A detailed comparison of optimality and simplicity in perceptual decision making.

Abstract

Two prominent ideas in the study of decision-making have been that organisms behave near-optimally, and that they use simple heuristic rules. These principles might be operating in different types of tasks, but this possibility cannot be fully investigated without a direct, rigorous comparison within a single task. Such a comparison was lacking in most previous studies, because a) the optimal decision rule was simple; b) no simple suboptimal rules were considered; c) it was unclear what was optimal, or d) a simple rule could closely approximate the optimal rule. Here, we used a perceptual decision-making task in which the optimal decision rule is well-defined and complex, and makes qualitatively distinct predictions from many simple suboptimal rules. We find that all simple rules tested fail to describe human behavior, that the optimal rule accounts well for the data, and that several complex suboptimal rules are indistinguishable from the optimal one. Moreover, we found evidence that the optimal model is close to the true model: first, the better the trial-to-trial predictions of a suboptimal model agree with those of the optimal model, the better that suboptimal model fits; second, our estimate of the Kullback-Leibler divergence between the optimal model and the true model is not significantly different from zero. When observers receive no feedback, the optimal model still describes behavior best, suggesting that sensory uncertainty is implicitly represented and taken into account. Beyond the task and models studied here, our results have implications for best practices of model comparison.

Using the past to estimate sensory uncertainty

“…Critically, we also show that for continuous and discontinuous changes in sensory uncertainty that are common for perception in our natural environment this optimal Bayesian inference can be approximated by more simple strategies of exponential discounting that could be performed by slowly updating the sensory weights without explicit representation of uncertainty 3,9 . Future experiments that increase observers' uncertainty about their visual uncertainty estimate may potentially be able to dissociate Bayesian from approximate exponential discounting.…”

“…Most prominently, in multisensory perception the most reliable or 'Bayes-optimal' percept is obtained by integrating sensory signals weighted by their reliability (i.e., precision or inverse of variance) with less weight assigned to less reliable signals. Indeed, accumulating evidence suggests that human observers are close to optimal in many perceptual tasks (though see [1][2][3] and weight signals according to their sensory reliabilities [4][5][6][7][8] .…”

The influence of evidence volatility on choice, reaction time and confidence in a perceptual decision

“…This does not mean humans and monkeys are incapable of using information about stimulus reliability or difficulty to adjust their decision policy, and perhaps they would have in other circumstances (Qamar et al, 2013; Shen and Ma, 2016). For instance, had we used only a very difficult and a very easy condition, there would be a stronger incentive to ascertain the difficulty of the decision online and use different termination criteria for each condition.…”

Causal inference and temporal predictions in audiovisual perception of speech and music

“…Yet, they are likely to face difficulties accommodating the numerous potential causal structures underlying the sensory richness and complexity that is a characteristic of real‐world situations, such as an orchestra performance with many different players and instruments. This suggests that parametric models of Bayesian causal inference may define the normative and computational principles underlying multisensory integration and audiovisual scene analysis, yet the brain will need nonparametric or approximate inference mechanisms or even simple heuristics to solve causal inference problems facing the brain in our natural environment . Critically, irrespective of the exact computational algorithms, the brain may use multisensory causal inference that relies on a range of correspondence cues that indicate whether signals in different senses are attributable to common events in the environment.…”

“…This suggests that parametric models of Bayesian causal inference may define the normative and computational principles underlying multisensory integration and audiovisual scene analysis, yet the brain will need nonparametric or approximate inference mechanisms or even simple heuristics to solve causal inference problems facing the brain in our natural environment. 30,31 Critically, irrespective of the exact computational algorithms, the brain may use multisensory causal inference that relies on a range of correspondence cues that indicate whether signals in different senses are attributable to common events in the environment. Given the importance of temporal information for music and speech processing, we will next discuss predictive coding and internal sensorimotor forward models as two complementary mechanisms that may allow the brain to determine whether sensory signals come from a common source based on cross-sensory or sensorimotor temporal predictions.…”