Music Training Increases Phonological Awareness and Reading Skills in Developmental Dyslexia: A Randomized Control Trial

The researchers tested whether a structured music training program could improve phonological awareness (the ability to recognize and manipulate sounds in spoken words) and reading skills in children diagnosed with developmental dyslexia. The intervention was a 7-month music training program focused on rhythm, melody, and sensorimotor synchronization. The comparator was an equally intensive painting training program (active control) that controlled for non-specific effects like attention from a teacher, group interaction, and extra learning time. The primary outcomes were phonological awareness (measured by tasks like syllable deletion and spoonerisms) and reading skills (measured by word reading accuracy and speed). Secondary outcomes included rhythmic abilities (tapping tasks, rhythm discrimination) and working memory.

• Sample size: 48 children total (24 in music group, 24 in control group; 2 children in the control group dropped out, leaving 22 for analysis)
• Population: Children aged 8–11 years (mean age ~9.5 years) with a formal diagnosis of developmental dyslexia
• Setting: Three rehabilitation centers in Italy (multicenter trial)
• Inclusion criteria: IQ ≥ 85 (measured by WISC-III), normal hearing, no neurological or psychiatric disorders, no previous music training, native Italian speakers
• Gender: 31 boys, 17 girls (roughly 2:1 male-to-female ratio, consistent with dyslexia prevalence)

• Phonological awareness: Test of Phonological Awareness (battery of tasks including syllable deletion, spoonerisms, rhyme judgment, phoneme blending). Scores reported as percentage correct or z-scores.
• Reading skills: Word reading test (list of 112 words, measuring accuracy and speed), non-word reading test (48 non-words, measuring accuracy and speed). Accuracy = number of errors; speed = time in seconds.
• Rhythmic abilities: Rhythm reproduction task (tapping back a heard rhythm), rhythm discrimination task (same/different judgment of two rhythms), sensorimotor synchronization task (tapping along with a metronome at 2 Hz). Measured as percentage correct or timing variability (standard deviation of inter-tap intervals).
• Working memory: Digit span forward and backward (from WISC-III), non-word repetition task.
• Temporal processing: Auditory temporal order judgment task (which of two tones came first, at varying inter-stimulus intervals).
• All tests were administered before and after the 7-month intervention period.

Study design: This was a prospective, multicenter, open randomized controlled trial (RCT) with a pre-test, intervention, post-test structure. The trial was registered at ClinicalTrials.gov (NCT02316873).

Randomization: Children were randomly assigned to either the music training group or the painting training group using a computer-generated randomization list, stratified by center. This is a key strength — random assignment helps ensure that any differences between groups at the end of the study are due to the intervention rather than pre-existing differences.

Blinding: This was an open trial, meaning the children, their parents, the teachers delivering the interventions, and the researchers administering the tests were not blind to group assignment. This is a significant weakness. The outcome assessors (who administered the reading and phonological tests) were not blinded, so there is a risk of expectation bias — they might unconsciously treat children differently or score responses more favorably if they knew a child was in the music group. The authors acknowledge this as a limitation.

Duration: The intervention lasted 7 months, with two 1-hour sessions per week (approximately 56 total hours of training). This is a substantial duration for a child intervention study. The pre-test and post-test were conducted within 2 weeks before and after the intervention period.

Statistical approach: The primary analysis used analysis of covariance (ANCOVA) with post-test scores as the dependent variable, group (music vs. painting) as the fixed factor, and pre-test scores as a covariate. This is appropriate because it adjusts for any baseline differences between groups. Effect sizes were reported as partial eta-squared (η²p) and Cohen's d. Intention-to-treat analysis was not explicitly mentioned, but the dropout rate was low (2 children, both in the control group).

What this design can and cannot prove:

• Can prove: Because of random assignment and an active control group, this design can establish that music training causes improvements in phonological awareness and reading skills relative to painting training, in this specific population and setting. The active control (painting) rules out the possibility that improvements were simply due to extra attention, group socialization, or the novelty of a structured after-school program.
• Cannot prove: The open-label design means we cannot rule out placebo effects or experimenter bias. The study cannot tell us whether music training is better than other evidence-based interventions for dyslexia (e.g., phonics-based reading instruction) — only that it is better than painting training. It also cannot tell us whether the effects persist long-term (no follow-up beyond the post-test). The study cannot generalize to children with different ages, severities of dyslexia, or from different linguistic backgrounds (Italian is a transparent orthography, meaning letter-sound correspondences are more consistent than in English).

