Rasch Model Logits: Interpretation, Use, and Transformation

Abstract

This article presents (a) a nontechnical explanation of Rasch model "logit" estimates of pediatric functional ability, (b) an illustration of two alternative transformations of logit estimates of functional ability, and (c) a pragmatic solution to the problem of obtaining a functional ability estimate for a zero or perfect performance score. An explanation of logits is offered because their presentation to practitioners in the clinic is often complicated and confusing. The transformations are performed because they remove negative and decimal valued estimates of functioning. The issue of zero and perfect scores is addressed because these are often the scores obtained by the youngest and oldest children in a clinical setting.