• 沒有找到結果。

In this chapter, the overview of the descriptive statistics of this study is provided. In the first section, it shows the sample characteristics of the respondents under this study. And in the second section, the researcher shows descriptive statistics of the result which includes mean and standard deviation of every item.

Sample Characteristics

The characteristics of the respondents are shown in Table 4.1. It consist of respondents’

age, gender, educational background, tenure, department, and job position. The largest group of the respondents in this study was at the age between 26 to 30 years old (32 percent). And the second largest groups of the respondent were at the age between 20 to 25 years old, and above 40 years old. Besides, the response shows that the majority of the respondents in this study were female (77 percent) against male (23 percent).

Concerning the respondent’s educational background, up to 78% have bachelor degree, and 11% are high school graduate. Besides, 9% held master degree, while 2% had doctorate degree. Regarding their length of service, the result showed that the majority has been working more than five years (34 percent). While there are 31% of the respondent have worked one to two years.

About the department they are in, the majority is from the department of Dispatching

Services (44 percent). There are 17% from the department of Foreign Workers Service, 16%

from Pan-Asia (headquarter), 15% from the department of Education, 6% from Head-hunting, and only 2% from the department of Corporate Training.

46

Regarding their job position, the majority is administrative staff in the organization. 22%

of the respondents are consultant/ marketing staff, 18% are managers, and 15% are assistants.

Table 4.1.

Summary Distribution of Respondents Based on Demographic Variables

Variables Entries Percentage

Age 20-25

47

Table 4.1. (continued)

Variables Entries Percentage

Department Head-hunting

Job Position Manager

Administrative staff

48

Descriptive Statistics Analysis

In this section, a summary of the responses to the questions relates to the variable in this research is provided. It consist of each construct’s min., max., mean and standard deviation which indicate how well the mean represents the data. A 5-point Likert scale was used. The respondent are asked to express their opinions ranging from “strongly disagree=1” to “strongly agree=5”. Besides, the averages of each variable are higher than the midpoint (3) of the 5-point Likert scale. It can conclude that the respondents showed some degree of agreement, because their answer mostly ranged from neutral to agree.

Trust

In table 4.2, the respondents showed the highest agreement on TiP2, which states that “I can trust the people I work with to lend me a hand if I needed it”. And the second highest was TiP1 noted that “If I got into difficulties at work I know my colleagues would try and help me out”. However, the respondents showed the lowest agreement on TiM2, which indicated that the employee feel quite confident that the firm will always try to treat them fairly.

Table 4.2

Trust via a 5-point Likert Scale

Survey Questions Min. Max. Mean SD

TiP1 If I got into difficulties at work I know my colleagues would try and help me out.

1 5 4.17 .700

TiP2 I can Trust the people I work with to lend me a hand if I needed it.

1 5 4.18 .692

(continued)

49

Table 4.2. (continued)

Survey Questions Min. Max. Mean SD

TiP3 Most of my colleagues can be relied upon to do as they say they will do.

1 5 4.09 .715

TiM1 Management at my firm is sincere in its attempts to meet the employees’ point of view.

1 5 3.91 .747

TiM2 I feel quite confident that the firm will always try to treat me fairly.

1 5 3.75 .851

Note: N=128; SD= Standard Deviation

Knowledge Sharing (KS)

Regarding knowledge sharing, table 4.3 showed the respondents had the highest agreement on both Intra1 and Intra2, which indicate that the employees frequently share knowledge and information with their work teammates, and they usually involve themselves in discussions of various topics rather than specific topics with their teammates. However, the lowest score noted that the employees usually involve themselves in discussions of various topics rather than specific topics with colleagues from other work teams.

50

Table 4.3

Knowledge Sharing via a 5-point Likert Scale

Survey Questions Min. Max. Mean SD

Intra1 I frequently share knowledge and information with my work teammates.

1 5 4.09 .732

Intra2 I usually involve myself in discussions of various topics rather than specific topics with my teammates.

1 5 4.09 .743

Inter1 I frequently share knowledge with people even though he (or she) is not in my team.

1 5 3.79 .800

Inter2 I usually involve myself in discussions of various topics rather than specific topics with colleagues from other work teams.

1 5 3.76 .849

Note: N=128; SD= Standard Deviation

Customer Relationship Management (CRM)

In the construct of customer relationship management exposed in table 4.4, the respondents had the highest agreement on CIM1, which indicated that their department has processes (or tools) to get in connection with potential customers using various channels (e.g., e-mail, customer service center, phone, FAX, face-to-face, etc.). However, the respondents showed the lowest agreement on WbM3 that noted that their department has processes (or tools) to evaluate the cost of wining back the lost customers.

