Word retrieval treatment for aphasia.

Non-brain-injured participants provided data for development of treatment stimuli.

SFA entails having the speech-language pathologist (SLP) guide the participant through generation of pertinent semantic features for pictured treatment items (e.g., category membership, physical description, location of item in context, personal associations, action associated with item). For some participants, treatment items were grouped according to typicality of category membership (e.g,, a robin-typical bird and penguin-atypical bird). Training of atypical items may stimulate a broader semantic activation of the category and thus, may promote greater generalization. Treatment was applied sequentially to sets of items in the context of single-subject, multiple baseline designs. In this way, replication of treatment effects could be evaluated within and across participants. Treatment was administered by certified SLPs three times per week until prescribed accuracy levels were met during nontreatment probes or a maximum number of treatment sessions was completed.

Accuracy of naming of pictured treated and untreated items was assessed in probes conducted separate from treatment. Probes were conducted repeatedly throughout the study, from baseline (prior to treatment) to follow-up (6 weeks following treatment). All naming responses were scored using a 0-10 scale reflecting promptness and presence of errors; scores of 8-10 received an "accuate" score and scores of 0-7 received an "inaccurate" score. A percentage accuracy score was calculated for each experimental set of items for every probe session. Baseline probe scores were compared to end of treatment and follow-up probe scores to obtain individual effect sizes for each experimental list of items for each participant (i.e., several effect sizes were calculated for each participant). All effect sizes were utlized to obtain an average effect size for each participant; these averages were then utlized to obtain a group average.

Semantic Feature Analysis (SFA) is a word-retrieval treatment for aphasia. SFA entails having the speech-language pathologist (SLP) guide the participant through generation of pertinent semantic features for pictured treatment items (e.g., category membership, physical description, location of item in context, personal associations, associated actions). For some participants, treatment items were grouped by typicality of category membership (e.g, robin-typical bird and penguin-atypical bird). Training of atypical items may stimulate a broader semantic activation of the category and thus, may promote greater generalization. Treatment was applied sequentially to sets of items in single-subject, multiple baseline designs. Thus, replication of treatment effects could be evaluated within and across participants. Treatment was administered by SLPs three times per week until prescribed accuracy levels were met during probes or a maximum number of treatment sessions was completed.

Accuracy of naming of pictured treated and untreated items was assessed in probes conducted separate from treatment. Probes were conducted repeatedly throughout the study, from baseline (prior to treatment) to follow-up (6 weeks following treatment). All naming responses were scored using a 0-10 scale reflecting promptness and presence of errors; scores of 8-10 received an "accuate" score and scores of 0-7 received an "inaccurate" score. A percentage accuracy score was calculated for each experimental set of items for every probe session. Baseline probe scores were compared to end of treatment and follow-up probe scores to obtain individual effect sizes for each experimental list of items for each participant (i.e., several effect sizes were calculated for each participant). All effect sizes were utlized to obtain an average effect size for each participant; these averages were then utlized to obtain a group average.

SFA Treatment Participants were stroke-survivors with chronic aphasia who had significant word retrieval difficulties. Non Treatment Stimuli Development Participants were only enrolled in the study to provide data for the development of treatment stimuli. As such, they were not assessed for the outcome measure.

Stroke-survivors who received Semantic Feature Analysis therapy for aphasic word-retrieval deficits

Semantic Feature Training: The treatment is designed to stimulate the semantic feature network so that it may serve as not only a mechanism for improving disrupted lexical semantic processing, but also as a compensatory strategy during word retrieval failures.