Detail and Gestalt Focus in Spontaneous Descriptions by Individuals with Optimal Outcomes From ASD

Background: Individuals with high-functioning autism (HFA) have consistently been found to focus on details, sometimes at the expense of the whole, or gestalt (Mottron et al., 2003; Happé & Frith, 2006). The tendency to focus on non-central details rather than the gestalt impacts story-telling (Barnes & Baron-Cohen, 2012), verbal memory (Hermelin & O’Connor, 1967), and other communication skills central to social interactions. Detail focus is of particular interest in our study of individuals previously diagnosed with HFA, who no longer meet criteria -- who have achieved a so-called "optimal outcome" (OO). 

Objectives: We aimed to identify differences in detail focus for individuals with HFA, OO, and typical development (TD), by evaluating the production, in spontaneous speech, of statements that included person-centered details and gestalt statements (likely to be frequent in the TD group data) as well as non-central details and unrelated statements (likely to be frequent in the HFA data). A central question was whether the OO group's speech would pattern more similarly to the TD or to the HFA group, because the degree of "detail focus" is non-diagnostic but quite prevalent in autism. 

Methods: A total of 45 children and adolescents with HFA, OO and TD (15 per group) described a rich visual stimulus (a painting) for ten seconds, while tapping their index finger; there were six such trials. Responses were recorded and transcribed, and each utterance was coded for focus: person, non-central, extrapolation, gist, or nonsense/unrelated. To date, data from 27 participants has been coded (5 HFA, 8 OO, 14 TD).

Results: In this preliminary analysis of a small sample, results nevertheless clearly indicated that the HFA group produced significantly fewer gist statements than the OO group, p = .008; the contrast with TD approached significance, p = .06, with no difference between the OO and TD groups.  There was a similar pattern for person-centered details:  the HFA group produced significantly fewer than the TD group, p = .002; the contrast with OO approached significance, p = .06, and there was no difference between the OO and TD groups. We also found that both the HFA group and the OO group produced significantly fewer non-central details than the TD group, p< .05, contrary to our predictions.

Conclusions:  Results suggest that OO individuals have largely "normalized" in the degree to which they focus on person-centered details and gestalts, such that they are indistinguishable from their TD peers.  This is a particularly stringent test of detail/gestalt focus, because participants were describing pictures during a "dual-task" paradigm, which involves a significant cognitive load. These results cannot distinguish between two possibilities: 1) individuals with OO had a more detail-oriented focus, which changed over development, or 2) their different early cognitive style was already less detail-focused, with the implication that this may related to their reduction in symptomatology.