BackgroundChildren are diagnosed as dyslexic when their reading performance is much below that which could be expected for their educational level and cannot be explained by a sensory, neurological or psychiatric deficit or by a low IQ. Although poor reading is a major obstacle to school and career achievement, the causes of dyslexia are unclear and traditional therapies are often unsuccessful. To determine the causes of dyslexia, experiments must demonstrate under which conditions a reading disorder occurs and whether the reading performance improves if these conditions are abolished or compensated. To avoid irreproducible results, experiments must be repeated and the effect size must be calculated.ObjectivesThe aims of the study were to investigate the rate and location of misread letters within pseudowords, prove the effectiveness of compensatory reading therapy and demonstrate the reproducibility of the experimental results. The influence of reading therapy on the rate of eye movements opposite to the reading direction was investigated and causes of a poor reading performance were identified.MethodsThe rate and location of misread letters were investigated by tachystoscopic presentation of pseudowords containing between three and six letters. Presentation time, fixation time, and the time it takes to begin pronouncing the words (speech onset latency) were changed until 95% of the pseudowords were recognized correctly. On the basis of these results, the children learned a reading strategy that compensated the causes of the reading disorder. The therapy was demonstrated to be highly effective and it was shown that the results of the therapy were reproducible.ResultsIt was shown that misread letters occurred at all locations in pseudowords, regardless of the word's length. Inadequate fixation, excessively large saccadic amplitudes, reduced ability to simultaneously recognize a sequence of letters, a longer required fixation time and a longer required speech onset latency were all identified as causes of dyslexia. Each of the studies included in the meta-analysis were much more efficient than conventional therapeutic methods. The overall effect size with a value of Hedges' G = 1.72 showed that the therapy had a reproducible and stable effect.ConclusionsThe causes of dyslexia can be revealed by a dual-intervention approach consisting of a pseudoword experiment and learning a compensatory reading strategy. Reading performance improves immediately if the identified causes of dyslexia are compensated by an appropriate reading therapy.