Abstract
How is what you see influenced by what you saw? The visual system may use recent perception to inform responses to current stimuli. This can cause the perception of current stimuli to be attracted toward previous observations, an effect termed serial dependence (Fischer & Whitney, 2014). This misperception might well be useful in a noisy visual environment, where minor image distortions over time may not actually represent meaningful change (Burr & Cicchini, 2014; Fischer & Whitney, 2014). Previous work has suggested that Bayesian perceptual inference may underlie serial dependence (Cicchini et al., 2018). For this to be true, the relative uncertainty associated with both prior and current sensory input should be taken into account. In an experiment manipulating the level of noise present in orientation stimuli, we found an effect of current stimulus uncertainty on serial dependence. We found no good evidence for an effect of previous stimulus uncertainty. Our results provide only partial evidence for the Bayesian interpretation of serial dependence. Non-Bayesian models (Ceylan et al., 2021) may provide a better account of the phenomenon.
| Original language | English |
|---|---|
| Article number | 6 |
| Pages (from-to) | 6 |
| Number of pages | 14 |
| Journal | Journal of Vision |
| Volume | 22 |
| Issue number | 1 |
| Early online date | 12 Jan 2022 |
| DOIs | |
| Publication status | Published - Jan 2022 |
Bibliographical note
Funding Information:Supported by a UK Engineering and Physical Sciences Research Council (EPSRC) studentship award, number 2120278.
Publisher Copyright:
© Copyright 2022 The Authors
Fingerprint
Dive into the research topics of 'Stimulus uncertainty predicts serial dependence in orientation judgements'. Together they form a unique fingerprint.Datasets
-
Noise and serial dependence
Benton, C. (Creator) & Gallagher, G. (Creator), University of Bristol, 5 Jan 2022
DOI: 10.5523/bris.3lf3skkfzen4a2ajz1l6bv05gv, http://data.bris.ac.uk/data/dataset/3lf3skkfzen4a2ajz1l6bv05gv
Dataset
Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver