2.2. Elaboración
2.2.6. diseño de PRUEBAS
The aim of this hypothesis was to find out if there were any difference between the two ethnic groups’ purchasing decisions from retailers with different sized wine product ranges and whether there was a relationship between whether the consumer will consider the wine product range and the importance of the retailer. Two sub-hypotheses were constructed to examine the relationships.
First, to analyze attitudes of the two ethnic groups in regard to whether they will consider the size of the retailer’s wine range, a hypothesis was constructed as follows:
H4a0: Perceptions of whether the consumer will consider the size of the retailer’s
wine range are not affected by ethnicity.
H4a1: Perceptions of whether the consumer will consider the size of the retailer’s
Figure 5.13 indicates that there were more respondents, in total, who do not consider the retailers’ wine range before they choose where to buy wine than those who do, especially for other Ethnic Groups, with 61.2% and 55.3%, for European.
Figure 5.13 Respondents’ perceptions of whether they will consider the retailer’s wine range
Hypothesis Test
In this case, in order to test the relationship between “the ethnic groups” and “whether people will consider the retailers’ wine range or not” as the two categorical variables so the Chi-square test was applied. The Chi-Square test for independence was suitable for determining whether there was a significant association between the two categorical variables. Categorical variables take on values that are names or labels. Both events considered in the hypothesis must be mutually exclusive and have total
probability (Plackett, 1983). In this case, the two variables in this hypothesis were both mutually exclusive, so Chi-Square can be applied. The result showed that there was no statistically significant relationship between the ethnic groups and whether people will consider the retailers’ wine range or not as the Asymp. Sig. value was 0.47, which is over the significance level of a=0.05 (Table 5.7). Therefore, the null hypothesis was not rejected and we can conclude that there were no statistically significant differences between the two ethnic groups’ perceptions of whether they will consider the size of the retailer’s wine product range before buying wine or not.
Value df Asymp. Sig. (2-sided) Exact Sig. (2-sided) Exact Sig. (1-sided) Pearson Chi-Square 0.522(b) 1 0.470 Continuity Correction(a) 0.310 1 0.578 Likelihood Ratio 0.526 1 0.468
Fisher's Exact Test 0.510 0.290
Linear-by-Linear
Association 0.520 1 0.471
N of Valid Cases 199
a Computed only for a 2x2 table
b No cells (.0%) have an expected count of less than five. The minimum expected count is 21.18.
Table 5.7 Chi-Square Tests: Relationship between ethnicity and consumers’ perceptions of whether they will consider the size of the retailer’s wine range
Secondly, to analyze whether there was a relationship between whether the consumer will consider the wine product range and the importance of the retailer, a sub-hypothesis was constructed as follows:
H4b0: There is no relationship between the two ethnic groups’ perceptions of whether
retailer.
H4b1: There is a relationship between the two ethnic groups’ perceptions of whether
consumers will consider the size of the retailer’s wine range and the importance of the retailer.
Of the respondents who said they do not consider the size of the retailer’s wine range, most perceived the retailer as not very important. It was interesting that most of the people who considered the size of the retailer’s wine range also perceived the retailer as not very important (Figure 5.14).
Figure 5.14 Relationship between whether respondents consider the size of the retailer’s wine range and the perceived importance of the retailer
Hypothesis Test
To further investigate the relationship between “whether respondents consider the size of the retailer’s wine range” and “the perceived importance of the retailer”, a Chi-Square test to compare these two variables was carried out for each ethnic group (Table 5.8 & 5.9). Chi-Square analysis is suitable for evaluating whether or not the two variables within a contingency table are independent, or that there is no association between them. Both events considered must be mutually exclusive and have total probability (Plackett, 1983). In this case, the two variables in this hypothesis were both mutually exclusive, so Chi-Square can be applied. Results revealed that for both ethnic groups, there was no statistically significant relationship between whether consumers consider the retailer’s wine range and the perceived importance of the retailer. Asymp. Sig. for Europeans and other ethnic groups equal to 0.143 and 0.902, respectively, which were all greater than the significance level 0.05, so it did not reject the null hypothesis and confirmed that the perceived importance of the retailer was not affected by the two ethnic groups’ perceptions of whether they consider the retailer’s wine range.
Value df
Asymp. Sig. (2-sided)
Pearson Chi-Square 6.870(a) 4 0.143
Likelihood Ratio 8.001 4 0.092
Linear-by-Linear Association 2.152 1 0.142
N of Valid Cases 150
a Three cells (30.0%) have expected count less than five. The minimum expected count is 1.34. Table 5.8 Chi-Square Test: Relationship between whether European respondents consider
Value df
Asymp. Sig. (2-sided)
Pearson Chi-Square 1.048(a) 4 0.902
Likelihood Ratio 1.065 4 0.900
Linear-by-Linear Association .022 1 0.882
N of Valid Cases 49
a Six cells (60.0%) have an expected count of less than five. The minimum expected count is 1.55. Table 5.9 Chi-Square Test: Relationship between whether other ethnic respondents consider
the size of the retailer’s wine range and the perceived importance of the retailer