Q-Bank: Using Information Technology for Collaboration to Enhance Survey Methodology

Surveys and questionnaires are a vital tool professionals use to understand the needs of the community. Questionnaire designers and evaluators devote great effort to testing particular survey questions to ensure they are effective and reliable. Often similar questions are evaluated again, because past evaluation results are filed away and forgotten after their initial use. Survey testing methodologies also occur in different forms, which lead to diverse terminologies and guidelines for practice. Therefore, survey testing and interviewing has evolved differently among various organizations with different objectives, different backgrounds of practitioners, and different institutional memories. As a result, it has been difficult to know what someone means when they use certain terms relating to their specific method. Methodological attempts to demonstrate what various pre-testing methods do and do not accomplish, are criticized for not reflecting usual practices. Practices across organizations are more likely to diverge. This makes it difficult to advance practices in the field when there is no consensus regarding what practices are.  The Q-Bank application addresses these problems in several ways. First, it serves as a central repository for pre-tested questions and research findings from participating government agencies. This avoids “reinventing the wheel” each time a questionnaire is to be evaluated. Previous reports could be checked to see if similar questions had been evaluated before and whether useful findings or recommendations had already been assembled. The methodological details would allow users to evaluate findings and the rationales for earlier recommendations. Second, Q-Bank helps advance the practice of questionnaire design and survey methodology. The standardized terminologies make the methodology apparent to both professionals in the field, allow people to evaluate what characteristics of questions cause recurring problems across investigators, across institutions, or across different testing methodologies.