Development of Cue Integration in Human Navigation

Abstract

Mammalian navigation depends both on visual landmarks and on self-generated (e.g., vestibular and proprioceptive) cues that signal the organism's own movement [1-5]. When these conflict, landmarks can either reset estimates of self-motion or be integrated with them [6-9]. We asked how humans combine these information sources and whether children, who use both from a young age [10-12], combine them as adults do. Participants attempted to return an object to its original place in an arena when given either visual landmarks only, nonvisual self-motion information only, or both. Adults, but not 4- to 5-year-olds or 7- to 8-year-olds, reduced their response variance when both information sources were available. In an additional "conflict" condition that measured relative reliance on landmarks and self-motion, we predicted behavior under two models: integration (weighted averaging) of the cues and alternation between them. Adults' behavior was predicted by integration, in which the cues were weighted nearly optimally to reduce variance, whereas children's behavior was predicted by alternation. These results suggest that development of individual spatial-representational systems precedes development of the capacity to combine these within a common reference frame. Humans can integrate spatial cues nearly optimally to navigate, but this ability depends on an extended developmental process.

The development of path integration: Combining estimations of distance and heading

“…Previous research by Nardini et al (2008) supports the notion that adults may be more adept at combining spatial cues: their findings demonstrated that the ability to effectively integrate multi-modal cues develops over time in humans, and children of at least 8 years of age and below are unable to perform at a similar level to adults. It could be argued that responses in the present study required a combination of Development of path integration 22 information within modality (i.e.…”

“…As such, the theoretical account of integration provided by Nardini et al (2008) may be just as applicable to the present data, with adults combining these multi-modal sources of information more effectively than children. However, this analogy might not be entirely appropriate when it comes to the distinction between representations of distance and heading (and the necessity for them to be combined in the trianglecompletion task).…”

“…Bremner, Hatton, Foster, & Mason, 2011;Acredolo & Evans, 1980) and this literature has been complemented by very useful reviews of the relationship between vision, locomotion, and spatial orientation in childhood, as provided by Campos et al (2000) and Rieser & Pick (2007). More recently, a study by Nardini, Jones, Bedford, & Braddick (2008) has explored how information from path integration is combined with visual cues in order to support place memory. In their task, participants in a darkened arena were required to collect 3 objects Development of path integration 6 sequentially and then return the first object to its original position.…”

“…The ability to switch between strategies may itself be something that possesses an interesting developmental profile (c.f. Nardini et al, 2008).…”

The development of multisensory body representation and awareness continues to 10years of age: Evidence from the rubber hand illusion

“…There is also evidence that at this age multisensory processing develops to an adult-like state, in that children begin to combine and weight multisensory cues optimally (Nardini, Jones, Bedford & Braddick, 2008;Nardini, Bedford & Mareschal, 2010;Gori, del Viva, Sandini & Burr, 2008). Indeed, one study (Nardini et al, 2013) suggests that this adultlike optimal cue weighting did not occur for weighting of visual and proprioceptive cues to hand position until at least 10-12 years.…”