ABSTRACT
This study reports a comparison of screening tests for dyslexia, dyspraxia and Meares-Irlen (M-I) syndrome in a Higher Education setting, the University of Worcester. Using a sample of 74 volunteer students, we compared the current tutor-delivered battery of 15 subtests with a computerized test, the Lucid Adult Dyslexia Screening test (LADS), and both of these with data on assessment outcomes. The sensitivity of this tutor battery was higher than LADS in predicting dyslexia, dyspraxia or M-I syndrome (91% compared with 66%) and its specificity was lower (79% compared with 90%). Stepwise logistic regression on these tests was used to identify a better performing subset of tests, when combined with a change in practice for M-I syndrome screening. This syndrome itself proved to be a powerful discriminator for dyslexia and/or dyspraxia, and we therefore recommend it as the first stage in a two-stage screening process. The specificity and sensitivity of the new battery, the second part of which comprises LADS plus four of the original tutor delivered subtests, provided the best overall performance: 94% sensitivity and 92% specificity. We anticipate that the new two-part screening process would not take longer to complete.