A comparative analysis of naming errors made by subjects with naming impairment following stroke, Alzheimer's disease, and traumatic brain injury.

Abstract

In two studies, the naming errors of individuals with naming impairment following stroke, Alzheimer's dementia and traumatic brain injury (TEl) were compared. Study 1 compared the errors of mild to moderately naming impaired individuals following stroke (n= 20), AD (n = 20), and TEl (n = 20), who were given a confrontation naming test. An error analysis was used to determine whether patterns existed within or across groups. Although each of these groups showed patterns in their naming errors, the differences were subtle. The particular types of naming errors were distributed differently among the three groups. A more consistent finding was a trend in all three groups toward more semantic errors in those with mild naming impairment and fewer semantic error in those with more severe naming impairment. In the stroke and TEl groups there was a trend of more irrelevant errors in those with more severe naming impairment and fewer irrelevant errors 100 in those with mild naming impairment. In Study 2, four groups of individuals post-stroke were given the same naming test. These groups were labeled high comprehension fluent (n = 17), low comprehension fluent (n = 10), low comprehension nonfluent (n = 11), and high comprehension non fluent (n = 10). Few naming error differences were found between groups, and these few differences were subtle. However the pattern in Study 1 also occurred for the subjects in Study 2. Individuals with milder naming impairment had a high frequency of semantic errors across the four groups. Individuals with more severe naming impairment had a high frequency of irrelevant errors, across the four groups. In the naming error literature two main issues are addressed. First, are different populations of naming impaired individuals affected differentially or, alternatively, is the severity of naming impairment a more salient determinant of naming error pattern. For individuals with mild to moderate naming impairment, the results of the current study favor the latter conclusion.