Duration estimates within a modality are integrated sub-optimally

Cai, Ming Bo; Eagleman, David M.

doi:10.3389/fpsyg.2015.01041

Cited by 9 publications

(8 citation statements)

References 61 publications

(97 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Third, performance was better in 1-2-3-Go condition in which subjects made two measurements, as evidenced by a lower root-mean-square error (RMSE) in 1-2-3-Go compared to 1-2-Go condition ( Fig 2C; permutation test; p-value < 0.01 for all subjects). This observation indicates that subjects combined the two measurements to improve their estimates, corroborating reports from other behavioral paradigms [46][47][48][49][50][51][52][53][54]. Combined with the systematic bias toward the mean of t s , these results indicated that subjects integrated prior information with one or two measurements to improve their performance.…”

Section: Subjects Integrate Interval Measurements With Prior Knowledgesupporting

confidence: 85%

“…However, since priors and the cost functions impact both 1-2-Go and 1-2-3-Go conditions, moderate inaccuracies in modeling these components may not be able to explain optimal behavior in 1-2-Go and suboptimal behavior in the 1-2-3-Go condition simultaneously. Finally, recent results suggest the performance may be limited by imperfect integration [70,71] and imperfect memory [48,72], which future models of sequential updating should incorporate.…”

Section: /34mentioning

confidence: 99%

See 1 more Smart Citation

A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise

Egger

Jazayeri

2018

Preprint

View full text Add to dashboard Cite

Bayesian models of behavior have advanced the idea that humans combine prior beliefs and sensory observations to minimize uncertainty. How the brain implements Bayes-optimal inference, however, remains poorly understood. Simple behavioral tasks suggest that the brain can flexibly represent and manipulate probability distributions. An alternative view is that brain relies on simple algorithms that can implement Bayes-optimal behavior only when the computational demands are low. To distinguish between these alternatives, we devised a task in which Bayes-optimal performance could not be matched by simple algorithms. We asked subjects to estimate and reproduce a time interval by combining prior information with one or two sequential measurements. In the domain of time, measurement noise increases with duration. This property makes the integration of multiple measurements beyond the reach of simple algorithms. We found that subjects were able to update their estimates using the second measurement but their performance was suboptimal, suggesting that they were unable to update full probability distributions. Instead, subjects' behavior was consistent with an algorithm that predicts upcoming sensory signals, and applies a nonlinear function to errors in prediction to update estimates. These results indicate that inference strategies humans deploy may deviate from Bayes-optimal integration when the computational demands are high.

show abstract

Section: Subjects Integrate Interval Measurements With Prior Knowledgesupporting

confidence: 85%

Section: /34mentioning

confidence: 99%

A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise

Egger

Jazayeri

2018

Preprint

View full text Add to dashboard Cite

show abstract

“…The optimal integration of multiple sensory signals predicts an improvement in timing sensitivity. However, this prediction has not been confirmed experimentally in some studies 21 , 27 . Hartcher-O’Brien et al .…”

Section: Discussionmentioning

confidence: 79%

Optimal multisensory integration leads to optimal time estimation

Murai

Yotsumoto

2018

Sci Rep

View full text Add to dashboard Cite

Our brain compensates sensory uncertainty by combining multisensory information derived from an event, and by integrating the current sensory signal with the prior knowledge about the statistical structure of previous events. There is growing evidence that both strategies are statistically optimal; however, how these two stages of information integration interact and shape an optimal percept remains an open question. In the present study, we investigated the perception of time as an amodal perceptual attribute. The central tendency, a phenomenon of biasing the current percept toward previous stimuli, is used to quantify and model how the prior information affects the current timing behavior. We measured the timing sensitivity and the central tendency for unisensory and multisensory stimuli with sensory uncertainty systematically manipulated by adding noise. Psychophysical results demonstrate that the central tendency increases as the uncertainty increases, and that the multisensory timing improves both the timing sensitivity and the central tendency bias compared to the unisensory timing. Computational models indicate that the optimal multisensory integration precedes the optimal integration of prior information causing the central tendency. Our findings suggest that our brain incorporates the multisensory information and prior knowledge in a statistically optimal manner to realize precise and accurate timing behavior.

show abstract

“…In the previous study involving with the neural energy model, Eagleman and Pariyadath (2009) mainly used sequential stimuli and manipulated the repetition and expectation. Recently, Cai, Eagleman, and Ma (2015) presented several kinds of the stimulus sequences and measured the perceived duration of the last stimulus of the sequence. They also controlled the high-level expectation of the last stimulus by manipulating the probability of the stimulus sequence, regularity of the preceding stimuli in the sequence or whether the last stimulus broke the overlearned sequence.…”

Section: Discussionmentioning

confidence: 99%

“…Some of them explored the effect of numerical magnitude on time perception (e.g., Hayashi, Valli, & Carlson, 2013 ; Xuan, Zhang, He, & Chen, 2007 ). Other studies investigated how the perceived duration of a target stimulus was modulated by the other stimuli surrounding the target ( Ayhan, Revina, Bruno, & Johnston, 2012 ; Cai & Eagleman, 2015 ). In addition to these findings, we newly discovered that spatial patterns of the multiple objects on time perception.…”

Section: Discussionmentioning

confidence: 99%

Regular Is Longer

Sasaki

Yamada

2017

i-Perception

View full text Add to dashboard Cite

The current study examined whether regularity of dot patterns would influence time perception. We presented observers the dot patterns with three levels of regularity (high, middle and low) and measured the perceived duration of each pattern by bisection and rating methods. The results revealed that the perceived duration of high regular patterns was longer than that of middle and low regular patterns. Thus, we found that stimulus regularity is one of the factors that influence time perception.

show abstract

Duration estimates within a modality are integrated sub-optimally

Cited by 9 publications

References 61 publications

A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise

A nonlinear updating algorithm captures suboptimal inference in the presence of signal-dependent noise

Optimal multisensory integration leads to optimal time estimation

Regular Is Longer

Contact Info

Product

Resources

About