5. Data Analysis
5.2. Hypothesis Testing
5.2.5. Hypothesis 4b, 6 and 7
國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
70
choice but to use Chinese products. On the other hand, almost one third of respondents in this survey answered „I don‟t know‟ to questions inquiring about the frequency of use and purchase of Taiwanese products, implying that they are not aware of the origin of products they buy and use.
As pointed out in the literature review section, early studies on the topic of COO used economic development of the country as the main independent variable, and investigated its effect on the perception of country products. This thesis followed an example of more recent studies and included economic development in a multiple-item Country image measurement scale. A simple linear regression test of the relationship between country development and product perception was ran with each country‟s data, to assess country development as a predictor, compared to the multiple-item country image variable. For Taiwan, R2 = .231, F = 75.059, p < .001. For Japan, R2 = .005, F = 1.205, p = .273. For China, R2 = .104, F = 28.864, p < .001. Therefore, for product perception, country image is a better predictor than (perceived) economic development of a country in every instance. This finding justifies the use of a multiple-item country image scale. Quite surprising is the finding that in Japan‟s case, product perception is not predicted by economic development of the country on any level of significance.
5.2.5. Hypothesis 4b, 6 and 7
A hierarchical multiple regression test was used again to test the relationship between perceived country image and a perception of a specific product category (laptop computers) made in the country. Based on the literature review, variables which may influence this relationship and which were therefore included in the test were product category expertise, product category involvement and, like in the case of general product evaluations, user experience with country products and demographic variables.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
71
Summaries of the hierarchical regression analysis can be found in Tables 5.14, 5.15 and 5.16. For Taiwan, country image is the most significant predictor, p < .001. Product category expertise was also a significant predictor at the .05 level. A test with Japan‟s data reveals that country image was the most important variable, significant at the .001 level. Other predictors found to be significant in this relationship at the p < .05 level were user experience with Japanese products, product category involvement, and the respondent‟s gender.
The test of the data relevant to China reveals that country image was the only significant predictor, at the p < .01 level. No other variables played a role at any of the usual levels of confidence.
Hypothesis 4b is therefore accepted. Country image was the most important predictor in every test of evaluation of specific product categories, significant at the .001 or .01 level.
Hypothesis 6 is partially supported. Product category expertise was a significant predictor for the evaluation Taiwan‟s laptops, but not significant on any usual level of confidence for laptops produced in Japan or China. Hypothesis 7 is partially supported. Product category involvement was a significant predictor of laptop quality perception in the case of Japan, but not significant for Taiwan or China. Hypothesis 5 (summarized in chapter 5.2.4.) is also partially supported. User experience with country products was a significant predictor for the evaluation of Japanese laptops only.
Although the literature review does not provide any universally valid agreement on the effect of demographic variables on country-product perceptions, many studies show that perception of products from foreign countries can indeed vary across different population groups. In general, literature suggests that younger, better educated and more affluent consumers evaluate foreign products more positively. Some studies also propose that men tend to view foreign products more favorably than females. As for this study, it is noteworthy that in both regression tests regarding the perception of Japanese products (general and
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
72
category specific), some demographic variables (though different ones in every test) came out as significant. None of the demographics were significant in any of the tests examining Chinese and Taiwanese products. In the test of general perception of Japanese products, age and gender were significant at the .05 level. The relationship with age was positive, i.e. older respondents tend to have a better perception of Japanese products than younger ones. This might suggest that the perception of Japanese products was exceptionally positive in the Czech Republic in the past, but it is not as pronounced with younger consumers. However, the finding is in opposition with the effect of age proposed by previous studies. The relationship between general perception of Japanese products and gender was positive. Because females were coded as 0 and males as 1, the positive value of β indicates that male respondents evaluated Japanese products on average as better than female respondents.
In the evaluation of Japanese laptops, the consumers‟ gender also came out as a significant variable. However, the value of β was negative in this case, indicating that females evaluated Japanese laptops more positively than males did. This finding appears fairly odd, considering that males were found to evaluate Japanese products in general more favorably than females.
‧
Table 5.14: Summary of Hierarchical Regression Analysis for Variables Predicting Perception of Taiwanese Laptops (N = 252)
Model 1 Model 2 Model 3
Variable B SE B β B SE B β B SE B β
TWimage5 0.438 0.051 .477*** 0.416 0.051 .453*** 0.412 0.053 .448***
TWuserexperience 0.046 0.034 .077 0.049 0.034 .083
involvement 0.034 0.054 .035 0.033 0.055 .035
Product expertise 0.075 0.03 .147* 0.086 0.036 .167*
Gender -0.061 0.097 -.042
Table 5.15: Summary of Hierarchical Regression Analysis for Variables Predicting Perception of Japanese Laptops (N = 252)
Model 1 Model 2 Model 3
Variable B SE B β B SE B β B SE B β
JPimage5 0.448 0.097 .281*** 0.352 0.099 .221*** 0.397 0.100 .248***
JPuserexperience 0.124 0.043 .183** 0.103 0.043 .152*
Involvement 0.149 0.058 .156** 0.136 0.058 .142*
Product expertise -0.018 0.032 -.035 0.023 0.039 .046
Gender -0.211 0.103 -.148*
Age 0.004 0.007 .042
Education -0.088 0.045 -.133
Income 0.000 0.000 .052
R2 .079 .131 .159
F for change in R2 21.383 4.967 2.029
*p < .05. **p < .01. ***pp < .001.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
74
Table 5.16: Summary of Hierarchical Regression Analysis for Variables Predicting Perception of Chinese Laptops (N = 252)
Model 1 Model 2 Model 3
Variable B SE B β B SE B β B SE B β
CNimage5 0.234 0.082 .178** 0.227 0.084 .172** 0.242 0.085 .183**
CNuserexperience 0.072 0.053 .086 0.072 0.054 .086
Involvement -0.020 0.074 -.017 -0.013 0.074 -.012
Product expertise 0.062 0.04 .100 0.034 0.049 .054
Gender 0.171 0.132 .099
Age 0.001 0.009 .009
Education 0.056 0.057 .070
Income 0.000 0.000 -.031
R2 .032 .050 .061
F for change in R2 8.169 1.593 .715
*p < .05. **p < .01. ***pp < .001.
