The dynamic nature of value : a longitudinal study
Accepting the view that the marketing process is centred on exchange between two parties (Hunt, 2002), it follows that exchange will take place between two (or more) parties when each party trades something of value in return for something of greater value. Consequently the logical conclusion is that value is the cornerstone of marketing (see for example Eggert and Ulaga, 2002; Holbrook, 2005). Perceptions of value can vary over time and experience (Eggert and Ulaga, 2002; Woodall, 2003; Sánchez-Fernandez and Iniesta-Bonillo, 2007). However, even though the temporal nature of value is widely acknowledged, research in this area has been largely overlooked, and while there is limited investigation within the b2b domain (see Flint et al., 2002; Beverland and Lockshin, 2003; Eggert et al., 2006) a literature search has been unable to identify any research that examines actual changes in perceptions of value within consumer research. Consequently, the aim of this study is to empirically examine the temporal stability (i.e., the nature and strength) of the functional relationships between value and its antecedents and outcomes. In order to address the above aim a theoretically grounded model is proposed. Based on common acceptance among researchers (see review by Woodall, 2003) value is conceptualised as the result of a 'trade-off' between benefits (get) and sacrifices (give). However, instead of treating value as a composite higher order construct the behaviours of its two components (get and give) with the following constructs are examined separately: service quality and personal values (terminal and instrumental) are modelled as determinants while satisfaction and intention are the outcomes of value. In addition the impact of knowledge (cognitive; Woodruff, 1997) and emotions (affective; Richins, 1997) as direct determinants of value and additionally as moderators of the value to satisfaction relationship is tested. The research was-conducted within the Higher Education sector among consumers of postgraduate education at a London business school. To test the temporal stability and pattern of development of the functional relationships between the value components and their above defined nomologically related constructs, related data were collected longitudinally from two sample of cohorts at three points in time (i.e., the beginning, middle and end of their studies) via a personally (Times 1 and 2) and internet (Time 3) administrated questionnaire. A total of 34 and 45 usable responses were collected from Cohorts 1 and 2 respectively over the three time points. The data were analysed using Partial Least Squares. Analysis indicates that the give component of value should be separated into money, and time and effort (denoted in this study as give). There is support for knowledge and emotions as direct determinants of the now three value components rather than as moderators of the relationships between these components and satisfaction. Comparisons between the two cohorts reveal the existence of a number of significant differences in the relative strength of corresponding relationships. Finally, in terms of the focal interest of this study, there is substantial evidence of the temporal nature of the functional relationships of the value components. Four of the hypothesised relationships are supported only at a single time point, while a number of significant changes in the strength of the functional relationships between the three points in time are identified. The research is considered to make the following contributions to the subject matter. It confirms the idiosyncratic nature of the value components in terms of their functional relationship with antecedents and consequences. It highlights the need to consider the location of monetary sacrifice within the give component. The existence of a time lag before some determinants have a significant impact on the formation of value is identified. There is tentative evidence to suggest that as consumption progresses, value is formed by a larger number of determinants. For the get component, significant variations in the strength of its functional relationships over time are found to exist.