preview

4.4 Construct Validity

Decent Essays

4.4 Construct validity
In order to test the construct validity of each scale, bivariate correlation will be used, because this test is to find out the association among those different variables and how strong the association is. (Pallant,2010). At the same time, checking the association among the subscales is regarded as way to assess scale’s construct validity (ibid). In addition, most of the researchers believe that the dividing point of the statistical significance (p) to ensure the relationship between variables is .05. Namely, in 100 samples, there are 5 times chances to get relationship that are not appeared in the whole population. Specifically, p < 0.05 is set to ensure the expected relationship (Bryman and Bell, 2007). And all the …show more content…

Firstly, in terms of CSR-Environment, the mean score of consumers (3.94) is much higher than that of employees (3.57), and consumers’ SD is .672, and employees’ SD is .70. Specially, t (134) =3.033 and p=0.03. Secondly, in terms of CSR-Society, the mean score of consumers (3.37) is much lower than that of employees (3.87), and consumers’ SD is .785, and employees’ SD is .762. Similarly as CSR-Environment, t (134) =-3.615 and p=. 000. In this case, consumers have better perception of McDonald’s CSR-Environment activities rather than CSR-Society activities. Alternatively, employees have better perception of McDonald’s CSR-Society activities rather than CSR-Society …show more content…

From the regression analysis result (Figure 29a in Appendix F), the consumers’ perception of CSR-Society doesn’t have positive influence on the functional brand image. This is because, based on the coefficients in the figure 20b, β=-.084, t=-.765 and p=. 447, which means the H 2a is rejected.
In terms of H2b, consumers’ perception of CSR-Society has the positive influence on emotional brand image, because, based on the figure 29b in Appendix F, β= .369, t=3.613 and p=. 001. At the same time, consumers’ perception of CSR-Society accounts for 13.6% of variance in emotional brand image (R²=. 136), and the Adjusted R² amend the R² to .126, which makes a better estimate. And based on F (13.057) and P (.001), the independence has the good prediction of the dependence. Therefore, according to those coefficients, H2b is accepted.
4.63 The test of hypotheses 3
H3a and H3b assume that employees’ perception of CSR-Society has the positive influence on functional brand image and emotional brand image. In terms of H3a (See figure 30a in Appendix F), employees’ good perception of CSR-Environment doesn’t have the positive influence on functional brand image, because the coefficients (β= -.246, t=-.1777 and p=. 082). Hence, H3a is

Get Access