Research issue 07 · September 13, 2026

Weekly research brief · Dyslexia

What matters.
What to do with it.

A concise, evidence-ranked scan for educational therapists—without the research-paper fog.

7 minute read3 research resources3 practice prompts
This week’s signal

The strongest result this week links sustained decoding and oral-language work to small, durable comprehension gains. A promising spelling package still needs replication, while special-font marketing remains ahead of measured reading benefit.

What matters this week

Build both sides of reading—and make every add-on prove itself

Vote on each one. Your feedback shapes the next scan.

01
High confidenceRecent backlog · Preregistered multisite randomized trial · Published May 2026

Sustained decoding-plus-language instruction produced small, durable comprehension gains.

A Norwegian multisite trial randomized 415 children ages 7–9 at risk for reading disorders to usual instruction or four 45-minute UiT-ReadWell sessions weekly for 24 weeks. The program combined explicit decoding with vocabulary, grammar, listening, and narrative work. Postintervention standardized mean differences were 0.26 for decoding, 0.20 for language comprehension, and 0.26 for reading comprehension; effects remained small at six to eight months (0.22, 0.16, and 0.24). Word reading and language gains mediated about 27% and 34% of the comprehension effect. The sample fell below the planned 506, four protocol deviations were disclosed, and Norwegian's transparent orthography plus trained-teacher intensity limit transfer to Los Angeles. Norway's Research Council funded the study; authors declared no conflicts.

Why it matters at edcahoots

Plan and monitor decoding, oral language, and reading comprehension as connected but distinct targets. The trial supports sustained integration, not a promise that a shorter educational-therapy schedule will reproduce its effects.

Read the primary source Nergård-Nilssen et al. · Journal of Educational Psychology
02
Early/Uncertain confidenceNew development · Small randomized controlled trial · Published September 10, 2026

A large spelling gain came from a promising package—not proof of any single ingredient.

Fifty Saudi Arabic-speaking students ages 9–10 with learning disabilities were randomized to usual classroom instruction or twenty 45–50-minute English spelling sessions across 10 weeks. The package combined explicit rules, phoneme–grapheme mapping, morphology, word patterns, guided practice, self-monitoring, immediate feedback, and multisensory activities. Intervention scores rose from 60.18 to 80.13 versus 58.12 to 63.61 for controls; the reported gain comparison was very large (d = 1.92). Yet one teacher delivered the program, the same researcher-made 20-word test was reused, and only the intervention group had a two-month follow-up. The tiny English-as-a-foreign-language sample and bundled design prevent attributing gains to multisensory instruction or assuming transfer. Prince Sattam bin Abdulaziz University funded the study; the author declared no competing interests.

Why it matters at edcahoots

Build spelling plans from phonological, orthographic, and morphological error patterns, with cumulative practice and immediate feedback. Verify transfer using unfamiliar words rather than the taught list alone.

Read the primary source Alamri · Scientific Reports
03
High confidenceRecent backlog · Random-effects meta-analysis · Online July 31, 2026

Dyslexia-friendly fonts did not reliably improve reading speed or accuracy.

A random-effects meta-analysis combined 15 empirical studies, 91 effect sizes, and 688 readers with dyslexia to test fonts marketed as dyslexia-friendly against common fonts such as Arial and Times. The average effect on reading performance was essentially zero (Hedges' g = -0.04, 95% confidence interval -0.15 to 0.07; p = .50), providing no reliable speed or accuracy advantage. The evidence base is still modest and varies in typography, tasks, languages, and outcomes, so the null average does not mean every reader experiences identical comfort or legibility. It does separate an accessibility preference from remediation: a learner may prefer a font without it treating dyslexia. Authors declared no competing interests; the accessible record did not identify study funding.

Why it matters at edcahoots

Honor individual readability and comfort preferences, but do not market or prescribe a special font as reading treatment. Spend educational-therapy time on instruction unless typography measurably improves access or engagement.

Read the primary source Azzarello et al. · Annals of Dyslexia

Worth watching

Promising, but not practice-changing yet.

A July 2026 systematic review found that 20 of 21 controlled rhythmic-training studies reported some language or reading improvement. The evidence remains too heterogeneous for a practice change: control conditions, outcome measures, follow-up, and intervention designs varied widely. This returns to August's rhythm-game item because the new review broadens the evidence base, but it does not overturn that trial's null primary outcome. Watch for preregistered trials with active controls, durable reading outcomes, and transparent effect sizes.

Questions for the team

Turn evidence into a useful conversation.

01

Do treatment plans name and measure decoding, oral-language, and reading-comprehension targets separately?

02

Could spelling reviews classify phonological, orthographic, and morphological errors, then test transfer with unfamiliar words?

03

Are font choices framed as comfort and accessibility preferences rather than dyslexia treatment?

Educational therapy research, filtered for practice.

Research information only—not individualized medical, diagnostic, or legal advice. Please do not include student names in feedback.

Weekly dyslexia research briefSeptember 13, 2026

What matters this week

Build both sides of reading—and make every add-on prove itself

A strong multisite trial supports sustained decoding-plus-language instruction; a small spelling trial offers a useful error-profile model with major caveats; and a meta-analysis finds no reading advantage from dyslexia-friendly fonts.
1

High confidence · Recent backlog · Preregistered multisite randomized trial · Published May 2026

Sustained decoding-plus-language instruction produced small, durable comprehension gains.

Takeaway

A 415-student randomized trial found small, durable gains when intensive instruction combined decoding with language comprehension.

Practice implication

Plan and measure decoding, oral language, and reading comprehension as connected but distinct targets.
2

Early/Uncertain confidence · New development · Small randomized controlled trial · Published September 10, 2026

A large spelling gain came from a promising package—not proof of any single ingredient.

Takeaway

A 50-student spelling trial produced a large gain, but reused one word test and bundled many instructional components.

Practice implication

Classify spelling errors, teach cumulatively, and verify transfer with unfamiliar words.
3

High confidence · Recent backlog · Random-effects meta-analysis · Online July 31, 2026

Dyslexia-friendly fonts did not reliably improve reading speed or accuracy.

Takeaway

Across 15 studies and 688 readers, dyslexia-friendly fonts produced no reliable reading-speed or accuracy advantage.

Practice implication

Honor readability preferences without presenting a special font as remediation.

Questions for the team

• Do treatment plans name and measure decoding, oral-language, and reading-comprehension targets separately?

• Could spelling reviews classify phonological, orthographic, and morphological errors, then test transfer with unfamiliar words?

Primary sources

1. Nergård-Nilssen et al. · Journal of Educational Psychology: https://doi.org/10.1037/edu0001017

2. Alamri · Scientific Reports: https://doi.org/10.1038/s41598-026-70926-3

3. Azzarello et al. · Annals of Dyslexia: https://doi.org/10.1007/s11881-026-00389-8

Built by Brantley