitness club employees due to the importance and the size of the fitness industry. Fitness clubs are among the largest and fastest growing businesses in the sport industry. According to the International Health, Racquet and Sportsclub Association (IHRSA) (2006), there has been a steady growth in the number of fitness clubs all around the world. In the United States, there were 29,069 clubs with 41.3 million members in 2005, an increase of 14% from the previous year (IHRSA, 2006). In addition, the total revenue for the fitness club industry was estimated at $15.9 billion in 2005, and it increased to $17.6 billion in 2006 (IHRSA, 2007). According to the report of U.S. Department of Labor (2007), fitness clubs employed approximately 205,000 workers in 2004.

The importance of human capital discussed earlier is magnified in fitness clubs as well because many of the employees in fitness clubs, such as activity instructors and personal trainers, interact directly with the customers over relatively longer periods. Such interactions may be task- related or social in nature (Chang & Chelladurai, 2003). Chang and Chelladurai (2003) found that these two forms of interactions were the most important determinants of customer perception of service quality. Along similar lines, Chelladurai, Scott and Haywood-Farmer (1987) found that employee’s instruction behavior was a very critical factor in the choice of fitness clubs.

In summary, the purpose of the study is to investigate the link between perceived support (i.e., from coworkers, supervisor, and organization), organizational commitment (affective, continuance, and normative), work effort, and intention to leave among fitness club employees. Also, the impact of organizational commitment on employees’ intention to leave and work effort is examined. Finally, the moderating effect of motivation in the relationship between perceived support and organizational commitment is explored.

3. Method

3.1. Participants

A random sample of 2,000 NETA (National Exercise Trainers Association) certified fitness instructors who have fulltime and/or part-time positions in fitness clubs in the United States was selected for this study. These employees were 18 years old or older and they worked for either profit or non-profit fitness clubs across the nation. Among the 2,000 questionnaires distributed to NETA members through email, 5% of them bounced back due to invalid email addresses. Therefore, total sample size was 1,900. Two hundred and sixty six people responded to the questionnaire representing a 14% response rate. However, 64 responses were not usable. Therefore, the remaining 202 responses were used in the analyses. Thirty-six of them were males (17.8%), and 165 of them were females (81.7%). Seventy-nine respondents (39.1%) had full time position at the fitness organization while 121 respondents (59.9%) had part time position. Further, 117 respondents (57.9%) worked for profit organizations, and 85 respondents (42.1%) worked for non-profit organizations. As for age, 43 respondents (21.3%) were between 21 and 30 years of age, 64 respondents (31.7%) were between 31 and 40 years, 50 respondents (26.2%) were between 41 and 50 years, and 42 respondents (20.8%) were over 50 years. The respondents came from different regions in the U.S. Eighty-six respondents (42%) were employed in the Midwest, 46 respondents (22.8%) were from Mid Atlantic, 37 respondents (18.3) were from the South, 15 respondents (7.4%) were from the West, 13 respondents (6.4%) were from the Southwest, and 5 respondents (2.5%) were from New England. The years the employees have been working in the profession ranged from five months to 30 years (M = 8.65) whereas the years the employees have been working in the current organization ranged from five months to 28 years (M = 4.91).

3.2. Data Collection Procedures

The researchers contacted the director of National Exercise Trainers Association (NETA) to recruit participants for the present study. NETA is an organization that offers certification in group exercise, personal training, pilates, and yoga. The program is offered across the country, and NETA has certified over 120,000 fitness professionals. NETA randomly selected 2,000 of their members holding valid certificates. Then, email messages to these randomly selected members requesting them to participate in the study were sent. The message included a link to a website for the online survey. A reminder email about the study and the survey link was sent out two weeks after the initial email. The online survey was so designed that when an employee responds to the questionnaire, the responses would go directly to the researchers and not to NETA. In addition, the responses were anonymous. In other words, there was no way for the researchers to find out who filled out the survey. The researchers used Selectsurvey.net software to distribute the questionnaire and collect the responses.

The survey consisted of three sections. First section included informed consent to participate in the study. The message stated that the participation was voluntary, and they could skip any questions they feel uncomfortable answering. By completing the survey, the participants consented to participate. The second section consisted of demographic information about the participants (i.e., gender, employment status, type of organization they work for, region, the years worked for the organization, and the years worked in the profession). Finally, the third section included 57 items about perceived support available at work and the fitness employees’ work attitudes. The items were jumbled so the participants would not know which domain each item belonged to. The questionnaire comprised of subscales on perceived support, organizational commitment, work effort, intention to leave, and motivation.

