Magnitude-sensitivity: rethinking decision-making

Pirrone, Angelo; Reina, Andreagiovanni; Stafford, Tom; Marshall, James A. R.; Gobet, Fernand

doi:10.1016/j.tics.2021.10.006

Cited by 23 publications

(21 citation statements)

References 140 publications

(272 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This nonlinear relationship between perceptual noise and the magnitude of the stimulus is also similar to scalar variability, which has been observed in numerical cognition, timing and other magnitude descrimination studies (Fernandes & Church, 1982; Gallistel & Gelman, 2000; Gibbon, 1977; Mechner, 1958; Nieder & Dehaene, 2009). Given these similarities some researchers have suggested the possibility of shared mechanisms between evidence accumulation and the estimation of number, time and magnitude in general (Gebuis et al, 2017; Howard & Hasselmo, 2020; Pirrone et al, 2022; Simen et al, 2011, 2016).…”

Section: Discussionmentioning

confidence: 99%

Assessing evidence accumulation and rule learning in humans with an online game

Kane

McGuire

et al. 2022

Preprint

View full text Add to dashboard Cite

Evidence accumulation, how the brain integrates sensory information over time, is an essential component of perception and decision making. In humans, evidence accumulation is commonly modeled as a diffusion process in which noise accumulates linearly with the incoming evidence. However, recent studies in rodents have shown that during perceptual decision making, noise scales non-linearly with the strength of accumulated evidence. The question of whether nonlinear noise scaling also holds for humans has been clouded by differences in the methodologies typically used to collect and analyze human and rodent data. For example, whereas humans are typically given explicit instructions in these tasks, rodents are trained using feedback. Therefore, to evaluate how perceptual noise scales with accumulated evidence, we developed an online evidence accumulation game and nonverbal training pipeline for humans inspired by pulse-based evidence accumulation tasks for rodents. Using this game, we collected and analyzed behavioral data from hundreds of participants trained either with an explicit description of the relevant decision rule or merely with experiential feedback. Across all participants, performance was well described by an accumulation process, in which stimuli were integrated equally across time. Participants trained using feedback alone learned the game rules rapidly and used similar strategies to those who received explicit instructions. Decisions in both groups were influenced in similar ways by biases and perceptual noise, suggesting that explicit instructions did not reduce bias or noise in pulse-based accumulation tasks. Finally, by leveraging data across all participants, we show that perceptual noise during evidence accumulation was best described by a non-linear model of noise scaling, consistent with previous animal studies, but inconsistent with diffusion models widely used in human studies. These results challenge the conventional description of humans' accumulation process and suggest that online games inspired by evidence accumulation tasks provide a valuable large-scale behavioral assessment platform to examine perceptual decision making and learning in humans. In addition, the feedback-based training pipeline developed for this game may be useful for evaluating perceptual decision making in human populations with difficulty following verbal instructions.

show abstract

Section: Discussionmentioning

confidence: 99%

Assessing evidence accumulation and rule learning in humans with an online game

Kane

McGuire

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Additive integration explains notional distractor effects in binary-choice trials Notably, DV -HV also negatively covaries with HV + LV (Pearson's r(148) = -.78, p < .001), which potentially leads to another explanation of why binary choice accuracy seems lower as the matched DV -HV variable decreases. This explanation appeals to the divisive normalisation (DN) model based on EV 5,33 . Imagine choosing between two prospects, H and L, in two different choice-sets, H1 vs. L1, and H2 vs. L2, with the EV difference between H and L being the same across the two choice-sets (Fig.…”

Section: Integrating Reward and Probability Information Additively In...mentioning

confidence: 99%

Clarifying the role of an unavailable distractor in human multiattribute choice

Cao

Tsetsos

2022

Preprint

View full text Add to dashboard Cite

Decisions between two alternatives can be influenced by a third irrelevant decoy alternative. Although decoy effects typically depend on the decoy's relative position in a multiattribute choice space, recent studies using risky prospects (i.e., varying in reward and probability) reported a novel 'positive' decoy effect operating on a single value dimension: The higher the expected value of an unavailable (distractor) prospect was, the easier the discrimination between two available target prospects became, especially when their expected value difference was small. Here we ask whether this previously reported distractor effect affords alternative interpretations. Indeed, we find that it occurred because the distractor's expected value covaried positively with the subjective utility difference between the two targets. Looking beyond this covariation, we report a modest negative distractor effect, which in turn could conceivably be attributed to classic multiattribute decoy effects. Our findings clarify the modulatory role of an unavailable distracting option in multiattribute decisions.

show abstract

“…[1]), it has only recently been applied to value-based decisions [2][3][4]; such decisions differ from perceptual decisions because decision makers are rewarded by the value of the selected option, rather than whether or not they selected the best option (e.g. [3][4][5][6][7]). Recently researchers have analysed multi-alternative value-based decisionmaking [4], building on earlier work in optimal decision policies for binary value-based choices [3].…”

Section: Introductionmentioning

confidence: 99%

“…Interestingly, the theoretically optimal policy for the binary decision case [ 3 ] is inconsistent with empirical observations of magnitude-sensitive reaction-times ([ 5 , 10 – 14 ], but see [ 15 ]), unless assumptions are made that subjective utilities for decision-makers are nonlinear, or decisions are embedded in a fixed-length time period with known or learnable distributions of trial option values, so that a variable opportunity cost arises from decision time [ 3 ]. Furthermore, even single-trial dynamics have been observed to lead to magnitude sensitive reaction times [ 16 ].…”

Section: Introductionmentioning

confidence: 99%

“…Furthermore, even single-trial dynamics have been observed to lead to magnitude sensitive reaction times [ 16 ]. While some descriptive models of decision-making can capture aspects of magnitude-sensitivity in the binary case [ 10 , 17 , 18 ], those models have not yet been extended to the multi-alternative case [ 5 ], hence their performance in the multi-alternatives case is unknown. Here we do not consider which extension of descriptive models could account for magnitude-sensitivity with multiple alternatives.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Magnitude-sensitive reaction times reveal non-linear time costs in multi-alternative decision-making

et al. 2022

Self Cite

View full text Add to dashboard Cite

Optimality analysis of value-based decisions in binary and multi-alternative choice settings predicts that reaction times should be sensitive only to differences in stimulus magnitudes, but not to overall absolute stimulus magnitude. Yet experimental work in the binary case has shown magnitude sensitive reaction times, and theory shows that this can be explained by switching from linear to multiplicative time costs, but also by nonlinear subjective utility. Thus disentangling explanations for observed magnitude sensitive reaction times is difficult. Here for the first time we extend the theoretical analysis of geometric time-discounting to ternary choices, and present novel experimental evidence for magnitude-sensitivity in such decisions, in both humans and slime moulds. We consider the optimal policies for all possible combinations of linear and geometric time costs, and linear and nonlinear utility; interestingly, geometric discounting emerges as the predominant explanation for magnitude sensitivity.

show abstract

Magnitude-sensitivity: rethinking decision-making

Cited by 23 publications

References 140 publications

Assessing evidence accumulation and rule learning in humans with an online game

Assessing evidence accumulation and rule learning in humans with an online game

Clarifying the role of an unavailable distractor in human multiattribute choice

Magnitude-sensitive reaction times reveal non-linear time costs in multi-alternative decision-making

Contact Info

Product

Resources

About