Attention in Motion: Using Scroll Tracking to Measure the Dynamics of Preferential Choice

Understanding how preferences are formed in commonplace and natural situations is highly valuable for researchers and practitioners. Here, I develop scroll tracking, a process-tracing method that uses a touch scroll–based responsive mobile web design well suited for studying the cognitive processes behind decision making. The method follows the decision maker’s scrolling behavior and records a response curve in the pixel–millisecond coordinate space. To demonstrate the value of response curve analysis in decision research, different metrics are proposed that are derived from the response curve and rooted in the motor response and gaze dwell metrics presented in the literature. An experimental study framed as a consumer choice problem is used to validate the method and test predictions made by attentional evidence accumulation models using the response curve metrics. The way the decision maker interacts in the app can be used to predict subjective valuations as well as latent individual difference measures and risk attitudes as a complementary risky choice study shows. I then compare the predictive ability of different drift diffusion models stocked with the scroll-tracking response metrics. The scroll-tracking metrics improve all the models. These findings show the potential of naturalistic process-tracing methods in responsive digital applications. These designs can be used to covertly study decision processes and are applicable to different practical settings.

This paper was accepted by John Beshears, behavioral economics and decision analysis.

Supplemental Material: The online appendix and data files are available at https://doi.org/10.1287/mnsc.2024.06264.