The purpose of this study was to gain an understanding of the structure and antecedents of cruise passengers' loyalty. Specifically, the study examined the dimensionality of the loyalty construct. Moreover, the study investigated the utility of applying the Investment Model (Rusbult 1980, 1983) to reveal the psychological processes underlying loyalty formation. The study also attempted to, guided by the Investment Model, integrate the seemingly segregated findings of loyalty antecedents from marketing and leisure/tourism literature. Based on the Investment Model and other marketing and leisure/tourism studies on loyalty, a conceptual framework was established for this study. An online panel survey was conducted to examine this model. Subjects (N = 554) were online panelists who were repeat cruisers and who have cruised at least once in the past 12 months. In this study, loyalty was conceptualized as a four-dimensional construct: cognitive loyalty, affective loyalty, conative loyalty, and behavioral loyalty. Further, the first three components were postulated as three subdimensions of a higher order construct, attitudinal loyalty. However, this conceptualization was not supported by the data. Alternatively, post-hoc analyses revealed that attitudinal loyalty was a first-order one-dimensional construct, containing cognitve, affective, and conative components. Moreover, behavioral loyalty was positively and significantly influenced by attitudinal loyalty. In sum, this study supported the traditional two-dimensional conceptualization of loyalty, which argues that loyalty has an attitudinal and a behavioral component. Following the Investment Model, this dissertation suggested that satisfaction, quality of alternatives, and investment size were three critical antecedents of consumers' attitudinal loyalty. These theoretical relationships were supported by the present study, and collectively, the three predictors accounted for over 74 percent of the variance in attitudinal loyalty. Finally, this dissertation hypothesized that quality and value, two constructs related to loyalty, served as antecedents of satisfaction, with quality also leading to value. Results of the study supported all these hypotheses, and satisfaction was found to partially mediate the quality-attitudinal loyalty, and value-attitudinal loyalty relationships. Results of the present study provide important direction for the development of a holistic theoretical framework to explain the formation and structure of customers' brand loyalty.

13 | P a g e SUMMARY In South Africa’s rapidly changing social and cultural context, where family dynamics and relationships are changing just as quickly, it is essential to discover which factors contribute to successful relationships that persist over time. Previous research has utilised the Investment Model Scale (Rusbult, 1980) in order to understand commitment processes and the constructs satisfaction, quality of alternatives and investment have been identified as key determinants of commitment in romantic contexts. In the present study however I investigated the primary research question, “How can commitment be measured in the context of family relationships?” Data collection took place through the Family Commitment Scale (which was adapted from the Investment Model Scale) and a new meaningfulness scale was added in an attempt to explore whether meaningfulness would be a better predictor of family commitment than satisfaction, quality of alternatives and investment. The statistical analyses were conducted with the purpose of examining the research question and hypotheses. The findings of the present study contribute to commitment literature by underlining the utility of the Invest Model Scale in nonromantic contexts and providing an instrument which can reliably measure family commitment. Meaningfulness is also highlighted as a motivational feature behind family commitment processes. / Dissertation (MEd)--University of Pretoria, 2013. / gm2014 / Educational Psychology / unrestricted