CoChat
Sign Up
Cited by 161Multisensory perception and integrationVisual perception and processing mechanismsTactile and Sensory Interactions

Fusion of visual cues is not mandatory in children

Open access
Nardini, Marko, Bedford, Rachael, Mareschal, Denis·2010·Proceedings of the National Academy of Sciences
DOI 10.1073/pnas.1001699107PMC PMC2947870
Open this paper in CoChat — free account

Abstract

Human adults can go beyond the limits of individual sensory systems’ resolutions by integrating multiple estimates (e.g., vision and touch) to reduce uncertainty. Little is known about how this ability develops. Although some multisensory abilities are present from early infancy, it is not until age ≥8 y that children use multiple modalities to reduce sensory uncertainty. Here we show that uncertainty reduction by sensory integration does not emerge until 12 y even within the single modality of vision, in judgments of surface slant based on stereoscopic and texture information. However, adults’ integration of sensory information comes at a cost of losing access to the individual estimates that feed into the integrated percept (“sensory fusion”). By contrast, 6-y-olds do not experience fusion, but are able to keep stereo and texture information separate. This ability enables them to outperform adults when discriminating stimuli in which these information sources conflict. Further, unlike adults, 6-y-olds show speed gains consistent with following the fastest-available single cue. Therefore, whereas the mature visual system is optimized for reducing sensory uncertainty, the developing visual system may be optimized for speed and for detecting sensory conflicts. Such conflicts could provide the error signals needed to learn the relationships between sensory information sources and to recalibrate them while the body is growing.

AI snapshot

Written from the paper’s full text.

TL;DR

Nardini et al. show that children do not integrate stereoscopic and texture cues to surface slant until about age 12, and that 6-year-olds do not experience adult-like sensory fusion. Instead, 6-year-olds keep the cues separate, which lets them detect cue conflicts and gain speed from multiple cues.

Key findings

  • Adults and 12-year-olds had significantly lower slant-discrimination thresholds when disparity and texture were presented together (DT) than with either cue alone, matching optimal integration predictions; 6- and 10-year-olds’ DT thresholds differed significantly from ideal-observer predictions (6 y: t(18) = 3.1, P = 0.006; 10 y: t(17) = 2.8, P = 0.012).
  • Reliability-based reweighting was present at 10 and 12 years and in adults, but not at 8 years; 6-year-olds weighted texture more when it was less reliable, opposite to optimal weighting.
  • Adults showed fusion: in the conflicting two-cue condition T+D−, their mean d′ was significantly lower than for either single cue (t(19) = 4.5, P < 0.001 vs. T+; t(19) = 3.6, P < 0.01 vs. D−). Six-year-olds were not subject to fusion and performed consistent with probability summation from single cues.
  • Six-year-olds were significantly faster with two cues than with either single cue at the hardest level, 5° (t(14) = 3.0, P = 0.02 vs. D; t(14) = 2.5, P = 0.03 vs. T), whereas adults were never faster given two cues than either one alone.

Methods

  • Three psychophysical experiments measured surface-slant judgments from stereoscopic disparity and texture cues in children aged 6, 8, 10, and 12 years and adults. Experiment 1 used staircases to measure 75% discrimination thresholds and point-of-subjective-equality weightings; Experiment 2 used same/different elliptical-disc judgments and d′ with congruent and conflicting cues; Experiment 3 used constant stimuli to measure accuracy and response latency.
  • Samples: Experiment 1 included n = 19 (6 y), 17 (8 y), 18 (10 y), 19 (12 y), and 19 adults; Experiment 2 included 20 6-year-olds and 20 adults; Experiment 3 included 15 6-year-olds and 17 adults. Participants were screened with the TNO stereo test, and some were excluded for failing stereo, chance performance, or unmeasurable thresholds.
  • Analyses included ANOVA, t-tests, psychometric function fits, ideal-observer predictions for cue integration, and probability-summation models for single-cue responding.

Limitations

  • The paper states that the hypothesis that integrating cues precludes recalibration “still needs to be tested directly,” and that an explicit model of speed and accuracy for decisions based on integrated vs. single cues “needs to be developed.” It also notes that developmental trajectories for cue weighting and cue integration may be uneven and should be investigated further with psychophysics and neuroimaging.
  • Evident from the text: the design is cross-sectional rather than longitudinal, fusion and latency experiments tested only 6-year-olds versus adults, and the findings are specific to visual slant judgments from disparity and texture.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1073/pnas.1001699107

Cites

38 of 38 references verified0 citation errors · 0 not checkable

Statistics

28 tests recomputed · 3 with the reported and recomputed p on opposite sides of .05

Claims the paper makes

0 of 34 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § Results

    Stimuli in the conflicting condition were marginally more discriminable than complete fusion would predict, suggesting that adult observers retained some access to single cues.

