Chapter 4 Research Methodology
4.1 Research Setting
In this thesis, the research design, data collection, and data analysis were devised to enable an exploratory test to be made to study the sources and consequences of trust in the Malaysian Chinese community.
In this thesis, two separate tests were carried out at country as well as ethnic level.
The Western, Japan and Malaysia were included at the country level; on the other hand, Indian, Chinese and Native (Malay) were included for the ethnic level within-Malaysia.
The research attempts to study the differences in sources and consequences of trust relationship of Malaysian Chinese not only within-Malaysia but also in a cross-country setting to allow us to determine which factors influence the sources and consequences of trust as well as which are country specific or ethnic specific.
In this thesis, Japan was chosen because it is the largest import partner of Malaysia as described in the earlier section. As for Western, it was chosen because United States is the largest export partner of Malaysia as described in the earlier section. In addition, Malaysia, once upon a time, was one of the former British colonies. Given close relationship between Malaysia and West countries through the abovementioned economic exchanges relationship and membership of British colonies, Malaysia is in the best position to utilize its social capital to attract mutually beneficial economic activities.
Second, these two region countries have been especially aggressive in capitalizing on overseas market opportunities, particularly in the Asia–Pacific region such as Malaysia.
In addition, there is a large cultural difference exists between the two region countries, for
example, they both have differences in language, consumer attitudes, law and regulations, and business practices. Therefore, making such a selection allows this study to focus on the role of cultural dissimilarity as a key factor in international transactions.
The research frame is showed in the figure 1.
Figure 1 : Research Framework
Sources of Trust:
Economic-based of trust
Process- based of trust
Characteristic-based of trust
Institutional-based of trust
Consequences of Trust:
Contract adherence
Flexibility beyond contract
Flexibility outside contract
Control variables:
Firm size
Respondents’ year of working experience
30 4.2 Sample and Data Collection
In this chapter, sample, data collection and analytic methods will be explained and to show the reliability of the construct.
As it was mentioned in the earlier section, the research examined the Malaysian Chinese’s internationalization experiences, both in term of sources of trust as well as consequences of trust. Both country level and ethnic level tests provide an ideal context for a study such as this.
A two-stage research design was employed. The first stage involved in-depth interview with several Malaysia Chinese-businessmen (whom were currently involved in international market). The purpose of this phase was to gain a first-hand understanding of the factors that motivated the “production of trust” or so called the sources of trust as well as the consequences of trust.
Stage 2 involved a cross-sectional survey of the firms in Malaysia and from this resultant sampling frame, 86 firms from the list agreed to participate in the survey were involved in international transactions or market. Using the pre-administered Chinese questionnaire as the point of departure, the same was translated into English and thus, e-mails containing both pretested Chinese and English questionnaires were mailed to all the aforementioned 86 firms in Penang, a state of Malaysia. Penang was chosen because it was experiencing a very huge sectoral transformation from the last few decades. Manufacturing is the most important component of the Penang economy, contributing 45.9% of the State's GDP (2000). It is highly industrialized with high-tech electronics plants (such as Dell, Intel,
AMD, Altera, Motorola, Agilent, Hitachi, Osram, Plexus, Bosch and Seagate) located within the Bayan Lepas Free Industrial Zone.
Of the 86 firms who agreed, 63 questionnaires were returned, however, usable responses were only 53(61.63% response rate). The rest of the 10 questionnaires were dropped due to the incomplete information. The data collection was done between April and May 2009.
The sample had a wide variety of the nature of business, such as legal consultant/services and manufacturing industry ranging from simple technology products ( such as frozen food, air-grilled system ), through medium technology products ( such as capacitators, voltage transformers, electro-chemical plant equipment, computer), to higher technology products (such as sonic products). The simple parameter used to distinguish the size of the firm was the total number of employees for that particular firm in Malaysia and the cut-off point is 1000. In this thesis, 34 of the sample firms had less than 1000 employees; on the other hand, 19 of the firms were with more than 1000 employees.
In the email, the covering letter attached herein directed the questionnaire to the senior executive officer or key person that responsible for the firm’s international transaction or market. If they were not so involved, it was requested that it be forwarded to the superiors. Thus, the presence of a single underlying organizational trait or construct was assumed rather than multiple individual realities.
From a methodological viewpoint, relying on a single person (key informant) for a study such as this has advantages and disadvantages (Philips, 1981). A question may arise whether a single respondent from each firm has sufficient knowledge and ability to assess
32
the questionnaires. Although responses from multiple informants would have been preferred, however, two problems were identified in using multiple informants such as the selection problem as well as the perceptual agreement problem (Kumar, Stern and Anderson, 1993).
