Alternative outcomes create biased expectations regarding the received outcome: Evidence from event-related potentials

Marciano, Déborah; Bentin, Shlomo; Deouell, Leon Y.

doi:10.1016/j.neuropsychologia.2018.03.010

Cited by 17 publications

(6 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The full results are presented in Supplementary Tables 7 and 8 and in Supplementary Figure 1. In line with past studies, we found we found that the FRN and the P3 were larger for Wins than No-Wins (all p's<, p<.001) [42][43][44] , [45][46][47][48] . Notably, NWB elicited larger P3 than the other No-Wins (all p's<=.001).…”

Section: Eeg Activity During Deceleration Varies By Outcome Consisten...supporting

confidence: 92%

Dynamic expectations: Behavioral and electrophysiological evidence of sub-second updates in reward predictions

Marciano

Bellier

Mayer

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Expectations are often dynamic: any sports fan knows that expectations are rapidly updated as games unfold. Yet expectations have traditionally been studied as static. Here, using slot machines as a case study, we provide parallel behavioral and electrophysiological evidence of sub-second moment-to-moment changes in expectations. In Study 1, we show that the dynamics of the EEG signal before the slot machine stopped differed depending on the nature of the outcome, including not only whether the participant won or lost, but also how close they came to winning. In line with our predictions, Near Win Before outcomes (the slot machine stops one item before a match) were similar to Wins, but different than Near Win After (the machine stops one item after a match) and Full Miss (the machine stops two or three items from a match). In Study 2, we designed a novel behavioral paradigm to measure moment-to-moment changes in expectations via dynamic betting. We found that different outcomes also elicited unique expectation trajectories in the deceleration phase. Notably, these behavioral expectation trajectories paralleled Study 1's EEG activity in the last second prior to the machine's stop. In Studies 3 (EEG) and 4 (behavior) we replicated these findings in the loss domain where a match entails a loss. Again, we found a significant correlation between behavioral and EEG results. These four studies provide the first evidence that dynamic sub-second updates in expectations can be behaviorally and electrophysiologically measured. Our findings open up new avenues for studying the ongoing dynamics of reward expectations and their role in healthy and unhealthy cognition.

show abstract

Section: Eeg Activity During Deceleration Varies By Outcome Consisten...supporting

confidence: 92%

Dynamic expectations: Behavioral and electrophysiological evidence of sub-second updates in reward predictions

Marciano

Bellier

Mayer

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…This fictive learning heuristic is rational when one option is systematically better than other one: a situation that was verified in many of the experiments considered here (but not all conditions, such as the 'symmetric' ones). Of note, behavioral studies showed that subjects tend to assume an anticorrelation between obtained and forgone outcomes, even when by both design and instructions are explicitly told not to do so [42,43]. We believe that this fictive update heuristic is plausible and may play a role, even if it suffers from a strong limitation: it only applies to partial feedback conditions (3 out of the 9 experiments under consideration involved complete feedback).…”

Section: Discussionmentioning

confidence: 89%

Choice-confirmation bias and gradual perseveration in human reinforcement learning

Palminteri¹

2021

Preprint

View full text Add to dashboard Cite

Do we preferentially learn from outcomes that confirm our choices? This is one of the most basic, and yet consequence-bearing, questions concerning reinforcement learning. In recent years, we investigated this question in a series of studies implementing increasingly complex behavioral protocols. The learning rates fitted in experiments featuring partial or complete feedback, as well as free and forced choices, were systematically found to be consistent with a choice-confirmation bias. This result is robust across a broad range of outcome contingencies and response modalities. One of the prominent behavioral consequences of the confirmatory learning rate pattern is choice hysteresis: that is the tendency of repeating previous choices, despite contradictory evidence. As robust and replicable as they have proven to be, these findings were (legitimately) challenged by a couple of studies pointing out that a choice-confirmatory pattern of learning rates may spuriously arise from not taking into consideration an explicit choice autocorrelation term in the model. In the present study, we re-analyze data from four previously published papers (in total nine experiments; N=363), originally included in the studies demonstrating (or criticizing) the choice-confirmation bias in human participants. We fitted two models: one featured valence-specific updates (i.e., different learning rates for confirmatory and disconfirmatory outcomes) and one additionally including an explicit choice autocorrelation process (gradual perseveration). Our analysis confirms that the inclusion of the gradual perseveration process in the model significantly reduces the estimated choice-confirmation bias. However, in all considered experiments, the choice-confirmation bias remains present at the meta-analytical level, and significantly different from zero in most experiments. Our results demonstrate that the choice-confirmation bias resists the inclusion of an explicit choice autocorrelation term, thus proving to be a robust feature of human reinforcement learning. We conclude by discussing the psychological plausibility of the gradual perseveration process in the context of these behavioral paradigms and by pointing to additional computational processes that may play an important role in estimating and interpreting the computational biases under scrutiny.

show abstract

“…This fictive learning heuristic is rational when one option is systematically better than other one: a situation that was verified in many of the experiments considered here (but not all conditions, such as the “symmetric” ones). Of note, behavioral studies showed that subjects tend to assume an anticorrelation between obtained and forgone outcomes, even when by both design and instructions are explicitly told not to do so (Marciano et al, 2018; Marciano-Romm et al, 2016). We believe that this fictive update heuristic is plausible and may play a role, even if it suffers from a strong limitation: It only applies to partial feedback conditions (three out of the nine experiments under consideration involved complete feedback).…”

Section: Discussionmentioning

confidence: 99%

Choice-confirmation bias and gradual perseveration in human reinforcement learning.

Palminteri¹

2023

Behavioral Neuroscience

View full text Add to dashboard Cite

Do we preferentially learn from outcomes that confirm our choices? In recent years, we investigated this question in a series of studies implementing increasingly complex behavioral protocols. The learning rates fitted in experiments featuring partial or complete feedback, as well as free and forced choices, were systematically found to be consistent with a choice-confirmation bias. One of the prominent behavioral consequences of the confirmatory learning rate pattern is choice hysteresis: that is, the tendency of repeating previous choices, despite contradictory evidence. However, choice-confirmatory pattern of learning rates may spuriously arise from not taking into consideration an explicit choice (gradual) perseveration term in the model. In the present study, we reanalyze data from four published papers (nine experiments; 363 subjects; 126,192 trials), originally included in the studies demonstrating or criticizing the choice-confirmation bias in human participants. We fitted two models: one featured valence-specific updates (i.e., different learning rates for confirmatory and disconfirmatory outcomes) and one additionally including gradual perseveration. Our analysis confirms that the inclusion of the gradual perseveration process in the model significantly reduces the estimated choice-confirmation bias. However, in all considered experiments, the choice-confirmation bias remains present at the meta-analytical level, and significantly different from zero in most experiments. Our results demonstrate that the choice-confirmation bias resists the inclusion of a gradual perseveration term, thus proving to be a robust feature of human reinforcement learning. We conclude by pointing to additional computational processes that may play an important role in estimating and interpreting the computational biases under scrutiny.

show abstract

Alternative outcomes create biased expectations regarding the received outcome: Evidence from event-related potentials

Cited by 17 publications

References 54 publications

Dynamic expectations: Behavioral and electrophysiological evidence of sub-second updates in reward predictions

Dynamic expectations: Behavioral and electrophysiological evidence of sub-second updates in reward predictions

Choice-confirmation bias and gradual perseveration in human reinforcement learning

Choice-confirmation bias and gradual perseveration in human reinforcement learning.

Contact Info

Product

Resources

About