    “Stimuli in this condition were marginally more discriminable than complete fusion would predict [Fig. 2B, yellow circle; t(19) = 2.0, P = 0.056], suggesting that adult observers retained some access to single cues (as in ref. 12).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites ref. 12

      Not checked in the free run.

  • Findingown data
    § Results

    At intermediate levels not near floor or ceiling, adults were significantly more accurate given both disparity and texture together (DT) than either D or T alone.

    “At intermediate levels not near floor or ceiling, adults were significantly more accurate given both disparity and texture together (DT) than either D or T alone [Fig. 3; at 5°, t(16) = 5.4, P < 0.001 vs. D, t(16) = 3.6, P < 0.01 vs. T; at 7°, t(16) = 2.7, P = 0.01 vs. D, t(16) = 2.3, P = 0.04 vs. T].”
    Quote in the paperSubstantial claim
  • Findingown data
    § Results

    8-y-olds showed no significant difference in weighting.

    “By contrast, 8-y-olds showed no significant difference in weighting [t(16) = 0.7, P = 0.51]”
    Quote in the paperSubstantial claim
  • Findingown data
    § Results

    Younger participants did not have significantly lower thresholds given DT vs. their best single cue.

    “Younger participants did not have significantly lower thresholds given DT vs. their best single cue, although all groups (especially 8-y-olds) showed a trend for improvement relative to both cues (Fig. 1B).”
    Quote in the paperSubstantial claim
  • Findingown data
    § Results

    Variability across observers was sufficiently low to show significant differences between measured DT thresholds and ideal observer predictions at 6 and 10 y.

    “Variability across observers was also sufficiently low to show significant differences between measured DT thresholds and ideal observer predictions at 6 and 10 y [at 6 y, t(18) = 3.1, P = 0.006 ; at 10 y, t(17) = 2.8, P = 0.012] (Fig. 1B).”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Sensory estimates are independent and have uncorrelated Gaussian noise.
  • May combine two claims. 29
  • Classified as probably not a substantial claim. 26

  • May combine two claims.

    Findingown data
    § Results

    Adults' mean d′ was significantly higher for T+D+ than for either single cue and matched model predictions for integration of cues.

    “Adults’ mean d′ was significantly higher for T+D+ than for either single cue [Fig. 2B; t(19) = 3.3, P < 0.01 vs. D+; t(19) = 3.6, P < 0.01 vs. T+] and matched model predictions for integration of cues (Fig. 2B, yellow circle).”
    Quote in the paperSubstantial claim
    • The authors state this as an assumption: Sensory estimates are independent and have uncorrelated Gaussian noise.

  • May combine two claims.

    Findingown data
    § Results

    Six-year-olds' mean d′ in T+D− was not lower than the worst single cue, but intermediate to the two, indicating that they were not subject to fusion.

    “Six-year-olds’ mean d′ in this same condition (T+D−) was not lower than the worst single cue, but intermediate to the two, indicating that they were not subject to fusion.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    Mean 75% discrimination thresholds declined with age, showing improving sensitivity to slant in all conditions.

    “Mean 75% discrimination thresholds declined with age, showing improving sensitivity to slant in all conditions [Fig. 1B; linear effects of age group in ANOVA, F(1, 4) = 3.4, P < 0.001 for D; F(1, 4) = 2.3, P < 0.01 for T; F(1, 4) = 5.8, P < 0.001 for DT].”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Results

    Six-year-olds' performance in T+D− was consistent with a probability summation model predicting responses based on the single-component cues, but not consistent with fusion.

    “Their performance was consistent with a probability summation model predicting responses based on the single-component cues [Fig. 2B, black circle; t(19) = 1.0, P = 0.32], but not consistent with fusion [Fig. 2B, yellow circle; t(19) = 3.3, P < 0.01].”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    Six-year-olds showed neither ability, whereas 8- and 10-y-olds were transitional in showing, respectively, a (nonsignificant) trend toward a lower threshold without reweighting, and reweighting without a lower threshold.