In the selection problem, response errors were likely to be higher for informants whose roles were not closely associated with the phenomena under study. However, in the perceptual agreement problem, the responses have often failed to demonstrate high levels of perceptual agreement and disagreements may arise because of differences in both knowledge and perceptions (Kumar, Stern and Anderson, 1993).
In view of the aforementioned problems, we believe that our respondents were well positioned to make this assessment for the several reasons in explaining our decisions not to use multiple informants. First, because informants were asked to report on their firms' relationships with other firms, it is vital that a tenure question should focus on how long informants have interacted with the others, not on the length of their current employment. Hence, our informant's involvement with the other both foreign and local firms and the length of time the informant has interacted with the other firms has taken into consideration (Kumar, Stern and Anderson, 1993).
In our research, the respondents had been employed for at least 3 years and had primary responsibility for managing the day-to-day relationship with both local and foreign customers or counterparts and they were well aware of the variety of interactions within the local as well as the foreign. Thus, they had a quite considerable working experience and knowledgeable enough to be asked about the level of trust in the firm's relationship with others.
4.3 Measurements of variables
In this research, constructs were the hypothetical variables that were being measured and they were the four sources and three consequences of trust. The items (questions) were generated based on the literature mentioned in the earlier chapter and each item was a statement followed by a seven-point Likert scale. Each respondent was asked to rate each item on one to seven scales where one represented strongly disagree, four represented neutral and seven represented strongly agree.
In addition to the aforementioned items, a number of demographic measures, such as firm size, years of establishment, respondents’ year of working experience as well as job position, were used to profile the firms and respondents that participated in our study.
In this research, the reliability and validity tests were especially important to determine if the items were a reliable measure of the constructs prior to the subsequent analyses. Hence, to ensure that the items(questions) asked related to the underlying construct(sources and consequences of trust) that we intended to measure and to determine the internal consistency or average correlation of items in the survey to gauge its reliability, Cronbach’s alpha (Cronbach, 1951)for each construct was obtained through SPSS software (Statistical Package for the Social Sciences). The reliability of those constructs in this research was satisfactory in view of the Cronbach alpha was all above 0.6 in general as shown in the table 1, higher than 0.50 for exploratory studies (Hair et al., 1995).
Therefore, it was an acceptable value for a research instrument indicating high reliability and the inter-item correlations were high in measuring the same underlying construct.
34
Furthermore, factor command was used to check the dimensionality of the data and each construct was with eigen value of more than 1.0. According to Hair et al. (1992), the most commonly used method of determining whether items are loading on one construct was the latent root criterion (eigen values). Only the factors having eigen values greater than 1 were considered significant; those with eigenvalues less than 1 were considered insignificant and were disregarded.
In summary, the items for measuring the constructs (sources and consequences of trust) were reliable and valid. The data obtained can be used in subsequent data analysis.
According to the literature we mentioned in the earlier chapter, the sources of trust were constructed into economic-based trust, process-based trust, characteristic-based trust and institutional-based trust.
Economic-based trust was operationalized by the following sub measures:
The inter-dependency level between both parties.
The input level of various resources (technology, capital, equity involvement)
The size/scale of the counterpart’s company/organization/corporation.
The cronbach alpha for this construct was 0.74 and eigen value was 1.98, indicating high reliability as shown by the Table 1.
Process-based trust was operationalized by the following sub measures:
The reputation/fame of the counterparts within the same industry.
The satisfaction levels of the previous transaction experiences.
The cronbach alpha for this construct was 0.60 and eigen value was 1.44, indicating high reliability as shown by the Table 1.
Characteristic-based trust was operationalized by the following sub measures:
Former acquaintance/personal relationship prior to the establishment of the contractual relationship.
The counterparts share a similar background (i.e. former classmate, former colleague/co-worker, social group/community organization etc.)
Contractual relationship established through a third-party recommendation, which is well acquainted with both parties
The qualification/experience/seniority/execution style of the personnel in counterpart’s company
The cronbach alpha for this construct was 0.76 and eigen value was 2.34, indicating high reliability as shown by the Table 1.
Institutional-based trust was operationalized by the following sub measures:
The counterpart’s recognition level in the business world (i.e. ISO Certification, IPO etc.)
The willingness of the industry’s player to abide by the law based upon the geographical area
The counterpart’s country’s level of law enforcement
The cronbach alpha for this construct was 0.79 and eigen value was 2.11, indicating high reliability as shown by the Table 1.
