Testing the Measurement Model
This chapter is different from OLS method including hypothesis and models. PLS method has five models and assumptions. Model 1 is trying to measure the leadership behavior dimension affect organizational performance. Model 2 is trying to measure the practice dimension affect organizational performance. Model 3 is trying to measure the culture dimension affect organizational performance. Model 4 is assume that culture, leadership and practice are the same construct and the average value for each dimensions were used to measure organizational performance. Model 5 is trying to measure the KM system influence to affect organizational performance.
This study uses Cronbach’s alpha in PLS approach to assess the measurement model and the Cronbach’s alpha are report at Table 5.2. All the Cronbach’s alpha values of the three constructs exceeded 0.79 (0.8667 for Model 1; 0.8819 for Model 2; 0.8548 for Model 3; 0.9327 for Model 4; 0.9181 for Model 5).
Table 5.4, 5.5, 5.6, 5.7, 5.8 shows the results of PLS loadings on all the items. Loadings of 0.5 or greater maybe acceptable if there are exists additional indicators for describing the latent construct (Chin, 1998). Therefore, items with loadings of 0.5 or greater are retained. There are other authors (Birkinshaw, Morrison, & Hulland, 1995) who have also followed this criterion in their exploratory studies.
66
Table 5.1
Research Hypothesis Hypothesis H1
H2
H3
H4
H5
Leadership has no effect on market leadership.
Leadership has no effect on financial performance.
Practice has no effect on market leadership.
Practice has no effect on financial performance.
Culture has no effect on market leadership.
Culture has no effect on financial performance.
KM system influence average has no effect onmarket leadership.
KM system influence average has no effect on financial performance.
KM system influence has no effect on market leadership KM system influence has no effect on financial performance.
Table5.2
Measurement Model Results (PLS)
Model Items Cronbach’s Alpha R² (%) Internal Consistency Model 1 18 0.8667 0.4353 0.9083 Model 2 27 0.8819 0.4942 0.9186 Model 3 18 0.8548 0.4877 0.9015 Model 4 49 0.9327 0.9060 0.9569 Model 5 49 0.9181 0.8457 0.9345
67
PLS Results
The data analysis method used in this paper is Partial Least Squares (PLS). PLS is used to analyze simultaneously the interrelationships among all the constructs. Additionally, in order to evaluate the statistical significance of the loadings and the path coefficients (standardized betas), a jackknife analysis was performed. By applying the jackknife formula, PLS estimates the parameters for each sample and compute the “pseudo values” (Table5.3). Three paths (leadership to organizational performance, practice to organizational performance, and culture to organizational performance) have shown significance at the p-value < 0.10.
Table 5.3
PLS Path Analysis Results (Standardized Beta Coefficients and Adjusted P-values)
Path Model β-path Adj.
Note: L= leadership, P=practice, C= culture, OP= organizational performance, KA= KM system influence’s average, KMS= knowledge management system influence.
* p<0.1, **p<0.05, *** p<0.001
Figure 5.1 demonstrates the results for the structural model. The results pinpoint that the three constructs that forms knowledge management model really effect organizational performance. One important benefit of the PLS methodology is that it makes it possible to separate direct and total effects of the variables included in the model (Cabrita & Bontis, 2008).
Figure 5.1 (Model 1) , according to the Smart PLS, from the leadership perspective (Table 5.4), the result show that key success factor of KM has a positive parameter (0.872) and t-ratio (9.584), which is significant at 1%. Trust has also a positive parameter (0.832) and t-ratio (4.224) which is significant at 1%. Authentic leadership has also a positive parameter (0.820) and t-ratio (3.115) which is significant at 1%. Resource integration has also a positive parameter (0.851) and t-ratio (22.656) which is significant at 1%.However, from organizational performance dimension, the result shows that marker leadership performance(parameter=0.952,
t-68
ratio=17.122), financial performance (parameter=0.955, t-ratio =13.937) have a positive parameter and t-ratio which are all significant at 1%.
Table 5.4
PLS Loadings of Model 1
Items Loading Cronbach’s Alpha
AF1 0.8716 0.8667
AF2 0.8325
AF3 0.8199
AF4 AY1
0.8515 0.9518
AY2 0.9546
Note: AY1= market leadership performance average, AY2= financial performance average, AF1= Key success factors of KM, AF2= trust, AF3=authentic leadership, AF4=resource integration.
Figure 5.1. Model 1 results
69
Figure 5.2 (Model 2), according to the Smart PLS, from the practice perspective (Table 5.5), the result show that key success factor of KM (parameter =0.852, t-ratio =7.403). Resource integration (parameter =0.867, ratio =13.845).Process effectiveness (parameter= 0.891, t-ratio=14.281).Knowledge sharing (parameter=0.891, 14.281). All of these four perspective significant at 1 %.However, from organizational performance dimension, the result shows that marker leadership performance (parameter=0.954, t-ratio =18.402), financial performance (parameter=0.953, t-ratio =8.272) have a positive parameter and t-ratio which are all significant at 1%.
Table 5.5
PLS Loadings of Model 2
Items Loading Cronbach’s Alpha
AF1 0.8516. 0.8819
AF4 0.8617.
AF5 0.8912
AF6 AY1
0.8260 0.9536
AY2 0.9529
Note: AY1= market leadership performance average, AY2= financial performance average, AF1= Key success factors of KM, AF4=resource integration, AF5= KM process effectiveness, AF6= knowledge sharing.
