Changes in the Focus of Attention Across Time in Individuals with Autism: The Effect of a Dual-Stream Paradigm

Background: The attentional blink (AB) paradigm has been used to explore the temporal and spatial dynamics of visual attention (Raymond, Shapiro, & Arnell, 1992). In a typical AB paradigm, two target letters are embedded in a rapid stream of digit distracters.  Although identification accuracy of the first target is normally very high, identification of the second target is impaired when it appears in close temporal proximity to the first. A notable exception to this occurs when the second target is presented directly after the first target (at the position known as Lag 1).  In this case, the AB is much reduced and second-target identification is spared – a phenomenon known as Lag-1 sparing (Potter, Chun, Banks, & Muckenhoupt, 1998). Lag-1 sparing can also occur to targets in different spatial locations (i.e., in dual streams), but only if the second target falls within the focus of attention (Jefferies, Ghorashi, Kawahara, & Di Lollo, 2007).

Objectives: In order to assess the difficulties displayed by persons with ASD in focusing on target stimuli (Ames & Fletcher-Watson, 2010; Burack, 1994), we used a dual-stream AB paradigm to further assess the nature of the narrowing of attention in autism.

Methods: Eleven high-functioning young adults with autism and 18 typically developing (TD) young adults who were matched on chronological age, were tested on a dual-stream AB task in which two target letters were embedded in streams of digit distracters.  One stream was presented to the left of a central fixation point, the second stream appeared to the right of fixation. The two targets appeared randomly in either the same stream as one another (same-stream condition) or in opposite streams (different-stream condition).  On a third of the trials, the second target appeared immediately after the first (Lag 1); on another third of the trials the targets were separated by a single digit distracter (Lag 3); on the remaining trials, the targets were separated by 8 distracters (Lag 9).  Each participant was tested with both an 80-ms and a 133-ms stimulus onset asynchrony (SOA). 

Results: The TD participants displayed the expected Lag-1 sparing at the short SOA (80 ms) but not at the long SOA (133 ms), consistent with a focus of attention that is initially broad and that narrows over the course of approximately 50 ms. For those participants with autism, however, no Lag-1 sparing was evident at either SOA. 

Conclusions: Previous research has reported that Lag-1 sparing occurs for individuals with autism in a single-stream AB paradigm (Amirault et al., 2009; Rinehart et al., 2010).  Our finding, therefore, that Lag-1 sparing does not occur in a dual-stream paradigm is somewhat surprising. We hypothesize that the combination of task difficulty and dynamic changes to focused attention in the dual-stream task, prevented Lag-1 sparing in individuals with autism, but not in TD individuals. Future research will investigate whether an unusually efficient attentional gate underlies the observed differences between TD individuals and those with autism.