Using the binomial distribution to assess effort: forced-choice testing in neuropsychological settings

Area of Science:

• Neuropsychology
• Psychometrics
• Statistics

Background:

• The binomial distribution is often overlooked for effort assessment in neuropsychological testing.
• Previous research has explored its utility, but further investigation is warranted.

Purpose of the Study:

• To evaluate the efficiency of binomial distribution cut points for the Test of Memory Malingering (TOMM).
• To compare these cut points with the existing manual recommendations and a novel error statistic.

Main Methods:

• Utilized published clinical and simulated data for the TOMM.
• Employed inverse binomial distribution functions to derive eight new cut points.
• Conducted repeated measures, multivariate, and univariate ANOVAs with Bonferroni correction.
• Assessed group homogeneity using normal curve density functions.

Main Results:

• Eight inverse binomial distribution cut points demonstrated comparable or superior efficiency to the TOMM manual's cut point for correct classifications.
• The new statistic, total number of errors, also showed potential for improved classification.
• Analysis supported recommendations for revised TOMM decision rules.

Conclusions:

• The binomial distribution is a viable and efficient statistical tool for effort assessment in neuropsychological tests.
• Revised decision rules for the TOMM could enhance classification accuracy.
• Strategies for improving norms in neuropsychological assessments were discussed.