Abstract
Patterns of reading disability were examined in 2 longitudinal studies. The major findings were (a) that on the basis of the observed data, remission of reading disability was relatively common with up to 37% of reading-disabled children showing remission of this disability within a 2-year period, and (b) when the data were analyzed with a latent Markov model that took account of measurement errors in test scores, the estimated true rate of remission of reading disability was between 15% and 19% over a 2-year period. The presence of measurement error in reading disability classifications may lead to an inflated and misleading impression of the rate of remission of these problems. General implications of these findings for interpreting patterns of stability and change in longitudinal-developmental data were discussed.