5.3. Sources of Country Information
Because the literature review provides only a limited insight about information sources that people (Czech people in particular) use to gather information about East Asian countries, an exploratory research on the topic was conducted as a part of the thesis survey. The aim was to determine how much information about Taiwan, China and Japan the Czech respondents receive from each of the following sources: Personal experience with the country (travel in the country, meeting people from the foreign country), traditional media, new media and second-hand information (word of mouth, from friends, colleagues, peers etc.). Furthermore, respondents were asked to indicate the overall favorability of information they had received about each of the countries.
Respondents in the survey were asked to self-assess the amount of information about each country under study that they receive from each of the information sources. They
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
75
disclosed this information on the scale from 1 (no information received from that particular source) to 5 (a lot of information received from that particular source). The results summarized in Table 5.17 reveal that for each of the three countries included in the research, people get most of their information from traditional media, closely followed by new media.
Second-hand information and finally personal experience with the country are next in order.
Quite remarkably, the order of information sources is exactly the same for each country in research.
The aggregated data in this section confirm that the information Czech respondents receive about Taiwan (from any information source) is on average very limited, with the amount of information coming from personal experience being especially low. Like in the case of subjective and objective country knowledge, all four information sources are significantly correlated with Taiwan‟s perceived country image, with Pearson correlation coefficient ranging from .339 to .427. This information confirms the finding that perceived Taiwan image improves with more knowledge and information about the country (regardless of the information source).
As for Japan, respondents feel on average that they receive a fair amount of information from traditional and new media, some information from second-hand information and little information from personal experience. All four information sources are again positively correlated with country image, though the relationships are somewhat weaker than in Taiwan‟s case. Pearson correlation coefficient for Japan‟s information sources ranges from .215 to .247.
Respondents of the survey report that the amount of information they receive about China from all four sources is virtually the same as with Japan, i.e. a fair amount of information from traditional and new media sources, some from second-hand sources and little from personal experience. However, unlike the other two countries, the amount of
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
76
information received from any source is not significantly correlated with perceived country image. Pearson correlation coefficient ranges from .024 to .080. These results support the findings of hypothesis 3. Unlike the other two countries, having more knowledge and being better informed about China does not necessarily lead to a positive perception of the country image, suggesting that the information people receive about China is probably not very favorable.
The previous assumption is confirmed in Table 5.18, which summarizes the overall favorability of information that the respondents of the study receive about each country (from any source). The overall favorability was self-reported, on a scale from 5 (very positive information) to 1 (very negative information). Information about Japan was on average positive to very positive. Information about Taiwan was on average neutral to positive.
Information about China that respondents received was on average negative to neutral. This data provides a good explanation for the observation that having more knowledge and being better informed about China was in most instances not positively correlated with perceived country image of China, unlike to the other two countries.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
77 Table 5.17: Respondents‘ Sources of Country Information
N Mean Std. Deviation
TW personal INFO 247 1.68 0.99
TW traditional media INFO 247 2.51 1.02
TW new media INFO 246 2.46 1.09
TW second hand INFO 246 2.17 1.02
JP personal INFO 247 2.38 1.16
JP traditional media INFO 247 3.26 0.95
JP new media INFO 246 3.16 1.09
JP second hand INFO 247 2.79 1.03
CN personal INFO 245 2.44 1.2
CN traditional media INFO 245 3.25 1.01
CN new media INFO 246 3.09 1.04
CN second hand INFO 246 2.81 1.07
Table 5.18: Favorability of Received Country Information
N Mean Std. Deviation
TW favorability 199 3.58 0.85
CN favorability 239 2.33 0.91
JP favorability 243 4.12 0.76
‧
This chapter is going to integrate the results of the data analysis, summarize the findings and discuss possible implications, and also touch upon limitations and suggestions for future research. Table 6.1 lists the research hypotheses of this study together with the outcome of the hypothesis testing.
Table 6.1: Summary of Research Hypotheses and Outcomes
Research Hypothesis Outcome
H1: Japanese products are rated as the best of the country products in this study, while Chinese products receive the worst ranking. Taiwanese products are in between them in terms of product perception.
Supported
H2: Japan‟s country image is evaluated as the best, followed by Taiwan, and China with the worst ranked country image.
Supported
perception of country products in general.
Supported
H4b: Better image of a country in the eyes of Czech consumers will lead to a better perception of country products in a specific product category (laptop computers).
Supported
H5: Czech consumers‟ user experience of products from Taiwan, China and Japan is a significant predictor in the relationship between the country image and the consumers‟
evaluation of products made in the country.
Partially image and the consumers‟ evaluation of laptop computers produced in corresponding countries.
Partially supported
6.1. Taiwan Country Image
One of the main interests at heart of this thesis is the perception of the Taiwan country image abroad. As a country that is extremely restricted in participation in international