36
Table 1: Measurement of the sources of Trust
Variables Operational items
Economic-based The inter-dependency level between both parties.
The input level of various resources (technology, capital, equity involvement)
The size/scale of the counterpart’s company (organization corporation)
0.74 1.98
Process- based The reputation/fame of the counterparts within the same industry.
The satisfaction levels of the previous transaction experiences.
0.60 1.44
Characteristic-based Former acquaintance/personal relationship prior to the establishment of the contractual relationship.
The counterpart shares a similar background (i.e. former classmate, former colleague/co-worker, social group/community organization etc.)
Contractual relationship established through a third-party recommendation, which is well acquainted with both parties
The qualification, experience, seniority, execution style of the personnel in counterpart’s company
0.76 2.34
Institutional-based The counterparts’ recognition level in the business world (i.e ISO Certification, certification, IPO, etc.)
The willingness of the industry’s player to abide by the law based upon the
geographical area
The counterpart’s country’s level of law enforcement
0.79 2.11
On the other hand, within the literature, the consequences of trust were contractual adherence, flexibility beyond contract and relationship outside contract.
Contractual adherence was operationalized by the following sub measures:
I will strictly follow the contract based upon the exact terms and conditions.
I will definitely maintain company trade secrets and confidences.
The cronbach alpha for this construct was 0.83 and eigen value was 1.71, indicating high reliability as shown by the Table 2.
Flexibility beyond contract was operationalized by the following sub measures:
If abrupt changes occurred, I will remain flexible and adjust to the changes by modifying the contract to fit the needs
I will allow my counterpart to make certain judgment calls, if expediency is
necessary.
If the contract needs to be amended, I will follow the conventional rules and
regulations.
The cronbach alpha for this construct was 0.71 and eigen value was 1.90, indicating high reliability as shown by the Table 2.
Relationship outside contract was operationalized by the following sub measures:
Aside from the agreements made within the contract, any verbal agreements made will also be honored.
I value long-term relationships over contracts that are solely based upon profit with a counterpart that is unfamiliar to me.
38
I will voluntarily provide important market information to my counterparts
The chronbach alpha for this construct was 0.67 and eigen value was 1.83, indicating high reliability as shown by the Table 2.
Table 2: Measurement of the consequences of trust
Variables Operational items
adherence I will strictly follow the contract based upon the exact terms and conditions.
I will definitely maintain company trade secrets and confidences.
0.83 1.71
Flexibility beyond contract
If abrupt changes occurred, I will remain flexible and adjust to the changes by modifying the contract to fit the needs
I will allow my counterpart to make certain judgment calls, if expediency is necessary.
If the contract needs to be amended, I will follow the conventional rules and regulations.
0.71 1.90
Relationship outside contract
Aside from the agreements made within the contract, any verbal agreements made will also be honored.
I value long-term relationships over contracts that are solely based upon profit with a counterpart that is unfamiliar to me.
I will voluntarily provide important market information to my counterparts
0.67 1.83
4.4 The Empirical Model
Subsequently, multiple regression analysis was conducted in order to examine how each of the four sources of trust may explain on the consequences of the trust and the impact of various simultaneous sources of trust upon each of the consequences of trust.
Four independent variables and three dependent variables were used for this empirical study in which the sources of trust were independent variables, however, the consequences of trust were dependent variables. On top of that, two control variables in this study were firm size and respondents’ year of working experience.
The linear OLS regression model for each relationship between sources and consequences of trust was being written as following:-
Model 1:
Contract adherence = β0 + β1ECON + β2PRO + β3 CHRC + β4INST + β5logSIZE + β6YEAR +εi
Model 2:
Flexibility beyond contract = β0 + β1ECON + β2PRO + β3 CHRC + β4INST + β5logSIZE + β6YEAR + εi
Model 3:
Relationship outside contract = β0 + β1ECON + β2PRO + β3 CHRC + β4INST + β5logSIZE + β6YEAR + εi
where β0 through β6 are the coefficients of the model, ECON is the Economic-based trust
PRO is the Process-based trust
40 CHRC is the Characteristic-based trust INST is the Institutional-based trust SIZE is the firm size
YEAR is the respondents’ year of working experience εi is the error term.
Chapter 5 Data Analysis
5.1 Inter-country Analysis
As we mentioned in the chapter of introduction, the main purpose of this study is to examine how the Malaysian Chinese’s economic behavior responds to the trust building process in inter-firm relationships and to explore how the sources and consequences of trust differ among international transacting parties of various nationalities and within-Malaysia ethnic groups. Does the trust building process differ significantly among the parties from various nationalities and ethnical groups, whether it is inter-country or inter-ethnic discrimination?