3.3. Perceived Support

Perceived support was indicated by three factors: coworker support, supervisor support, and organizational support.

3.3.1. Coworker Support

Coworker support was measured by a six-item scale modified from Ducharme and Martin’s (2000) ten-item scale. The original 10-item scale included affective (α = .85) and instrumental (α = .76) facets of coworker support. For the purposes of the current study, the three items with the highest loading in each facet were selected. The items were reworded to replace the words ‘you’ and ‘your’ to ‘I’ and ‘my’ to be consistent with the items in other subscales.

3.3.2. Supervisor Support

Supervisor support was measured by a six-item scale proposed by Anderson, Coffey, and Byerly (2002). The Cronbach’s alpha for the scale was .89 (Anderson et al., 2002).

3.3.3. Organizational Support

Organizational support was measured by the eight-item scale of perceived organizational support (Eisenberger, Cummings, Armeli & Lynch, 1997), which is a shorter version of the original 36 items suggested by Eisenberger et al. (1986). In order to increase reliability of the scale, two more items that had high factor loadings in the original study (Eisenberger et al., 1986) were added. That is, organizational support was measured using 10 items (five for valuation of employees’ contribution and five for care about employees’ well-being).

3.4. Organizational Commitment

Commitment to the organization was measured by 18- item version of commitment scale suggested by Meyer, Allen and Smith (1993). This version of scale measures affective (6 items), continuance (6 items), and normative commitment (6 items). Construct validity and internal reliability (ranging from .73 to .85) of the scale are well established in previous studies (e.g. Allen & Meyer, 1996; Meyer et al., 1993).

3.5. Intention to Leave

Employees’ intention to leave was measured by two items suggested by Turner and Chelladurai (2005). The Cronbach’s alpha of the scale was .80. Wording of the items was slightly modified from the original version to fit the overall questionnaire. The word ‘university’ is replaced with ‘organization’.

3.6. Work Effort

Work effort was measured by Chang’s three-item (2003) modified version of Brockner, Grover, Reed, and Dewitt’s (1992) scale. Cronbach’s alpha for these items was reported to be .91.

3.7. Motivation

Intrinsic motivation was measured by a six-item scale suggested by Oliver and Anderson (1994). Extrinsic motivation was measured with six items—three from Oliver and Anderson (1994) and three from Teresa, Elizabeth, Karl, and Beth’s (1994). The Cronbach alpha for the intrinsic motivation scale was reported to be 0.82, and the Cronbach alpha for the Oliver and Anderson’s three-item scale of extrinsic motivation was 0.83. The items in Oliver and Anderson’s (1994) scales were reworded to fit into fitness club settings since the original version was developed for the sales profession. The word of ‘sell’ was replaced with ‘work’.

3.8. Data Analysis

We performed a confirmatory factor analysis (CFA) with AMOS 7 to confirm the factor structure of the proposed measurement model and structural equation modeling (SEM) to test the proposed relationships among the constructs. The internal consistency (Cronbach’s alpha) of each subscale was also estimated. The Average Variance Extracted (AVE) values were used to measure construct reliability of the scales. AVE values indicate the amount each item contributes to explaining the specified construct. AVE values higher than .50 are considered to have a good construct reliability (Fornell & Larcker, 1981).

AMOS 7 provides the following measures of fit: Normed Fit Index (NFI), Incremental Fit Index (IFI), Comparative Fit Index (CFI), Root Mean Square Error of Approximation (RMSEA), and chi-square value divided by degrees of freedom. For NFI, IFI, and CFI, values higher than .90 is considered to have a close fit (Hair, Anderson, Tatham & Balck, 1998; Kline, 1998). However, as the goodness of fit indices are likely to be depressed as a function of the complexity of the model, the use of .90 guideline must be used with caution (Cheung & Rensvold, 2002). It has been noted that RMSEA is the only goodness-of-fit index (GFI) that is not influenced by the number of items in each subscale and the number of factors in the measurement model (Browne & Cudeck, 1992; Cheung & Rensvold, 2002). Accordingly, RMSEA value was used to determine a model fit. RMSEA values less than .06 indicates close fit of the model while values less than .08 indicates a reasonable fit (Hu & Bentler, 1999).

To test the moderating effect of motivation on the relationship between perceived support and organizational commitment, regression analyses were used. In these analyses, the composite scores (i.e., the mean of the items in a subscale) were used. According to Baron and Kenny (1986), moderating effect can be detected by examining the relationships between predictor and outcome variable, between moderator and outcome variable, and between the interaction of predictor and moderator on outcome variables.

