Explore the Difficulty of Applying Meta-Analysis to Taiwanese Customer Satisfaction

• Main Authors: Shen I Hong, 沈益宏
• Other Authors: 劉原超
• Format: Others
• Language: zh-TW
• Published: 2004
• Online Access: http://ndltd.ncl.edu.tw/handle/09586648943520881378

碩士 === 大葉大學 === 事業經營研究所 === 92 === The purpose of this study was using the qualitative research method to explore the difficulty of applying the Hunter meta-analysis to Taiwanese customer satisfaction based on Oliver''s (1997) conceptual model of the antecedents and consequences of customer satisfaction. We obtained 28 studies matched with the key word “customer satisfaction” or “consumer satisfaction”, published in 2003, B to C (business to ending user) and services from the Dissertation and Thesis Abstract System of National Central Library. Results showed that the following problems must be considered when conducting a meta-analysis in the field of customer satisfaction at service in Taiwanese business. 1. How to search generally and gather economically these studies on customer satisfaction. 2. The decision of the respectively theoretic framework of the field for the meta-analysis had a magic impact on abstracting research results. Therefore, it was necessary to adjust framework sometimes. 3. How to classify satisfaction-related variables correctly according to their labels. 4. To discover potential moderators of customer satisfaction. 5. To select referent population for adjusting range variation. 6. The different patterns of customer satisfaction and satisfaction- related variables in testing for association. 7. How to transfer these statistic values of tests (etc. t-test, X2 test) to a common scale of product-moment correlation. 8. Absence of some variables’ reliability values and standard deviation values. In this study we attempted to resolve above problems. To this end, we offered several suggestions for future research. 1. What information has to be included when using Oliver''s conceptual model of customer satisfaction for conducting a meta-analysis. 2. In the phase of data analysis, when one study owns more than one correlation coefficient to adjust the weighting value is necessary to avoid over estimated the contribution for general conclusion from the meta-analysis. 3. Transfer each customer satisfaction measure value into a common scale and ‘meta-analyze’ an indicator of Taiwanese studies of customer satisfaction. Further, compare the difference among industries.