Chapter 4 Research Methodology
4.6 Data Collection and Analysis Techniques
464 questionnaires were sent to the respondents working in the different departments such as manufacturing, marketing, technical and management in 116
54
companies. The questionnaire was distributed and collected by postal mail during 1 month. Because of time and budget constraints, the sample size was determined by the judgment method.
The analysis process was carried out in several steps: First, the data was coded and tested on the reliability and the validity of measurements of each constructs. Next, the descriptive analysis was conducted. Then the correlation analysis and regression analysis were performed. The correlations were performed to explore bi-variate relationships among variables. The regression analysis was used to test all research hypotheses. The Statistical Package for the Social Sciences (SPSS) computer software package was used for data analysis.
The multiple regression analysis can determine the effect of a set of independent variable on the dependent variable. The mediating variable was also examined by the regression analysis. In this research, the intervening variable of middle manager is the mediating variable. Mediation is a hypothesized causal chain in which the predictor variable affects an outcome variable and in turn affects the third variable.
According to Baron and Kenny (1986), there are four steps to test the research model which includes mediating variables. In order to test the mediation effect of middle manager, the analysis process has been conducted by three panels A, B and C in Figure 8 and corresponding with four steps.
Step 1: Conducting a simple regression analysis with independent variables to test dependent variable.
Step 2: Conducting a simple regression analysis with independent variables to test moderator variable.
Step 3: Conducting a simple regression analysis with moderator variable to test dependent variable.
55
Step 4: Conducting a multiple regression analysis with independent variables and moderator variable to test dependent variable.
Figure 3.Mediating variable analysis steps
56
Chapter 5 Findings and Discussion
5.1. Descriptive analysis
Respondents in this study are individuals, who work at 45 companies Taiwanese companies locating at some provinces and cities in Vietnam (see Table 5). Most companies where the valid respondent have been operating in the industrial and processing field such as precision, mechanic, electrical cable, about 88.9%. There were only 11.8% companies operating in electronic field. All the companies are small-medium companies with the number of employees from 120 to 1300 persons.
Among 182 of feedbacks, there were 172 valid respondents, equivalent 38% total of sent out questionnaires. The analysis was conducted based on valid respondents.
Table 5
Location of surveyed companies City/Province Number of
company
Number of respondent
City/Province Number of company
Number of respondent
Ha Noi 2 8 Ho Chi Minh 12 45
Bac Giang 3 12 Dong Nai 14 53
Ha Tay 1 4 Binh Duong 11 42
Thai Binh 2 8
In terms of gender, 61% of respondents were male and 38% of them were female. Among 172 respondents, 17.4% were top managers, 18.0% were middle managers and the rest were employees. As for the working department, the technical staffs were 19.8%, manufacturing staffs were 30.8%, purchasing and sales staffs were 19.2%, and the only 8.7% staffs were working in production
57
research and development. All respondents working in general management were top managers and middle managers.
Table 6 shows that respondents have work experiences of more than 2 years, 26.2% have had experiences less than 2 years and only 5.8% respondents have been working more than 10 years and all of them are Taiwanese managers.
Table 6
Profile of respondents
Demographic variables Frequency Percentage
Gender
Product research and development Purchase and sale More than 10 years
45
58
Most of respondents agreed that employees’ ability to learn and to apply knowledge from the parent firms was at medium level. Knowledge of management and manufacturing were 2 types of knowledge (mean=3.23; 3.27; 3.22; 3.18, SD= .854; 0.872; .771; .883) which the respondents can learn and tend to apply in daily work much more than the other types of knowledge. This result has suggested that there were existing factor influencing and moderating the knowledge process from the parent firms to subsidiaries overseas (see Table 7).
Table 7
Mean results of degree of knowledge transfer
Scale items Mean SD
Leaning of marketing knowledge 2.92 .776
Learning of manufacturing knowledge 3.23 .854
Learning of purchasing knowledge 3.01 .697
Learning of R&D knowledge 2.88 .727
Learning of managing knowledge 3.27 .872
Applying of marketing knowledge 3.04 .695
Applying of manufacturing knowledge 3.22 .771
Applying of purchasing knowledge 2.97 .764
Applying of R&D knowledge 2.95 .748
Applying of managing knowledge 3.18 .883
Note. Scale measurements are from 1 (very low) to 5 (very high), n = 172
Table 8 shows the means and standard deviations of independent variables in this research. All variables of characteristics of knowledge source, characteristics of knowledge recipient and relationships between parent firms and subsidiaries in Vietnam had good variance, with variable means generally in the middle of their range.
