Does stimulus order affect central tendency and serial dependence in vestibular path integration?

Reproduction of perceived stimuli, such as distances or durations, is shaped by two biases: central tendency, where reproductions are biased toward the mean of the stimulus distribution, and serial dependence, where the reproduction of the current stimulus is influenced by the previous stimulus. We examined whether stimulus autocorrelation affects these biases in path integration. Twenty-four participants performed a vestibular distance reproduction task, actively replicating passively experienced distances. Two conditions were tested: high autocorrelation, where stimulus distances followed a random walk, and no autocorrelation, where distances were randomly shuffled. Both biases were quantified using separate simple linear regressions or a joint multiple regression model accounting for stimulus autocorrelation, derived from causal graphs. Simple regressions showed weaker central tendency and reversed serial dependence under the high-autocorrelation condition, but these differences disappeared in the multiple regression analysis. Thus, both biases persist regardless of stimulus autocorrelation, indicating robust neural strategies in self-motion perception.