Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Music education for improving reading skills in children and adolescents with dyslexia

This is not the most recent version

Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To study the effectiveness of music education on the spectrum of reading skills in children and adolescents with dyslexia.

Background

Description of the condition

Dyslexia is a specific learning disability "with a neurobiological origin marked by difficulties with accurate and/or fluent recognition and poor spelling" (World Health Organization 1992; Lyon 2003). Also known as developmental dyslexia, specific reading disability or specific learning difficulty, dyslexia is a difficulty in learning present in people with normal or higher than normal intelligence and not resulting from poor vision, hearing difficulty or a lack of socio‐environmental opportunities, lack of motivation, or even lack of adequate instruction (Shaywitz 2003). This review does not include dyslexia that is acquired, for example, due to brain injury or due to a condition such as glue ear. Dyslexia is a persistent difficulty, such that "children who fail to read adequately in 1st grade have a 90% chance of reading poorly in 4th grade and a 75% chance of reading poorly in high school" (Gabrieli 2009). Gabrieli 2009 also points out that the roots of dyslexia begin before initial reading instruction, but it is commonly diagnosed (in the United States) in children aged seven to eight years, by which time the reading difficulties are clearly measurable. According to Lyon 1996, 75% of children with reading disabilities not identified before the third grade continue to have reading disabilities into the ninth grade, and fewer than two per cent go on to participate in a four‐year higher education program after high school.

Dyslexia is strongly heritable (54% to 75%), occurring in up to 68% of identical twins and 50% of individuals who have a parent or sibling with dyslexia (Pennington 1996), and its prevalence is around 4% to 10% of school‐aged children (Shaywitz 1998; Bishop 2004; Blomert 2005), depending on the criteria applied (for example, Zoccolotti 2010 identified 17 types of developmental dyslexia). The main theory about its cause is related to a deficit in phonological processing (Bradley 1978; Wagner 1987; Ramus 2003), and more recent functional studies have revealed a hypo‐activation of the left temporo‐parietal cortex when dyslexic children are compared with typically developing readers (Hoeft 2007). Other authors, such as Kronbichler 2008 and Pernet 2009, have reported a reduction in gray and white matter in children and adults with dyslexia.

Reading can be thought of as having different components (or 'domains') and dyslexic children have problems with some of these components, such as phonological awareness and reading fluency. Phonological awareness is the ability to attend to and manipulate the sounds in words (Stanovich 1986), segmenting each individual speech sound. This awareness is important when learning to read in alphabetic script such as Spanish, German or English (Goswami 1990; Hulme 2002; Muter 2004). Reading fluency is most often defined as the ability to read text quickly, accurately, and with appropriate expression (National Reading Panel 2000; Kuhn 2003) and oral reading fluency has been shown to be related to comprehension (Fuchs 1988; Fuchs 2001; Jenkins 2003). Dyslexic children often have difficulty with this reading fluency domain, which results in poor comprehension. Recently, this domain was identified as an area of difficulty for individuals with dyslexia by the International Dyslexia Association (Lyon 2003). The reading process is slow and laborious for dyslexic children (Chall 1990) and may be followed by an avoidance of reading and general frustration (Pinnell 1995; Leinonen 2001).

Description of the intervention

This review is concerned with the potential effect that school‐based musical training (or music education) can have on children's reading. Here we consider music education as any methodology of teaching music, whether in a specific center of music teaching (for example, conservatory, Music School) or a single music class included in the syllabus of a kindergarten or elementary school. Generally, musical training or music education (both terms will be used interchangeably in this review) can be described as the process of learning music, supervised by a music teacher or a specialist in music education. A music teacher is defined as someone who has an understanding of the elements of music context, score analysis, musical style and aesthetics and is competent to teach this to children. There are different approaches to music education (Suzuki, Dalcroze, Orff and Kodály) and students are generally exposed to key elements of music such as rhythm, melody, harmony and timbre regardless of the approach, although each one has its own particular form. The main approaches to music education currently are:

  • Suzuki Method: developed by Shin´ichi Suzuki (1898‐1998), a Japanese educationist and violin teacher, who believed that children could learn musical skills at an early age, initially through listening, and observed the way young children acquire language through hearing others speak (Mills 1973). Just as children learn to speak before learning to read, Suzuki advocated a delay in teaching musical notation until there was an adequate grounding in playing skills and the development of musical memory. In the Suzuki Method, the children are taught in groups rather than individually, encouraging cooperation and teamwork.

