Further probing of higher order in satisfaction construct Academic Article uri icon

abstract

  • PurposeThis paper aims to bring a new insight by examining the satisfaction model in a more complex manner to capture the variations of the satisfaction construct better.Design/methodology/approachThe paper presents a subsequent expansion from a previous research. The methodology used is similar, as the researcher believes that it is the best way to measure and to capture the curvilinear relationship of the variables under study. Regression of a higher form has been discussed and established and three variables were tested, i.e. word‐of‐mouth referrals (dependent variable), satisfaction (independent variable), and ownership (moderating variable). In order to test for the moderating effect, the hierarchical moderated regression analysis has been employed.FindingsThis study contributes to the body of knowledge in two ways. First, a literature review and exploratory study suggest that marketers should consider a higher‐order model in the models of satisfaction. Second, further considerations should be made to determine the curvature of the relationship where the researcher suggested that there might be a “U”‐shaped relationship between satisfaction perception and word‐of‐mouth referrals when there is a change in ownership.Research limitations/implicationsCustomer satisfaction should be seen as a non‐linear construct. In fact, in most industries there are tendencies to build marketing strategies around their core services. However, this research suggests that customers in Malaysia view beyond satisfaction in choosing their banks. Malaysian bank customers prefer patronizing banks from the same ethnic group as theirs, where banks owned by the same ethnic group as the customers are able to moderate the satisfaction level.Originality/valueThe paper shows and suggests how satisfaction construct should be measured in a non‐linear form with interaction in a different cultural context as in the banking sector in Malaysia.

publication date

  • 2006

number of pages

  • 13

start page

  • 98

end page

  • 111

volume

  • 24

issue

  • 2