اثرات تیم و تعهد سازمانی - مطالعه طولی

Retention management, i.e., keeping qualified employees, is a top priority for contemporary organizations. Commitment, and especially team commitment, can be the key to mastering this challenge. There is a lack of longitudinal research concerning the development and the direction of the effects of team commitment over time. In a longitudinal field-study design with three points of measurement, a total of 360 employees in 52 semi-autonomous industrial teams were surveyed over a period of three years. On the one hand, organizational commitment showed stronger effects on organization-related criteria (job satisfaction and intention to leave). These effects were consistent over the three points of measurement. Team commitment, on the other hand, affected team-related criteria (team performance and altruism). Longitudinal analyses confirmed the effects of organizational commitment on job satisfaction and intention to leave, and of team commitment on team performance and altruism. Moreover, these effects increased over time. Theoretical and practical implications of these findings are discussed.

Keeping employees commited to the organization is a top priority for many contemporary organizations (Hausknecht et al., 2009, Hunziger and Biele, 2002 and Reiche, 2008). Especially in times of crises and job cuts, committing top performers to the organization becomes a challenge (Hunziger & Biele, 2002). Organizations which fail to accomplish this will have reduced resources for the capability of competing in the future (Rappaport, Bancroft, & Okum, 2003). Top performers are not limited to higher management, but can be found at all levels of an organization. Organizational commitment is one of the main reasons for these employees to stay (Hausknecht et al., 2009). However, in large organizations or after mergers or acquisitions, employees’ commitment to the organization as a whole may be questionable (Riketta and Van Dick, 2005, Van Dick, 2004 and Van Dick and Riketta, 2006). In these cases, team commitment can be a key to retention management (cf. Hausknecht et al., 2009). While organizational commitment has been subject to extensive research, team commitment has rarely been investigated. There is a particular lack of research on longitudinal effects of team commitment (Riketta & Van Dick, 2005). Individuals identify with social groups that are familiar and similar to them (Van Knippenberg & Van Schie, 2000). Team members spend most of their organizational lives in the context of their team, which leads to higher familiarity and cohesion within rather than between teams (Moreland & Levine, 2001). Moreover, teams have a stronger direct influence on their members than does the organization (Anderson & Thomas, 1996). This makes teams be more salient in employees’ everyday lives than the organization as a whole (Riketta & Van Dick, 2005). Research has shown that employees are more committed to their team than to the organization (Riketta & Van Dick, 2005). Organizational commitment affects relevant outcomes, e.g., employees’ turnover intentions, organizational citizenship behavior (OCB), and job satisfaction (Mathieu and Zajac, 1990 and Meyer et al., 2002). Some studies hint at a link between organizational commitment and performance; however, several meta-analyses have shown that this link is rather weak (Cohen, 1991, Mathieu and Zajac, 1990, Randall, 1990 and Riketta, 2002). A recent meta-analysis shows relationships between team commitment and team satisfaction, organizational citizenship behavior toward colleagues, and team climate (Riketta & Van Dick, 2005). However, some questions remain. Does team commitment contribute to explaining changes in relevant organizational variables such as performance? How strong are the effects of team commitment in comparison with organizational commitment?

