Chapter 5 The Case Studies, Result Analysis
5.2. The analysis on the survey questionnaires
5.2.6. Perceived Local Government Performance
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
146
We assume that when people said that religiosity is important it was because being religious means also being honest.
5.2.6. Perceived Local Government Performance
cation and the local economic development performance. A discussion of the performance of these aspects follows. When the respondents agree that they are satisfied, it may indicate that performance is good. Four indicators are proposed in this study to identify the performance: 1) I am satisfied by the quality or skills of the medical labors; 2) I am satisfied by the quality of healthcare facilities; 3) I am satisfied by the quality of educational services provided and 4) Under his leadership, the economy of the city has improved.
5.2.6.1. City A performance
The score of Public Satisfaction Index for City A (Table 5.18) is between 2.48 and 2.71 which indicate that the performance is categorized between fair and good. The performance of healthcare facilities is the best, followed by the educational service and then the economy improvement and finally the performance of healthcare worker the last. The results corresponds the percentage of respondent satisfaction. The result shows that 73.2 percent respondents agree that they were satisfied by the quality of medical labor, 82.6 percent agree that they were satisfied by the quality of healthcare facilities, 79 percent agree that they were satisfied by the quality of educational services and finally 76.1 percent respondent agree that the economy of the city had improved. The percentage and the perception index results are corresponding.
‧
Table 5-18 Perceived city A performanceNo Indicator Mean of
3 I am satisfied by the quality of educational facilities
2.66 Good 79%
4 I believe that the economy of the city has improved
2.61 Good 76.1%
The respondents who give “do not know” response for the healthcare services are those who never had health problems actually. In other words, they are people who never had experiences with medical treatment.
Meanwhile, the respondents who give “do not know” response for the educational services, they were questioning whether the curriculum also included as the services. Therefore, it is only a matter of having more specific indicators.
Correlation between the City A’ mayor leadership behaviors and the government performance.
Table 5.19 shows the correlation analysis between the independent and dependent variables. The tables show the observed leader‟s behaviors have significant correlation with government performance satisfaction. However, the strength of correlation is not too strong because most of it is far from coefficient 1. The strongest correlation is shown by the mayor‟s skill behaviors whereas the weakest is personal integrity behaviors.
Source: author
‧
Table 5-19 Correlation analysis mayor A perceived leadership behaviors and city A perceived government performance
**. Correlation is significant at the 0.01 level (2-tailed).
To know how leader‟s behaviors affect the performance, we run SPSS factor analysis and regression to identify behavior‟s variance. Table 5.32 shows there are three components of behaviors and we name it CityA style1, CityA style2 and CityA style3. City A style1 component consists of managing resources, managing people, listening and oral communication skills, technical competence, impersonality and consistency. City A style 2 consists of non-verbal and writing communication skill and exchange tactics. Finally, city A style3 consists of honesty and coalition tactics.
The SPSS regression result shows that with F (3,134) = 18.598, p < 0.05, R Square = 0.294, we can reject the null hypothesis. Therefore, there is a significant correlation between the mayor leadership behaviors and City A performance quality.
‧
Table 5-20 Factor analysis result of perceived leadership behaviors of mayor AVariable
writing communication skill 3.78 .87 .730 .202
exchange tactics 3.91 .70 .193 .445 .338
Honesty 3.69 .950 .151 .822
coalition tactics 3.8 .80 .149 .434 .465
n = 138, p < 0.50
N items = 12, Cronbach‟s Alpha .723
Rotation Method: Varimax with Kaiser Normalization .815 a. Rotation converged in 5 iterations.
The R Square indicates that 29.4 percent of the total variable in the city performance statistically has significant relationship with the leadership behaviors while 71.6 percent is influenced by other variable. Based on the weights assigned to each leadership behavior components, the local government performance could be estimated on the basis of the following equation: Performance = 15.391 + 0.758(cityAstyle1) + 1.036 (cityAstyle2) + 0.680(cityAstyle3).
The regression model indicates that for every one person who identifies leader displays the leader‟s style 1, the four aspects performance satisfaction or quality increases 0.758 point controlling for leader‟s style 2 and style 3. Moreover, for every one person who identifies leader display the leader‟s style 2, the performance quality increases 1.036 controlling for leader‟s style 1 and style 3. While for every one person who identifies leader display the leader‟s style 3, the performance quality increases 0.680 controlling for leader‟s style1 and style2. The regression analysis indicates that all predictors contributed positively to the prediction of the government performance and city A leader‟s style2 is preferred to contribute to performance. The summary of
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
150
this regression analysis is presented in Table 5.21 whereas the scatterplots is shown by Figure 5.19.
