Investigating Two Contradictory Views of Formative Measurement in Information Systems Research

Initially candidate survey items were compiled from existing literature and presented on a seven-point Likert scale. Then the survey items were filtered through several iterations of pretests. First, these items were examined by a focus group. This group consisted of three colleagues who are knowledgeable about the research subject as well as the measurement theory, and five senior IT managers with practical knowledge in IT infrastructure. This group of people met three times to examine the content validity of the research instrument during the two week period. Each time they met, the participants gradually reduced the items through intensive discussions. Second, the revised questionnaire was pretested with 20 senior IT managers. Each participant was asked to complete the questionnaire and, during the debriefing period, to offer any suggestions for improvement. Again, from this process, four survey items were dropped and several minor refinements on the remaining items were made. Third, the revised questionnaires were mailed to another group of 20 senior managers for a pilot test. Follow-up interviews with 20 respondents did not indicate any need for substantive changes to the questionnaire. This seemed to indicate strong preliminary evidence of construct validity. Survey items used for this research and their sources are listed next.