Figure 5.2. Model 2 results
70
Figure 5.3(Model 3), according to the Smart PLS, from the culture perspective (Table 5.6), the result show that key success factor of KM (parameter=0.859, t-ratio =5.988). Trust (parameter =0.796, t-ratio=5.465). Knowledge sharing (parameter=0.823, t-ratio=8.126).
Knowledge compatibility (parameter=0.857, t-ratio =12.315). All of these four perspectives are significant at 1 %. However, from organizational performance dimension, the result shows that marker leadership performance (parameter=0.952, t-ratio=32.967), financial performance (parameter=0.954, t-ratio=8.272) have a positive parameter and t-ratio which are all significant at 1%.
Table 5.6
PLS Loadings of Model 3 Items
(M3)
Loading Cronbach’s Alpha
AF1 0.8593 0.8548
AF2 0.7962
AF6 0.8227
AF7 AY1
0.8557 0.9521
AY2 0.9544
Note: AY1= market leadership performance average, AY2= financial performance average, AF1= Key success factors of KM, AF2= trust, AF3=authentic leadership, AF6= knowledge sharing, AF7= knowledge compatibility.
Figure 5.3. Model 3 results
71
Figure 5.4 (Model 4), according to the Smart PLS, from the KM system influence average (Table 5.7), the result shows that average culture dimension has a positive parameter (0.940) and t-ratio (28.764). Average leadership dimension has also a positive parameter (0.921) and t-ratio (4.496). Average practice dimension has also a positive parameter (0.955) and t-ratio (4.690) which is significant at 1%. However, from organizational performance dimension, the result shows that marker leadership performance (parameter=0.970, t-ratio =78.740), financial performance (parameter=0.967, t-ratio =45.451) have a positive parameter and t-ratio which are all significant at 1%.Besides, the path from KM system influence average to organization performance (parameter= 0.952, t-ratio =17.854).
Table 5.7
PLS Loadings of Model 4 Items
(M4)
Loading Cronbach’s Alpha
AC 0.9402 0.9327
AL 0.9209
AP 0.9546
AY1 AY2
0.9697 0.9671
Note: AY1= market leadership performance average, AY2= financial performance, AC= culture dimension’s average, AL=leadership dimension’s average, AP= practice dimension’s average AC, AL and AP are culture, leadership and practice’s average. There are four factors in each dimension, researcher get average from each dimension from execel file, AC including key success factors of KM, trust, knowledge sharing and knowledge compatibility, AL including key success factors of KM, trust, authentic leadership and resource integration., AP including key success factors of KM , resource integration, KM process effectiveness and knowledge sharing.
72
Figure 5.4. Empirical results of model 4
73
Figure 5.5(Model 5), according to the Smart PLS, from the KM system influence perspective (Table 5.8), the result show that key success factor of KM has a positive parameter (0.834) and t-ratio (6.611). Trust has also a positive parameter (0.780) and t-ratio (6.271).
Authentic leadership has also a positive parameter (0.772) and t-ratio (5.651). Resource integration has also a positive parameter (0.810) and t-ratio (4.688) KM process effectiveness (parameter=0.854, t-ratio=7.981).Knowledge sharing (parameter=0.810, t-ratio=4.668).
Knowledge compatibility (parameter= 0.835, t-ratio = 9.326) which is significant at 1%.
However, from organizational performance dimension, the result shows that marker leadership performance (parameter=0.969, t-ratio=27.511), financial performance (parameter=0.968, t-ratio=32.706) have a positive parameter and t-ratio which are all significant at 1%.Besides, the path from KM system influence perspective to organization performance (parameter= 0.920, t-ratio=21.816).
Table 5.8
PLS Loadings of Model 5
Items Loadings Cronbach’s Alpha
AF1 0.8366 0.9181
Note: AY1= market leadership performance average, AY2= financial performance average, AF1= Key success factors of KM, AF2= trust, AF3=authentic leadership, AF4=resource integration, AF5= KM process effectiveness, AF6= knowledge sharing, AF7= knowledge compatibility.
74
75
PLS Results Discussions
As the matter of fact, the origin theoretical framework has problem of multicollinearity in OLS multiple regressions. Multicollinearity is a statistical phenomenon in which two or more predictor variables in a multiple regression model are highly correlated. In this situation the coefficient estimates may change erratically in response to small changes in the model or the data.
A system equation model was established by using PLS method. The empirical results of PLS method for different assumptions and hypothesis shows that leadership behavior, knowledge management practice and corporate culture have positive and significant effect on organizational performance for banking institution in Taiwan.
The findings indicate that knowledge management system (leadership, practice, culture) do have a significant effect on organizational performance separately. KMS (knowledge sharing, knowledge compatibility, trust, resource integration, authentic leadership, KSF of KM and process effectiveness) have positively influence organizational performance including tangible and nontangible asset. Furthermore, the empirical findings of this research are also in support of the fact that not only leadership, but also organization’s culture and the practice of knowledge management have positively influence on organizational performance. That is, if Taiwan banks want to raise their organizational performance they need to put more focus on supervisor’s leadership behavior, organization’s culture and the practice of knowledge management. Table 5.9 shows the result of hypothesis.
76
Table 5.9
Research Hypothesis Results
Hypothesis Outcome H1
H2
H3
H4
H5
Leadership has no effect on market leadership. Rejected Leadership has no effect on financial performance. Rejected Practice has no effect on market leadership. Rejected Practice has no effect on financial performance. Rejected Culture has no effect on market leadership. Rejected Culture has no effect on financial performance. Rejected KM system influence average has no effect onmarket leadership. Rejected KM system influence average has no effect on financial performance. Rejected KM system influence has no effect on market leadership Rejected KM system influence has no effect on financial performance. Rejected
77