Assessment of Construct Dimensionality and Consistency
In earlier chapters, eWOM trust has been theorized to be a complex, second-order construct consisting of several discriminating trust aspects. According to literature and qualitative scale development procedures, five sub-dimensions which are also assumed to mirror cognitive, affective as well as conative elements of trust, are most likely to comprehensively capture the nature of trust in OCR. During the identification stage, 53 items were identified that are supposed to capture the diverse eWOM sub-dimensions. A critical question is whether the hypothesized structure of the construct is also evident in empirical data. In order to empirically verify the adequateness of the advocated operationalization, the following multi-part research question was proposed:
© Springer Fachmedien Wiesbaden 2017
W. Weitzl, Measuring Electronic Word-of-Mouth Effectiveness,
RQ 1: (a) Is eWOM trust a complex, higher-order construct best described by several
facets or dimensions? (b) Are cognitive, affective, and conative elements suited to measure consumers’ generalized tendency to trust customer reviews?
In order to improve reliability in this research’s results, data for answering this question were gathered from diverse samples during the sequential stages of the research process. Initial data was collected during the reliability stage by means of an online survey among university students. This process enabled this research to initially assess the psychometric properties of the scale but also refine the set of observable indicators to a more manageable but statistically adequate number of items. To improve the general confidence in the new scale, additional data which was used to validate and cross-validate the measure were collected during the validation stage, resulting finally in a valid, reliable and practicable measurement instrument for eWOM trust.