Validation of the sentence repetition test as a measure of suspect effort

It has been suggested that the Sentence Repetition Test (SRT) could serve as an adequate embedded symptom validity measure identifying suspect effort during neuropsychological testing. However, very little research has examined sensitivity and specificity rates when using this measure in a variety of clinical settings. The SRT was administered to 1031 patients referred for neuropsychological assessment in outpatient, inpatient, and independent medical evaluation settings. These patients were diagnosed with a wide range of psychiatric, developmental, and neurological disorders. The results of this study reveal that the SRT is a valid measure of suspect effort for the vast majority of these patients (sensitivity = 56.8% and specificity = 95.8% in combined clinical settings). However, analyses also indicate that the SRT is not a valid effort measure for individuals with mental retardation or dementia due to specificity rates falling well below 90% for both groups. Furthermore, the validity of the SRT as an effort measure is questionable for individuals with English as a second language, with a verbal learning disability, with a left cerebrovascular accident, or with an expressive-receptive language disorder due to small sample sizes or borderline specificity rates. Sensitivity, specificity, positive predictive accuracy, and negative predictive accuracy rates are provided for varying cutoff scores in inpatient, outpatient, and IME settings. The results of this extensive study confirm that the SRT can be a useful measure in detecting suspect effort in neuropsychological testing while also providing valuable clinical information.