Kids See Human Too: Adapting an Individual Differences Measure of Anthropomorphism for a Child Sample

Abstract

The study of anthropomorphism in adults has received considerable interest with the development of the Individual Differences in Anthropomorphism Questionnaire (IDAQ; Waytz, Cacioppo, & Epley, 2010). Anthropomorphism in children—its development, correlates, and consequences—is also of significant interest, yet a comparable measure does not exist. To fill this gap, we developed the IDAQ-Child Form (IDAQ-CF) and report on 2 studies. In Study 1A, adults (N = 304) were administered the IDAQ and IDAQ-CF to directly assess comparability between the measures. In Study 1B, an additional 350 adults were administered the IDAQ-CF to confirm that the new measure had the same underlying structure as the original IDAQ when the measures were not administered together. In Study 2, children (N = 90) in 3 age groups—5, 7, and 9 years old—were administered the IDAQ-CF and an Attribution Interview, which probed their conceptions of a robot and puppet. Results indicated the IDAQ-CF a) is comparable to the original IDAQ in adult (Studies 1A and 1B) and child (Study 2) samples, and b) predicts children’s tendency to attribute animate characteristics to inanimate entities (Study 2). This research provides strong evidence that the IDAQ-CF is an effective adaptation of the original IDAQ for use with children.

Notes

¹ Robust statistics could not be computed because the input data set was a correlation matrix.

² Analyses on the child sample were also conducted without the polychoric correlation adjustment. The pattern of results was identical but with weaker factor loadings.

³ An additional regression was conducted reversing the entry order of the IDAQ-CF subscales. The pattern of results was the same, with the Animal subscale having a negative relationship with robot attributions predicting 7% of the variance beyond age and gender in Step 2 and the Technology/Nature subscale having a positive relationship predicting an additional 22% of the variance in Step 3.

⁴ As with robot attribution analysis, an additional regression was conducted reversing the entry order of the IDAQ-CF subscales. The pattern of results did not change, with the Animal subscale having a nonsignificant relationship with puppet attributions (0% variance explained) in Step 2 and the Technology/Nature subscale having a positive, significant relationship predicting an additional 23% of the variance in Step 3.

Additional information

Funding

This work was supported, in part, by a Grant-in-Aid (No. 693177) from Western Washington University to R. L. Severson.