51

Table 4.4

Customer Relationship Management via a 5-point Likert Scale

Survey Questions Min. Max. Mean SD

WbM1 Our department has processes (or tools) to win-back valued lost customer.

1 5 3.52 .784

WbM2 Our department has processes (or tools) to determine the value of lost customers.

1 5 3.49 .813

WbM3 Our department has processes (or tools) to evaluate the cost of wining back the lost customers.

1 5 3.41 .846

P/SC1 Our department has processes (or tools) to differentiate the customer acquiring efforts based on their value.

1 5 3.57 .781

P/SC2 Our department has processes (or tools) to provide customized product/service to customers based on their value.

1 5 3.76 .729

P/SC3 Our department has processes (or tools) to manage expectations of high valued customers.

1 5 3.67 .744

CIM1 Our department has processes (or tools) to get in connection with potential customers using various channels (e.g., e-mail, customer service center, phone, FAX, face-to-face, etc.).

1 5 3.92 .809

(continued)

52

Table 4.4. (continued)

Survey Questions Min. Max. Mean SD

CIM2 Our department has processes (or tools) to trace the status of our relationship with customers.

1 5 3.66 .855

CIM3 Our department has processes (or tools) to identify and affect different potential customers segments.

1 5 3.57 .810

CIM4 Our department has processes (or tools) to identify non-profitable customers.

1 5 3.55 .841

CIM5 Our department has processes (or tools) to evaluate the cost of retaining each customer.

1 5 3.50 .813

CIM6 Our department has processes (or tools) to capture and integrate customer data from whole contact points (email, call-center, web site, FAX, face-to-face, etc.).

1 5 3.79 .848

RefM1 Our department has processes (or tools) to manage customers’ referrals.

1 5 3.76 .781

RefM2 Our department has processes (or tools) to trace customer referrals.

1 5 3.74 .825

RefM3 Our department has processes (or tools) to motivate customers’ referrals.

1 5 3.64 .801

RefM4 Our department has processes (or tools) to reward our customers based on their referrals.

1 5 3.63 .762

(continued)

53

Table 4.4. (continued)

Survey Questions Min. Max. Mean SD

EM1 Our department has processes (or tools) to maintain up-sell activities (providing higher level service, or more customized service) with customers.

1 5 3.68 .822

EM2 Our department has processes (or tools) to maintain cross-sell activities (providing a variety of services) with customers.

1 5 3.70 .807

EM3 Our department has processes (or tools) to increase the sales from potential customers with high value.

1 5 3.63 .792

EM4 Our department has processes (or tools) to improve the relationship with potential customers with high value in order to provide cross-sell (providing a variety of services) and up-sell (providing higher level service, or more customized service) possibilities.

1 5 3.65 .838

TM1 Our department has processes (or tools) to finalize the relationship with non-profitable customer (e.g. the cost of maintaining customer relationships is too high to have profit).

1 5 3.56 .858

(continued)

54

Table 4.4. (continued)

Survey Questions Min. Max. Mean SD

TM2 Our department has processes (or tools) to orientate the non-profitable customer to terminate their relationship with our company (e.g. delay in service).

1 5 3.52 .823

Note: N=128; SD= Standard Deviation

Innovation Capabilities (IC)

In the construct of innovation capabilities, table 4.5 showed the respondents had the highest agreement on ProdI1, indicating that Pan Asia Human Resources Consulting Company often develops new services well accepted by the market. And the second highest was ProdI5, indicating that Pan Asia Human Resources Consulting Company has better capability in R&D of new services than its competitors. However, the respondents showed the lowest agreement on MI4, indicating that Pan Asia Human Resources Consulting Company doesn’t really put much emphasis on innovative and creative capability when recruiting staff.

55

Table 4.5

Innovation Capabilities via a 5-point Likert Scale

Survey Questions Min. Max. Mean SD

ProdI1 Our company often develops new services well accepted by the market.

1 5 3.89 .734

ProdI2 A great majority of our company’s profits are generated by the new services developed.

1 5 3.63 .850

ProdI3 The new services developed by our company always arouse imitation from competitors.

1 5 3.66 .818

ProdI4 Our company can often launch new services faster than our competitors.

1 5 3.76 .811

ProdI5 Our company has better capability in R&D of new services than our competitors.

1 5 3.88 .774

ProcI1 Our company often tries different operation procedures to hasten the realization of the company’s goals.

1 5 3.87 .736

ProcI2 Our company always acquires new skills or equipment to improve the manufacturing operation process.