  • Dalcroze Method: created by the Swiss musician and educator Emile Jaques‐Dalcroze (1869‐1950) who developed a system of group music teaching through ‘gymnastique rythmique’(eurhythmics). This system uses physical exercises together with the music class to help students respond physically and aesthetically to music (Henry 1958).

  • The German Orff‐Schulwerk system, created by Carl Orff, which combines choral singing, aural training, movement, improvisation and activities that use specially designed pitched and non‐pitched percussion instruments. Five books of teaching materials, Musik für Kinder (1950‐54), exemplify ways of making what is called "elementary music". Various arrangements of folk songs and traditional melodies are intended as models or suggestions for teachers rather than a comprehensive scheme (Keller 1963).

  • Kodály's principles of music teaching (Hungarian Method) are in many ways similar to those of Orff, but the Hungarian approach is more fundamentally choral. Concerned with the development of inner hearing and musical literacy, and determined to improve the musical life of the nation, Kodály drew on his country's folk song tradition, which he combined with art music using the pitch teaching principles of Curwen, hand signs and the rhythmic language of the Galin‐Paris‐Chevé movement (Sándor 1975).

In music education, each one of the musical elements (rhythm, timbre, aesthetic, harmony, pitch) can be studied and developed separately through different strategies in order to stimulate and develop children’s perception of them as shown by Moreno 2009. Studies vary in the frequency and length of sessions, for example, Overy 2003 studied 20 minute singing‐based music lessons delivered three times a week. Research has been conducted using different methodologies of musical teaching as a basis for musical practice, where multisensory and developmental group activities are presented to the children. Some activities may involve singing (all the children together singing the same musical passage, in unison, or in canon structure), rhythm (via corporal movement or corporal percussion) and instrumental practice (either highly technical in learning a specific musical instrument or using the instrument only as a form of interaction between the musical sheet system and the children).  

The National Association for Music Education, whose mission is to advance music education by encouraging the study and making of music by all, has established nine national standards for Music Education (MENC 2010).

1. Singing, alone and with others, a varied repertoire of music.
2. Performing on instruments, alone and with others, a varied repertoire of music.
3. Improvising melodies, variations, and accompaniments.
4. Composing and arranging music within specified guidelines.
5. Reading and notating music.
6. Listening to, analyzing, and describing music.
7. Evaluating music and music performances.
8. Understanding relationships between music, the other arts, and disciplines outside the arts.
9. Understanding music in relation to history and culture.

It is important to emphasize that musical training is not the same as music therapy, despite the fact that both use music as an intervention and both are often provided in the same setting. Music therapy is a psychotherapeutic method that uses musical interaction as a means of communication and expression and aims to help people with mental illness to develop relationships and to address issues they may not be able to address using words alone (Gold 2004; Gold 2009).

How the intervention might work

Studies correlate children's ability to read with their ability to distinguish pitches accurately (Fisher 2001; Hansen 2002; Schön 2004; Magne 2006; Besson 2007; Marques 2007; Nikjeh 2009), arguing for a strong link between basic auditory perception abilities and reading abilities. Anvari 2002 found significant correlations between music skills, phonological awareness and the reading development of four‐ and five‐year‐old children. The connections between reading achievement and two distinct styles of music education have been investigated: (a) Orff, Kodály or Dalcroze instruction, which stresses multisensory, developmental group activities emphasizing singing or playing percussion (Hurwitz 1975), or (b) participation in choral, band, or orchestral ensembles requiring music reading skills and extensive practice to achieve competence (Douglas 1994). Nikjeh 2009 observed that trained musicians, when compared with non‐musicians, showed more efficient neural detection of pure tones and harmonic tones, and demonstrated superior memory for acoustic features of pure tones, harmonic tones, and speech. Growing evidence from a range of research disciplines suggests that musical experience can have a positive effect on language and literacy abilities (Douglas 1994; Sutton 1995; Kilgour 2000). In addition, there is increasing recognition of the numerous shared features of music and language, from developmental characteristics to perceptual processes and common neural substrates (Sloboda 1985; Patel 1998).

According to Schlaug 2005, playing an instrument requires a range of skills, including reading a symbolic system (musical notation) and translating it into sequential, bimanual motor activity dependent on multisensory feedback. Learning to read, according to this point of view, could be compared to playing or learning a musical instrument, because it requires coordination of the eye muscles to follow a single line of printed musical notation (for example, when playing the violin, viola, bass, or flute) or a conjunction of lines (for example, when playing the piano, organ, or harpsichord). It is necessary to develop a spatial orientation to play each musical note from the sheet music and correlate it with a specific position on the instrument, which is comparable in this case with words and letters that together form a larger structure.

