Unpacking the subjective norm: applying structuration theory to traditional measures of social influence

This study took place within the context of a larger project seeking to identify the factors that are associated with the acceptance and use of electronic recordkeeping systems in public sector organizations. This class of system relies on ordinary end-users to choose to select and file appropriate records into the system in order to comply with organizational and legislative recordkeeping requirements. The use of such a system necessarily includes a social or organizational context dimension to explain the intention to use. This paper focuses on the construct validity associated with social influence within the research model. This particular construct traces its roots back to the 1970s and appears in much of the technology acceptance literature as the subjective norm. The measures and techniques for identifying construct validity prior to the development of a survey instrument, and subsequent measures and techniques for detecting construct validity after gathering the detailed survey dataset, are discussed. The techniques used to group the measurement items by construct, and thus into a survey instrument, included card sorting and the use of cluster analyses and dendrograms. After the survey instrument data collection activity, the detailed construct validity analysis utilized the circle of correlations based on a principal component analysis. The research found that the traditional measures of social influence constructs cluster well, but are in fact multidimensional. Furthermore, the dimensionality revealed by the statistical analysis aligns with and supports the predicted interactions of society as put forward in Giddens' Structuration Theory. This finding lends empirical support to Structuration Theory and underscores the importance of construct validity, particularly in the current push to increase the "social" aspects of technology.