Premorbid intelligence score in Neuropsychological assessment

Abstract

Background: Tests to assess cognitive impairments differ in dimensions covered, number and format of items, scoring methods, score ranges, etc. and are not comparable.

Method: The paper converts discrete raw scores of an item to continuous score following normal distribution in a desired score range and proposes equating premorbid estimates of tests using area under Standard Normal Curve. Dimension score of an individual is taken as sum of normally distributed item-wise P-scores and test score is the sum of sub-scale scores.

Results:  Normally distributed proposed scores (P-scores) facilitate meaningful arithmetic aggregation, parametric statistical analysis including estimation and testing hypothesis, assessing progress/deterioration of cognitive ability of a patient or a group of patients across time.

Conclusions: Proposed method to find equivalent scores may be preferred over conversion tables generated from the Crosswalk Studies which may vary by several points. Considering the theoretical advantages, use of the proposed method of equating in terms of P-scores is recommended.