    “Six-year-olds showed neither ability, whereas 8- and 10-y-olds were transitional in showing, respectively, a (nonsignificant) trend toward a lower threshold without reweighting, and reweighting without a lower threshold.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    Six-year-olds' mean d′ was similar to that for the single best cue, did not match predictions for integration, but was consistent with a probability summation model predicting responses based on single cues.

    “Six-year-olds’ mean d′ was similar to that for the single best cue, did not match predictions for integration (Fig. 2B, yellow circle), but was consistent with a probability summation model predicting responses based on single cues [Fig. 2B, black circle; t(19) = 0.6, P = 0.59].”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    Whereas adults were never faster given two cues (DT) than either one alone, 6-y-olds were significantly faster given two cues than either single cue at their most difficult stimulus level, 5°.

    “However, whereas adults were never faster given two cues (DT) than either one alone, 6-y-olds were significantly faster given two cues than either single cue at their most difficult stimulus level, 5°; t(14) = 3.0, P = 0.02 vs. D, t(14) = 2.5, P = 0.03 vs. T.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    Ten- and 12-y-olds and adults weighted texture significantly more for high base slants, consistent with taking reliability into account.

    “Ten- and 12-y-olds and adults weighted texture significantly more for high base slants (Fig. 2C), consistent with taking reliability into account [10 y, t(17) = 2.9, P < 0.01; 12 y, t(18) = 2.5, P = 0.02; adults, t(18) = 3.1, P < 0.01].”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Results

    At easier levels, both groups showed significant speed gains for DT relative to D, but not T, consistent with simply relying on texture.

    “At easier levels, both groups showed significant speed gains for DT relative to D, but not T (Fig. 3), consistent with simply relying on texture.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data
    § Materials and Methods

    Some participants were distractable, with long response latencies due to inattention to the task.

    “Under these circumstances some participants were distractable, with long response latencies due to inattention to the task.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data
    § Materials and Methods

    In pilot work with adults, fusion could be found only with planes whose edges were visible.

    “In pilot work with adults we could find fusion only with planes whose edges were visible; therefore we presented elliptical discs cut from planes similar to those in the first experiment (Fig. S2).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Discussion

    6-year-old children were not confused by conflicting visual stimuli that produce sensory fusion in adults, and they remained able to make judgments based on the component cues.

    “First, 6-y-olds were not confused by conflicting stimuli with which adults experience sensory fusion, but remained able to make judgments on the basis of the component cues.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data
    § Discussion

    A major finding across all three experiments is that whereas adults gained sensitivity to slant given disparity and texture together vs. either one alone, 6-y-olds did not.

    “A major finding across all three experiments is that whereas adults gained sensitivity to slant given disparity and texture together vs. either one alone, 6-y-olds did not.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation
    § Abstract

    Some multisensory abilities are present from early infancy, but children do not integrate information across modalities to reduce sensory uncertainty until age ≥8 y.

    “Although some multisensory abilities are present from early infancy (6–8), in recent studies, children did not integrate information across modalities for shape discrimination, spatial localization, or detection of visual–auditory events until age ≥8 y (9–11).”
    • Also stated in: § Abstract · plain
      “Although some multisensory abilities are present from early infancy, it is not until age ≥8 y that children use multiple modalities to reduce sensory uncertainty.”
    Quote in the paperSubstantial claim0 of 6 sources checked
    • cites [6] Kuhl PK, Meltzoff AN. The bimodal perception of speech in infancy. Science. 1982;218:1138–1141. doi: 10.1126/science.7146899. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [7] Bahrick LE, Lickliter R. Infants’ perception of rhythm and tempo in unimodal and multimodal stimulation: A developmental test of the intersensory redundancy hypothesis. Cogn Affect Behav Neurosci. 2004;4:137–147. doi: 10.3758/cabn.4.2.137. [DOI] [PMC free article] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [8] Sann C, Streri A. Perception of object shape and texture in human newborns: Evidence from cross-modal transfer tasks. Dev Sci. 2007;10:399–410. doi: 10.1111/j.1467-7687.2007.00593.x. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [9] Nardini M, Jones P, Bedford R, Braddick O. Development of cue integration in human navigation. Curr Biol. 2008;18:689–693. doi: 10.1016/j.cub.2008.04.021. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [10] Gori M, Del Viva M, Sandini G, Burr DC. Young children do not integrate visual and haptic form information. Curr Biol. 2008;18:694–698. doi: 10.1016/j.cub.2008.04.036. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [11] Barutchu A, Crewther DP, Crewther SG. The race that precedes coactivation: Development of multisensory facilitation in children. Dev Sci. 2009;12:464–473. doi: 10.1111/j.1467-7687.2008.00782.x. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Results

    Adults' performance indicated integration of cues, whereas 6-y-olds' performance was consistent with responding on the basis of single cues.

