Abstract
In this study, we investigated relations between infants' understanding of intentional actions and measures of social responsiveness during a transitional period, 9- to 11-months. Infants (N = 52) were tested in visual habituation paradigms tapping their understanding of the relation between a person and the object of her attention. Measures of social responsiveness included orienting to the target of another's attention, point production, and supported joint attention in parent---child play. Infants' responses to the habituation events were related to their social responsiveness. Distinct factors for understanding actions and social responsiveness as relational were revealed. Infants who produced object-directed points were more likely to understand pointing as relational, and infants who engaged in high amounts of shared attention were more likely to understand gaze. Infants' tendency to orient in response to an adult's gaze shifts and points was unrelated to their understanding of gaze and pointing. These findings elucidate the ways in which social cognition and social responsiveness, although distinct, are related in development.
Notes
1To control for the possibility that the actor performed differently across trials, a coder, whowas unaware of which condition infants were assigned to, guessed the test trial order by watching a videotape of the actor. A sample of about 20% (n = 10) of the sessions for both event types was coded. The coder was correct on 40% and 53% of the pairs for gaze and pointing events, respectively; for the first pair, she was correct 50% and 60%. The coder's accuracy did not differ significantly from chance, t(9) = –1.20, p = .26, and, t(9) = .25, p = .81, respectively; first pair, t(9) = 0, p = .99 and t(9) = .61, p = .56.
aNumber of children out of 52 in study who had complete data for measure.
∗p < .05, two-tailed t test compared to chance (.50).
2Secondary analyses revealed no significant correlations between the measures of habituation (i.e., number of test trials to habituate, total habituation time, proportion decrement in looking across trials) and the measures of action understanding or the measures of social responsiveness.
3To rule out age as an explanation for the effect of point production on point understanding, we conducted an alternative analysis after reducing the sample's age range. This eliminated 15 younger infants (14 nonpointers, 1 pointer), and 1 older pointer. The remaining 19 girls and 17 boys had an average age of 10;11. Pointers' (n = 18) age (M = 10;13, range 9;2–10;29) did not differ from nonpointers' (n = 18) age (M=10;8, 9;27 − 10;29), t(34) = −1.55, p = .13. Point understanding scores were significantly above chance for the pointers, t(17) = 3.47, p < .005, but not for the nonpointers, t(17) = –.57, p = .58. Gaze understanding scores did not differ significantly from chance for either group (ts < 1.00). Pointers' scores were significantly greater than nonpointers' for point understanding, t(34) = 2.51, p < .05, but not for gaze understanding, t(34) = .30, p = .76.
To explore this interaction, we conducted post hoc ANOVAs separately on the pointer and nonpointer groups. Point understanding was entered as the dependent variable, with goal (ball or teddy bear in habituation) as the dependent variable, and age as a covariate. For the pointers, there was no significantmain effect of goal. For the nonpointers, there was a significant effect of goal, F(1, 32) = 4.78, p = .04. Post hoc comparisons of pairs revealed a mean difference of .15 (SE = 0.07), p = .04; nonpointers who saw the teddy bear in habituation had higher point understanding scores than nonpointers who saw the ball in habituation, 0.57 (SD = 0.18) and 0.42 (SD = 0.20), respectively. Planned comparisons against chance (.50) revealed that pointers looked reliably longer on new object compared to new side trials, t(19) = 2.56, p = 0.02. Nonpointers did not differ from chance as a group, t(31) = 0.21, p = 0.83, or when divided into the groups that saw the teddy bear versus the ball as the goal object, t(18) = 1.70, p = 0.11 and t(12) = –1.55, p = 0.15, respectively.
4Orienting was an average of gaze orienting and point orienting because they were positively related, rs = 0.26, p = .08, after removing two outliers who were greater than 2 standard deviations above the mean for gaze orienting; when separate analyses were conducted with gaze orienting and point orienting the results did not change.
N = 52.
∗p < .01.
5Mean substitutions were used to replace missing data points for two infants (one for attention following, the other for orienting in habituation).When these infants were excluded from the analysis instead, the pattern of findings did not change. One infant was excluded from the analysis because the amount of time he spent in shared attention was greater than 3 standard deviations above themean. This inflated the relation between shared attention and the other variables within the factors. When he was excluded, the overall pattern of findings (i.e., factor structure, variance) did not change.