Table 5-21 Regression analysis results for the mayor‟s A leadership behaviors and the city A performance
Coefficients
Component Weight Std. Error t Sig.
Constant 15.391 .194 79.453 .000
CityA Style1 .758 .194 3.896 .000
CityA Style2 1.036 .194 5.327 .000
CityA Style3 .680 .194 3.498 .001
a. Dependent Variable: City A Performance
Figure 5-19 Normal P-P Plot regression of city A
5.2.6.2 City B performance
Based on the Public Satisfaction Index measurement, City B performance is categorized as having good performance on healthcare, education and economy aspects. The scores ranged between 2.62 and 3.02 of
‧
the maximum score of 4. The economy had the best performance compared to other sectors while the performance of healthcare workers is the lowest. The satisfaction index result and the quality percentage are corresponding. The result shows that 76.6 percent respondents agree that they were satisfied by the quality of medical labor, 78.8 percent agree that they were satisfied by the quality of healthcare facilities, 87.2 percent agree that they were satisfied by the quality of educational services and finally 91.5 percent respondent agree that the economy of the city had improved.Table 5-22 Perceived city B performance
Indicator PSI Criteria Percentage 1 I am satisfied by the quality of the
The Table 5.23 shows correlation between the independent and dependent variables. Among the observed behaviors, we find the mayor influence tactics has no significant correlation with the performance. The mayor‟s skills show the strongest correlation followed by the mayor personal integrity. The results might indicate that skills have better influence compared to personal integrity.
To see the relationship between the perceived leader‟s behavior and the perceived government performance quality, we run factor analysis and regression
Source: author
‧
analysis test. The factor analysis results three components of leader‟s behaviors variances. We name the groups as City B style 1, City B style 1 and City B style 3.
Table 5-23 Correlation analysis mayor B perceived leadership behaviors and the city B performance
Correlations
1 2 3 4
Mayor personal integrity Pearson Correlation 1 Sig. (2-tailed)
Mayor skills Pearson Correlation .480** 1 Sig. (2-tailed) .001
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
Table 5-24 Factor analysis city B mayor perceived leadership behaviors
Variable
Rotation Method: Varimax with Kaiser Normalization .726 a. Rotation converged in 4 iterations.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
153 City B style1 component consists of impersonality, honesty, exchange tactics, writing communication skill and consistency. City B style2 component consists of managing person, managing resource, monitoring, listening and oral communication skill. Finally, city B style3 consists of nonverbal communication skill and coalition tactic. The SPSS regression result shows that with F (3,43) = 6.355, p < 0.05, R Square = 0.307, we can reject the null hypothesis. Therefore, there is a significant correlation between the mayor leadership behaviors and the City B performance quality. The R Square indicates that 30.7 percent of the total variable in the city performance statistically has significant relationship with the performed leadership behaviors. Based on the weights assigned to each component, the city B performance could be estimate on the basis of the following equation: City B performance = 15.851 + 0.717(cityB style1) + 0.787(cityB style2) + 0.695(cityB style3). The regression analysis indicates that all predictors contributed positively to the prediction of the government performance and city B leader‟s style2 is preferred to contribute to performance.
The above equation shows that for every one person who identifies leader displays the leader‟s style 1 the four aspects performance quality increases 0.717 controlling for leader‟s style 2 and style 3. Moreover, for every one person identifies leader display the leader‟s style 2, the performance quality increases 0.787 controlling for leader‟s style 1 and style 3. While for every one person identifies leader display the leader‟s style 3, the performance quality increases 0.695 controlling for leader‟s style1 and style 2. The summary of this regression analysis is presented in Table 5.25 While for the scatterplots for the mayor B leadership behavior and the city B performance is shown by Figure 5.20.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
154
Table 5-25 Regression analysis results including component variable weights and t test value for the mayor B leadership behaviors and city B performance
Coefficients
Component Weight Std. Error t Sig.