Temporal cues are important in speech perception (Martin 1986), and temporal fluency is a key factor in reading proficiency (Hanes 1986). More specifically, according to Huss 2010, musical metrical sensitivity has been found to be a predictor of phonological awareness in reading development. In this context, Besson 2007 observed that, as result of musical practice, a set of common processes may be responsible for pitch processing in music and in speech. Accordingly, it would be reasonable to assume that if pitch ‐ the perceptual attribute that corresponds to sound frequency ‐ is an important acoustic parameter for both music and speech perception, then increased efficiency in pitch processing due to musical expertise should also improve pitch perception in speech. Overy (Overy 2000; Overy 2003), on the other hand, focused her hypothesis on the temporal processing component. According to this author, dyslexic people have particular difficulties with skills involving accurate or rapid timing, including musical timing skills. It has been hypothesized that musical training may be able to remediate such timing difficulties, and have a positive effect on fundamental perceptual skills that are important in the development of language and literacy skills. Gabrieli 2009 has pointed out that dyslexic children who retain their benefits after systematic instruction in phonological awareness and decoding strategies improve from year to year, but they do not catch up with the typical reader; according to the same author, improvements are more likely to occur in children who are beginning to read (ages six to eight) than in older children.

Why it is important to do this review

Several studies have reported positive associations between music education and enhanced abilities in non‐musical (for example, linguistic, mathematical, and spatial) domains in children (Schellenberg 2004; Piro 2009), although the knowledge about this topic is non‐specific and contradictory. It is therefore important to critically analyze and synthesize the evidence for the effectiveness of musical training as a means of improving reading in those children with dyslexia.

Objectives

To study the effectiveness of music education on the spectrum of reading skills in children and adolescents with dyslexia.

Methods

Criteria for considering studies for this review

Types of studies

We will include randomized controlled trials (including quasi‐randomized or cluster‐randomized) in this review.

Types of participants

Dyslexic children and adolescents attending public and private schools.

Types of interventions

Any musical training approach, as defined below, compared with waiting list or no treatment control group.

Eligible forms of musical training are: individual or group music lessons or musical training for children with a music advisor or teacher at music school (extracurricular) or at the school where the children are receiving their formal instruction, as part of the general curriculum or as additional tuition. Children may be exposed to music in a natural musical setting such as song and tonal or atonal and rhythmic patterns or a specific musical methodological approach (for example, Dalcroze Method, Kodaly Method, Suzuki Method or Orff Approach) and may be encouraged to practice music in small or large groups.

Types of outcome measures

Primary outcomes

Because this review is concerned with the impact of musical training on reading skills, we will only include studies that report one of the following outcomes. These outcomes of oral reading skills, reading comprehension, reading fluency, phonological awareness and spelling can be measured through validated instruments, such as the following.

  • Oral reading skills (for example, Gray Oral Reading Test, Safety Word Inventory and Literacy Screener, Get Ready to Read! (GRTR), Woodcock Reading Mastery Tests (Woodcock 1987)).

  • Phonologic awareness (for example, Comprehensive Test of Phonological Processing).

  • Reading fluency and comprehension (for example, Test of Oral Reading Fluency (TORF), Retell Fluency (RTF), maze (MZ), written retell (WRT), and sentence verification technique (such as SVT)).

  • Nonword reading and spelling (for example, The Graded Nonword Reading and Spelling Test (Snowling 1996)).

  • Expressive and receptive vocabulary (for example, Clinical Evaluation of Language Fudamentals (Semel 1986)).

  • Phonological processing (for example, Comprehensive Test of Phonological Processing ‐ CTOPP (Wagner 1999) or Rapid Automatized Naming Test (Wolf 2005)).

Secondary outcomes

  • Self‐esteem.

  • Improved academic performance.

We will examine the outcome data in the short term (up to six months), medium term (between six and 12 months) and long term (more than 12 months).

Search methods for identification of studies

Electronic searches

We will search the following electronic databases for all available years: The Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, CINAHL, LILACS, PsycINFO and ERIC. We will seek information about ongoing clinical trials by searching Current Controlled Trials, ClinicalTrials.gov, the WHO trials register (ICTRP ) and the National Research Register Archive. We will not apply any date or language limits.

We will use a sensitive search strategy to search MEDLINE and modify it as necessary to search the other databases mentioned above (see Appendix 1).