    “In sum, adults’ performance indicated integration of cues, whereas 6-y-olds’ performance was consistent with responding on the basis of single cues.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Results

    Integrating cues conferred an advantage on adults when the cues agreed, whereas not integrating cues conferred an advantage on 6-y-olds when the cues disagreed.

    “Whereas integrating cues conferred an advantage on adults when the cues agreed (congruent condition T+D+), not integrating cues conferred an advantage on 6-y-olds when the cues disagreed (conflicting condition T+D−).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationreasoning
    § Abstract

    The mature visual system is optimized for reducing sensory uncertainty, whereas the developing visual system may be optimized for speed and for detecting sensory conflicts.

    “Therefore, whereas the mature visual system is optimized for reducing sensory uncertainty, the developing visual system may be optimized for speed and for detecting sensory conflicts.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductown data
    § Results

    Mean overall proportions of correct responses were similar for adults and 6-y-olds, showing that the tasks were comparable in difficulty.

    “Mean overall proportions of correct responses were similar for adults (mean = 0.80, SEM = 0.01) and 6-y-olds [mean = 0.76, SEM = 0.02; t(30) = 1.8, P = 0.09], showing that the tasks were comparable in difficulty.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Results

    Adults improved in accuracy given two information sources, consistent with integration, whereas children showed no improvement, consistent with responding on the basis of single cues.

    “Thus, adults improved in accuracy given two information sources, consistent with integration, whereas children showed no improvement, consistent with responding on the basis of single cues.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S1 (asserted), S4 (based).

    Premisereasoning
    § Results

    If children do not integrate visual information, they should not be subject to fusion.

    “However, if children do not integrate visual information, they should not be subject to fusion.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Premisecitation
    § Results

    Binocular disparity provides a second source of slant information.

    “A second source of slant information comes from binocular disparity (16).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [16] Rogers BJ, Bradshaw MF. Disparity scaling and the perception of frontoparallel surfaces. Perception. 1995;24:155–179. doi: 10.1068/p240155. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § Results

    The gradient of change in element size and density in a homogenously tiled surface's projection ("texture") provides information about its slant.

    “The gradient of change in element size and density in a homogenously tiled surface's projection (“texture”) provides information about its slant (15).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [15] Knill DC. Surface orientation from texture: Ideal observers, generic observers and the information content of texture cues. Vision Res. 1998;38:1655–1682. doi: 10.1016/s0042-6989(97)00324-6. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation
    § Results

    Adults integrate texture and disparity cues to reduce the uncertainty of their estimates of slant, adjusting their relative weightings as they change in reliability.

    “Adults integrate texture and disparity cues to reduce the uncertainty of their estimates of slant, adjusting their relative weightings as they change in reliability (3).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [3] Hillis JM, Watt SJ, Landy MS, Banks MS. Slant from texture and disparity cues: Optimal cue combination. J Vis. 2004;4:1–24. doi: 10.1167/4.12.1. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Results

    Adults but not 6-y-olds showed improved slant discrimination consistent with integrating congruent cues.

    “Thus, adults but not 6-y-olds showed improved slant discrimination consistent with integrating congruent cues.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation
    § Abstract

    Human adults can reduce sensory uncertainty by integrating estimates, both across modalities and within a modality.

    “Human adults can reduce sensory uncertainty by integrating estimates, both across modalities (1, 2) (e.g., integrating vision and touch to judge size) and within a modality (3) (e.g., integrating visual stereoscopic and texture information to judge surface slant).”
    • Also stated in: § Abstract · plain
      “Human adults can go beyond the limits of individual sensory systems’ resolutions by integrating multiple estimates (e.g., vision and touch) to reduce uncertainty.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [1] Ernst MO, Banks MS. Humans integrate visual and haptic information in a statistically optimal fashion. Nature. 2002;415:429–433. doi: 10.1038/415429a. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [2] Alais D, Burr D. The ventriloquist effect results from near-optimal bimodal integration. Curr Biol. 2004;14:257–262. doi: 10.1016/j.cub.2004.01.029. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [3] Hillis JM, Watt SJ, Landy MS, Banks MS. Slant from texture and disparity cues: Optimal cue combination. J Vis. 2004;4:1–24. doi: 10.1167/4.12.1. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Results

    These younger groups' behavior indicates that they either misperceived the cues' reliabilities or used reliability information in a manner different from that required to weight cues for optimal uncertainty reduction.

