Internet usage for information provisioning: theoretical construct development and empirical validation in the clinical decision-making context

Despite substantial discussion of the Internet's impact on individual activities, there is an absence of a theoretically grounded measure of Internet usage for the provisioning of information required by decision-makers. Our research addresses this void in the literature. We conceptualize Internet Usage for Information Provisioning !IUIP) as the degree to which the Internet is used to meet information requirements of individual decision-makers engaged in diagnostic decision-making tasks. Drawing on the information processing theoretical framework, three dimensions (usage for breadth of content, usage for depth of content, and usage for interaction dynamism) of IUIP are identified. The construct was validated through a two-phase empirical study conducted in the physician clinical decision-making context. The first phase involved interviewing physicians to establish face validity of the dimensions and to generate multi-item measures for each dimension. The second phase consisted of surveying physicians to evaluate measurement properties and nomological validity. Principal components analysis, confirmatory factor analysis, and partial least squares analysis were the primary statistical techniques used to analyze the data and evaluate construct validity. Our results suggest that decision-makers operating in uncertain and equivocal decision contexts use the Internet to a greater degree for their content needs and selective attention requirements. Our results also suggest that a fit between the user's information processing requirements and technology, and selected social norms, lead to higher levels of IUIP. Implications for practice and future research are discussed. Index Terms--Information processing, information systems (IS) success, IS usage, Internet, partial least squares.