In order to test for a scale’s criterion-related validity, its operationalization has to be assessed against some criterion (Trochim, 2012). Therefore, based on trust theory, some predictions are proposed about how the eWOM trust scale will perform in the presence of such a criterion. In order to be in accordance with the standards of judgment (Churchill, 1979; Churchill & Iacobucci, 2005), multiple research questions are used to target different sub-types of criterion- related validity.
An important issue of concern is concurrent validity. According to Netemeyer et al. (2003), evidence for concurrent validity is provided by considerably strong statistical correlations between the construct measure under development and one or more criteria, which are measured at the same point of time or “concurrently”. A measure’s concurrent validity is assumed to be present in the case where the newly developed measure exhibits statistically significant relations with a measure that previously has been validated. These two measures may quantify the same construct or two distinct but presumably related constructs. This thesis assumes that evidence for concurrent validity is provided in cases where the new measure is able to distinguish between groups of consumers that it should be able to distinguish (Soh, 2007). The fact that the new scale is able to distinguish consumers that indicated high levels of eWOM Trust on a single-item scale from consumers that stated a low degree of trust on the same scale, should provide confidence that the scale possesses validity. Thus, the following research question is addressed:
RQ 2: Does the developed measure of eWOM show the ability to distinguish groups of consumers that have indicated varying levels of eWOM trust on a single-item scale?