Beyond Percent Correct: Measuring Change in Individual Picture Naming Ability

Area of Science:

• Neuroscience
• Speech and Language Pathology
• Psycholinguistics

Background:

• Picture naming deficits are common in individuals with stroke-induced aphasia.
• Standard accuracy measures may not fully capture the nuances of anomia therapy outcomes.
• Assessing the quality and type of naming errors is crucial for understanding treatment efficacy.

Purpose of the Study:

• To evaluate a novel distance measure for assessing picture naming abilities in stroke aphasia.
• To compare the sensitivity of the distance measure against traditional accuracy measures in detecting treatment effects.
• To explore how item difficulty and response types influence the assessment of naming improvements.

Main Methods:

• Data from 72 participants with stroke aphasia in an anomia therapy trial were analyzed.
• Four summary statistics were derived: standard accuracy, accuracy adjusted for item difficulty, accuracy adjusted for item difficulty and response type, and a distance measure adjusted for item difficulty and response type.
• Reliability among five raters was assessed.
• Treatment effects were compared between treatment and control groups using these statistics.

Main Results:

• Interrater and intrarater reliability were high, with minor impact from phonological errors.
• The distance measure demonstrated greater sensitivity to treatment effects in both frequency and magnitude compared to accuracy measures.
• Additional treatment effects were identified using the distance measure that were not apparent with standard accuracy.

Conclusions:

• The distance measure, considering item difficulty and response types, offers a more sensitive assessment of naming improvements in stroke aphasia.
• Findings support the decomposability of naming ability into subabilities, suggesting tailored outcome measurement.
• This approach provides new avenues for evaluating the effectiveness of anomia therapies.