    “These younger groups’ behavior indicates that they either misperceived the cues’ reliabilities or used reliability information in a manner different from that required to weight cues for optimal uncertainty reduction.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given
    § Materials and Methods

    Participants with 5% of responses >5 s and one child at chance on stereo judgments were excluded from analysis.

    “We excluded from analysis those with 5% of responses >5 s (3 children, 2 adults) and one child who was at chance on stereo judgments.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning
    § Discussion

    Normal early sensory experience is likely also important for development of cue integration within vision.

    “Normal early sensory experience appears crucial for development of cross-modal perception (28, 29) and is likely also to be important for development of cue integration within vision.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data
    § Discussion

    This result was found using a staircase procedure (experiment 1), same/different judgments (experiment 2), and the method of constant stimuli (experiment 3).

    “This result was found using a staircase procedure (experiment 1), same/different judgments (experiment 2), and the method of constant stimuli (experiment 3).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given
    § Materials and Methods

    The estimate of 6-year-olds' lapse rate related to interpretation of results from experiment 1 was made using data from all trials irrespective of their latency, before excluding any trials or participants.

    “The estimate of 6-y-olds’ lapse rate related to interpretation of results from experiment 1 was made using data from all trials irrespective of their latency, i.e., before excluding any trials or participants.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given
    § Materials and Methods

    Participants with d′ scores ≤0 on more than one condition were excluded from analysis.

    “Participants with d′ scores ≤0 on more than one condition were excluded from analysis (n = 4; two children, two adults); in all four excluded participants the ≤0 scores were in single-cue conditions.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § Materials and Methods

    Conditions were mixed so that observers would not learn the strategy of looking for one direction of slant difference during a block.

    “Conditions were mixed so that observers would not learn the strategy of looking for one direction of slant difference during a block.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Discussion

    The coupling of vision and action must be fast enough to keep up with events in the world.

    “The coupling of vision and action must be fast enough to keep up with events in the world (30).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [30] Ballard DH, Hayhoe MM, Pook PK, Rao RP. Deictic codes for the embodiment of cognition. Behav Brain Sci. 1997;20:723–742;. doi: 10.1017/s0140525x97001611. discussion 743–767. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • May combine two claims.

    Interpretationown data
    § Results

    6-y-olds exploited multiple cues for speed, tending to follow the fastest-available single cue on each trial.

    “Therefore, 6-y-olds exploited multiple cues for speed, tending to follow the fastest-available single cue on each trial.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given
    § Discussion

    Developmental trajectories for cue weighting and cue integration between 6 and 12 years may be uneven.

    “The developmental trajectories for cue weighting and cue integration between 6 and 12 y may be uneven (Fig. 1)”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationreasoning
    § Discussion

    The developing visual system may be optimized for speed and for detecting sensory conflicts.

    “Whereas the mature visual system is optimized for reducing sensory uncertainty, the developing visual system may be optimized for speed and for detecting sensory conflicts.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Interpretationreasoning
    § Discussion

    Sensory conflicts provide an error signal that could be used to calibrate sensory information sources against each other, including while the body is growing.

    “Such conflicts provide an error signal that could be used to calibrate sensory information sources against each other.”
    • Also stated in: § Discussion · hedged
      “Such conflicts may provide the error signals needed to learn the relationships between sensory information sources and to recalibrate them while the body is growing.”
    Quote in the paperSubstantial claim

  • Classified as probably not a substantial claim.

    Study design and conductreasoning
    § Materials and Methods

    Participants with an estimated threshold >23.5° in any condition were excluded from analysis.

    “We therefore excluded participants with an estimated threshold >23.5° in any condition from analysis (n = 9; six aged 6 y, four high disparity-only and two high texture-only thresholds; two aged 8 y, both high texture-only thresholds; and one aged 12 y, high disparity-only threshold).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Discussion

    Normal early sensory experience appears crucial for development of cross-modal perception.

    “Normal early sensory experience appears crucial for development of cross-modal perception (28, 29)”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [28] Schorr EA, Fox NA, van Wassenhove V, Knudsen EI. Auditory-visual fusion in speech perception in children with cochlear implants. Proc Natl Acad Sci USA. 2005;102:18748–18750. doi: 10.1073/pnas.0508862102. [DOI] [PMC free article] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [29] Putzar L, Goerendt I, Lange K, Rösler F, Röder B. Early visual deprivation impairs multisensory interactions in humans. Nat Neurosci. 2007;10:1243–1245. doi: 10.1038/nn1978. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Interpretationreasoning
    § Materials and Methods

    The strategy of looking for one direction of slant difference could artificially lower performance on fusion condition T+D− that includes both directions.

