About this report

An overview of the project, its methodology and those who contributed to it

ViolinMaking Music is the most comprehensive collaborative survey of its kind yet undertaken. ABRSM had previously identified a knowledge gap in the learning and teaching of instruments and singing and in 1993 conducted Making Music, a large-scale learner and teacher survey repeated in 1996 and 1999. Its key finding – that music learning was on the decline – contributed to securing wider provision and funding for music education at state level.

This 2014 report represents a major collaboration with partners from across the music education sector. Its recommendations have been informed not only by the survey’s statistical data but also by a roundtable discussion and a series of one-to-one interviews with many organisations and individuals. ABRSM would particularly like to thank the following for their commitment, advice and input:

Deborah Annetts (Incorporated Society of Musicians); Nick Beach (Trinity College London); Matthew Chinn (Royal Conservatoire of Scotland); Dónal Doherty (Heads of Music Services Northern Ireland); Hannah Fouracre (Arts Council England); Miranda Francis (Royal College of Music); Matt Griffiths (Youth Music); Richard Hallam MBE (Music Education Council); Professor Susan Hallam (Institute of Education); Robin Hammerton HMI (Ofsted); Fran Hanley (Musician’s Union); Darren Henley OBE (Managing Director, Classic FM); Karen Humphreys (Royal Northern College of Music); Howard Ionascu (Royal Academy of Music); Rachel Kilby (CAGAC, Wales); Douglas Lonie (Youth Music); Graeme Smith and Nigel M Taylor (Music Mark); Diane Widdison (Musicians’ Union); Crispin Woodhead (musicteachers.co.uk) and Paul Wood (Heads of Instrumental Teaching Scotland).

Research objectives and methodology

London-based research agency, Critical Research, worked with ABRSM to find out more about attitudes and behaviour towards learning and teaching in the UK.

The research was conducted in two stages:

Adults (aged 18+) and children (aged 5–17) were interviewed via an online panel partner. The final data was re-weighted back to a nationally representative profile on age and gender, in line with statistics obtained from the Office of National Statistics1.

In total 1,726 children and 1,255 adults were interviewed.

In order to obtain views from teachers throughout the UK, a number of partner organisations that represent the diversity of music education were also invited to participate in the research. Critical Research designed an online survey that was emailed to the full databases of ABRSM and Trinity College London teachers. The following partners also distributed the survey link to their teacher networks:

CAGAC; Heads of Instrumental Teaching Scotland; Heads of Music Services Northern Ireland; Incorporated Society of Musicians; Music Education Council; Music Mark; musicteachers.co.uk; Musicians’ Union; National Foundation for Youth Music.

In total 4,491 teachers were interviewed.

Where possible, results have been tracked back to data available from child learners from research conducted in 1993, 1996 and 1999 for previous ABRSM Making Music reports.

The previous data for adult and child learners was collected via a nationwide omnibus survey. In 2014, an online methodology was used to enable more in-depth questioning, a more robust sample frame and greater control over data quality. Data from 1999 and before was not re-weighted to a nationally representative sample and details of the exact sampling are no longer available. Therefore we are unable to apply statistical significance testing to tracking data and whilst the original question wording and bases have remained consistent, any tracking comparisons can only be seen as indicative as opposed to conclusive.

Number of interviews (5-14 year olds)


Number of interviews









It should be noted that the 1993-1999 data segmented children as aged 5 to 14 and adults as 15 years or older. In 2014, children were reclassified as 5 to 17 and adults as 18 years or older, as this was deemed more appropriate. However, any data used for tracking purposes was re-weighted to reflect the previous age profiles.

Figure 1 - respondent profile

Figure 3 - respondent profile

Figure 4 - respondent profile

Figure 5 - respondent profile

Figure 6 - respondent profile

Figure 7 - respondent profile

Figure 8 - respondent profile

This website uses cookies to improve your experience. By using our website, you are agreeing to our cookie policy and consent to our use of cookies. Find out more.