4. Results

4.1. Confirmatory Factor Analysis

Confirmatory factor analysis (CFA) of the 57 items and 10 factors revealed that some of the items did not load high enough on their factors meaning that these items did not represent the underlying construct well. After thorough examination of these items, the items that did not enhance the psychometric property of the construct were removed. One item from each of affective commitment and organizational support subscales, two items from each coworker support and intrinsic motivation subscales, and three items from each continuance commitment, normative commitment, and extrinsic motivation subscales were deleted. Further, the construct of normative commitment was eliminated due to lack of discriminant validity of the construct. The correlation between affective commitment and normative commitment was extremely high (r = .97), and the AVE for affective commitment and normative commitment were much less than the squared correlation between these constructs and any other. These indicate lack of discriminant validity. Those who have discussed the discriminant validity of normative commitment (e.g. Cohen, 2007; Jaros, 2007; Meyer et al., 1993; Meyer & Herskovitch, 2001) claim that normative commitment is a part of affective commitment because many items in the former dimension have content of affective commitment (Jaros, 2007). Therefore, the construct of normative commitment was eliminated in the model.

As a result, the final measurement model included nine constructs and 39 items. The modified model of the relationship among support, organizational commitment, motivation, and work outcomes is shown in Figure. 2. The results of the CFA on the modified measurement model showed a reasonable fit (RMSEA = .069; CI = .063-.075; p_close = 0.0, χ²/df = 1304/666 = 1.96, NFI = .788, IFI = .884, CFI = .881). The NFI, IFI, and CFI values were slightly lower than recommended values of .90. However, as discussed earlier, these values are depressed when many items and factors are included in the model (Cheung & Rensvold, 2002).

The factors, items, loadings, alpha coefficients, and AVEs are shown in Table 1. The alpha coefficients were adequate ranging from .69 to .95 for a mean of .81. The AVE values were over the recommended .50 (Fornell & Larcker, 1981) with three exceptions: continuance commitment (.44), intrinsic motivation (.37), and extrinsic motivation (.46). The means, standard deviations, and correlations among the constructs are displayed in Table 2.

4.2. Structural Equation Modeling

The goodness-of-fit statistics indicated that the structural model showed a reasonable fit (RMSEA = .074; CI = .067-.080; p_close = 0.0, χ²/df = 959/457 = 2.098, NFI = .818, IFI = .896, CFI = .894). The second order factor of perceived support was well represented by the first order factors, coefficients ranging from β = .788 to β = .981. However, continuance commitment did not load highly on organizational commitment (β = .382), and affective commitment loaded perfectly (β = 1.00) on organizational commitment. The former indicates that continuance commitment is not a good representation of organizational commitment for this data set while the latter indicates a boundary parameter violation for this data set.

The relationships among the latent variables were significant and explained a significant amount of variance. As hypothesized, Perceived Support was significantly positively related with Organizational Commitment (β = .891) explaining 79% of the variance in organizational commitment. In addition, Organizational Commitment was significantly negatively correlated with Intention to Leave (β = -.781) and explained 61% of the variance in Intention to Leave. Organizational Commitment was also positively associated with Work Effort (β = .404) explaining about 16% of the variance in Work Effort.

4.3. Moderating Effects of Motivation

Due to the perfect loading of affective commitment on organizational commitment and poor loading of continuance commitment on organizational commitment, the moderating effect of motivation between Perceived Support and two dimensions of organizational commitment was separately examined.

The predictor variable of Perceived Support was significantly correlated with the dependent variable of Affective Commitment (r = .818; p <.001), and with the Continuance Commitment (r = .172; p <.05). The moderator variable of Intrinsic Motivation was significantly correlated with Affective Commitment (r = .287; p <.001) but not with Continuance Commitment (r = .042; n.s.). The moderator variable of Extrinsic Motivation was significantly but negatively correlated with Affective Commitment (r = -.326; p <.001) and positively correlated with Continuance Commitment (r = .165; p <.05). To test the moderator hypotheses with reference to Affective Commitment, we carried out two regression analyses where the predictor variable (i.e., Perceived Support) and the moderator variable (Intrinsic Motivation or Extrinsic Motivation) were entered in the first step and the interaction of the predictor and the relevant interaction term was entered second. Similarly, two other regression analyses were carried out with reference to Continuance Commitment. The results are presented in Table 3.