    “This strategy could artificially lower performance on fusion condition T+D− that includes both directions.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning
    § Discussion

    Remaining able to detect sensory conflicts into late childhood could be adaptive for growing children.

    “Remaining able to detect sensory conflicts into late childhood could also be adaptive for growing children.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: Changing interocular distance alters the relationship between disparity and texture cues to slant.

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductcitation
    § Materials and Methods

    Participants who failed the TNO stereo test, performed at chance, or had thresholds above the measurable range were excluded.

    “Participants who failed the TNO stereo test (33), performed at chance, or had thresholds above the measurable range were excluded.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [33] Cooper J, Feldman J, Medlin D. Comparing stereoscopic performance of children using the Titmus, TNO, and Randot stereo tests. J Am Optom Assoc. 1979;50:821–825. [PubMed] [Google Scholar]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation
    § Discussion

    Recent models propose that uncertainty reduction depends on neural computations implementing weighted averaging of sensory estimates.

    “Recent models propose that uncertainty reduction depends on neural computations implementing weighted averaging of sensory estimates (24, 25).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [24] Gu Y, Angelaki DE, Deangelis GC. Neural correlates of multisensory cue integration in macaque MSTd. Nat Neurosci. 2008;11:1201–1210. doi: 10.1038/nn.2191. [DOI] [PMC free article] [PubMed] [Google Scholar] source

      Not checked in the free run.

    • cites [25] Beck JM, et al. Probabilistic population codes for Bayesian decision making. Neuron. 2008;60:1142–1152. doi: 10.1016/j.neuron.2008.09.021. [DOI] [PMC free article] [PubMed] [Google Scholar] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data
    § Materials and Methods

    On this criterion, 2.4% of 6-year-old and 1.2% of adult trials were excluded.

    “On this criterion, 2.4% of 6-y-old and 1.2% of adult trials were excluded.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data
    § Discussion

    Adults are optimized for accuracy, whereas children may be optimized for speed.

    “Thus, whereas adults are optimized for accuracy, children may be optimized for speed.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductcitation
    § Materials and Methods

    Trials that were extreme outliers, defined as latencies ≥3× the third quartile + the interquartile range of latencies for each observer, were excluded.

    “We excluded any trials that were “extreme outliers” (generally due to inattention to the task, e.g., stopping to ask the experimenter a question), defined as latencies ≥3× the third quartile + the interquartile range of latencies for each observer (36).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [36] Tukey JW. Exploratory Data Analysis. Reading, MA: Addison–Wesley; 1977. [Google Scholar]

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductreasoning
    § Materials and Methods

    To make results comparable with the integration experiment, speed was not emphasized; participants were only encouraged to concentrate for each set of 10 trials without talking.

    “To make results comparable with the integration experiment, we did not emphasize speed, but only encouraged participants to “concentrate” for each set of 10 trials without talking.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Study design and conductown data
    § Materials and Methods

    All participants passed the TNO test for stereo vision or were correct more often than chance on disparity-only trials.

    “All participants passed the TNO test for stereo vision (33) or were correct more often than chance (binomial test) on disparity-only (D) trials.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [33] Cooper J, Feldman J, Medlin D. Comparing stereoscopic performance of children using the Titmus, TNO, and Randot stereo tests. J Am Optom Assoc. 1979;50:821–825. [PubMed] [Google Scholar]

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductcitation
    § Materials and Methods

    The study's assumptions and approach are based on previous more detailed validation of similar stimuli.

    “Our assumptions and approach are based on previous more detailed validation of similar stimuli (3).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [3] Hillis JM, Watt SJ, Landy MS, Banks MS. Slant from texture and disparity cues: Optimal cue combination. J Vis. 2004;4:1–24. doi: 10.1167/4.12.1. [DOI] [PubMed] [Google Scholar] source

      Not checked in the free run.

49 of 49

  • The authors state this as an assumption.

    Sensory estimates are independent and have uncorrelated Gaussian noise.

    “Given independent estimates with uncorrelated Gaussian noise, the optimal reduction in uncertainty (variance) is obtained by a weighted averaging of estimates, in which each estimate is weighted in proportion to its relative reliability (1/variance) (4, 5).”

