Limits...
Evidence against continuous variables driving numerical discrimination in infancy.

Starr A, Brannon EM - Front Psychol (2015)

Bottom Line: Seven-month-old infants detected a threefold change in contour length but failed to detect a twofold change.These results, in conjunction with previously published data on numerosity discrimination using the same experimental paradigm, suggest that infants are not more sensitive to changes in contour length compared to changes in numerosity.Consequently, these findings undermine the claim that attention toward contour length is a primary driver of numerical discrimination in infancy.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychology and Neuroscience, Duke University, Durham, NC USA ; Center for Cognitive Neuroscience, Duke University, Durham, NC USA.

ABSTRACT
Over the past decades, abundant evidence has amassed that demonstrates infants' sensitivity to changes in number. Nonetheless, a prevalent view is that infants are more sensitive to continuous properties of stimulus arrays such as surface area and contour length than they are to numerosity. Very little research, however, has directly addressed infants' sensitivity to contour. Here we used a change detection paradigm to assess infants' acuity for the cumulative contour length of an array when the array's surface area and number were held constant. Seven-month-old infants detected a threefold change in contour length but failed to detect a twofold change. These results, in conjunction with previously published data on numerosity discrimination using the same experimental paradigm, suggest that infants are not more sensitive to changes in contour length compared to changes in numerosity. Consequently, these findings undermine the claim that attention toward contour length is a primary driver of numerical discrimination in infancy.

No MeSH data available.


Related in: MedlinePlus

Example stimuli. Smaller contour lengths are displayed on the left and larger contour lengths are displayed on the right. (A) Condition 1A (twofold change in contour with a 1.5-fold change in area). (B) Condition 1B (twofold change in contour with a constant area). (C) Condition 2 (threefold change in contour with a constant area).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4488599&req=5

Figure 1: Example stimuli. Smaller contour lengths are displayed on the left and larger contour lengths are displayed on the right. (A) Condition 1A (twofold change in contour with a 1.5-fold change in area). (B) Condition 1B (twofold change in contour with a constant area). (C) Condition 2 (threefold change in contour with a constant area).

Mentions: We used a custom Matlab (Mathworks) program that utilized a random walk technique to create shapes that varied in contour length while holding surface area constant1. In Condition 1A, the images in both streams contained arrays of 6 black irregularly shaped items on a white background (see Figure 1). The items in each array were constructed with an area of 6 square pixels and a contour length of 12 pixels or an area of 9 square pixels and a contour length of 24 pixels. These shapes were then scaled up to a mean diameter of 2 cm. Note than in Condition 1A, the twofold change in contour length was accompanied by a 1.5-fold change in surface area. Prior research indicates that this change in surface area is below the discrimination threshold for 7-month-old infants (Brannon et al., 2006; Cordes and Brannon, 2008, 2011; Libertus et al., 2014), though it is possible that there could be an additive effect in combination with contour length changes. This concurrent change in surface area was eliminated in Conditions 1B and 2. In Condition 1B, there were 10 (rather than 6) items in each array, and the items in each array were larger in size and red in color to increase their saliency. The items in each array had a constant area of 60 square pixels and contour lengths of either 60 or 120 pixels. The shapes were then scaled up to a mean diameter of 2.5 cm. In Condition 2, the arrays contained arrays of 10 irregularly shaped red items on a white background. The shapes were constructed with a constant area of 60 square pixels and had contour lengths of either 40 or 120 pixels. The shapes were then scaled up to a mean diameter of 1.3 cm. In all conditions, each of the items in a given array had the same contour length but a unique shape. Accordingly, both the contour length of the individual elements and the cumulative contour length of the array varied by a factor of two (Conditions 1A,B) or three (Condition 2) while surface area remained constant (Conditions 1B and 2) or varied by a factor of 1.5 (Condition 1A).


Evidence against continuous variables driving numerical discrimination in infancy.

Starr A, Brannon EM - Front Psychol (2015)

Example stimuli. Smaller contour lengths are displayed on the left and larger contour lengths are displayed on the right. (A) Condition 1A (twofold change in contour with a 1.5-fold change in area). (B) Condition 1B (twofold change in contour with a constant area). (C) Condition 2 (threefold change in contour with a constant area).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4488599&req=5

Figure 1: Example stimuli. Smaller contour lengths are displayed on the left and larger contour lengths are displayed on the right. (A) Condition 1A (twofold change in contour with a 1.5-fold change in area). (B) Condition 1B (twofold change in contour with a constant area). (C) Condition 2 (threefold change in contour with a constant area).
Mentions: We used a custom Matlab (Mathworks) program that utilized a random walk technique to create shapes that varied in contour length while holding surface area constant1. In Condition 1A, the images in both streams contained arrays of 6 black irregularly shaped items on a white background (see Figure 1). The items in each array were constructed with an area of 6 square pixels and a contour length of 12 pixels or an area of 9 square pixels and a contour length of 24 pixels. These shapes were then scaled up to a mean diameter of 2 cm. Note than in Condition 1A, the twofold change in contour length was accompanied by a 1.5-fold change in surface area. Prior research indicates that this change in surface area is below the discrimination threshold for 7-month-old infants (Brannon et al., 2006; Cordes and Brannon, 2008, 2011; Libertus et al., 2014), though it is possible that there could be an additive effect in combination with contour length changes. This concurrent change in surface area was eliminated in Conditions 1B and 2. In Condition 1B, there were 10 (rather than 6) items in each array, and the items in each array were larger in size and red in color to increase their saliency. The items in each array had a constant area of 60 square pixels and contour lengths of either 60 or 120 pixels. The shapes were then scaled up to a mean diameter of 2.5 cm. In Condition 2, the arrays contained arrays of 10 irregularly shaped red items on a white background. The shapes were constructed with a constant area of 60 square pixels and had contour lengths of either 40 or 120 pixels. The shapes were then scaled up to a mean diameter of 1.3 cm. In all conditions, each of the items in a given array had the same contour length but a unique shape. Accordingly, both the contour length of the individual elements and the cumulative contour length of the array varied by a factor of two (Conditions 1A,B) or three (Condition 2) while surface area remained constant (Conditions 1B and 2) or varied by a factor of 1.5 (Condition 1A).

Bottom Line: Seven-month-old infants detected a threefold change in contour length but failed to detect a twofold change.These results, in conjunction with previously published data on numerosity discrimination using the same experimental paradigm, suggest that infants are not more sensitive to changes in contour length compared to changes in numerosity.Consequently, these findings undermine the claim that attention toward contour length is a primary driver of numerical discrimination in infancy.

View Article: PubMed Central - PubMed

Affiliation: Department of Psychology and Neuroscience, Duke University, Durham, NC USA ; Center for Cognitive Neuroscience, Duke University, Durham, NC USA.

ABSTRACT
Over the past decades, abundant evidence has amassed that demonstrates infants' sensitivity to changes in number. Nonetheless, a prevalent view is that infants are more sensitive to continuous properties of stimulus arrays such as surface area and contour length than they are to numerosity. Very little research, however, has directly addressed infants' sensitivity to contour. Here we used a change detection paradigm to assess infants' acuity for the cumulative contour length of an array when the array's surface area and number were held constant. Seven-month-old infants detected a threefold change in contour length but failed to detect a twofold change. These results, in conjunction with previously published data on numerosity discrimination using the same experimental paradigm, suggest that infants are not more sensitive to changes in contour length compared to changes in numerosity. Consequently, these findings undermine the claim that attention toward contour length is a primary driver of numerical discrimination in infancy.

No MeSH data available.


Related in: MedlinePlus