Research That Matters (January 17 - 20, 2008)

Purpose: The empowerment of social workers is a crucial factor affecting client satisfaction and organizational effectiveness in social work agencies. Recent studies demonstrate the empowerment of an organization's members is closely related to the type of organizational culture. Thus, finding the desirable organizational culture is an important task for social work organizations. This study examined the dynamic between the organizational culture of social work organizations and the empowerment of social workers. Although several studies investigated this dynamic, their analytical methods had statistical drawbacks. Prior studies depended on conventional multivariate models (e.g., regression and SEM) which assume all individuals are independent. However, social workers from the same organization share common experiences and therefore violate the independence assumption. When conventional multivariate models are applied to non-independent data (or nested data), they usually lead to biased estimations. Moreover, testing individual-organizational level interactions is limited by conventional methods. This study overcomes the methodological weaknesses of previous studies by employing multilevel modeling analysis. This study examined (1) to what extent does the type of organizational culture explain the degree of empowerment of social workers? and (2) which type of organizational culture is most effective in enhancing the empowerment of social workers?

Method: 244 social workers from 62 community welfare centers were surveyed using multistage cluster sampling. The type of organizational culture (organizational-level variable) was measured by the Quinn and Kimberly's (1994) organizational culture model where four types of organizational culture (group, rational, development, and hierarchical culture) were identified. The empowerment of social workers (individual-level variable) was measured by Leslie, Holzhalb & Holland's (1998) empowerment index. Other individual-level covariates (e.g., gender and wage etc.) and organizational-level covariates (e.g., size and type of operational body etc.) were used as control variables. Power analysis showed that this sample size (244 in individual-level and 62 in organizational-level) produces a beta of .83 with the .05 alpha level. We conducted multilevel modeling with LINEAR MIXED MODEL of SPSS 14.0.

Results: The analysis yielded the intra-class correlation (ICC) equal to .32, indicating the nested data structure with non-independency. The result demonstrated that while individual-level factors do not affect social workers' empowerment, three organizational-level factors (type of operational body, number of staff, and organizational culture) significantly influence empowerment. Social workers belonging to small-sized agencies were more empowered. When controlling for other individual and organizational level variables, organizational culture solely explained 40.5% of the total variability of social workers' empowerment. While group and developmental culture increased the empowerment of social workers, hierarchical and rational culture decreased their empowerment. There was no cross-level interaction effect between organizational culture and individual factors on empowerment.

Implications: This study illustrated the effects of organizational culture on social workers' empowerment in social work organizations and provided a practical strategy for enhancing the empowerment of social work administrators and researchers. Additionally, this study demonstrated that multilevel modeling sophisticatedly deals with nested data structures and analyzed the dynamics between individual-organizational level factors in organizational research.