Classification accuracy of multiple visual spatial measures in the detection of suspect effort

Area of Science:

• Neuropsychology
• Cognitive Assessment
• Psychometrics

Background:

• Cognitive measures, especially memory tests, are used to detect suspect effort in neuropsychological assessments.
• Visual spatial measures have been less explored for their ability to identify biased responding.

Purpose of the Study:

• To evaluate the classification accuracy of common visual spatial measures in detecting suspect effort.
• To determine if combining visual spatial measures enhances the detection of biased responding.

Main Methods:

• Utilized data from 491 neuropsychological referrals, classifying them into unbiased (N=415) and biased responding (N=30) groups based on effort measures.
• Administered and analyzed performance on the Judgment of Line Orientation Test, Benton Facial Recognition Test, Hooper Visual Organization Test, and Rey Complex Figure Test.
• Employed receiver operating characteristic (ROC) analysis to assess classification accuracy of individual and combined measures.

Main Results:

• All individual visual spatial measures demonstrated acceptable classification accuracy in detecting biased responding.
• A combined score from all visual spatial measures showed excellent classification accuracy.
• Individual measures achieved ~32-46% sensitivity at >=87% specificity; combined measures increased sensitivity to 57% while maintaining 87% specificity.

Conclusions:

• Visual spatial measures are valuable tools for detecting suspect effort in neuropsychological evaluations.
• Combining multiple visual spatial measures significantly enhances the ability to identify biased responding.
• The findings support the use of multiple effort measures in clinical practice for improved diagnostic accuracy.