5. Results and discussion
5.7. SEM results for Model 2 – Leckie, Munyaradzi & Johnson’s model
do not think about the brand, unless something goes wrong with their service.
Third, two of the CBE dimensions (cognitive processing and activation) had positive effect on brand loyalty, whereas affection showcased a negative impact on brand loyalty. The significance of the relationships found in the study has to be verified due to the poor model fit.
Table 9. Structural model results.
5.7. SEM results for Model 2 – Leckie, Munyaradzi & Johnson’s model
Structural equation modelling (SEM) was used to test both direct effects of CBE antecedents on brand loyalty, as well as the indirect effect using the mediation of CBE. The group of antecedents incorporated in this model was based on the theoretical foundations of consumer culture theory and included involvement, participation and self-expressive brand. Table 10
Hyphotesis β S.E. C.R. Conclusion
H1a: Satisfaction → Cognitive processing -0.011 0.091 -0.107 Not supported H1b: Satisfaction → Affection 0.516 0.072 7.830 Supported H1c: Satisfaction → Activation -0.384 0.143 -3.460 Supported H8a: Satisfaction → Brand loyalty 0.265 0.107 2.944 Supported H2a: Trust → Cognitive processing 0.393 0.089 3.353 Supported
H2b: Trust → Affection 0.384 0.067 5.478 Supported
H2c: Trust → Activation 0.710 0.140 5.763 Supported H8b: Trust → Brand loyalty 0.205 0.106 2.018 Supported H3a: Commitment → Cognitive processing 0.304 0.055 4.353 Supported H3b: Commitment → Affection 0.075 0.039 1.874 Not supported H3c: Commitment → Activation 0.269 0.081 3.895 Supported H8c: Commitment → Brand loyalty 0.462 0.059 8.479 Supported H7a: Cognitive processing → Brand loyalty -0.120 0.061 -2.719 Supported H7b: Affection → Brand loyalty -0.055 0.104 -0.584 Not supported H7c: Activation → Brand loyalty 0.282 0.042 6.227 Supported
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
presents the model fit obtained from the structural model. GFI and AGFI have not reached the desired .95, thus indicating poor model fit.
Table 10. GOF indices for Model 2.
The SEM results are shown in Table 11. The results don’t support the relationship between the
consumer involvement and cognitive processing or activation. Thus, H4a and H4c are rejected.
However, consumer involvement is positively related to affection (β=.185, t=3.372) in support
of H4b. The results support H5a-c as participation is positively related to cognitive processing (β=.265, t=4.349), affection (β=.221, t=3.82) and activation (β=.211, t=3.243), respectively.
The results also support H6a-c as self-expressive brand has a positive relationship with cognitive processing (β=.423, t=6.858), affection (β=.344, t=6.051) and activation (β=.204, t=3.214),
respectively. In terms of the direct relationship of antecedents on brand loyalty, the result link positively involvement (β=.090, t=2.012) and self-expressive brand (β=.153, t=2.909) with
brand loyalty, thus supporting H8d,f. suggesting that CBE only partially mediates the relationship between involvement, self-expressive brand and brand loyalty. The H8e is rejected, as there is no positive relationship between participation and brand loyalty (β=-.054, t=1.118) indicating
GOF statistics df p-value CMIN/DF GFI AGFI RMSEA NFI TLI CFI
Results 1085.646 332 <.001 3.27 0.800 0.755 0.081 0.883 0.903 0.915
‧
proving a positive link between affection (β=.453, t=9.188), activation (β=.472, t=10.895) andbrand loyalty. On the other hand, H7a is not supported as the relationship between cognitive processing (β=.472, t=10.895) and brand loyalty is negative.
Table 11. Structural model 2 results and comparison.
While the results confirm most of the findings by Leckie et al. (2016), some outcomes are interestingly different. Firstly, in both studies all three CBE dimensions were found to have an effect on brand loyalty. In both cases, affection and activation resulted in having a positive relationship with brand loyalty, while cognitive processing had a negative impact on brand loyalty. The authors offer their explanation to the negative impact of cognitive processing on
Hyphotesis β S.E. C.R. Conclusion Leckie, Munyaradzi
& Johnson’s study H4a: Involvement → Cognitive processing 0.091 0.050 1.67 Not supported Supported
H4b: Involvement → Affection 0.185 0.061 3.372 Supported Supported
H4c: Involvement → Activation -0.018 0.085 -0.287 Not supported Supported
H8d: Involvement → Brand loyalty 0.090 0.053 2.012 Supported Supported
H5a: Participation → Cognitive processing 0.265 0.080 4.349 Supported Supported
H5b: Participation → Affection 0.221 0.093 3.82 Supported Not supported
H5c: Participation → Activation 0.211 0.127 3.243 Supported Not supported
H8e: Participation → Brand loyalty -0.054 0.083 -1.118 Not supported Not supported H6a: Self-expressive brand → Cognitive processing 0.423 0.054 6.858 Supported Supported
H6b: Self-expressive brand → Affection 0.344 0.061 6.051 Supported Supported
H6c: Self-expressive brand → Activation 0.204 0.083 3.214 Supported Not supported H8f: Self-expressive brand → Brand loyalty 0.153 0.061 2.909 Supported Not supported H7a: Cognitive processing → Brand loyalty -0.046 0.066 -0.899 Not supported Not supported
H7b: Affection → Brand loyalty 0.453 0.053 9.188 Supported Supported
H7c: Activation → Brand loyalty 0.472 0.038 10.895 Supported Supported
Note: β= standardised regression weights, S.E.= standardised estimates, C.R.=critical ratio.
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
brand loyalty based on proposition by Hollebeek (2011a) suggesting that the impact of CBE dimensions on brand loyalty is likely to reach an optimal point as it is based on a non-linear relationship. The authors indicate that the negative effect can be caused by fatigue and burnout resulting in lower levels of attitudinal loyalty that highly engaged customers’ exhibit once exceeding the optimal point of engagement with the brand (Leckie et al. 2016). Also, Leckie et al. (2016, p.) mention the research on cognitive response to advertising repetition by Cacioppo
& Petty, (1979), which found that with high levels of exposure to message repetition, receivers rebel against the message due to boredom.
Secondly, the antecedents presented a diverse impacts on CBE dimensions. Involvement was found to have a positive effect on affection, what is inconsistent with findings by Leckie et al.
(2016) and Hollebeek et al. (2014) where involvement showed a positive effect on all three CBE dimensions. As involvement showed also a positive direct effect on brand loyalty suggesting that CBE dimensions only partially mediate the relationship, this negative effect between involvement and cognitive processing and activation is not surprising. In case of mobile service providers, highly involved customers may tend to ‘demonstrate they loyalty in the form of repeat purchase intention’ (Leckie et al., 2016, p.570). Furthermore, in this study
participation was found to have a positive effect on all aspects of CBE but not with brand loyalty directly, what supports the mediating role of CBE dimensions. Finally, self-expressive brand
‧
國立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
was found to have positive effect on all aspects of CBE and a direct positive effect on brand loyalty, what is consistent with prior studies (e.g. Algesheimer et al., 2005; Jahn & Kunz, 2012;
Sprott et al., 2009). While a study by Leckie et al. (2016, p.571) found a positive link between self-expressive brand, cognitive processing, and affection the authors indicate that ‘consumers do care about their self-image and how congruent this is with the brand image of the mobile phone service provider.’
The differences in the research findings can be due to the poor model fit or different mobile market characteristics, consumer behavior in a diverse culture setting, or cognitive processes in a different language context. While the nature of those differences requires further attention, the significance of this research’s findings should be considered with caution.