Presented at the 9th European Conference on Psychological Assessment. Thessaloniki, Greece, 3–6 May.

Specific learning disability in written language is associated with difficulties in learning to read and write, expressed as poor reading and spelling performance. In English, the high complexity and inconcistency of the orthography allows assessment of reading focused on accuracy measures. However, in orthographically more transparent languages, such as Greek, deficits in reading accuracy, despite being statistically significant, are not very powerful indices of reading problems, because the overall performance of even disabled readers is rather high. In such cases, fluency emerges as the most reliable indicator or reading disability, as attested in German, Spanish, Finnish etc. In our studies of Greek schoolchildren we have contrasted clinical samples, rich in children diagnosed with dyslexia (specific reading disability), with the general school population. Reading speed was found to exhibit the largest between-group effect size, both for 3rd-4th grade (11 vs 167 children, respectively) and for 7th grade (28 vs 185 children). For both age groups, the effect sizes were the same whether we compared pure reading speed, measured as the time to complete reading of a word list or a passage, or reading fluency, measured as the number of correct words per unit time (based on the same word list and passage readings). In the former case, reading speed measures loaded on a "speed" factor distinct from reading accuracy, phonological awareness and spelling, resulting in a 3-factor structure (the 3rd factor taking up intelligence and comprehension measures), whereas in the latter case reading fluency loaded on the same factor as accuracy and spelling, resulting in a 2-factor structure.