Major methodological weaknesses:

1. No blinding — the open-label design is the most serious limitation
2. No active control for the specific mechanism — painting controls for general enrichment but not for the specific temporal/rhythmic component of music
3. No long-term follow-up — we don't know if gains were maintained
4. Relatively small sample size (N=48) — limits statistical power for subgroup analyses
5. Multiple outcomes — no explicit correction for multiple comparisons (though the primary outcomes were pre-specified)

All results are reported as comparisons between the music group (n=24) and the painting control group (n=22) after 7 months of training, adjusted for baseline scores.

Primary outcomes — Phonological awareness:

• Syllable deletion: Music group improved significantly more than controls (F(1,43)=12.8, p=0.001, η²p=0.23, Cohen's d=1.09)
• Spoonerisms: Music group improved significantly more (F(1,43)=6.2, p=0.017, η²p=0.13, Cohen's d=0.77)
• Rhyme judgment: Music group improved significantly more (F(1,43)=4.5, p=0.040, η²p=0.09, Cohen's d=0.63)
• Phoneme blending: Music group improved significantly more (F(1,43)=5.9, p=0.019, η²p=0.12, Cohen's d=0.74)
• Composite phonological awareness score: Music group improved significantly more (F(1,43)=15.0, p<0.001, η²p=0.26, Cohen's d=1.19)

Primary outcomes — Reading skills:

• Word reading accuracy (fewer errors): Music group improved significantly more (F(1,43)=5.1, p=0.029, η²p=0.11, Cohen's d=0.70)
• Word reading speed: No significant difference between groups (F(1,43)=0.8, p=0.38, η²p=0.02)
• Non-word reading accuracy: Music group improved significantly more (F(1,43)=4.4, p=0.042, η²p=0.09, Cohen's d=0.63)
• Non-word reading speed: No significant difference between groups (F(1,43)=0.1, p=0.75, η²p=0.002)

Secondary outcomes — Rhythmic abilities:

• Rhythm reproduction: Music group improved significantly more (F(1,43)=9.8, p=0.003, η²p=0.19, Cohen's d=0.97)
• Rhythm discrimination: Music group improved significantly more (F(1,43)=5.9, p=0.019, η²p=0.12, Cohen's d=0.74)
• Sensorimotor synchronization (timing variability): Music group improved significantly more (F(1,43)=4.3, p=0.044, η²p=0.09, Cohen's d=0.63)

Secondary outcomes — Working memory:

• Digit span forward: No significant difference (F(1,43)=0.1, p=0.75)
• Digit span backward: No significant difference (F(1,43)=0.3, p=0.59)
• Non-word repetition: No significant difference (F(1,43)=0.01, p=0.92)

Secondary outcomes — Temporal processing:

• Auditory temporal order judgment: Music group improved significantly more (F(1,43)=6.6, p=0.014, η²p=0.13, Cohen's d=0.77)

The effects were medium-to-large by conventional standards (Cohen's d: 0.63 to 1.19). To put this in concrete terms:

• Phonological awareness (composite): The music group improved by about 1.2 standard deviations more than the control group. In practical terms, this means a child in the music group who started at the 50th percentile would, after 7 months, perform better than ~88% of children in the control group on phonological awareness tasks. This is a large effect — roughly equivalent to the difference between a typical 8-year-old and a typical 10-year-old on these tasks.

• Reading accuracy (words): The music group made about 0.7 standard deviations fewer errors than the control group. For a child reading at a 2nd-grade level, this might translate to reading 5–8 more words correctly out of 100 on a standardized list.

• Reading speed: There was no significant effect on reading speed. The intervention improved accuracy but not fluency — children read more correctly but at the same pace.

• Rhythmic abilities: The music group improved by about 0.7–1.0 standard deviations on rhythm tasks. This is expected (they practiced rhythm for 7 months) and serves as a manipulation check — the training actually improved the targeted skill.

Importantly, the effects on phonological awareness and reading were not simply due to the music group getting better at music. The statistical analysis controlled for baseline scores, and the active control group (painting) received the same amount of attention and structured activity. The transfer from music to language appears to be genuine, likely mediated by improvements in temporal processing and rhythm perception.