Figure 2. Modified model of the relationship among perceived support, organizational commitment, motivation, and work outcomes

Table 1. Items, Factor Loadings (β), Cronbach’s alpha (α), and Average Variance Explained Values (AVE) for the Subscales

Table 2. Means, Standard Deviations, and Correlations for the Subscales

Table 3. Moderating Effects of Intrinsic and Extrinsic Motivation

As shown in Table 3, the interaction of Perceived Support and Intrinsic Motivation was significant only in the case of Affective Commitment (β = .834; p <.01) explaining 1.2% of its variance.

The nature of the interaction is shown in Figure. 3. When perceived support was low, those high on intrinsic motivation were less affectively committed to the organization than those low on intrinsic motivation. However when perceived support was high, those high on intrinsic motivation were higher on affective commitment than those low on intrinsic motivation.

Figure 3. Interaction effect of perceived support and intrinsic motivation on affective commitment

5. Discussion and Conclusions

The purpose of the current study was to examine the impact of perceived support on organizational commitment, and the impact of organizational commitment on work effort and intention to leave. In addition, the moderating role of motivation in the relationship between perceived support and organizational commitment was also investigated.

The first comment relates to the measurement of the variables of the study. We expected that organizational commitment would be represented by three dimensions of affective commitment, continuance commitment, and normative commitment as was suggested by some scholars (e.g., Eisenberger et al., 2001; Meyer & Allen, 1990; Pack et al., 2007; Turner & Chelladurai, 2005; Wasti, 2005). Contrary to these earlier studies, our measurement model failed to distinguish between affective and normative commitment (r = .97). The lack of discriminant validity of normative commitment has been pointed out in previous research (Ko et al., 1997; Meyer et al., 1993). In Ko et al.’s (1997) study, the correlation between affective commitment and normative commitment was .82, and in Meyer et al. (1993) the correlation was .82 between the two dimensions. These values were still in acceptable range of less than .85 (Kline, 2005). However, in this study, the correlation between affective commitment and normative commitment was .97 among fitness employees indicating these two factors are almost the same. This may be because, as Jaros (2007) noted, some of normative commitment items in Meyer and Allen’s commitment scale reflect affective commitment. Further, as argued by Meyer et al. (1993), the high correlation between affective commitment and normative could be due to the sharing of common antecedents. It might be that perceived support available in fitness clubs is a very strong determinant of both affective and normative commitment of employees of fitness organizations. Therefore, these forms of commitment were very highly correlated. Although this finding deterred us from using both measures in subsequent analyses, it does signify the importance of support in cultivating either or both forms of commitment.

While it is encouraging that the three forms of support were shown to be distinct factors, the high correlation between supervisor support and organizational support (r = .85) is worthy of note. As the supervisor is the linking between the organization and the employees, he or she acts as the conduit through which much of organizational support flows to employees. Thus, employees’ perceptions of organizational support could accentuate their perceptions of supervisor support or vice versa.

The antecedent role of perceived support on organizational commitment found in this study is consistent with the findings from the previous literature (Eisenberger et al., 2001; Pack et al., 2007; Thompson, Jahn, Kopelman, & Prottas, 2004). When fitness club employees felt that they were supported by the organization, supervisor, and/or coworker, they were more committed to the organization. In addition, the work outcomes of organizational commitment were also in the expected direction confirming that those employees who are highly committed to the fitness clubs are more likely to put forth more effort into work and less likely to leave the fitness club. These attitudinal outcomes can be explained by social exchange theory (Blau, 1964). Fitness employees who perceive they are supported at their work place reciprocate such favorable treatment through enhanced commitment to the organization, greater effort and reduced intention to leave the organization.

Cumulatively, our results show that the experiences of the fitness industry workers of this study are very similar to the workers in other industries. Workers in service industries compared to those in goods manufacturing industries face different sets of contingencies in dealing directly with the customers. Further, fitness workers differ from conventional service workers in that their clients are in contact with the employees for a considerably longer duration than in conventional services, and such involvement is agonistic in nature. Therefore, one could expect that fitness workers’ experiences and reactions would be different from those of workers in other industries. This was not the case in the present study of perceived support from the organization, supervisor, and coworkers. It could be argued that irrespective of the type of work, employee commitment to the organization and subsequent attitudinal outcomes are mainly a function of people and processes within the organization. Future studies may test this proposition by incorporating the nature of employee-customer interface as a predictor of work effort and turnover intentions.

