Quantifying uncertainty in time perception: A modified reproduction method DOI Creative Commons
Jaume Boned, Joan López‐Moliner

i-Perception, Journal Year: 2024, Volume and Issue: 15(6)

Published: Sept. 1, 2024

In time perception research, we typically measure how an observer perceives intervals by collecting data from multiple trials with a single estimate recorded on each. However, this gives us limited information about the observer's uncertainty for each estimate, which usually variability across trials. Our study tested potential of modified reproduction task to provide duration as well single-trial basis. Participants were instructed press and hold key temporally bracket end learned (0.6–4 s) narrowly possible. Therefore, expected bracket's length indicate level uncertainty. We compared method conventional task. Taking mid-point found that both methods produced equivalent data. Critically, predicted variability, indicating obtained in individual trial could potentially much reproductions. Additionally, relative start positions suggests combination additive multiplicative noise components. findings highlight more efficient nuanced approach estimates their associated uncertainty, expanding methodological toolkit opening new avenues research.

Language: Английский

Estimating the contribution of early and late noise in vision from psychophysical data DOI Creative Commons
Jesús Malo, José J. Esteve‐Taboada, Guillermo Aguilar

et al.

Journal of Vision, Journal Year: 2025, Volume and Issue: 25(1), P. 12 - 12

Published: Jan. 13, 2025

Human performance in psychophysical detection and discrimination tasks is limited by inner noise. It unclear to what extent this noise arises from early (e.g., the photoreceptors) or late (at immediately prior decision stage, presumably cortex). Very likely, behaviorally limiting a nontrivial combination of both Here we propose method quantify contributions purely data. Our approach generalizes classical results for linear systems combining theory propagation through nonlinear network with expressions obtain perceptual metric network. We show that threshold-only data, relative can only be disentangled when experiments include substantial external When full psychometric functions are available, sources quantified even absence estimate magnitude noise—assuming standard cascade model stages—is substantially lower than cone photocurrents computed via an accurate retinal physiology, ISETBio. This consistent idea one fundamental vision reduce comparatively large

Language: Английский

Citations

0
A comparative analysis of perceptual noise in lateral and depth motion: Evidence from eye tracking DOI Creative Commons
Joan López‐Moliner

Journal of Vision, Journal Year: 2025, Volume and Issue: 25(1), P. 15 - 15

Published: Jan. 24, 2025

The characterization of how precisely we perceive visual speed has traditionally relied on psychophysical judgments in discrimination tasks. Such tasks are often considered laborious and susceptible to biases, particularly without the involvement highly trained participants. Additionally, thresholds for motion-in-depth perception frequently reported as higher compared lateral motion, a discrepancy that contrasts with everyday visuomotor In this research, rely smooth pursuit model, based Kalman filter, quantify observational uncertainties. This model allows us distinguish between additive multiplicative noise across three conditions motion dynamics within virtual reality setting: random walk, linear nonlinear incorporating both depth components. We aim assess tracking performance perceptual uncertainties versus motion-in-depth. alignment prior our results indicate diminished walk condition, characterized by unpredictable positioning. However, when velocity information is available facilitates predictions future positions, become more consistent in-depth motion. consistency noticeable ranges where retinal speeds overlap these two dimensions. Significantly, emerges primary source uncertainty, largely exceeding noise. predominance computational accounts Our study challenges earlier beliefs marked differences processing motions, suggesting similar levels uncertainty underscoring significant role

Language: Английский

Citations

0
Introspective psychophysics for the study of subjective experience DOI
Megan A. K. Peters

Cerebral Cortex, Journal Year: 2024, Volume and Issue: unknown

Published: Nov. 21, 2024

Abstract Studying subjective experience is hard. We believe that pain not identical to nociception, nor pleasure a computational reward signal, fear the activation of “threat circuitry”. Unfortunately, introspective self-reports offer our best bet for accessing experience, but many still introspection “unreliable” and “unverifiable”. But which introspection’s faults do we find most damning? Is it provides imperfect access brain processes (e.g. perception, memory)? That objectively verifiable? hard isolate from non-subjective processing capacity? Here, I argue none these prevents us building meaningful, impactful psychophysical research program treats as valid empirical target through precisely characterizing relationships among environmental variables, behavior, self-reported phenomenology. Following recent similar calls by Peters (Towards canonical computations generating phenomenal experience. 2022. Neurosci Biobehav Rev: 142, 104903), Kammerer Frankish (What forms could systems take? A programme. 2023. J Conscious Stud 30:13–48), Fleming (Metacognitive psychophysics in humans, animals, AI. 30:113–128), “introspective psychophysics” thus apparent features, bugs—just noise distortions linking environment behavior inspired Fechner’s over 150 years ago. This next generation will establish powerful tool testing precise explanatory models phenomenology across dimensions—urgency, emotion, clarity, vividness, confidence, more.

Language: Английский

Citations

3
Bayes vs. Weber: how to break a law of psychophysics DOI Creative Commons
Arthur Prat-Carrabin, Samuel J. Gershman

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 9, 2024

Abstract A classic result of psychophysics is that human perceptual estimates are more variable for larger magnitudes. This ‘Weber behavior’ has typically not been the focus prominent Bayesian paradigm, which models perception as an optimal statistical inference conducted on basis noisy internal signals. Here we examine variability a observer, in comparison with subjects. In two preregistered experiments, manipulate prior distribution and reward function numerosity-estimation task. When large numerosities frequent, when they rewarding, observer exhibits ‘anti-Weber behavior’, magnitudes results less responses. Human subjects exhibit similar pattern, thus breaking long-standing by showing opposite behavior. allows to minimize errors make about frequent or rewarding Nevertheless, model fitting suggests subjects’ responses best captured features logarithmic encoding, proposal Fechner often regarded accounting Weber We obtain anti-Weber behavior together encoding. Our suggest increasing may be primarily due skewness natural priors.

Language: Английский

Citations

1
Quantifying uncertainty in time perception: A modified reproduction method DOI Creative Commons
Jaume Boned, Joan López‐Moliner

i-Perception, Journal Year: 2024, Volume and Issue: 15(6)

Published: Sept. 1, 2024

In time perception research, we typically measure how an observer perceives intervals by collecting data from multiple trials with a single estimate recorded on each. However, this gives us limited information about the observer's uncertainty for each estimate, which usually variability across trials. Our study tested potential of modified reproduction task to provide duration as well single-trial basis. Participants were instructed press and hold key temporally bracket end learned (0.6–4 s) narrowly possible. Therefore, expected bracket's length indicate level uncertainty. We compared method conventional task. Taking mid-point found that both methods produced equivalent data. Critically, predicted variability, indicating obtained in individual trial could potentially much reproductions. Additionally, relative start positions suggests combination additive multiplicative noise components. findings highlight more efficient nuanced approach estimates their associated uncertainty, expanding methodological toolkit opening new avenues research.

Language: Английский

Citations

0