Abstract
Online diagnostic measurement of musical abilities in a school environment is a new way to enhance pedagogic effectiveness. In our study we developed a technology-based musical perception test which examines basic musical hearing abilities among 1st to 11th graders in Hungarian schools (N = 2961). The instrument was composed of tasks measuring the discrimination of rhythm, tempo, melody, pitch, harmony, timbre and dynamics, and we also tested the connection between auditory and visual modality. The results indicated an increasing developmental tendency during the period under examination. Students who attend classes with a special music curriculum achieved better results, except in the first years of school. A weak relationship was found between socio-economic background variables and test achievement; however, parents’ educational attainment and students’ school marks have a stronger relationship. Similarly to previous research findings, no significant gender differences were found.
Keywords
The impacts of musical training have been examined in numerous contexts, and results show that it not only develops musical abilities, but also has a significant effect on the development of a number of cognitive and affective characteristics as well. It has a positive impact on emotional development (Hunter & Schellenberg, 2010; Koelsch, 2010); on social skills (Creech & Hallam, 2011; Garland & Kuhn, 1995; Gerry, Unrau, & Trainor, 2012; Hagen & Bryant, 2003), among other areas; and on numerous cognitive domains, including reading (Corrigal & Trainor, 2011; Herrera, Lorenzo, Defior, Fernandez-Smith, & Costa-Giomi, 2010), mathematical abilities (Hallam, 2010; Schmithorst & Holland, 2004), and foreign language skills (Milovanov, Huotilainen, Valimaki, Esquef, & Tervaniemi, 2008; Posedel, Emery, Souza, & Fountain, 2012; Tallal & Gaab, 2006). Previous studies have revealed the link between musical abilities and higher-order cognitive areas, including general intelligence; for example, a positive association was found between musical aptitude and IQ, especially in childhood (Schellenberg & Weiss, 2013). A significant correlation was found between musical practice and intelligence, but there was no difference between musically trained and untrained twins (Mosing, Madison, Pedersen, & Ullén, 2015). Despite the mounting evidence of the positive impact of musical training, it does not receive enough attention in schooling.
Although research on the assessment of musical abilities started decades ago, regular monitoring of their development in educational practice requires further research; specifically, it requires valid instruments which are broadly available and easy to administer. Objective and standardized tests can support high-quality teaching and educational accountability as well. As in the case of other domains (Kromann, Jensen, & Ringsted, 2008; McDaniel, Roediger, & McDermott, 2007), frequent feedback on musical development aids teachers in planning further steps in educational processes, provides evidence-based information for education policy, and may enhance students’ aptitude.
To use objective tests in music education is of fundamental importance because the current forms of evaluation are often subjective and informal. Furthermore, regular national and international school monitoring processes mostly focus on disciplines which are assumed to have direct economic potential (e.g. OECD-Programme for International Student Assessment, Trends in International Mathematics and Science Study, Progress in International Reading Literacy Study and the National Assessment of Basic Competencies in Hungary), while less attention is devoted to other subjects. For example, in the National Assessment of Educational Progress in the United States (NAEP), music is included, but it has only been assessed four times since 1971 (Scott, 2009). As schools have to cope with the growing amount of learning material, areas not tested receive less space in the curricula. Enrichment of the evaluation culture with easy-to-use and user-friendly tools could be a solution for this problem.
Rapid development of computer-based assessment facilitates more effective data collection from the viewpoint of the quality and quantity of data alike. Technology not only offers more efficient and less expensive instruments, but also forms measurable constructs that are otherwise difficult to measure with traditional paper-based instruments (Csapó, Ainley, Bennett, Latour, & Law, 2012). Musical abilities fall within this group of constructs and online instrument assessment may easily be extended to the area of musical abilities as well.
In the present study, we utilize the benefits of technology provided by the eDia platform (Electronic Diagnostic Assessment), which was devised for online diagnostic assessments. We have developed items for delivery via this system to measure musical hearing abilities. The online platform makes it possible to use innovative items (including multimedia features), store data and meta data, and administer complex background questionnaires. The innovative features of technology-based assessment not only promote higher test objectivity and better validity, but also arouse students’ interest and motivation (McCarrick & Xiaoming, 2007).
For this study, an online test and a questionnaire were developed to examine musical hearing abilities, their structure, developmental trends, and their relationships with other cognitive and social variables among 7- to 17-year-old students. In this paper, we outline the theoretical background of testing musical ability, introduce the measurement instrument, and present the results of large-scale, cross-sectional assessments.
Assessment of musical development
In defining what to measure and in outlining an assessment framework in school contexts, we may use three different resources in general. We may use the curricula for the given discipline or school subject (the disciplinary dimension), we may explore how and where the knowledge mastered at school is expected to be utilized outside of school (application and social and cultural dimensions), and we may examine how certain abilities and skills develop as personal psychological attributes (the psychological dimension, Csapó, 2010). In the area of musical tests, Hargreaves isolated three main forms: tests of achievement, attitude, and ability. The last type is designed to assess musical aptitude regardless of previous musical learning or experience (Hargreaves, 1986/2001). In our present study, we have focused on measuring the psychological dimension: specifically, musical hearing abilities regardless of previous experience or learning.
