You are on page 1of 2

3.3 Reliability and validity Composite reliability assesses the inter-item consistency, which is estimated using Cronbachs alpha.

Cronbachs alpha values of all factors were well above the cutoff point 0.7 (Cronbach, 1971), which suggested that the theoretical constructs exhibited good psychometric properties. Confirmatory factor analyses (CFA) for each set of focal constructs were taken to further test the composite reliability and construct validity. At 0.772 or higher, the composite reliability (CR) for each construct exceeded the 0.7 benchmark (Fornell and Larcker, 1981). Further, the average variance extracted (AVE) for each construct was 0.503 or higher, exceeding the 0.5 benchmark (Fornell and Larcker, 1981). Thus, composite reliability and convergent validity were supported by the CFA results. Discriminant validity is the degree to which measures of different latent constructs are unique enough to be distinguished from other constructs (Hatcher, 1994). The discriminant validity of the measures was assessed by running chi-square difference tests for all the
constructs in pairs to determine if the restricted model (correlation fixed as 1) performed significantly worse than the freely estimated model (correlation estimated freely). All the chi-square differences were highly significant, providing evidence of discriminant validity (Anderson and Gerbing, 1988). Dependent and independent variables were measured by two respondents assessments and evaluation of their company in this study, thus, the potential dangers of common method variance should not exist in this study (Zhou and Wu, 2010). Additionally, the possibility of common method variance was also examined via Harmans one-factor test (Podsakoff and Organ, 1986). Significant common method bias would result in one general factor accounting for the majority of covariance in the variables. The Harmans one-factor test on all the multi-item variables extracted seven distinct factors that accounted for 71.2 percent of the total variance, with the first factor explaining 24.9 percent. Thus, no general factor was apparent, and common method variance was not a serious issue (Podsakoff and Organ,

1986).
The score for each item was the average of those from two interviewers, thus the inter-rater reliability should be tested in this study. Inter-rater reliability refers to the degree of agreement among raters, and it can be tested by correlation coefficients (Shrout and Fleiss, 1979). All the variables in this study showed high correlation coefficients. Thus, the data had good inter-rater reliability.

3.4 Findings
Table I shows basic information on each factor and correlations among them. Following existing studies, the regression by the ordinary least squares method (OLS) was used to test the hypotheses in three steps. First, the control variables were added into the model, next three independent variables individualism-collectivism, uncertainty avoidance, and power distance were individually added, and finally the full model. The values of the variance inflation factor (VIF) of all the models were well below the cut-off of ten recommended by Neter et al. (1985). Table II reports the results of the regression analysis. Model 2a indicates the impact of

collectivism on knowledge creation capability is positive (b 0:514, p , 0:001), which supports H1. Model 2b indicates the impact of uncertainty avoidance on knowledge creation capability is negative (b 20:675, p , 0:001), which supports H2. Model 2c indicates the impact of power distance on knowledge creation capability is negative (b 20:138, p , 0:05), which supports H3. In addition, the full model shows the same results as Models 2a, 2b, and 2c. Thus, H1, H2, and H3 all were supported by the regression results.

4. Discussion and conclusion 4.1 Contributions


This study offers distinct contributions to knowledge management research by identifying a key antecedent of knowledge creation capability organizational culture. Investigating the antecedent of knowledge creation capability not only helps to enrich the knowledge on how firms create new knowledge, but also advances knowledge creation. As a result, what are the antecedents of knowledge creation capability is a crucial research issue of knowledge management research (Smith et al., 2005). This study finds that organizational culture plays a critical role on knowledge creation capability. Specially, collectivism has a positive impact on knowledge creation capability, while power distance and uncertainty avoidance have negative effects. Thus, this study contributes to knowledge management research by clarifying the antecedent role played by organizational culture on knowledge creation capability. In addition, the findings also provide some support to the proposition of Smith et al. (2005) that organizational routines are critical antecedents of an organizations knowledge creation capability. Moreover, this study contributes to organizational culture literature as well. Existing research on organizational culture suggests that organizational culture is crucial for successful knowledge management and lack of focus on the cultural issues involved has led to the failure of many knowledge management initiatives (Pillania, 2006, p. 124). Although several studies have explored the impact of organizational culture, what is the appropriate organizational culture for knowledge creation remains an unsettled issue (Park et al., 2004). The findings of this study indicate that the organizational culture with the characteristics of high collectivism, low power distance, and low uncertainty avoidance contributed to knowledge creation capability. Thus, this study contributes to organizational culture literature by demonstrating the proper organizational culture for knowledge creation capability. Besides theoretical contributions, this study has strong empirical values as well. Most existing knowledge management research is developed in the developed economies and little examination is conducted on emerging economies, such as China (Farh et al., 2004). As the research horizon is now increasingly expanded to China, it is critical to know more about what is going on there if the field aspires to be globally relevant (Meyer, 2006). Thus, this study offers a good threshold to extend knowledge management literature to an emerging economy context. Second, China has the potential to further delineate the theoretical boundaries of organizational culture, which have been largely developed in Western contexts (Peng, 2003). This study finds that the dimensions of organizational culture developed from Western contexts can be generalized to the Chinese context and that different dimensions function diversely, which help to enhance the theoretical completeness of organizational culture literature.

You might also like