  • The paper relies on this without support.

    Changing interocular distance alters the relationship between disparity and texture cues to slant.

    “For example, changing interocular distance alters the relationship between disparity and texture cues to slant.”

  • The authors state this as an assumption.

    Texture and disparity noise is uncorrelated.

    “The ideal observer prediction for condition DT (Fig. 1B, circles) was formulated assuming that (i) texture and disparity noise is uncorrelated and (ii) texture and disparity information available in single-cue conditions D and T provides as good a basis for judging slant as the information available in two-cue condition DT.”

  • The authors state this as an assumption.

    Texture and disparity information available in single-cue conditions D and T provides as good a basis for judging slant as the information available in two-cue condition DT.

    “The ideal observer prediction for condition DT (Fig. 1B, circles) was formulated assuming that (i) texture and disparity noise is uncorrelated and (ii) texture and disparity information available in single-cue conditions D and T provides as good a basis for judging slant as the information available in two-cue condition DT.”
4 of 4

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Evidence from psychophysics suggests that some kind of training via detection of sensory conflicts may be occurring in children.

    “Evidence from psychophysics suggests that some kind of training via detection of sensory conflicts may be occurring in children (Nardini et al., 2010).”

    From Toward an integration of deep learning and neuroscience (2016) · cites this paper as Nardini et al., 2010

Statistics that don’t add up

Each reported p-value recomputed from its test statistic and degrees of freedom.

  • Inconsistent p-value across alpha = .05: F(1, 4) = 2.3, p < .01

    Major

    The paper reports p < .01 here; recomputing from the test statistic and its degrees of freedom gives p = .204. The two fall on opposite sides of alpha = .05. The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p < .01 · recomputed p = .204

  • Inconsistent p-value across alpha = .05: F(1, 4) = 3.4, p < .001

    Major

    The paper reports p < .001 here; recomputing from the test statistic and its degrees of freedom gives p = .139. The two fall on opposite sides of alpha = .05. The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p < .001 · recomputed p = .139

  • Inconsistent p-value across alpha = .05: F(1, 4) = 5.8, p < .001

    Major

    The paper reports p < .001 here; recomputing from the test statistic and its degrees of freedom gives p = .074. The two fall on opposite sides of alpha = .05. The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p < .001 · recomputed p = .074

  • Inconsistent p-value: t(14) = 3.0, p = .02

    Minor

    The paper reports p = .02 here; recomputing from the test statistic and its degrees of freedom gives p = .010 (two-tailed). The check cannot tell whether the statistic or the p-value is the misprint.

    Reported p = .02 · recomputed p = .010

4 of 4

Statistics recomputed

Every reported test result, recomputed from its test statistic and degrees of freedom. A result shows as not adding up only once an independent check has confirmed it.

  • Doesn’t add upF(1, 4) = 3.4reported p < .001 · recomputed p = .139The two fall on opposite sides of .05

    “F(1, 4) = 3.4, p < .001”

  • Doesn’t add upF(1, 4) = 2.3reported p < .01 · recomputed p = .204The two fall on opposite sides of .05

    “F(1, 4) = 2.3, p < .01”

  • Doesn’t add upF(1, 4) = 5.8reported p < .001 · recomputed p = .074The two fall on opposite sides of .05

    “F(1, 4) = 5.8, p < .001”

  • Doesn’t add upt(14) = 3reported p = .02 · recomputed p = .010

    “t(14) = 3.0, p = .02”

  • Consistentt(18) = 2.1reported p = .047 · recomputed p = .050

    “t(18) = 2.1, p = .047”

  • Consistentt(18) = 2.2reported p = .039 · recomputed p = .041

    “t(18) = 2.2, p = .039”

  • Consistentt(18) = 2.5reported p = .02 · recomputed p = .022

    “t(18) = 2.5, p = .020”

  • Consistentt(18) = 5.8reported p < .001 · recomputed p < .001

    “t(18) = 5.8, p < .001”

  • Consistentt(18) = 3.1reported p = .006 · recomputed p = .006

    “t(18) = 3.1, p = .006”

  • Consistentt(17) = 2.8reported p = .012 · recomputed p = .012

    “t(17) = 2.8, p = .012”

  • Consistentt(17) = 2.9reported p < .01 · recomputed p = .010

    “t(17) = 2.9, p < .01”

  • Consistentt(18) = 2.5reported p = .02 · recomputed p = .022

    “t(18) = 2.5, p = .02”

