When vision is not an option: children's integration of auditory and haptic information is suboptimal

Abstract

When visual information is available, human adults, but not children, have been shown to reduce sensory uncertainty by taking a weighted average of sensory cues. In the absence of reliable visual information (e.g. extremely dark environment, visual disorders), the use of other information is vital. Here we ask how humans combine haptic and auditory information from childhood. In the first experiment, adults and children aged 5 to 11 years judged the relative sizes of two objects in auditory, haptic, and non-conflicting bimodal conditions. In Experiment 2, different groups of adults and children were tested in non-conflicting and conflicting bimodal conditions. In Experiment 1, adults reduced sensory uncertainty by integrating the cues optimally, while children did not. In Experiment 2, adults and children used similar weighting strategies to solve audio–haptic conflict. These results suggest that, in the absence of visual information, optimal integration of cues for discrimination of object size develops late in childhood.

Hearing Where the Eyes See: Children Use an Irrelevant Visual Cue When Localizing Sounds

“…The results indicate that children's weighting of the irrelevant visual cue was not driven by the reliability of the auditory cue. The fact that children do not weight multimodal cues proportionally to their reliability is consistent with previous findings using task‐relevant cues (e.g., Gori et al., ; Nardini et al., ; Petrini et al., ), and with findings demonstrating that decreasing sound reliability by adding noise does not facilitate integration (Barutchu et al. ().…”

“…Our findings show that acquiring mature cue combination mechanisms is a multifaceted process and that the late development of optimal cue combination for relevant information (e.g., Gori et al., ; Gori, Sandini, et al., ; Nardini et al., , ; Petrini et al., ) is paralleled by a late development of mechanisms selecting cues for combination. Being able to keep sensory information separate (Nardini et al., ) may be adaptive when the body is growing, as each sense needs to be continuously recalibrated (Burr, Binda, & Gori, ).…”

“…However, later in childhood multisensory processing continues to be refined in many ways before it becomes adult‐like. For example, children become able to weight cues proportionally to their reliability (Gori et al., ; Gori, Sandini, et al., Nardini et al., , ; Petrini et al., ), and to increase their multisensory selectivity (Hillock et al., ; Hillock‐Dunn & Wallace, ; Innes‐Brown et al., ) to optimize precision. Here, we show that despite the presence of individual differences among children and adults, which are very common in the developmental literature on cue combination (e.g., Gori et al., ; Nardini et al., ; Petrini et al., ), children's ability to ignore visually irrelevant but temporally redundant information is much poorer.…”

“…For example, children become able to weight cues proportionally to their reliability (Gori et al., ; Gori, Sandini, et al., Nardini et al., , ; Petrini et al., ), and to increase their multisensory selectivity (Hillock et al., ; Hillock‐Dunn & Wallace, ; Innes‐Brown et al., ) to optimize precision. Here, we show that despite the presence of individual differences among children and adults, which are very common in the developmental literature on cue combination (e.g., Gori et al., ; Nardini et al., ; Petrini et al., ), children's ability to ignore visually irrelevant but temporally redundant information is much poorer. Our findings suggest that during early childhood there is a shift from associating cues based on simple heuristics, such as temporal co‐occurrence (mechanism prioritized during infancy; Bahrick & Lickliter, , ; Hyde et al., ; Lewkowicz, , ), to combining cues using more complex integration models, and thus increasing the behavioral gain (Ernst & Banks, ) afforded by these previously learned associations.…”

Integration of audio‐visual information for spatial decisions in children and adults

“…It is interesting to compare how our evidence for multisensory speed gains via pooling of signals from at least 4 years of age stands with results from numerous developmental studies measuring multisensory precision, in tasks that were not speeded. These tasks have found absent or sub‐optimal pooling of signals until much later in childhood (Gori et al ., ; Nardini et al ., ; Nardini et al ., ; Gori et al ., ; Nardini et al ., ; Petrini et al ., ). In these studies, children aged 8 years or older did not improve their perceptual discrimination thresholds by taking correctly weighted averages of multiple estimates (e.g.…”