59
Table 8
Mean results of independent variables
Scale items Mean SD
Characteristics of Knowledge Source
New knowledge is difficult to understand 3.63 .733 Difference between existing and new knowledge 3.26 1.063 New knowledge is difficult to apply in daily work 3.55 .945 Knowledge is hard to encode and write down in manuals 3.21 1.016 Expatriate’s comprehension of host country culture 4.01 .612 Expatriate’s Vietnamese speaking ability 3.73 .693 Expatriate’s experience of training/ transfer knowledge 4.00 .582 Expatriate’s willingness to train/transfer knowledge 4.00 .582 Characteristics of Knowledge Recipients
Employees’ understanding and assimilating of parent firm’s knowledge
3.91 .633
Employees’ learning and applying of parent firm’s knowledge 4.01 .692
Subsidiary’s training plan 3.93 .645
Subsidiary’s encouragement policy of learning 3.94 .743 Parent Firms and Subsidiaries Relationship
Dependence of subsidiary’s marketing technique on parent firm 3.48 .841 Dependence of subsidiary’s manufacture technique on parent
firm
3.65 .813
Dependence of subsidiary’s purchasing technique on parent firm
3.57 .925
Dependence of subsidiary’s R&D technique on parent firm 3.60 .870 Dependence of subsidiary’s managing procedure on parent firm 3.62 .860 Difference of language between parent firm and subsidiary 3.47 1.051 Difference of political, social, economic system between host
countries and parent firm’s country
3.51 1.084
Difference of organization culture between parent firm and subsidiary
3.32 1.058
Frequency of connection between parent firm and subsidiary 3.84 .723 Note: Scale range from 1 (strongly disagree) to 5 (strongly agree), n = 172
60
Most respondents highly agreed with the statement that the middle managers play the role as knowledge translator, knowledge transfer facilitator and
communication enhancer (see Table 9). The means and standard deviations of variables present that all variables had good variance with the variable means generally in the relatively high of their range.
Table 9
Mean results of middle manager’s role
Scale items Mean SD
Receiving and translating knowledge materials from parent firm 4.00 .507 Editing and recording knowledge materials from parent firm 3.87 .558 Actively helping employee in learning process 3.94 .594 Pushing employee’s motivation in learning 3.95 .517 Making sense and dealing with the conflicting issues 3.92 .535 Sustaining momentum in team work as critical friend, supervisor and
advisor
3.92 .535
Sustaining the negotiation, discussion processes 3.92 .572 Actively supporting communication between knowledge transmitters
and recipients
3.95 .539
Actively building and maintain the cooperative environment within individual, group and organization
3.95 .539
Note. Scale are from 1 (strongly disagree) to 5 (strongly agree), n = 172
5.2. Validity and reliability results
The research uses the factor analysis and the reliability analysis to assess the measure of all variables. The reliability is defined as dependability and consistency of the measure of a variable. The validity refers to the truthfulness of the relation
between idea and actual reality. The validity analysis result can give the answer to the
61
question how well is the reality being measured through research matches with research construction (Neuman, 2006).
Factor analysis was used to test the validity of each constructs of conceptual model base on the collected data. The study applied the varimax rotation method. The result in Table 10, except knowledge absorptive capacity variable with the KMO = .5 other variable’s KMO is from .7 to .8. This result presents that the sampling adequacy is acceptable to determine the appropriateness in the factor analysis. Table 10 also shows that p- value of the Bartlett’s test of sphericity is equal to .000, therefore according to Hair (2006), the conceptual model is statistically significant and can be conducted for further analysis.
Table 10
KMO and Bartlett’s for variables
Variable KMO
measure
Approx.
Chi –Square Df Sig.
Degree of knowledge transfer .818 924.151 45 .000
Knowledge character .755 217.694 6 .000
Knowledge expatriate .716 239.801 10 .000
Knowledge absorptive capacity .500 90.418 1 .000
Learning capacity .718 257.873 6 .000
Parent firms-subsidiaries relationship .829 471.524 10 .000
Communication environment .707 340.614 6 .000
As presented in Table 11, the result of testing the validity and the reliability of measurement of constructs show that all of the measures of factors have the
62
Cronbach's coefficient alpha greater than .7. According to Kim, all items of the study were considered internally consistent (陳正昌, 程炳林, 陳新豐, & 劉子鍵著, 2009).
5.3. Correlation analysis
Pearson correlation analysis was performed on the independent and dependent variables. Pearson correlation coefficients were calculated for each pair of variables.
The result shown in Table 12 indicates that the degree of knowledge transfer is significantly associated with knowledge absorptive capacity (r = .164, p < .01) and learning capacity (r= .258, p< .001). The degree of knowledge transfer is also associated with the middle manager as translator (r= .321, p < .01), as a facilitator (r
= .241, p< .01, and as a communication enhancer (r = .159, p< .05).