Searching other resources

1. Reference lists: we will check references in the identified studies for additional citations.

2. Personal contact: we will email study authors and experts to request any unpublished data.

Data collection and analysis

Selection of studies

Two authors (HCM and RBA) will independently review all titles and abstracts identified through the search strategy. We will obtain full reports of any title or abstract that seems likely to meet our inclusion criteria. HCM and RBA will independently review all reports and determine their eligibility. If any doubt remains after reading the full report(s), we will contact the study authors for further information. In the event of disagreement we will consult with the CDPLPG Co‐ordinating Editor.

Data extraction and management

Two reviewers (HCM and RBA) will independently extract data from studies that meet the inclusion criteria, using a standard extraction form.

Details to be extracted will include:

  1. Study: information regarding the author(s); year of publication; source; country; and language.

  2. Characteristics of setting and participants: eligibility criteria for participants; explanation of recruitment procedures; setting (country, location, clinical or non‐clinical); demographic features of the sample.

  3. Sampling: sample sizes for treatment and control; whether power analysis was used to determine sample size; allocation to treatment and control; explanation of method used to generate the allocation.

  4. Research Design: type of design including major features such as random selection, random assignment, and non‐equivalent control group.

  5. Intervention Data: nature of interventions; for example, intervention focused on a specific musical instrument or methodological approach (for example, Dalcroze, Suzuki, Kodály or Orff Method).

  6. Outcome Data: primary and secondary outcomes; measures used; information on reliability and validity of measures.

  7. Results: attrition at post intervention and follow‐up; number excluded from the analysis; length of follow‐up; statistical methods; type of data effect size is based on; data needed for effect size calculations.

We will not be blind to the names of the study authors, institutions or journal of publication. We will resolve all disagreements by consensus amongst ourselves and planned referral to the editorial base of the Cochrane Developmental, Psychosocial and Learning Problems Group for arbitration when necessary.

Assessment of risk of bias in included studies

At least two review authors will independently assess risk of bias within each included study in accordance with guidance in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008). Review authors will independently assess the risk of bias within each included study in the following domains with ratings of low risk of bias, high risk of bias and unclear risk of bias.

Sequence generation

We will describe in detail  the method used to generate the allocation sequence so as to assess whether it should have produced comparable groups, and make a judgement on whether the allocation concealment sequence was adequately generated.

Allocation concealment

We will describe the method used to conceal allocation sequence in sufficient detail to assess whether intervention schedules could have been foreseen in advance of, or during, recruitment, and make a judgement on whether allocation was adequately concealed.

Blinding

We will describe any measures used to blind participants, personnel and outcome assessors so as to assess the knowledge of any group as to which intervention a given participant might have received, and make a judgement on whether knowledge of the allocated intervention was adequately prevented during the study. Blinding in the case of the participants (for example, children), in practise, is not possible, because of nature of the intervention. The children will know whether they are receiving music education or not (compared to non‐musical activity such as a control group would take part in).

Incomplete outcome data

If studies do not report intention‐to‐treat analyses, we will try to obtain missing data by contacting the study authors. We will extract and report data on attrition and exclusions as well as the numbers involved (compared with total randomized), reasons for attrition or exclusion where reported or obtained from investigators, and any re‐inclusions in analyses performed by review authors; we will then make a judgment on whether incomplete data were dealt with adequately by the study authors. (See also Dealing with missing data).

Selective outcome reporting

We will try to assess the possibility of selective outcome reporting by investigators and judge whether reports of the study free of any suggestion of selective outcome reporting.

Validity and reliability of outcome measures used

We will assess whether the outcome measures were standardised and validated for the population.

Other sources of bias

We will assess whether the study apparently free of other problems that could put it at a high risk of bias.

Measures of treatment effect

If participants, interventions and outcome measures are sufficiently similar, we will carry out meta‐analyses. We will enter data into an Excel spreadsheet and two authors will independently enter data into Review Manager 5 (RevMan 2008), each author entering data from another author's extraction sheets, using the double data entry facility in Review Manager 5. Where the same rating scale has been used for all studies, we will pool data using mean differences; where different rating scales have been used to measure the same outcome, we will use standardized mean differences.

