The Effect of Weighted Arms on Scale Errors

• In Event: 3-184 - Integrating Perception, Cognition, and Action: The Case of Scale Errors

Integrative Statement

We studied the effects of weighted arms on the frequency of scale errors as part of a larger research program designed to understand why toddlers make action errors. These errors are surprising given the substantial research showing that infants appear to have knowledge of objects and events. Scale errors are one example of such a mistake. Scale errors are an attempt to perform an action on an object that is commonly associated with the larger version of that object such as trying to put on doll’s clothes or insert their body in a toy-sized object (Ware, 2006; DeLoache et al., 2004; Rosengren, Schein, & Gutierrez, 2009). We do not think older children have forgotten the knowledge that they had as infants or that they need to relearn it. Instead, we propose that older children are having trouble integrating their knowledge with their action. Recent research found that weights placed on children’s arms during search tasks improved children’s performance, leading to the suggestion that by slowing down toddlers’ arm movements, they were better able to plan their action by considering the perceptual and cognitive information (Arterberry, Hespos, & Neal, 2017).
Children played with a disassembled Mr. Potato Head toy with a duplicate pair of glasses, shoes, and hat. Parents showed their children how to play with Mr. Potato Head, but without direct instruction regarding what to do with the duplicate pieces. All children wore wristbands, half of which were weighted (163 g). Two research assistants, naïve to the weight condition, coded the number of scale errors made with the Mr. Potato Head pieces. Figure 4.1 shows an example of participants committing scale errors with a Mr. Potato Head hat and glasses.
In Experiment 1, 27 children within one month of their 24-month birthday showed an effect of weights on scale errors. Specifically, children made fewer errors when wearing weighted wrist bands (M = .67, SD = .57) compared to unweighted wrist bands (M = 2.33, SD = .57, p = .05). This effect was driven by girls, who made significantly more errors than boys in the no weight condition did, and significantly fewer errors than boys in the no weight condition. In Experiment 2, a wider age range was tested (24 to 29 months; Mage = 27.48 months), and weight was manipulated within-subject such that children engaged with Mr. Potato Head with and without weighted arms within the same session. Among the 28 children, the rate of scale errors was low (M < 1.0), and incidence did not differ across weight condition (p = .656).
The results suggest that weighted wristbands can help children integrate visual information such that they commit fewer scale errors, at least within the youngest age group tested. The findings will be discussed in the context of (a) the time course of scale errors (appearance and decline) and (b) other tasks where weights affected toddlers’ performance, highlighting the role of feedback and spatial relations.

Authors

• Martha E Arterberry, Colby College
  Presenting Author

• Cole A. Walsh, Colby College
  Non-Presenting Author

• Susan Hespos, Northwestern University
  Non-Presenting Author