The finding that extrinsic motivation did not have a moderating effect on either affective commitment or continuance commitment can be set aside as an artifact of the inadequate internal consistency and construct reliability of the scale measuring extrinsic motivation. However, it is not clear why intrinsic motivation would moderate the relationship between perceived support and affective commitment. If support from other sources is perceived to be an extrinsic reward (Herzberg, 1987), those intrinsically motivated should not be swayed in their commitment by perceived support. The intrinsically motivated respondents of this study reacted differently. One possible explanation could be that some of items in the scale measuring organizational support refer to the organization recognizing and valuing employee effort and contribution. As these would reflect the motivators of achievement and recognition in Herzberg’s dual factor theory, the reactions of the intrinsically oriented respondents to perceived support is understandable. However, the issue becomes mute because the explained variance due to this interaction was rather meager (ΔR²= .012).

In the present study, we tested the relationship between support from three different sources (coworkers, supervisors, and organization) and commitment to a single entity—the organization. Research, however, shows that employees could be differentially committed to the organization and the supervisor (e.g., Stiglhamber & Vandenberhe, 2003). Given this possibility, future research in the sport and fitness context may assess employee commitment to the organization as well as commitment to the supervisor and, then, test their differential effects on attitudinal and behavioral outcomes.

The current study significantly contributes to the body of sport management literature in several ways. First of all, the results of the relationships between constructs indicate that the model is well applicable to fitness club employees. As mentioned earlier, although many studies about antecedents and consequences of organizational commitment have been conducted among coaches, studies on fitness employees are greatly limited despite the importance and the size of the industry. Therefore, the current study demonstrates that the previously established relationships among the constructs in other industries and samples are applicable to fitness employees.

In addition, the present study is the first attempt to examine moderating effects of work motivation in the relationship between perceived support and organizational commitment. As the moderating effect of intrinsic motivation was found to be significant, although the effect size was rather small, it adds to the limited literature and inform organizational behavior researchers about the necessity of studying employees’ motivation as a construct that alters the relationships among the work attitudes and behaviors.

As for managerial implications, this study identifies areas where fitness club managers could focus in order to keep valuable employees and to increase their work effort. Our findings suggest that by providing support, employees’ level of organizational commitment can be increased and that resultant high level of organizational commitment could contribute to reducing employees’ intention to leave the organization and to increasing their work effort. Therefore, managers of fitness clubs should realize that they can influence the commitment level of employees by increasing the support available to the employees. Obviously, the managers and/or supervisors can extend their personal support to the employees. In addition, they can also influence the higher-ups of their clubs to create policies and procedures that would be supportive of the employees. Finally, the managers and/or supervisors can create conditions in the workplace to be conducive of workers supporting each other in facilitating task performance and satisfying social needs.

An important point to note is that the support from all three sources (i.e., organization, manager/supervisor, and coworkers) does not cost anything in financial terms. Such support emanates from employee-oriented policies and practices focused on making the work experiences pleasant and positive. An equally important point is that the manager/supervisor can significantly control the conditions that facilitate the positive feelings among the employees that are antecedent to their commitment and work effort.

The current study has several limitations. First of all, the response rate was low; therefore, the results cannot be generalized. According to Kaplowitz, Hadlock, and Levine (2004), web based surveys typically yield a low response rate. In addition, although NETA was able to reach 1,900 members who are holding valid NETA certificates through email, NETA did not know whether these members were all employed by fitness organizations. In fact, we asked the participants to respond to the questionnaire only if they were currently employed by a fitness organization. Therefore, it is possible that many of the people who received the email did not respond simply because they were not employed by fitness organizations at that moment. However, a redeeming feature of our data set is that the respondents who provided the data were both males and females who held both part-time and full-time jobs in profit as well as nonprofit fitness organizations. Their ages ranged from 21 years to over 50 years, and they hailed from different regions of the country. Such diversity among the respondents lets us place confidence in our results.

Secondly, the internal consistency and construct reliability of some scales (continuance commitment, intrinsic motivation, and extrinsic motivation) used in this study were somewhat problematic. For continuance commitment, the alpha and AVE values were slightly lower (α = .69, AVE = .44) than the guidelines. Although the AVE value for extrinsic motivation was slightly low (.46), its alpha was satisfactory (α = .71). Similarly, the intrinsic motivation scale had a slightly low alpha (.69), but had a problematic AVE value (.37) meaning that the construct was not well represented by the items.

Future studies should replicate the current findings with better measures of intrinsic motivation and extrinsic motivation. In addition, it would be worthwhile to investigate separate effects of coworker, supervisor, and organizational support on the different dimensions of organizational commitment. In order to do this, the measure of organizational commitment needs to be refined. In particular, the normative commitment scale should be rewritten to reflect the definition of the construct. Moreover, future study should include more work outcome variables that are related to organizational commitment, such as organizational citizenship behavior and performance.