Search Abstracts | Symposia | Slide Sessions | Poster Sessions
A precision rehabilitation approach to naming treatment in aphasia: Retrieval practice versus errorless learning
Poster Session C, Saturday, September 13, 11:00 am - 12:30 pm, Field House
Yingxue Tian1, Marja-Liisa Mailend1,2, Daniel Mirman3, Erica Middleton1; 1Jefferson Moss Rehabilitation Research Institute, Elkins Park, PA, USA, 2University of Tartu, Tartu, Estonia, 3University of Edinburgh, Edinburgh, UK
Naming impairment is a prevalent and debilitating issue in post-stroke aphasia, often resulting from lexical access disorder, specifically, difficulty mapping between semantics and words (Stage 1) and/or between words and phonology (Stage 2). Two established naming treatments, retrieval practice (RP) and errorless learning (EL), leverage distinct learning principles to support language recovery. RP prompts retrieval from long-term memory (e.g., naming an object from its picture), engaging both Stage 1 and Stage 2 mappings during naming. In contrast, EL emphasizes the practice of Stage 2 mapping (e.g., repeating the name of a picture after hearing it), relying less on semantics-driven retrieval. Little is known about how individual differences in major processing stages of naming influence the benefits of RP and EL. RP and EL differentially engage the two stages of naming and may selectively improve errors from different stages. Through the lens of precision rehabilitation, the present study investigates person-specific and item-specific factors that impact differential treatment outcomes. A naming treatment study was conducted with 30 participants with chronic stroke-induced aphasia. For item selection, a 660-item common objects picture corpus was administered twice for naming assessment. Responses were coded into correct, phonological errors (Stage 2 error), semantic errors (Stage 1 error), and others. Each participant’s personalized 120-item set was evenly assigned to RP and EL treatment conditions in a matched fashion. Each item was treated twice after an initial learning trial, with treatment trials ending in correct-answer feedback. Naming performance for 120 treated and 30 untreated items was assessed the next day and one-week post-treatment. Six cognitive-linguistic variables were collected as person-specific factors, including input and output semantic and phonological processing abilities, naming severity, and verbal working memory (WM) capacity. For each treated word, the item-specific factor (item nature) was coded into six categories based on pre-treatment item selection performance: correct, pure Stage 1, blended Stage 1, pure Stage 2, blended Stage 2, and other. Naming accuracy at the next-day post-treatment test was used as the dependent variable. We used separate multilevel logistic mixed-effects models to investigate how the variability in participants’ characteristics and item nature impact naming treatments’ benefits. Both models evaluated the interaction between treatment type (RP versus EL) and person-specific or item-specific factors. Both treatments significantly improved naming accuracy compared to the untreated condition. In the person-specific model, higher verbal WM capacity was associated with greater benefit from RP relative to EL. In the item-specific model, RP outperformed EL for pure and blended Stage 1 items and blended Stage 2 items one-day post-treatment. At the one-week follow-up, RP maintained its advantage for pure Stage 1 and previously correct items. Theoretically, person-specific results suggest that WM may contribute to the effectiveness of RP by supporting semantic-driven retrieval, integrating successful retrievals into long-term representations, and/or facilitating learning from feedback after unsuccessful retrievals. Item-specific findings indicate that RP’s mechanism of action may be strengthening Stage 1 mappings. Clinically, these findings suggest that RP may be particularly effective for individuals with high WM and for items affected by Stage 1 (semantic) errors.
Topic Areas: Language Production, Disorders: Acquired