Hence, basically, both macro analysis and microanalysis were conducted through MANOVA analysis. The former is to see how trust differs at the inter-country level, however, the latter is to see how it differ at the inter-ethnic level within Malaysia itself. The simple descriptive statistics of the variables are shown in Table 3 as following:-
Table 3. Table of Descriptive Statistics (Note: N= 265)
Variables Minimum Maximum Mean Std.
Deviation Contract adherence 1.0000 7.0000 5.5340 1.3970 Flexibility beyond contract 1.0000 7.0000 5.1799 1.1374 Flexibility outside contract 1.0000 7.0000 4.6893 1.3002 Economic-based of trust 1.0000 7.0000 5.2943 1.1513 Process-based of trust 1.0000 7.0000 5.7075 1.1425 Characteristic-based of trust 1.0000 7.0000 4.8358 1.1394 Institutional-based of trust
42
Table 4. The MANOVA table of the inter-country analysis
Note:
a. BOLDED figures are the mean scores, figures in ( ) are the standard deviation.
b. N= 265
According to the result shown by table 4, basically, all the sources of trust differ significantly among three countries except for Characteristic-based of trust.
According to the Scheffe’s test that been used to determine the comparisons of all possible pairs of means, the score of Malaysian vs. Japan in the four sources and
Inter-country Western Japan Malaysian F value P Value Scheffe
Contract adherence 5.9906 6.1415 5.1792 14.2730 0.0000 1-3 2-3
(1.1158) (1.1239) (1.4542) Flexibility beyond contract 5.5094 5.5786 4.9371
9.7040 0.0000 1-3 2-3
(1.0533) (0.9678) (1.1572)
Flexibility outside contract 4.9623 5.1824 4.4340 8.5020 0.0000 1-3 2-3 (1.3199) (1.2736) (1.2431)
10.6070 0.0000
Economic-based of trust 5.6604 5.6981 5.0377 1-3 2-3
(0.9773) (0.9984) (1.1859)
Process-based of trust 6.0000 (0.9903)
Characteristic-based of trust 4.8726 (1.0926)
Institutional-based of trust 5.8365 (1.1687)
consequences of trust was the highest among the other 2 pairs, which is followed by Malaysian vs. Western and Malaysian vs. Malaysian.
Economic-based of trust
As for economic-based of trust, the pair of Malaysian vs. Japan has the highest score of mean (5.6981), follows by Malaysian vs. Western (5.6604) and Malaysian vs.
Malaysian (5.0377). According to Scheffe’s test, the pairs of Malaysian vs. Japan and Malaysian vs. Western are significantly different from Malaysian vs. Malaysian.
It is explainable that most of the Malaysian Chinese highly dependent on the technical resources or know-how knowledge of Japan. Japan is the primary provider of technology for Malaysia (Peng, 2002). Therefore, Japan and Malaysia have already cultivated positive and good partnership as well as strong bilateral trade and investment linkages for a long time.
Furthermore, the Look East Policy of Malaysia and direct investment of the Japanese firms in Malaysia contribute to the close relationship between the two countries.
For example, in order to attract higher foreign investment and to increase the skills of the industry workers to accommodate along with the needs of the era of high technology industries, positive strategies and ways have taken. One of those is the establishment of the higher technology training center which is the Japanese and Malaysian Technical Institute (JMTI).
In 1997, both governments have agreed to have a technical collaboration to establish it. The institute’s main objective is to generate quality skilled workers through the skilled training programme at the higher level as it will cater to the industries needs in
44
Malaysia and prepares the students to adapt and face the challenge of the technology’s rapid growth as well as the future globalization.
Hence, the level of dependency on Japan’s technology is high and the technical collaboration is always expected while having business contact with Japanese. The Japanese business sector introduced their efforts on the promotion of technology transfer in the electronics R&D field and activities for the development of medium and small enterprises’ ability conducted in collaboration between JACTIM (Japanese Chambers of Trade and Industry Malaysia) and SMIDEC (Small and Medium Industries Development
Hence, the level of dependency on Japan’s technology is high and the technical collaboration is always expected while having business contact with Japanese. The Japanese business sector introduced their efforts on the promotion of technology transfer in the electronics R&D field and activities for the development of medium and small enterprises’ ability conducted in collaboration between JACTIM (Japanese Chambers of Trade and Industry Malaysia) and SMIDEC (Small and Medium Industries Development