1 5 3.81 .781

ProcI3 Our company can develop more efficient manufacturing process or operation procedure.

1 5 3.80 .797

ProcI4 Our company can flexibly provide products and services according to the demands of the customers.

1 5 3.85 .700

(continued)

56

Table 4.5. (continued)

Survey Questions Min. Max. Mean SD

ProcI5 The new manufacturing process or operation procedure employed by our company always arouses imitation from competitors.

1 5 3.61 .776

MI1 Our company will change the division of work among different departments according to the needs of market management.

1 5 3.81 .718

MI2 Our company’s department heads will adopt new leadership approaches to lead all staff towards task completion.

1 5 3.78 .752

MI3 The new financial management system adopted by our company can effectively monitor the actual discrepancy between our performance and our goals.

1 5 3.71 .775

MI4 Our company emphasizes innovative and creative capability when recruiting staff.

1 5 3.60 .854

MI5 The new performance assessment method adopted by our company can enable

department heads to gain a better picture of how far the staff has achieved the company goal.

1 5 3.67 .852

Note: N=128; SD= Standard Deviation

57

Discussion for Descriptive Statistics Analysis

From all the tables above in this chapter, the research conclude the top five and bottom two responses. In table 4.6, we can identify the top five responses that the employees of Pan Asia Human Resources Consulting Company agree on the most. Besides, in table 4.7, we can identify the least agree responses from the research questionnaire.

Table 4.6

Top Five Responses

Item Description Score

TiP2 I can Trust the people I work with to lend me a hand if I needed it. 4.18 TiP1 If I got into difficulties at work I know my colleagues would try and help

me out.

4.17

Intra2 I usually involve myself in discussions of various topics rather than specific topics with my teammates.

4.09

Intra1 I frequently share knowledge and information with my work teammates. 4.09 TiP3 Most of my colleagues can be relied upon to do as they say they will do. 4.09 Note: N=128

Table 4.7

Bottom Two Responses (N=128)

Item Description Score

WbM3 Our department has processes (or tools) to evaluate the cost of wining back the lost customers.

3.41

WbM2 Our department has processes (or tools) to determine the value of lost customers.

3.49

58

Correlation Analysis

Pearson coefficient correlation is used to test the direction and the strength of linear relationship between variables in this study. According to the result, trust, knowledge sharing, customer relationship management, and innovation capabilities are significantly correlated.

Besides, the result shows that the relationships between dependent variables are positively related to the independent variables.

In the pilot study, there wasn’t multicollinearity problem, however, as the growing sample size, it appeared severe multicollinearity problem after conducting main study. One of the possible reasons led to this result might because of the respondent distribution. Since the respondents of pilot study were collected only from two departments of Pan Asia, and the questionnaire design was suitable for them. However, when conducting the main study, the respondents were mainly from other departments rather than those two departments for pilot study.

Table 4.8.

Correlation among All of the Constructs

Constructs Trust KS CRM IC

Trust 1

KS .631** 1

CRM .619** .588** 1

IC .752** .599** .793** 1

Note: N=128; **p<.05; KS= Knowledge Sharing; CRM= Customer Relationship

Management; IC= Innovation Capabilities

59

Table 4.9.

Correlation among All the Variables

Variables 1 2 3 4 5 6 7 8 9 10 11 12 13

1 TiP 1

2 TiM .654** 1

3 Intra .673** .499** 1

4 Inter .454** .466** .659** 1

5 WbM .559** .548** .456** .456** 1

6 P/SC .556** .543** .536** .529** .790** 1

7 CIM .486** .518** .506** .480** .818** .822** 1

8 RefM .440** .515** .486** .474** .659** .705** .802** 1

9 EM .510** .531** .494** .472** .800** .794** .865** .804** 1

10 TM .516** .516** .501** .495** .763** .720** .841** .733** .812** 1

11 ProdI .594** .686** .512** .472** .716** .677** .685** .621** .714** .663** 1

12 ProcI .620** .741** .537** .495** .703** .696** .679** .592** .688** .649** .874** 1

13 MI .623** .662** .545** .543** .721** .663** .738** .638** .744** .711** .816** .850** 1 Note: N=128; **p<.05

60

PLS FINDINGS

Partial Least Square (PLS) is a family of alternating least squares algorithms, or

‘‘prescriptions,’’ which extend principal component and canonical correlation analysis (Henseler, Ringle, & Sinkovics, 2009). An important benefit of the PLS methodology is that it’s possible to disentangle direct and total effects of the variables included in the entire model (Cabrita & Bontis, 2008). By using PLS methodology, the finding result of this study is provided in this chapter. There are three sections in this chapter. In the first section, the more accurate reliability and validity of the instrument is illustrated. And in the second section, R square shows the explanatory power among all the variables, as well as the bootstrapping method was conducted to test both the significance of the path coefficients and hypothesis. In the finally section, the researcher do further investigation by comparison of different departments: the department of Dispatching Services, Foreign Workers Services, and Corporate Training; the department of Head-hunting, Education, and Pan-Asia (headquarter).

