USING SYNTHETIC RESPONDERS FOR SCALE DEVELOPMENT IN INFORMATION SYSTEMS

This study explores the potential for using synthetic data in scale development knowledge in Information Systems (IS). If successful, the use of synthetic data can revolutionize how scholars conduct scale development, potentially eliminating the need to collect data from a human sample in the early stages of scale development. We use genAI tools (like GPT3.5) to generate synthetic responses to a previously validated IS measurement model and test whether correlation patterns can be replicated. Response patterns to validated scales could only be partially replicated. This calls for caution with respect to the use of synthetic responses use in scale development.