Constant 15.851 .288
.291 .291 .291
55.038 .000
City B style1 .717 2.463 .018
City B style2 .787 2.704 .010
City B style3 .695 2.386 .022
a. Dependent Variable: City B Performance
Figure 5-20 Normal P-P Plot regression of city B
5.2.6.3. County D performance
Based on the Public Satisfaction Index standard, County D was categorized as fair performance quality (Table 5.26). The economy performance is the highest while the healthcare worker performance quality is the lowest. In total, using the SPSS frequency, we found 25 percent respondents disagree that they were satisfied while 45 percent were agree. On the other hand, respondents who do not know if they are satisfy or not are
‧
relative high at 31 percent, the number is slightly higher than those disagree.In addition the rate satisfaction percentage below 50 percent may indicate that the government performance quality is deprived.
Table 5-26 Perceived county D performance
Indicator PSI Criteria Percentage
1 I am satisfied by the quality of the medical labor
1.48 Fair 42.8%
2 I am satisfied by the quality of healthcare facilities
1.78 Fair 44.4%
3 I am satisfied by the quality of educational facilities
1.78 Fair 42.9%
4 Under his leadership the economy has improved
2.24 Fair 47.7%
The correlation between the Bupati’s leadership behaviors and the government performance
Table 5.27 shows the Pearson correlation test between County D Bupati behavior and the county performance. Table 5.27 shows that the three groups of behaviors have significant correlation with the perceived government performance.
Table 5-27 Correlation analysis Bupati D perceived behaviors and the county D perceived performance
Leader skills Pearson Correlation .518** 1 Sig. (2-tailed) .000
Leader Influence Behaviors
Pearson Correlation .429** .524** 1 Sig. (2-tailed) .001 .000
County D Performance Pearson Correlation .509** .400** .296* 1 Sig. (2-tailed) .000 .001 .018
**. Correlation is significant at the 0.01 level (2-tailed).
*. Correlation is significant at the 0.05 level (2-tailed).
Source: author
‧
The strongest correlation is shown by leader‟s personal integrity followed by leader‟s kills and finally leader‟s influence behaviors competency. The strength of these correlations is not too strong since the coefficient is far from 1. As in City B case, we also use the SPSS regression analysis to find out how leader‟s behaviors may influence the performance quality. Three components of behaviors result when we run the factor analysis.
Table 5-28 Factor analysis county D Bupati perceived leadership behaviors
Variable
Mean Standard Deviation
Component
1 2 3
listening communication skill 3.79 1.182 .841 .105 .192
technical competence 3.76 .995 .802
nonverbal communication skill 3.68 1.013 .110 .761
coalition tactics 3.59 .854 .165 .739
exchange tactics 3.51 1.030 .201 .491 .620
n = 63, significant at p < 0.50
N items = 12, Cronbach‟s Alpha .851
Rotation Method: Varimax with Kaiser Normalization = .788 a. Rotation converged in 5 iterations.
County D style1 component consist of listening and oral communication, technical competency, managing resources and managing people behaviors. County D style2 consist of honesty, impersonality, consistency and writing communication skill.
Finally, county D style3 consist of nonverbal communication skill, coalition and exchange tactics. The SPSS regression result shows that with F (3, 58) = 6.648, p <
0.05, R2 = 0.256 we can reject the null hypothesis that there is a statistically significant relationship between the dependent variable and the independent variable.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
157 The R square indicates that 25.6 percent of the variation in performance is explained by leadership style County D 1, 2 and 3. Based on the weights assigned to each component, the count D performance could be estimated on the basis of the following equation: Performance = 12.516 + 0.892(county D style1) + 1.527(county D style2) + 0.175(county D style3). The regression analysis indicates that all predictors contributed positively to the prediction of the government performance and county D leader‟s style2 is preferred to contribute to performance.
The equation model indicates that trait, skills and influence behavior have positive affect on the performance quality. For every one person identifies county D style 1 behaviors were displayed by leader, the performance increases 0.892 controlling for county D style2 and county D style3. Moreover, for one person who perceives leader was displaying county D style 2 behaviors, the performance quality increases 1.527 controlling for county D style 1 and county D style 3. Finally, for one person identifies county D style 3 were performed by the leader the performance quality increases 0.073 controlling for county D style 1 and 2. The summary of this regression analysis is presented in Table 5.29 whereas Figure 5.21 shows the scatterplot of County D performance and the Bupati leadership behaviors.
Table 5-29 Regression analysis results including predictor variable weights and t-test values for the county D performance and Bupati D behaviors.
Coefficients
Component Weight Std. Error t Sig.
