Development of local-global preference in vision and haptics

Abstract

We aimed to advance our understanding of local-global preference by exploring its developmental path within and across sensory modalities: vision and haptics. Neurotypical individuals from six years of age through adulthood completed a similarity judgement task with hierarchical haptic or visual stimuli made of local elements (squares or triangles) forming a global shape (a square or a triangle). Participants chose which of two probes was more similar to a target: the one sharing the global shape (but different local shapes) or the one with the same local shapes (but different global shape). Across trials, we independently varied the size of the local elements and that of the global configuration-the latter was varied by manipulating local element density while keeping their numerosity constant. We found that the size of local elements (but not global size) modulates the effects of age and modality. For stimuli with smaller local elements, the proportion of global responses increased with age and was similar for visual and haptic stimuli. However, for stimuli made of our largest local elements, the global preference was reduced or absent, particularly in haptics, regardless of age. These results suggest that vision and haptics progressively converge toward similar global preference with age, but residual differences across modalities and across individuals may be observed, depending on the characteristics of the stimuli.

Figure 1.

Stimuli examples. (A) Three-dimensional small-, medium-, and large-size stimuli are represented by the blue, green, and red colors on the y-axis, respectively. The black and gray colors on the x-axis represent the two densities: high- and low-density stimuli (with 2 and 4 mm of interelement distance, respectively). (B) The four possible trial configurations, with inconsistent target and consistent responses (1 and 3) or vice-versa (2 and 4). Each configuration was presented for each size and density and the comparisons were randomly swapped (left/right or right/left).

Figure 2.

Proportion of global responses for low- and high-density stimuli, in vision and haptics (circles and asterisks) and for differently sized local elements (colors, see legend). Symbols show averages and SEM across all participants. Notice that low-density stimuli also take up larger global area and higher density stimuli take up a smaller global area.

Figure 3.

Individual proportion of global responses in small (A), medium (B), and large-size (C) stimuli for vision (open circles) and haptics (stars) as a function of age. Thin and dashed lines represent the running average of the proportion of global responses respectively for vision and haptics. (D) Correlation between global responses in vision (x-axis) and haptics (y-axis) for small (blueish dots), medium (greenish dots), and large size (reddish dots). The thick blue, green and red lines show the best-fitting linear regressions calculated on the average of individual proportion of global responses for small (blue line), medium (green line), and large (red line). The dashed black line shows the bisector of the axes. Text inset gives Pearson's correlation coefficient and associated P value and base-10 logarithm of the Bayes Factor (lgBF) for each size.

Figure 4.

Absolute values of the difference of the proportions of global responses between modalities (visuo-haptics mismatch) as a function of age. Colored filled dots show the individual level of modality mismatch for large (reddish), small (blueish) and medium-size (greenish) stimuli. Different lines represent the running average for the modality mismatch in large (red); medium (green) and small (blue) size stimuli by function of age.

Figure 5.

Proportion of global responses for different global shapes (squares and triangles) as a function of the number of local elements composing them. Notice that numerosity equal to 12.5 is the result of the average between 10 and 15 global triangles in small sizes stimuli. Examples of stimuli are given as icons under the numerosity line (x-axis). The figure reports three or four local items for large size stimuli, nine or 10 local items for medium size stimuli and 10, 15, or 16 local elements for small-size stimuli. Open and filled symbols refer, respectively, to visual and haptic stimuli and they show averages and SEM across all participants.