Do Children with Autism Understand Communicative Intentions in a Hiding-Finding Game?

Background:  Most autism research on understanding communicative intentions has focussed on verbal communication (e.g., Eales, 1993; Happé, 1993). We have reported provisional evidence that children with autism often fail to grasp a person’s head nods as intending to indicate location (Hobson, Garcia-Perez, & Lee, 2010).  Here we modified a hiding-finding game inspired by Tomasello, Call and Gluckman (1997), who reported that children under 3 years, but not great apes, understood someone’s conventional and unconventional actions as intending to inform about a hidden object’s location.

Objectives:  Our aim was to assess whether participants with autism would read one person’s unconventionally expressed communicative intentions to help someone else.  Participants were shown videotaped scenes involving a ‘hider’, ‘finder’ and ‘helper’ (played by trained child actors), and were asked to describe what happened in the vignette.  The vignettes included different conditions which might have facilitated or hindered participants’ understanding, such as whether the ‘finder’ was successful.

Methods: Participants were 15 children and adolescents with autism (5 girls, 10 boys: diagnosis confirmed with ADOS: Lord, Rutter & Goode, 1998) and a comparison group of 15 children without autism  (5 girls, 10 boys) either typically developing (n = 4) or with learning/developmental disabilities (n = 11), matched for chronological age (Autism M = 11 years; 5 months and Comparison M = 11 years; 10 months) and the Expressive and Receptive Vocabulary subscales of the Test of Word Knowledge. 

Participants viewed 11 videotaped scenarios of a hide-and-seek game.  In nine of the clips, a ‘helper’ used an unconventional means to communicate to the ‘seeker’ where a hidden object was located e.g., by ‘pointing’ with a leg, or throwing an object in the direction of the hidden object.  Two final videotapes showed conventional pointing-to-inform.  Following each clip, the tester asked the child to describe what happened, and then asked what each character said and did. Transcripts and the video-recording of participant responses were coded by judges unaware of diagnosis, for levels of (1) Understanding the Vignette, Kappa = .93; (2) Describing the Actions of the Helper, Kappa = .96; (3) Attributing Communicative Intent to the Helper, Kappa = .97, and (4) Need for Prompt to describe the sequence, Kappa = .97.

Results: Children with as well as without autism were able to interpret conventional points.   On the other hand, participants with autism found it difficult to discern the communicative intent when someone tried to indicate a location using unconventional gestures or actions.  Participants in the comparison group judged approximately two thirds (M = 5.27) of these vignettes in terms of the helper’s intention to communicate something to the seeker, whereas the participants with autism made this judgement for fewer than one-third of the video vignettes (M = 2.67), t(28) = 2.03, p < .05.  We shall provide further details on analysis as well as qualitative examples of responses.

Conclusions: Children with autism face difficulties in reading communicative intent to inform someone else, expressed in unconventional non-verbal communication,  even when they can understand conventional pointing-to-inform.