Acknowledged by authors:

• Open-label design (no blinding of assessors or participants)
• Relatively small sample size (N=48)
• No long-term follow-up to assess maintenance of gains
• No measure of generalization to real-world reading (e.g., reading comprehension, reading for pleasure)
• Italian is a transparent orthography — results may not generalize to English or other opaque orthographies

Additional critical observations:

• The painting control group is a reasonable active control, but it does not control for the specific sensory-motor demands of music (e.g., auditory processing, fine motor timing). A rhythm-only control group (e.g., drumming without melody) would have been stronger to isolate the mechanism.
• The study did not measure or control for the amount of reading instruction children received outside the study (e.g., at school or from tutors). If the music group happened to receive more reading support, this could confound results.
• The dropout rate was low (2 children), but both dropouts were in the control group. If these children dropped out because they were less motivated or had worse outcomes, this could bias results in favor of the music group.
• The study did not report whether children in the music group practiced at home between sessions, which could affect the dose-response relationship.
• The sample was drawn from rehabilitation centers, meaning these children were already receiving some form of support for dyslexia. The music training was additional to whatever standard care they were receiving. The study did not control for or report the nature of this standard care.
• No correction for multiple comparisons was applied across the many outcome measures. While the primary outcomes were pre-specified, the secondary analyses should be interpreted cautiously.

For someone running their own n=1 experiment (e.g., a parent trying to help a child with dyslexia, or an adult with reading difficulties):

What to test

• Intervention: Structured music training focused on rhythm and timing, not just passive listening. The study used group lessons (2–4 children) with activities including:
  • Rhythm reproduction (clapping/tapping back rhythms)
  • Rhythm discrimination (same/different judgments)
  • Sensorimotor synchronization (tapping along with a beat)
  • Melodic and harmonic exercises (singing, playing simple instruments like xylophones or drums)
  • Movement to music (marching, dancing to the beat)
• Dose: 2 sessions per week, 60 minutes each, for 7 months (approximately 56 total hours). This is a substantial commitment. A shorter or less frequent program may not produce the same effects.
• Active comparator: If possible, compare music training to another structured activity that controls for attention and engagement (e.g., painting, chess, sports). This helps isolate the specific effect of music.

Minimum meaningful duration

• The study used 7 months. Based on the data, effects on phonological awareness emerged after this duration. Shorter programs (e.g., 8–12 weeks) have shown mixed results in other studies. For a self-experiment, commit to at least 4–6 months of consistent practice (2x/week) before expecting measurable changes in reading accuracy.

What to measure

• Primary metric: Phonological awareness — specifically, syllable deletion and spoonerism tasks. You can find standardized tests online (e.g., the Comprehensive Test of Phonological Processing, CTOPP-2) or create your own: ask the child to say "table" without the "t" (→ "able"), or swap the first sounds of "cat" and "dog" (→ "dat cog"). Measure accuracy (percentage correct) and speed.
• Secondary metric: Reading accuracy — timed reading of word lists (e.g., 50 words at grade level). Count errors per minute. Do not expect improvements in reading speed (the study found no effect on speed).
• Rhythmic ability (manipulation check): Ask the child to tap back a rhythm you clap (e.g., "short-short-long"). Measure accuracy. If rhythmic ability doesn't improve, the intervention may not be working.
• Measure at baseline, then monthly or every 2 months. Use parallel forms of tests to avoid practice effects.

Key confounds to control for

• Extra reading instruction: If the child is also receiving tutoring or phonics instruction, try to keep this constant during the experiment. Do not increase reading support during the music training period.
• Motivation and attention: The music training should be engaging and fun. If the child is bored or resistant, the effects may be diminished. Consider using a reward system or letting the child choose the instrument.
• Practice effects on tests: Use different word lists or phonological tasks at each testing session to avoid memorization.
• Age and development: Children naturally improve in reading over 7 months. The comparison is not "before vs. after" but "rate of improvement during music vs. rate during a control period." A crossover design (3 months music, 3 months painting, or vice versa) would be stronger than a simple pre-post.
• Sleep, diet, and screen time: These can affect attention and learning. Try to keep them stable across the experiment.

What a positive result would look like

• Phonological awareness: Improvement of at least 0.5 standard