However, the degree of knowledge transfer is not significantly associated with the parent firms’ knowledge characters and knowledge expatriate. And the
communication environment negatively and not significantly correlates with the degree of knowledge transfer.
63
Table 11 Results of reliability of factors Factors Items LoadingAlpha Degree of knowledge transfer
Leaning of marketing knowledge.864 .870 Learning of manufacturing knowledge.856 Learning of purchasing knowledge.853 Learning of R&D knowledge.861 Learning of managing knowledge.856 Applying of marketing knowledge.848 Applying of manufacturing knowledge.862 Applying of purchasing knowledge.851 Applying of R&D knowledge.843 Applying of managing knowledge.856 Charateristic of knowledge source
New knowledge is difficult to understand .631 .705 Difference between existing and new knowledge.640 New knowledge is difficult to apply in daily work.659 Knowledge is hard to encode and write down in manuals.650 Expatriate’s comprehension of host country culture .710 Expatriate’s Vietnamese speaking ability.667 Expatriate’s experience of training/ transfer knowledge.682 Expatriate’s willingness to train/transfer knowledge.682 Charateristic of knowledge recipients
Employees’ understanding and assimilating of parent firms’ knowledge.761 .802 Employees’ learning and applying of parent firms’ knowledge.781 Subsidiary’s training plan .750 Subsidiary’s encouragement policy of learning .787 (continued)
64
Factors Items Loading Alpha Parent firms- subsidiaries relationship
Dependence of subsidiary’s marketing technique on parent firm.727 .781 Dependence of s subsidiary’s manufacture technique on parent firm.743 Dependence of subsidiary’s purchasing technique on parent firm.737 Dependence of subsidiary’s R&D technique on parent firm.744 Dependence of subsidiary’s managing procedure on parent firm.743 Difference of language between parent firm and subsidiary.770 Difference of political, social, economic system between host countries and parent firm’s country.764 Difference of organization culture between parent firm and subsidiary.754 Frequency of connection between parent firm and subsidiary.770 Middle manager’s roleReceiving and translating knowledge materials from parent firm.893 .899 Editing and recording knowledge materials from parent firm.884 Actively helping employee in learning process .884 Pushing employee’s motivation in learning.895 Making sense and dealing with the conflicting issues .883 Sustaining momentum in team work as critical friend, supervisor and advisor.891 Sustaining the negotiation, discussion processes .892 Actively supporting communication between knowledge transmitters and recipients .886 Actively building and maintain the cooperative environment within individual, group and organization .891 Receiving and translating knowledge materials from parent firm.881
65
Table 12 Correlation analysis Variable1234567891011121314151617 KS.01 KC- .07.844** EX.11.566**.04 KR.239**.276**.00.521** KA.164*.245**.01.444**.881** LC.258**.245**- .02.479**.892**.572** SPD.233**.164*-.04.370**.459**.432**.803** CE-.12.544**.646**.02.09.09.707**.15 MT.321**-.212**-.386**.197**.408**.312**.06.350**.833** MF.241**-.233**-.428**.223**.526**.517**.07.351**.895**.625** MC.159*-.169*-.347**.219**.336**.378**-.03.202**.843**.542**.641** MTKC.195*.800**.880**.14.206**.170*.420**.14-.06.10- .14.845** MTEX.257**.304**-.183*.847**.607**.497**.240**.464**.636**.689**.504**.579**.158* MFKA.239**-.04-.282**.359**.769**.820**.221**.440**.824**.566**.914**.185*-.01.951** MFLC.294**-.04-.298**.393**.806**.639**.209**.432**.814**.631**.889**.215**.00.962**.897** MCSPD.259**.08-.161*.398**.517**.509**.685**.936**.607**.497**.533**.279**.08.599**.598**.813** MCCE.280**.469**.483**.13.260**.05.707**.250**.07-.05.02.454**.498**.178*.15.744**.293** **. Correlation is significant at the .01 level (two-tailed). *. Correlation is significant at the .05 level (two-tailed).
66
5.4. Hypotheses testing
The regression analysis was applied to the moderation model. As mentioned in chapter 3, the statistical testing of research model in panel A, B and C hypothesis was conducted. The research also conducted the calculation of variable inflation factor (VIF) for each of the independent variables in the regression models. The results show that the VIF of independent variable in all regression models is from 1.059 to 3.861.
According to Hair (2006), the VIF which is less than to 10 can be acceptable. Thus, there is no problem of multicollinearity for independent variables.
