In a panel survey where the same set of questions is asked repeatedly over time, the need to update the survey questionnaire may arise. Knowledge of the quality of earlier responses can be used to improve the questionnaires for subsequent waves. Response rate and item non-response can be improved through information obtained in earlier waves. Changes in layout, appearance, number of questions and wording, influence the quality of the responses. However, the need for continuity of the questionnaire's content and appearance across waves, substantially limits the choices for redesign. Therefore, the redesign of a questionnaire is a constrained problem with an ultimate objective of increasing the response rate, item response, and accuracy of the responses, within the limits imposed by the need to maintain continuity. A travel impact analysis conducted in connection with the State of California Telecommuting Pilot Project offers an opportunity to study panel questionnaire updating. The process of updating a three-day travel diary is documented, and its effects on trip reporting and reporting accuracy are discussed in this paper. The results indicate that improved wording and layout contribute to improved data quality, but do not help to prevent under-reporting of trips due to panel fatigue.
