[P2] Poster: Day 2

[P2-18] Serial dependence between duration and numerosity perception

*Takuma Hashimoto^1,2, Yuko Yotsumoto¹ (1. The University of Tokyo (Japan), 2. Research Fellow of Japan Society for the Promotion of Science (Japan))

Keywords:

serial dependence,duration,numerosity,ATOM

Serial dependence refers to the phenomenon where current perception is biased toward previous perception. This effect has been observed across various features, including magnitude dimensions such as numerosity. Converging evidence suggests overlapping mechanisms among magnitude dimensions, and cross-dimensional adaptation has been reported. However, whether cross-dimensional serial dependence occurs remains unclear. Togoli et al. (2021) investigated serial dependence between duration and numerosity perception but failed to observe such an effect. One possible cause is the absence of feature-based attention to the inducing feature (i.e., duration or numerosity). Therefore, the present study examines whether cross-dimensional serial dependence occurs under condition that require feature-based attention.In line with the previous study, we will recruit 28 participants. Participants will compare sequentially presented dot arrays and indicate which stimulus had a longer duration or a greater number of dots. The two tasks –duration comparison and numerosity comparison will be conducted on separate days. Inducer stimuli will be presented prior to the comparison task to influence the reference stimuli. To ensure feature-based attention to the inducing feature, we will include catch trials at unpredictable intervals. In these trials, participants will classify the inducer stimuli according to the dimension not used in the main task. For example, when the main task involves duration comparison, participants will classify the inducer as “few” or “many” in catch trial, and vice versa. If feature-based attention to the inducing feature is critical for cross-dimensional serial dependence, we expect to observe the effect. Such a finding would suggest that serial dependence occurs more broadly than previously thought and supports the view that higher-level cognitive processes contribute to serial dependence. It would also further support the idea of shared processing mechanisms for duration and numerosity. Conversely, if feature-based attention is not critical, or cross-dimensional serial dependence is inherently absent, the effect will not be observed.