Attention and executive control during lexical processing in aphasia

The goal of this project was to investigate the relationship between executive attention and specific linguistic and control processes during goal-directed tasks in aphasia. Its central premise was that PWA often possess dissociable impairments in linguistic processes and in the mechanisms that control and efficiently utilize those processes. The motivation for this claim was based on observations that PWA often present with deficits in the online processing of linguistic information, which in some instances have been interpreted as evidence for impaired linguistic operations, but in others has been interpreted as evidence for impaired control of language processing due to more general cognitive constraints. The current work tested claims regarding the Executive Attention model (Engle and Kane, 2004) in aphasia and its relation to varying task sets in linguistic and nonlinguistic tasks. 20 PWA and 23 matched controls were tested on four tasks measuring executive attention in verbal and nonverbal domains using word-picture interference, semantic and perceptual go/no-go, and spatial Stroop designs. Participants were also tested on lexical decision and numerosity judgment tasks with varying speed and accuracy-focused instructions, with performance modeled using the Diffusion Model (Ratcliff, 1978). Overall, the current work found evidence for the predicted domain-general and domain-specific impairments in executive attention at the level of individual PWA. However, these executive attention deficits did not appear to be associated with difficulties adapting to shifting speed-accuracy constraints. In addition, group-level patterns of performance across experiments suggest an additional related executive control deficit in the area of generating and maintaining arbitrary stimulus-response mappings. This study also demonstrated the appropriateness and potential applicability of the diffusion model in aphasia research, and diffusion model analyses found that PWA had difficulty adjusting their nondecision times in response to speed constraints, had lower drift rates in lexical decision, which reflected inefficient processing of lexical information, and had a disproportionately difficult time efficiently processing easy stimuli in lexical and numerosity tasks