5.4.1. Characteristics of knowledge source and degree of knowledge transfer Table 13 shows although the result indicate that knowledge characteristics positively impact on degree of knowledge transfer but the result was not statistically significant (R2= .004, F = .743, p> .05). The results did not support the hypothesis H1.
And as the same as knowledge characteristics variable, knowledge expatriate variable which were measured by items such as the comprehension of host country culture, Vietnamese speaking ability, experience of training/transfer knowledge , willingness to train/transfer knowledge is not significant associated with degree of knowledge transfer. The result fails to support hypothesis H2 (see Table 13).
67
Table 13
Result of characteristics of knowledge source, parent firms- subsidiaries relationship, characteristics of knowledge recipient and degree of knowledge transfer
R2 Adj. R2 SE F df1 df2 Sig. F Durbin-W atson Knowledge
Characters .004 -.002 5.283 .743 1 170 .390 1.798 Knowledge
Expatriate .012 .007 5.262 2.146 1 170 .145 1.807 Subsidiary
Dependence .054 .049 5.149 9.787 1 170 .002 1.866 Communication
Environment .015 .010 5.254 2.650 1 170 .105 1.800 Knowledge
Absorptive Capacity
.027 .021 5.224 4.670 1 170 .032 1.834
Learning capacity .066 .061 5.116 12.093 1 170 .001 1.858 Note. Independent variable: Degree of Knowledge transfer
* p< .05 **p< .01 ***p< .001
5.4.2. Parent firms- subsidiaries relationship and degree of knowledge transfer The value Durbin-Watson 1.856, between 1.5 and 2.5, and the Adj. R2 = .049, F
= 9.787, p < .05 shows that the predictability of subsidiary dependence from parent firms for all type knowledge which parent firms often transfer to their subsidiary was significant. The result supports hypothesis H3 (see Table 13).
In contrast of the subsidiary dependence, the result in Table 13 suggest that communication environment variable which is measured by the difference of
language, difference of political, social, economic system, difference of organization culture and frequency of connection between parent firms and subsidiary as
significantly associated with the degree of knowledge transfer with Adj.R2= .010, F=
68
2.650, p> .05. This result show that the hypothesis H4 was not supported.
5.4.3. Characteristics of knowledge recipients and degree of knowledge transfer The statistical results presented in Table 13 suggest that both knowledge
absorptive capacity of employee and subsidiary’s learning capacity strongly and positively affects the degree of knowledge transfer. The value of Durbin- Watson = 1.843, between 1.5 and 2.5, the Adj. R2= .021 show that 2.1% of variance in degree of knowledge transfer can be predicted from variables employees’ understanding and assimilating of parent firms’ knowledge and employees’ learning and applying of parent firms’ knowledge combined. The F= 4.670 and is statistically significant with p< .05 indicate that knowledge absorptive capacity significantly influences the degree of knowledge. On the other hand, the result also shows that the learning capacity attributes is significant (Adj. R2= .061, F= 5.116 and p< .001) and suggest that the learning capacity is positively associated with the degree of knowledge transfer. This result supports the hypothesis H5 and H6.
5.4.4. Middle manager’s role
The results presented in Table 14, 15, and 16 suggest that there are significant relationships among the three independent variables (i.e., the characteristics of knowledge source, the characteristics of knowledge recipient, and the parent
firms-subsidiaries relationship), and the mediating variable (i.e., the middle managers).
The middle managers positively relate to the characteristics of knowledge source as the role as translator (Adj. R2= .272, F =64.891, p< .001), to the characteristics of knowledge recipients as the role as facilitator (Adj. R2= .237, F=27.519, p< .001), to the parent firms-subsidiaries relationship as the role as communication enhancer
69
(Adj.R2= .281, F= 34.393, p< .001).
Table 14
Result of characteristics of knowledge source and middle manager as translator
Variable R2 Adj. R2 SE F df1 df2 Sig. F Note. Independent variable: middle manager as translator
* p< .05 **p< .01 ***p< .001
Table 15
Result of parent firms-subsidiaries relationship and middle manager as communication enhancer Note. Independent variable: middle manager as communication enhancer
* p< .05 **p< .01 ***p< .001
Table 16
Result of characteristics of knowledge recipients and middle manager as facilitator
Variable R2 Adj.
Note. Independent variable: middle manager as facilitator
* p< .05 **p< .01 ***p< .001
70
5.4.5. Middle managers and degree of knowledge transfer
The result presented in Table 17 suggests that both three roles of middle
managers have positive impact on the degree of knowledge transfer as translator (Adj.