All hypotheses were tested by means of regression analysis. To examine the strength of the effects of organizational and team commitment, we entered both types of commitment as predictors. In addition, the control variables age, organizational tenure, and education were entered as predictors. To find hints at the direction of the effects, we applied the “Granger test” (cf. Finkel, 1995 and Granger, 1969). This test requires two regression analyses including the preceding value of the independent variable (X1) and the preceding value of the dependent variable (Y1). The latter (Y1) is treated as an additional predictor to control for potential effects of preceding levels of the dependent variable (cf. Finkel, 1995 and Granger, 1969). A second regression analysis examines the direction of the effects. In the present study, organizational commitment and team commitment at t1 as well as the respective outcome variable were entered as predictors of both types of commitment at t2. In accordance with Granger (1969), causal inferences may be drawn by means of the following two equations: View the MathML sourceY2=β1X1+β2Y1+U1X2=β3Y1+β4X1+U2 Turn MathJax on U1 and U2 represent error variables with an expected value of zero. The coefficients β1 and β2 describe the strength of the effects of X1 and Y1 at t1 on the outcome variable Y2 at t2. Likewise, the coefficients β3 and β4 describe the strength of the effects of Y1 and X1 at t1 on the outcome variable X2 at t2 (cf. Finkel, 1995). An effect of the commitment foci was present when (a) in the first regression analysis, the coefficients concerning the effects of the two commitment foci at t1 on the respective outcome at t2 were significant, and (b) in the second regression analysis, there were no significant effects (or less strong effects than produced by the commitment foci at t1) of the outcomes at t1 on the commitment foci at t2. According to Granger, “one could then talk of causality existing at this moment of time” (Granger, 1969, p.429). The same procedure was followed for the t1–t3 and t2–t3 periods. To examine whether the effects of team and organizational commitment are stable over a longer period of time, we tested effects from t1 to t2 and from t1 to t3, respectively. Table 1 shows the means, standard deviations, intercorrelations, and reliabilities (internal consistency) of the above-named variables. As depicted diagonally in Table 1, the internal consistency values were good (ranging between α = .81 and α = .86 at t1 through t3 for team performance, altruism, and overall job satisfaction). Only the turnover intentions scale had a lower reliability at t1 (α = .58). An initial look at the intercorrelations over time indicates that organizational commitment showed higher correlations with job satisfaction and turnover intentions, while team commitment showed higher correlations with team performance and altruism. This already suggests stronger relationships between variables at the same level. Team performance was significantly correlated only with team commitment at t2, but with both commitment foci at t1 and t3. There was no statistically significant relationship between team commitment at t1 and turnover intentions at t3. This corresponds to previous research assuming little contribution of team commitment to the prediction of turnover intentions (cf. Vandenberghe et al., 2004). The regression analysis reported in Table 2 through 7 compared the power of organizational commitment (t1) and team commitment (t1) to explain the variance in t2 and t3 outcomes. Table 2 shows the effects on organization-related outcomes and Table 3 shows the effects on team-related outcomes between t1 and t2. The control variables company, age, professional education, and organizational tenure had no incremental predictive value over the commitment foci in the context of turnover intentions (see Table 2), team performance, and altruism (see Table 3). However, for the criterion of overall job satisfaction, there were some incremental effects (see Table 2). While organizational tenure was only marginally significant in this context (β = −.14, p < .10), company (β = .21, p < .01) and the degree of professional education (β = −.17, p < .01) increased variance explanation for overall job satisfaction significantly. This indicates that more qualified employees tend to be less satisfied with their jobs. A possible explanation is that more qualified employees might see more opportunities for changing their jobs, comparing their working conditions with those in other organizations, or might have higher expectations due to previous experience in other organizations (cf. Felfe & Six, 2005). The control variables were irrelevant for testing the direction of the effects in the second regression analysis, and were therefore not considered.Organizational commitment at t1 was a statistically significant predictor of job satisfaction at t2 (β = .21, p < .05) and of turnover intentions at t2 (β = −.30; p < .001). By contrast, team commitment showed no significant relationship with these outcomes. On the other hand, team commitment at t1 did tendentially help predict team performance at t2. Both organizational and team commitment showed no significant relationship with altruism at t2. The outcomes at t1 (job satisfaction, turnover intentions, team performance, and altruism) had no incremental predictive value for organizational commitment and team commitment at t2. This sustains the influence of organizational commitment on job satisfaction and turnover intentions. Altogether, these findings support our first hypothesis, since organizational commitment at t1 did show a stronger influence on job satisfaction at t2 (H1a) and on turnover intentions at t2 (H1b) than team commitment. Hypotheses H3a and H3b were rejected. H2 and H4 hypothesized about the stability of the effects from t1 to t3. Table 4 shows the effects on organization-related outcomes and Table 5 shows the effects on team-related outcomes between t1 and t3. Among the control variables, the only marginally significant influence concerns professional education as an additional predictor of altruism at t3 (β = −.13, p < .10; see Table 5).Both organizational and team commitment showed significant effects on the hypothesized outcome variables at t3 (see Table 4 and Table 5). These results correspond to or even exceed the effects found at t2. Organizational commitment at t1 was a significant predictor of job satisfaction (β = .38, p < .001) and turnover intentions (β = −.23, p < .05) at t3. Contrarily, team commitment had no influence on these outcomes. Job satisfaction at t1 had a significant negative effect on organizational commitment at t3 (β = −.28, p < .01). Although H2a concerning stronger effects of organizational commitment than team commitment on job satisfaction was supported, causality may not be inferred from these findings. In line with H2b, turnover intentions at t1 showed no significant effect on organizational commitment at t3. Team commitment at t1 turned out to be a significant predictor of team performance (β = .29, p < .01) and altruism (β = .31, p < .001) at t3. By contrast, organizational commitment showed no significant relationship on these outcomes. Additionally, team performance at t1 and altruism at t1 had no effect on team commitment at t3, respectively. These results support H4a and H4b concerning stronger effects of team rather than organizational commitment on team performance and altruism at t3. An additional look at the t2–t3 period supported the majority of the t1–t2 findings (cf. Table 6 and Table 7). Table 6 shows the effects on organization-related outcomes and Table 7 shows the effects on team-related outcomes between t2 and t3. Control variables had no effect on job satisfaction and team performance. The organization had a marginal effect on turnover intentions at t3 (β = .16, p < .10; cf. Table 6). Organizational tenure had a negative effect on altruism at t3 (β = −.23, p < .05; cf. Table 7). A possible explanation could be that employees with more organizational tenure had been working for the organization before team work was implemented, thus feeling responsible for equipment and technology (such as machine maintenance) but not as much for their co-workers.Contrary to the t1–t2 findings, neither organizational nor team commitment at t2 had an effect on job satisfaction at t3 (cf. Table 6). However, job satisfaction did not affect organizational commitment at t3 either (cf. Table 6). Thus, these additional results do not actually contradict the effects found for the t1–t2 and t1–t3 period. As found for the t1–t2 and t1–t3 period, organizational commitment at t2 showed a negative effect on turnover intentions at t3 (β = −.20; p < .05). Turnover intentions at t2 did not affect organizational commitment at t3 (cf. Table 6). Compared to organizational commitment, team commitment at t2 showed a considerably weaker effect on turnover intentions at t3 (β = .16; p < .10), which lends support to hypothesis H1b. As was the case for the t1–t2 period, neither organizational nor team commitment at t2 had a significant impact on team performance and altruism at t3 (cf. Table 7). On the other hand, team performance at t2 showed a marginally negative relationship with team commitment at t3 (β = −.15; p < .10). However, this result appears negligible considering the much stronger relationship between team commitment at t2 and team commitment at t3 (β = .63; p < .001). Altruism at t2 did not affect team commitment at t3 (cf. Table 7). Thus, these additional results do not contradict our hypotheses.