Derivation and validation of a quantitative definition of specific reading disability for adults.

One hundred and fifty-four adult relatives of children with specific reading disability, and a group of 90 adults matched for age, sex, educational level and IQ, were given a battery of intelligence, reading and spelling tests and a questionnaire on their reading history, habits and attitudes. Data obtained from the latter group were used to derive multiple regression equations for the prediction of reading and spelling scores. An index was derived for each adult, expressing actual scores relative to those predicted. A definition of severe and borderline reading disability was validated, using the scores from the questionnaire.