Social Evaluation Is Reduced in Toddlers with ASD and Other Developmental Delays Compared to Toddlers with Typical Development

Background: Studies utilizing a nonverbal social attribution paradigm show that typically developing infants make dispositional state attributions about ambiguous stimuli as young as ten months of age and make goal state attributions and social evaluations by three months of age (Hamlin, Wynn, & Bloom, 2007, 2010). Given that social attribution has shown to be reduced in older individuals with ASD, early disruptions in social attribution may serve as a behavioral marker for early detection.

Objectives: This study examined whether a puppet-based social attribution paradigm developed for use in preverbal infants is appropriate for use in toddlers with and without delays and whether performance differentiates toddlers with ASD from those with non-ASD delays (DD) and typical development (TD).

Methods: The sample included 57 toddlers (N_ASD=15, N_DD=20, N_TD=22; M_age=21.61 mo, SD=2.98) who were presented with a nonverbal puppet-based social attribution paradigm in which a shape tries to climb a hill and another shape either helps or hinders its climb (Hamlin et al.). Toddlers were habituated to alternating helping and hindering social scenes until habituated (max. 12 trials). Toddlers were then presented with a Looking Time Test (LTT), a violation of expectation-based measure of dispositional attribution; longer looking time towards an unexpected social event (i.e., novel scene of climber moving towards hinderer vs. helper) was considered evidence that children made dispositional attributions during habituation. Toddlers also completed a Preference Test (PT) to measure social evaluation; choosing helper helper over hinderer was interpreted as evidence that children made goal attributions during habituation and integrated that knowledge in their social evaluation.

Results: Children habituated to the social scenes at similar rates across diagnosis (X²(2,N=57)=1.966, p=.374); however, toddlers within the TD group looked longer per habituation exposure (M=10.39s, SD=5.85) compared to the DD group only (M=7.89s, SD=3.30; F(2,54)=.0368, p=.047, partial η² =.107). Toddlers in the ASD (M_exp= 5.36s, M_unexp=5.88s) and DD (M_exp = 7.00s, M_unexp=8.29s) groups did not demonstrate increased looking time towards the unexpected event during the LLT, and in contrast to studies in infants, neither did those in the TD group (M_exp=5.84s_, M_unexp=5.82s; F(2,52)=0.564, p=.572, partial η²=.021). The TD group demonstrated preference for the helper shape during the PT (71.4%_helper)_, whereas the ASD (28.6%_helper) and DD groups (40%_helper) did not (X²(2,N=55)=7.213, p=.027).

Conclusions: This is the first study utilizing a nonverbal social attribution paradigm in toddlers with and without developmental delays. Contrary to expectations, TD toddlers did not show looking time differences on the LLT bringing into question whether looking time is an appropriate measure of dispositional attribution at this age. Despite this, toddlers in the TD group demonstrated evidence of social evaluation on the PT, whereas toddlers within the ASD and DD groups did not. While all toddlers habituated at similar rates, TD toddlers looked longer towards each habituation exposure compared to the DD group, which suggest that other factors may have impeded success in the DD group. Future research is warranted to examine whether deficient social evaluation is specific to ASD or characterizes developmental delays more broadly.