  • Consistentt(18) = 3.1reported p < .01 · recomputed p = .006

    “t(18) = 3.1, p < .01”

  • Consistentt(16) = 0.7reported p = .51 · recomputed p = .494

    “t(16) = 0.7, p = .51”

  • Consistentt(18) = -3reported p < .01 · recomputed p = .008

    “t(18) = −3.0, p < .01”

  • Consistentt(19) = 3.3reported p < .01 · recomputed p = .004

    “t(19) = 3.3, p < .01”

  • Consistentt(19) = 3.6reported p < .01 · recomputed p = .002

    “t(19) = 3.6, p < .01”

  • Consistentt(19) = 0.6reported p = .59 · recomputed p = .556

    “t(19) = 0.6, p = .59”

  • Consistentt(19) = 4.5reported p < .001 · recomputed p < .001

    “t(19) = 4.5, p < .001”

  • Consistentt(19) = 2reported p = .056 · recomputed p = .060

    “t(19) = 2.0, p = .056”

  • Consistentt(19) = 1reported p = .32 · recomputed p = .330

    “t(19) = 1.0, p = .32”

  • Consistentt(30) = 1.8reported p = .09 · recomputed p = .082

    “t(30) = 1.8, p = .09”

  • Consistentt(16) = 5.4reported p < .001 · recomputed p < .001

    “t(16) = 5.4, p < .001”

  • Consistentt(16) = 3.6reported p < .01 · recomputed p = .002

    “t(16) = 3.6, p < .01”

  • Consistentt(16) = 2.7reported p = .01 · recomputed p = .016

    “t(16) = 2.7, p = .01”

  • Consistentt(16) = 2.3reported p = .04 · recomputed p = .035

    “t(16) = 2.3, p = .04”

  • Consistentt(30) = 0.8reported p = .46 · recomputed p = .430

    “t(30) = 0.8, p = .46”

  • Consistentt(14) = 2.5reported p = .03 · recomputed p = .025

    “t(14) = 2.5, p = .03”

28 of 28

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Self-check results appear here once an independent check has confirmed them.

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. JJ Clark, AL Yuille Data Fusion for Sensory Information Systems (Kluwer Academic, Boston, 1990).

Cited by

Sorted by citations

Related

Authors
Nardini, M., Bedford, R., & Mareschal, D.
Affiliations
  • Department of Visual Neuroscience, University College London Institute of Ophthalmology, London EC1V 9EL, United Kingdom; and
  • Birkbeck, University of London (GB)
  • Centre for Brain and Cognitive Development, Birkbeck, University of London, London WC1E 7HX, United Kingdom
ORCID
Bedford, Rachael
Venue
Proceedings of the National Academy of Sciences
Publisher
National Academy of Sciences
Published
2010-09-13
Volume
107
Issue
39
Pages
17041-17046
ISSN
  • 0027-8424 (ISSN-L)
  • 1091-6490
DOI
10.1073/pnas.1001699107
PMC
PMC2947870
OpenAlex
W2129943232
Semantic Scholar
37f267bdff77e5dea6cbe04a4f6f25e5c581d6f3
Cited by
161
References
38
Citations by year
2026: 6 · 2025: 5 · 2024: 9 · 2023: 8 · 2022: 14 · 2021: 21 · 2020: 8 · 2019: 10 · 2018: 13 · 2017: 9 · 2016: 17 · 2015: 17 · 2014: 8 · 2013: 5 · 2012: 9
FWCI
6.70
Type
journal article
Language
en
Open access
Locations
  • Proceedings of the National Academy of Sciences (publishedVersion)
  • PubMed Central (publishedVersion)
  • PubMed (publishedVersion)
  • BIROn (Birkbeck, University of London) (submittedVersion)
  • UCL Discovery (University College London) (submittedVersion)
  • Durham Research Online (Durham University) (submittedVersion, cc-by)
  • Research Portal (King's College London) (submittedVersion)
Funders
  • Wellcome Trust
  • Economic and Social Research Council (ES/F017154/1)
Topics
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Multisensory perception and integration
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Visual perception and processing mechanisms
  • Life Sciences › Neuroscience › Cognitive Neuroscience › Tactile and Sensory Interactions
Keywords
multisensory integration, sensory fusion, surface slant perception, texture cues, sensory uncertainty, children, perceptual speed, sensory conflict, perceptual development
MeSH terms
Adult, Child, Child Development, Cues, Depth Perception, Female, Humans, Male, Visual Perception

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review