Constant 12.516 .424
.427 .427 .427
29.541 .000
County D Style1 .892 2.087 .041
County D Style2 1.527 3.576 .001
County D Style3 .715 1.673 .100
a. Dependent Variable: county D performance
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
158
Figure 5-21 Normal P-P Plot regression county D performance
5.2.6.4. City C performance
Table 5.30 shows the Public Satisfaction Index and the satisfaction percentage in City C. Generally, the city has good criteria for all aspects except economy performance quality. The table shows that 74.5 percent respondents agree that they were satisfied by the quality of healthcare workers and 72.7 percent agree that they were satisfied by the quality of healthcare facilities. Moreover, regarding the quality of education facilities performance, 80.5 percent was satisfied by the performance quality. While regarding the economy performance, 78.6 percent respondent believes the economy performance quality was improved under mayor C ruling. The number of respondent who give a “don‟t know” response was 31 percent which is relatively close to disagree. The table shows that there is difference between perception index score and the percentage of positive perception. In terms of
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
159 percentage the economy performance of City C is 2nd, in the PSI standard it is the last.
Table 5-30 Perceived city C performance
Indicator PSI Criteria Percentage
1 I am satisfied by the quality of the healthcare worker
2.60 Good 72.1%
2 I am satisfied by the quality of healthcare facilities
2.52 Good 76.7%
3 I am satisfied by the quality of educational facilities
2.60 Good 79.1%
4 Under his leadership the economy has improved
2.40 Fair 74.5%
Correlation between the City C mayor’s leadership behaviors and the government performance
The table 5.31 shows the correlation between City C mayor behaviors and the city performance. It shows that all leadership behaviors have significant correlations with the government performance perceived. The strongest correlation is shown by leader skills performance followed by leader personal integrity and the last is leader influence behaviors competency. To see how City C mayor‟s leadership behaviors affect the City C performance we run factor analysis and regression analysis. The factor analysis test results three variances of leadership behaviors. We name these leadership groups as City C style 1, City C style 2 and City C style 3 (Table 5.32).
Source: Author
‧
Table 5-31 Correlation analysis mayor C leadership behaviors and the city C performance
1 2 3 4
Mayor Personal integrity
Pearson Correlation 1 Sig. (2-tailed)
Mayor skills Pearson Correlation .674** 1 Sig. (2-tailed) .000
**. Correlation is significant at the 0.01 level (2-tailed).
Table 5-32 Factor analysis city C mayor perceived leadership behaviors
Variable
nonverbal communication skill 4.12 .762 .795
exchange tactics 3.70 .741 .429 .597
n = 43, significant at p < 0.50
N items = 12, Cronbach‟s Alpha .788
Rotation Method: Varimax with Kaiser Normalization .762 a. Rotation converged in 8 iterations.
City C style 1 component consist of impersonality, honesty, consistency, managing people, managing resources, writing and communication skill, and technical competency. City C style2 consist of coalition tactics and oral communication skill and city C style 3 consist of exchange tactics and nonverbal
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
161 communication skill. The regression analysis shows that with F (3,39) = 8.276, p <
005, R2 = 0.389, we can reject the null hypothesis that there is significant correlation between the City C style1, style 2 and style 3. Based on the weights assigned to each component, the city C performance could be estimated on the basis of the following equation: Performance = 15.279 + 1.151(city C style1) + 0.762(city C style2) + 0.645(city C style3).
The equation model indicates that for every one person who perceives city C style1 displayed by the observed mayor, the performance quality increases 1.151 controlling for city C style2 and city C style3. For every one person who perceives city C style2 behaviors are displayed by the mayor, the performance quality increases 0.762 controlling for C style1 and C style3. Finally, for every one person who perceives that C style3 behaviors are performed by the mayor, the performance quality increases 0.645 controlling for city C style1 and city C style 2 behaviors. The regression analysis indicates that all predictors contributed positively to the prediction of the government performance and city C leader‟s style1 is preferred to contribute to performance. The summary of this regression analysis is presented in Table 5.33 while the scatterplots for the mayor C leadership behaviors and the city C performance is shown by Figure 5.22.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
162
Table 5-33 Regression analysis results including component weights and t test value for the mayor C leadership behaviors and city C performance
Coefficients
Component Weight Std. Error T Sig.
(Constant) 15.279 .302
.306 .306 .306
50.559 .000
City C style1 1.151 3.764 .001
City C style2 .762 2.491 .017
City C style3 .645 2.110 .041
a. Dependent Variable: City C Performance
Figure 5-22 Normal P-P plot regression the city C performance