The Direct and Indirect Effect of Antecedent Variables on Customer Loyalty Toward Commercial Bank in Southern Thailand

The main objective of this study is to empirically determine the major predictors that directly or indirectly affect customer loyalty in commercial banking sector in South Thailand (Songkla, Krabi, Nakronsrithamarad, Suradthanee) based on underlying concept of Oliver's (1999). The banking sector continued to playa vital role in Thailand's economy despite world financial crisis. The banking industry contributes 8.1% to the country's GDP, the highest contributor for the service sector. However, stiff competition in the banking industry in this difficult economic time is the compelling reason to conduct this research whereby bankers have to attract loyal customers for their continuous performance and profit. The banking industry in Thailand consists of fourteen local and five international banks with over 5,550 branches nationwide in 2009, an increase from 3,554 branches in 2002 (50% increase),indicating increased competition among banks (Chaoprasert & Elsy, 2004). Furthermore, performing research on customer loyalty is a necessity especially in Thailand where empirical research is still a scarcity. The research framework consists of three exogenous latent (perceived service quality, customer satisfaction and trust) and two endogenous latent variables (commitment and customer loyalty). Perceived service quality is measured using 9-item; customer satisfaction (7-item); trust (5-item); commitment (6-item) and customer loyalty (9-item). This study determines 384 samples (bank customers) based on 1.5 million employed population in the four areas specified in the South of Thailand (Krejcie & Morgan, 1970). Since, this study intends to investigate causal and mediating relationships between latent variables, structural equation modeling (SEM) was the main analysis method used. The findings indicate that customer satisfaction, trust, commitment and perceived service quality are the direct causal antecedents of customer loyalty, whereby perceived service quality is a new found significant predictor in the generated structural model of this study. Similarly, commitment was significantly related to satisfaction and trust. A new direct predictor was also suggested to influence commitment in the generated structural model i.e. perceived service quality. As for the antecedents of satisfaction, perceived service quality remains the significant and positive predictor of customer satisfaction
. The finding also concludes that commitment do not mediate the linkages between trust nor satisfaction on customer loyalty. Additionally, commitment only partially mediates the relationship between perceived service quality and customer loyalty. As a contribution to new knowledge, two new mediating paths are produced by the generated structural model of this study. It shows that satisfaction partially mediates two newly found linkages: perceived service quality to loyalty and perceived service quality to commitment. The squared multiple correlations (SMC) or the R2 for loyalty (79%) and commitment (82%) shows the highest variance explained in the generated structural model. Finally, the underlying conceptual model underpinned by Oliver (1999) proves its relevance by the achievement of model's goodness of fit. The thesis ends the report by discussing the pros and cons of the findings.