Research Instrument Validity and Reliabilities

To examine the validity and validity of the constructs in this study, the researcher use PLS to see the composite reliability, Cronbach’s alpha, Average Variance Extracted (AVE), as well as factor loading related to individual items.

To ensure the internal consistency, a minimum value of .70 for both composite reliability and Cronbach’s alpha is recommended (Nunnally, 2010). In Table 5.1. the composite reliability of Trust (.916), Knowledge Sharing (.889), Customer Relationship Management (.951), and Innovation Capabilities (.933); the Cronbach’s alpha of Trust (.818), Knowledge Sharing (.752), Customer Relationship Management (.938), and Innovation Capabilities (.893) are all

61

above .752.Accordingly, the internal consistency of all the constructs in this study is high, and it indicated that each of the designed questions accurately direct to each construct.

Average Variance Extracted (AVE) value should be greater than.50, which is the minimum requirement for justifying the variance captured by the indicators relate to the measurement error. In Table 5.1., the AVE values of Trust (.845), Knowledge Sharing (.800), Customer Relationship Management (.765), and Innovation Capabilities (.824) are all above .765.

Therefore, it can conclude that all the constructs meet convergent validity, and the relationship between measures and constructs are appropriate.

Table 5.1.

Main Study Constructs’ Reliability Analysis

Constructs Composite Reliability Cronbach Alpha AVE

Trust 0.916 0.818 0.845

KS 0.889 0.752 0.800

CRM 0.951 0.938 0.765

IC 0.933 0.893 0.824

Note: N=128; KS= Knowledge Sharing; CRM= Customer Relationship Management; IC=

Innovation Capabilities

Another method to assess individual item reliability, factor loading is used to examine how much shared variance between the constructs. According to Chin (1998), the acceptable criterion for factor loading is .50. In Table 5.2., all the values are greater than .831; therefore, all the constructs meet the required criteria. And it can be concluded that all the minimum requirement of reliability and validity tests in this main study are passed.

62

Table 5.2.

PLS Outer Loadings

Items Loadings Items Loadings Items Loadings

TiP 0.935 P/SC 0.876 ProdI 0.904

TiM 0.903 CIM 0.909 ProcI 0.929

Intra 0.914 RefM 0.831 MI 0.889

Inter 0.875 EM 0.904

WbM 0.847 TM 0.879

Note: N=128; TiP= Trust in Peers; TiP= Trust in Management; Intra= Intra-groups; Inter=

Inter-groups; WbM= Win-back Management; P/SC= Production/ Service Customization;

CIM= Customer Information Management; RefM= Referrals Management; EM= Expansion Management; TM= Termination Management; ProdI= Product Innovation; ProcI= Process Innovation; MI= Management Innovation

Testing the Measurement Model

The TKICC model of this research is assessed by using Smart PLS 2.0 software.In order to test the effects between each variables, the path coefficients and t-values are assessed.

R squared (R2) is used to examine the extent of explanatory power. R2 measures were originally developed for linear regression models with homoscedastic errors (Cameron &

Windmeijer, 1996). According to Chin (1998), in PLS path models, R2 values of 0.67, 0.33, and 0.19 is considered as substantial, moderate, and weak, respectively. In Table 5.3., the researcher provide the results of PLS internal consistency and R2 of this study. The results showed that Trust can explain 37.2% of Knowledge Sharing, Knowledge Sharing can explain 32.4% of Customer Relationship Management, and Knowledge Sharing can also explain 33%

63

of Innovation Capabilities. It conclude that all the variables have moderate explanatory power, and it also indicate their relevance in TKICC model.

Table 5.3.