R2= .103, F =19.589, p< .001), as facilitator (Adj. R2= .058, F= 3.367, p< .001 and as communication enhancer (Adj. R2= .025, F =4.255, p< .05). Comparing with the role as communication enhancer and knowledge transfer facilitator, the effect of the role as translator is significantly higher.
Table 17
Result of middle managers’ roles and degree of knowledge transfer
Variable R2 Adj.
R2 SE F df1 df2 Sig. F
Durbin-Wa tson Translator .103 .098 5.014 19.589 1 170 .000 1.956
Facilitator .058 .053 5.139 10.502 1 170 .001 1.857 Communication
Enhancer .025 .020 5.227 4.425 1 170 .037 1.883 Note. Independent variable: degree of knowledge transfer
* p< .05 **p< .01 ***p< .001
71
Table 18
Summary of simple regression
Variables
Degree of Knowledge
transfer Middle Manager's Role Adj.
72
5.4.6. Characteristics of knowledge source, recipients, parent firms-subsidiaries relationship and middle manager’s role and degree of knowledge transfer
The statistical result in Table 19 present the significant relationship between the characteristics of knowledge source and the degree of knowledge transfer with the intervention of middle manager as translator (Adj.R2= .098, F= 10.341, p< .001).
However, they affect indirectly and significantly the degree of knowledge transfer . The result shows that the parent firms- subsidiaries relationship through the middle manager as communication enhancer significant influenced on the degree of knowledge transfer (Adj. R2= .023, F =3.017, p< .05) (see Table 19). The effect of middle manager also presented in the relation with characteristics of knowledge recipients and degree of knowledge transfer. The results suggest that the intervention of managers as facilitator combining with the absorptive and learning capacity
contribute significantly to degree of knowledge transfer with Adj. R2= .065, F= 6.904, p< .001 (see Table 19). These results present a support to the hypothesis H7, H8, and H9.
Table 19
Result of characteristics of knowledge source, middle manager and degree of knowledge transfer
R2 Adj.
source, Middle manager as translator
.109 .098 5.013 10.341 2 169 .000 1.965
Parent firms - subsidiaries relationship, communication enhancer
.035 .023 4.208 3.017 2 169 .000 1.904
Characteristics of Knowledge
Recipients, Facilitator .076 .065 5.106 6.904 2 169 .001 1.886 Note. Independent variable: degree of knowledge transfer
73
* p< .05 **p< .01 ***p< .001 Table 20
Summary of hypothesis testing
Hypothesis Contents Result
H1
The lower the complexity and the tacitness of knowledge of parent firms, the higher the degrees of
knowledge transfer to subsidiaries is expected Not supported
H2
The higher of experience of expatriate, the higher degree of knowledge transfer to subsidiaries is
expected Not supported
H3
The better the communication between on between knowledge transmitters and recipients, the higher the degree of knowledge transfer to the subsidiary is expected.
Not supported
H4 The more dependence of subsidiary on parent firms,
the higher degree of knowledge can be expected Supported H5 The more the employee’s absorptive capacity
increase, the higher the degree of knowledge transfer Supported H6 The more the building learning capacity, the higher
the degree of knowledge transfer is expected Supported
H7
The more the subsidiary adopts middle manager as translator, which affects characteristics of knowledge source, the higher degree of knowledge transfer to the subsidiary is expected
Supported
H8
The more the subsidiary adopts middle manager as communication enhancer, which affects relationship between parent firm and subsidiary, the higher degree of knowledge transfer to the subsidiary is expected
Supported
H9
The more the subsidiary adopts middle manager as facilitator, which affects characteristics of knowledge recipient, the higher degree of knowledge transfer to the subsidiary is expected
Supported
74
5.5. Discussion
This study proposed and conducted the testing of a model linking knowledge transfer’s determinants, the role of middle manager and the degree of knowledge transfer in the Taiwanese FDI companies in Vietnam. Based on the suggestion of Minbaeva (2007), Martins and Antonio (2010) on the relationship between knowledge transfer process factors and the effectiveness of knowledge transfer, a research model and hypothesis were presented to address the effect of the characteristics of the knowledge source, the characteristics of knowledge recipients and the relationship between transmitters and recipients on the degree of knowledge transfer.
Parent firms’ expectations of high degree of knowledge transfer to the
subsidiaries are not always achieved. Most respondents agreed that knowledge which they receive from the parent firms and then apply in their daily task is limited.
However, most of them agreed that the tacitness and the differences of the knowledge characteristics are not main factors affecting the low degree of knowledge transfer.
The analysis result showed that knowledge characteristics positively and significantly correlated with all the other independent variable except the degree of knowledge
The analysis result showed that knowledge characteristics positively and significantly correlated with all the other independent variable except the degree of knowledge