Unit of analysis issues

We will follow the guidance on statistical methods for cluster‐randomized trials described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2008, Section 16.3). We will seek direct estimates of the effect (for example, an odds ratio with its confidence interval (CI)) from an analysis that properly accounts for the cluster design; alternatively, we will extract or calculate effect estimates and their standard errors as for a parallel group trial, and adjust the standard errors to account for the clustering (Donner 1980). This requires information on an intraclass correlation coefficient (ICC), which describes the relative variability in outcome within and between clusters (Donner 1980). We will extract this information from the articles if available, and otherwise we will contact the authors or use external estimates obtained from similar studies. We will find closest‐matching scenarios (with regard to both outcome measures and  types of clusters) from existing databases of ICCs (Ukoumunne 1999), and if we are unable to identify any, we will perform sensitivity analyses using a high ICC of 0.1, a moderate ICC of 0.01 and a small ICC of 0.001. We recognize that these values are relatively arbitrary, but prefer to use them to adjust the effect estimates and their standard  errors due to the implausibility that the ICC is actually 0. Subsequently, we will combine the estimates and their corrected standard errors from the cluster randomized trials with those from parallel designs using the generic inverse variance method in Review Manager 5.

Dealing with missing data

We will contact the original investigators to request any missing data and information, in order to decide whether or not missing data can be assumed to be ‘missing at random'. For dichotomous data, we will report missing data and dropouts for each included study and will report the number of participants who are included in the final analysis as a proportion of all participants in each study. We will provide reasons for missing data in the narrative summary and will assess the extent to which the results of the review could be altered by the missing data by, for example, a sensitivity analysis based on consideration of 'best‐case' and 'worst‐case' scenarios (Gamble 2005). Here, the 'best‐case' scenario is that where all participants with missing outcomes in the experimental condition had good outcomes, and all those with missing outcomes in the control condition had poor outcomes, and the 'worst‐case' scenario is the converse (Higgins 2008, section 16.2.2).

For missing continuous data, we will provide a qualitative summary. The standard deviations of the outcome measures should be reported for each group in each trial. If these are not given, we will impute standard deviations using relevant data (for example, standard deviations or correlation coefficients) from other, similar studies (Follmann 1992), but only if we decide, after seeking statistical advice, that to do so is practical and appropriate.

Assessment of heterogeneity

We will assess the extent of heterogeneity using the three methods suggested by the Cochrane Handbook for Systematic Reviews of Interventions (Deeks 2008): visual inspection of forest plots, the Chi2 statistic (increasing the level of significance to 0.10 to avoid underestimating heterogeneity) and using the I2 statistic designed to assess the impact of heterogeneity on the meta‐analysis. It describes the “percentage of the variability in effect estimates that is due to heterogeneity rather than sampling error (chance)" (Higgins 2002; Higgins 2003). However, it is advised that the thresholds of the I2 statistic might be misleading and the following guide is offered:

  • 0% to 40%: may not be important;

  • 30% to 60%: may represent moderate heterogeneity;

  • 50% to 90%: may represent substantial heterogeneity;

  • 75% to 100%: considerable heterogeneity.

We will bear in mind that the importance of the observed value of I2 depends on (i) magnitude and direction of effects and (ii) strength of evidence for heterogeneity (for example, P value from the Chi2 test, or a CI for I2 (Higgins 2008)).

Assessment of reporting biases

We will draw funnel plots (effect size versus standard error) to assess publication bias if sufficient studies are found. Asymmetry of the plots may indicate publication bias, although they may also represent a true relationship between trial size and effect size. If we identify such a relationship, we will further examine the clinical diversity of the studies as a possible explanation (Egger 1997).

Data synthesis

We may conduct meta‐analyses to combine comparable outcome measures across studies. In any meta‐analysis, the weight given to each study will be the inverse of the variance so that the more precise estimates (from larger studies with more events) are given more weight. We will use random‐effects models because studies may include somewhat different treatments or populations. We will group outcome measures by length of follow‐up.

Subgroup analysis and investigation of heterogeneity

We are planning to carry out subgroup analyses to explore the possible differential effect of the intervention depending on:

1) duration of musical training: short term (up to six months); medium term (between six and 12 months) and long term (more than 12 months);

2) type of musical training (for example, Dalcrose Method, Suzuki Method or even a specific instrumental practice such as keyboard or string);

3) age range of the participants (for example, children (six to 12 years) versus adolescents (13 to 18 years)).

Sensitivity analysis

We will conduct the following sensitivity analyses.

1) The removal of studies with inconsistencies in the definition, measurement, or reporting of results (for example, if the number of participants varies in the report or if measures were not taken at consistent time points for all participants).

2) Changing the way that values are imputed for missing data (for example, last value carried forward versus mean scores for missing values).

3) Reanalysing the data using different statistical approaches (for example, using a fixed‐effect model instead of a random‐effects model) (Higgins 2008).