PLS Internal Consistency and R2 in this study

Constructs No. of items Internal Consistency R2(%)

Trust 5 0.916 0.000

KS 4 0.889 0.372

CRM 22 0.951 0.324

IC 15 0.933 0.330

Note: N=128; KS= Knowledge Sharing; CRM= Customer Relationship Management; IC=

Innovation Capabilities

According to Hair et al (2011), individual path coefficients in the PLS structural model can be interpreted as the standardized beta coefficient in ordinary regressions. To determine the path coefficients significance value, a bootstrapping procedure using 40 random resampling of the original data was performed. The result of PLS testing in Table 5.4 demonstrate all path coefficients for variables are significant at the p-value < .001. The results showed that there is a positive significant effect between Trust to Knowledge Sharing (ß= .610, t= 5.025, p < .001), thus null hypothesis one was rejected. Knowledge Sharing has a positive significant effect to Customer Relationship Management (ß= .570, t= 4.580, p < .001), and Innovation Capabilities (ß= .574, t= 4.303, p < .001) as well. Therefore, null hypotheses two and three were both rejected. It can conclude that all of them have positive influences, and rejected all the null hypotheses of this study (see Table 5.5).

64

Table 5.4.

PLS Path Analysis Results (Standardized Beta Coefficients and Adjusted T-values)

Path Hypotheses β-path Adj. t-value Sig. Direction

TKS H1 0.610 5.025 *** +

KSCRM H2 0.570 4.580 *** +

KSIC H3 0.574 4.303 *** +

Note: N=128; ***p<.001

Table 5.5.

Research Hypothesis Results

Research Hypothesis Results

H1 Trust has no effect on Knowledge Sharing. Rejected

H2 Knowledge Sharing has no effect on Customer Relationship Management. Rejected H3 Knowledge Sharing has no effect on Innovation Capabilities. Rejected

The results of the multivariate test of the TKICC structural model are clearly demonstrate in Figure 5.1.

65

Figure 5.1. TKICC Structural Path

*p<.10. **p<.05. ***p<.001.

Trust

1. Trust in Peers 0.935***(34.604) 2. Trust in Management

0.903***(11.058)

3. Customer Information Management 0.909***(19.004)

66

Comparisons of Categorized Departments

According to 128 total questionnaires collected from respondents, the researcher conduct further investigation by comparison of different departments according to their similar functional properties. Six departments of Pan Asia were divided into two group, which are group A, department of Dispatching Services, Foreign Workers Services, and Head-hunting, and group B, department of Corporate Training, Education, and Pan-Asia headquarter. For the composition of group A, the function of the departments are related to manpower agency for various levels and nations jobs. And for the composition of group B, the function of the departments are related to human resource development. There were 86 respondents collected from group A (67.2 percent), while there were 42 respondents collected from group B (32.8 percent).

In Table 5.6., the responses from the employees in the department Dispatching Services, Foreign Workers Services, and Head-hunting (group A) show the composite reliability of trust (.907), knowledge sharing (.833), customer relationship management (.931), and innovation capabilities (.922); the Cronbach’s alpha of trust (.796), knowledge sharing (.606), customer relationship management (.911), and innovation capabilities (.874) are all above the acceptable criteria .70, only except the Cronbach’s alpha of knowledge sharing, indicating that it didn’t achieve the internal consistency.

The criteria of AVE value should be greater than .50, which is the minimum requirement for justifying the variance captured by the indicators relate to the measurement error. In Table 5.6., the AVE values of group A of trust (.830), knowledge sharing (.715), customer relationship management (.692), and innovation capabilities (.798) are all above the acceptable criteria.

67

And in Table 5.7., the AVE values of department Corporate Training, Education, and Pan-Asia headquarter (group B) of trust (.849), knowledge sharing (.901), customer relationship management (.767), and innovation capabilities (.802) are all above the acceptable criteria.

Thus, it can conclude that all the constructs meet convergent validity, and the relationship between measures and constructs in different departments are appropriate.

Table 5.6.

Group A Constructs’ Reliability Analysis

Constructs Composite Reliability Cronbach Alpha AVE

Trust 0.907 0.796 0.830

KS 0.833 0.606 0.715

CRM 0.931 0.911 0.692

IC 0.922 0.874 0.798

Note: N=86; KS= Knowledge Sharing; CRM= Customer Relationship Management; IC=

Innovation Capabilities

In comparison with the results of group A, Table 5.7 shows the responses from the employees in the group B. The composite reliability of trust (.919), knowledge sharing (.948), customer relationship management (.952), and innovation capabilities (.924); the Cronbach’s alpha of trust (.826), knowledge sharing (.890), customer relationship management (.939), and

In comparison with the results of group A, Table 5.7 shows the responses from the employees in the group B. The composite reliability of trust (.919), knowledge sharing (.948), customer relationship management (.952), and innovation capabilities (.924); the Cronbach’s alpha of trust (.826), knowledge sharing (.890), customer relationship management (.939), and

相關文件