Models and tests of musicality reflect the complexity of musical activity. Research on intelligence has had a remarkable impact on measurement instrument frameworks in the 20th century; therefore, one- and multiple-factor models can be distinguished among models of musical ability. One-factor models approach musicality as a general, homogeneous ability. Tests based on this paradigm examine holistic, high-level knowledge and generally require preliminary knowledge or musical training. Consequently, the primary target group for these tests consists of musicians or those who have learnt or are learning how to play an instrument (Bentley, 1966; Révész, 1946; Wing, 1939). The multiple-factor approach is based on the assumption that musicality is defined by diverse musical abilities. The first prominent representative of this view was Carl Seashore (1919), whose endeavors in measuring musical abilities are still decisive in music psychology. Gordon’s (1986) musical aptitude tests can also be classified as examples of this genre. The content of multiple-factor tests is basically focused on cognitive capacity to process musical information as discrimination and judgment of musical parameters, and is thus applicable to average test-takers.
Recent tests that measure individual differences in musical perception skills are mostly designed for adult participants and focus on a comparison of musicians’ and non-musicians’ aptitude (e.g. Law & Zentner, 2012; Ullén, Mosing, Holm, Erikkson, & Medison, 2014). Among musical ability tests developed during the past few years, several have to be highlighted.
The Musical Ear Test (MET) was designed by Wallentin, Nielsen, Friis-Olivarius, Vuust, and Vuust (2010). Following the traditions of musical ability testing, it contains same/different questions in the areas of melody and rhythm discrimination with 104 items in total. This paper-and-pencil test was created for a normal adult population. Its main advantage is that it provides objective and reliable information with a short testing time; however, due to the paper-and-pencil method participants have no direct feedback on their performance (Wallentin et al., 2010).
In contrast to the MET, the Profile of Musical Perception Skills (PROMS) examines the discrimination of melody, pitch, timbre, tuning, rhythm, rhythm-to-melody, accent, tempo, and loudness with fewer items and within a more holistic framework. Each subtest contains 18 items. This test was also developed to measure a general adult population. The test battery is based on stylistically neutral, relatively basic and abstract musical stimuli to avoid the advantage of certain participants being more familiar with a particular genre. Unlike other well-known tests of musicality, PROMS provides multiple-answer options involving levels of confidence. The measurement tool was one of the first musical tests realized in an online environment (Law & Zentner, 2012).
The Swedish Musical Discrimination Test (SMDT) also examines musical abilities via the Internet with simple discrimination tasks related to melody, rhythm, and pitch. The objectives of the design were to develop an online test which has a short test-taking time and is optimal for average, musically untrained populations in industrialized countries (Ullén et al., 2014).
Although tests with a psychological approach provide essential information for researchers, the number of musical ability tests applicable in educational contexts is – with a few significant exceptions (e.g. Bentley, 1966; Gordon, 1979, 1982, 1989) – minimal. Musical testing with a diagnostic purpose may open new perspectives for educational practice as well as cognitive psychology and educational research.
In this article, we introduce a new musical ability test developed to measure individual differences in musical hearing among elementary and secondary school students and to provide objective feedback about their developmental level. In order to collect detailed information, we used tasks to measure a wide range of abilities. Conceptions of musical ability (Erős, 1993; Hargreaves, 2012; Karma, 2007; Moles, 1966) and cognitive models of musical processing (Koelsch, 2012; Peretz & Coltheart, 2003) provided valuable sources to develop the test framework. The modular model of music processing (Peretz & Coltheart, 2003) was also our starting point for the confirmatory factory analysis, as we built the factors of the hypothetic model on the main organization areas, like temporal and pitch organization, including contour and interval analysis as well. Tasks cover the abilities of temporal organization (rhythm and tempo discrimination) and melodic organization (pitch, harmony, and melody discrimination). To cover a holistic system of musical parameters, we supplemented the framework with tasks involving timbre, dynamics, and visual connection. The latter measures the degree of association between auditory and visual modality using pictures of schematic melody contours.
The musical ability test is followed by a questionnaire about students’ school achievement and attitudes, socio-economic status, and musical experience. We used part of the Home Musical Environment Scale (Brand, 1985) to examine musical background variables.
Computer-based assessments and the context of the present study
The online testing of musical abilities has several innovative features that are impossible to realize with traditional paper-and-pencil tests. In this case, participants only need Internet access, a computer with a browser and earphones for the administration of the tests. The process of evaluating and coding students is automated, thus providing much higher precision and objectivity and offers immediate feedback in contrast to paper-and-pencil tests (Csapó, Molnár, & Tóth, 2009). The reduced costs of data collection and administration are also important advantages over traditional tests (Price et al., 2009). Technology-based assessment also provides an opportunity to answer the questions at the participants’ own pace, so stimuli can be heard when they are prepared to pay attention to them. As volume can be changed according to individual needs, current health status or hearing acuity does not hinder the test-taker. Although it may cause validity problems when online testing is not organized and when participants complete the test in their own uncontrolled environment (Ullén et al., 2014), in our study this was prevented with the teachers supervising the classroom measurements.
