The Effect of Doolexia – a Digital Training Method for Students with Dyslexia

Doolexia is a computer based learning program where the students daily, either in school or at home, train with custom reading and spelling exercises. An adaptive algorithm selects exercises to fit the proficiency of the student. A teacher at the school, who is also trained to use Doolexia, further supports the intervention with at least two weekly lessons throughout the intervention period.

Doolexia is developed by Jan Bay Kønig and Peer Kromann Nielsen and pilot tested in eight schools (Kønig & Nielsen, 2021).

2023-01-11

2023-12-15

The primary outcomes are students’ spelling test and reading test.

The spelling and the reading tests are widely used in Danish primary and lower secondary school to assess and analyze students’ spelling and reading proficiency. They are from the publishing firm Hogrefe Psykologisk Forlag, and the spelling tests (in Danish: Hogrefe ST prøver) are developed by Inger-Lise Heinze and Poul Erik Jensen, and the reading tests (in Danish: Hogrefe læseprøver) are developed by Lene Møller and Holger Juul.

We test the effect on students’ average ‘C-level’ on the spelling test and average ‘K-level’ on the reading test.

We conduct a cluster randomized controlled trial (C-RCT).

The randomization unit is the school. The intervention takes place and is implemented at school level and thus we randomize at school level. All students at the school receives the intervention at the same time. The intervention period is about 16 weeks.

Schools register participating students in October 2022 and complete all pre-tests in October-November 2022. Randomization in November 2022 (conducted by researcher from VIVE). The randomization result is sealed until the day of announcement on November 30, 2022.

Using a waiting-room design, all schools (and thus all students) will at some point during the trial period receive the intervention. Schools assigned the treatment group will receive the intervention in January – June 2023 while the control schools teach as usual. The control schools will receive the intervention in August – December 2023.

Inclusion criteria: In order to participate, the school must sign up at least 7 students passing the inclusion criteria and assign at least one teacher to the intervention (at least two lessons per week throughout the intervention period). For a student to be included, the student must be assessed dyslexic in the Danish national dyslexic test (category ‘red’ or ‘yellow’). Students must be enrolled in grades 4-8 in the academic school year 2022/23.

Exclusion criteria: Students with ADHD or other behavioral or mental diagnosis.
Outcomes: Outcomes are measured at student level.
All participating students complete the spelling and reading test before randomization. These tests will constitute a pre-test. Students complete the same tests after the 16 weeks intervention period and a follow-up test about a half-year later, to evaluate long-term effetcs.

Analysis plan: We use results from spelling and reading tests before and after 16 weeks of using Doolexia (for the treatment group) and 16 weeks of usual dyslexic training (for the control group) to evaluate the average effect of Doolexia.

We test heterogeneous effects with respect to the students’ initial placement on the pre-tests (levels C0-C4 on the spelling test and K1-K5 on the reading test) and school type (public or private).

We test mediating effects with respect to the intervention ‘dosage’ received, meaning how often the students have used the computer program.

We test moderating effects with respect to the variables of stratification (group size and proportion of participants with severe learning difficulties).

Randomization done in office by a computer using Stata.

The randomization result is sealed until the day of announcement on November 30, 2022.

The randomization unit is the school. The intervention takes place and is implemented at school level and thus we randomize at school level. All students at the school receives the intervention at the same time.


Stratification: In order to ensure equal allocation of the participating schools into the treatment and control groups we use stratification before randomization. We stratify on two main characteristics: i) The number of participating students in the trial in each school (above or below 19 students) and ii) the proportion of the participating students who test ‘C0-C2’ on the initial spelling test, meaning that the test indicate they have “severe learning difficulties”. We stratify on these two characteristics to ensure that treatment and control groups are balanced based on how challenged the schools are with respect to dyslexic students.

Yes

Approx. 34 schools.

Approx. 500 students

17 schools control group, 17 schools treatment group.