Topics

• Introduction
• Random sample
• Course of the study
• Training
• Data collection
• Standardized series of tests
• Results
• Publications

Until the spring of 2006 Dybuster was mainly developed of a research team around Prof. Markus Gross of the computer science department at the ETH Zürich. The software is founded on a mathematical and psychological model. The present study from the period between April and December 2006 demonstrated the effectiveness of Dybuster. The study was supported by Prof. L. Jäncke and Dr. M. Meyer from the neuropsychological institute of the University of Zürich. On this page you'll find a description of the study, the random sample, the course and of the performed series of tests. The results of the study are presented at the end of the page.

Only native German speaking children were taken into the random sample. The random sample consisted of 43 dyslexic children, of which 15 were female, and 37 control children of the same age (17 female). Both groups were split up into a training and non-training group. Altogether, 4 distinct groups were formed. It should be noted that all four groups were training, but at different points in time. The groups described with 'training' trained during the first three month. The groups described with 'non-training' performed their training during the second three month study period. It was paid attention to the suspension of the training of the training group for the second period.

The children of the training group worked daily for about 15 - 20 minutes at home at their own computer for a period of 3 month. Once a week they performed their training under surveillance of psychologist and computer scientists at the ETH. However, they did not receive a therapeutical support, but were merely advised of the functionality and learn aids of Dybuster.

To measure the learn progress, all the children completed a spelling test at the beginning, after 3 month and at the end of the study. The test consisted of 100 words, containing 50 words which were trained in Dybuster. The 50 not trained words were used to objectify the transfer effect of the Dybuster training. The trained and not trained words are equal in terms of frequency of occurrence, difficulty and number of syllables.

Further, all the children were neuropsychologically examined. They were divided into the dyslexic and non-dyslexic group by classical spelling tests as well as by a standardized reading test. We used the Salzburger-reading-spelling-test (SLRT) or the diagnostic spelling test (DRT 5) for 5th grade children. As the reading test the Zürcher reading test (ZLT) was used. To ensure that no child has lower reading and writing skills due to minor intelligence, we performed an IQ test (HAWIK III). Further we tested the attention functions. The tonic and phaseal alterness was examined by a series of attention tests (TAP). Furthermore we collected the ADHD/ODD-parents-questionnaire (Steinhausen 2001). To test the frontal brain functions we used the digitalized card sorting procedure (rule and concept finding test) as well as the GoNogo from the TAP (impulse control test). The verbale working memory was determined on one hand by the HAWIK III (number repeating subtest) and on the other hand by the 1-back from the TAP.

The results showed that dyslexic children without the training could increase their writing skills only marginally (5%). This finding confirms the experience that dyslexic children can hardly benefit from regular school lessons. It's pleasant that dyslexic children were able to increase their writing skills by 27% during the 3 month of training. Thereby is important that the children did not only improve in writing learnt words (32%), but also in writing the not learnt words (23%). This shows that a transfer effect took place.
In addition, the results showed that dyslexic children can profit from the Dybuster training as much as non dyslexic children. The control group with training increased their writing skills by 26%. The difference between learnt (27%) and not learnt (26%) words, however, was marginal. This indicates that the healthy control children were able to draw analogy conclusions. Another remarkable finding is that the control children without training also improved by 17%. This points at a higher gain from the regular school lessons of healthy children.

Table 1: Relative changes first learn period

The results from the first period of the study could not only be replicated, but even be exceeded in the second period. The dyslexic group, training in the second phase, improved by 29%. The error reduction on the trained words was even 35%. Also this group showed transfer effects by an improvement of 23% on the not learnt words. Further, the results show a performance decrease (5%) of the dyslexic children, which had to stop their Dybuster training. The performance decreased by 11% on the learnt words and 4% on the not learnt. This result shows that dyslexic children need a continuous and intensive training for a long-term successful writing performance. The healthy children, however, who suspended their training, were able to increase their writing performance.

Table 2: Relative changes second learn period

The values in table 3 show that all groups could improve their writing skills significantly thanks to the Dybuster training. The control group increased their writing performance slightly more than the dyslexic group. This difference can either be due to a stronger gain of the school lessons or to the lower forgetting values of the control group.

Table 3: Relative changes entire study

• Kast M., Meyer M., Voegeli C., Gross M. and Jaencke L., Computer-based multisensory learning in children with developmental dyslexia, Restorative Neurology and Neuroscience, Vol. 25, Pages 355 – 369, IOS Press, 2007.

back to top