Purchase this article with an account.
Jozsef Fiser, Jozsef Arato, Abbas Khani, Gregor Rainer; Change-related weighting of statistical information in visual decision making. Journal of Vision 2016;16(12):574. doi: https://doi.org/10.1167/16.12.574.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Perceptual decisions are influenced not only by short-term history (e.g. adaptation, repetition, sequential effects), but also by statistics of events from many decisions ago. Nevertheless, such long-term effects are often ignored or, at best, described as positive long-term priming. Recent findings, however, suggest a complex interaction between short and long-term statistics in modulating perceptual decisions, but these effects are not well understood. We conducted six 2-AFC visual shape discrimination experiments, in which we not only independently manipulated the appearance probabilities (APs) of abstract shapes over short and long time ranges, but also tested the effect of dynamically changing these probabilities. To assess the interaction between subjective uncertainty and past information, one of two possible shapes appeared in varying levels of Gaussian noise on each trial. First, we report that, in ambiguous trials, instead of simply being primed by earlier APs, subject made decisions so that it would compensate the discrepancy between recent and earlier APs. This behavior led to the paradoxical result that a stimulus presented more frequently in recent past was significantly not preferred if it was less frequent in the distant past. Second, we found that this compensatory mechanism did not take effect when the difference in APs between long past and recent times was introduced gradually rather than abruptly. This led to the paradox of false and lasting negative compensation when there was no difference in APs between past and present due to a momentary abrupt shift followed by a gradual return. We replicated our key human finding with behaving rats, demonstrating that these effects do not rely on explicit reasoning. Our results suggest that instead simply following the rule of gradually collected event statistics, perceptual decision making is influenced by a complex process in which statistics are weighted by significance due to detected environmental changes.
Meeting abstract presented at VSS 2016
This PDF is available to Subscribers Only