Research questions
In the present study, we explore the possibilities of using an online diagnostic musical measurement tool in different educational contexts. For this purpose, we employed the test in primary and secondary schools and in schools with and without a special music curriculum. The sample allows us to analyze individual differences in musical abilities, differences between age groups and the impact of focused vocal or instrumental training. With the accompanying background questionnaire, we use the data to explore the relations between cognitive and social background variables. Our research questions are:
Is online testing a reliable method for measuring students’ musical abilities? What are the psychometric characteristics of the tests?
What are the developmental tendencies of some basic musical skills? What are the differences between the different age groups? Which abilities are more affected by age?
What is the structure of musical abilities; how are the subtests related?
Are there any differences between students in music and regular schools? Which musical abilities are more amenable to training?
How do school achievement and social backgrounds affect students’ achievement on the musical test?
Are there any differences between boys and girls?
Methods
Participants
Participants in the study were Hungarian primary and secondary school students from Grades 1 to 11. The sample included students from primary schools with a special music curriculum and from non-music primary schools (Table 1). The curricula in Hungarian schools are based on the principles of Zoltán Kodály and Jenő Ádám, but students in schools with a special music curriculum have four times more music classes a week, and they are taught by specialist music teachers. The music classes following the Kodály pedagogy are basically vocal-centered, and choral singing has a major role. Although the number of music schools grew during the second part of the 20th century – there were 50 primary music schools in operation in Hungary in 1959, with as many as 120 by 1969, and later this number reached more than 200 (Solymosi Tari, 2002; Szabó, 1989) – there are now only 20 primary music schools and a further 20 primary schools with special music programs. Secondary schools with music curricula refer here to music conservatories.
Samples of the study.
The study samples are not strictly representative, but participants cover the diversity of students in the Hungarian schools (Table 1).
The data collection was carried out in 2013, and students completed the tasks in the computer labs (classrooms equipped with computers with Internet connection) in their school. They used the computers and the browser installed originally on them to access the online test. From Grades 1 to 4, students had two classes (2 × 45 min) to complete the test and the questionnaire that followed; from Grade 5, they had one class (45 min). This proved to be sufficient time for students to go through all the items in every grade.
Instruments
The online data collection was carried out via the eDia (Electronic Diagnostic Assessment) platform through the Internet. We applied mostly tasks for testing cognitive operations related to musical phenomena. All the tasks contained closed questions, and the correct answer earned one point. To ensure objectivity in the musical excerpts that were played, we created the musical stimuli with Sibelius 7 software. The test consisted of 73 items, arranged in seven subtests, followed by an 80-item questionnaire. In order to prepare valid measurement tools for primary and secondary school students, we developed items which require no previous knowledge to complete. We also took care to avoid special musical terms during testing, or we gave a short audio explanation when a term was necessary. In contrast to most adult musical perception tests, we created a familiar musical environment through folk and classical excerpts, although we chose melodies which are not included in Hungarian curricula and schoolbooks. The instructions related to the tasks were voice-recorded, and students could listen to the instructions. Poor reading abilities, therefore, did not hinder their understanding of the tasks.
Rhythm
Rhythmic discrimination tasks are based on the ability to differentiate and memorize temporal patterns. We used samples played on bongos (a non-melodic percussion instrument from the Sibelius 7 sample directory) to measure if students could recognize small differences between two rhythmic patterns. Rhythmic patterns consisted of eight beats. We defined two difficulty levels of rhythmic context: first, when items had only quarter and eighth notes, and, second, when they consisted of sixteenths, triplets, a dotted quarter and a quarter rest (Figure 1). In these items, students had to decide whether the two rhythmic patterns they hear are the same or different. The possible small changes were calibrated by interchanging two beats with different durations or by transforming a complex rhythmic pattern (e.g. a dotted quarter and two sixteenths) to a simple pattern (e.g. a quarter and two eighths).
Rhythmic discrimination task items.
Tempo
To measure the ability to hear tempo, we developed tasks containing the same melody twice. The melodies could be in the same tempo, or there could be a 10, 20, or 30 BPM (beats per minute) difference between them. Students had to make a decision about which one is faster or if they are the same. In this task, we employed four monophonic and four polyphonic musical excerpts from the classical era (Figure 2).
Tempo discrimination task items.
Harmony
Harmony discrimination items examine how participants process the relations of pitches played simultaneously. As the basis for polyphonic hearing and interpreting, the perception of harmonies requires multimodal musical thinking. Piano timbre was used in the nine items. In this task, students heard two successive chords and had to decide whether they were the same or different. In every case, there was just one note difference between the two chords. As the interval between the original and the altered note was a minor or major second, the manipulation altered the trial from major to minor or from minor to diminished. For example a major root position was followed by a minor first inversion, or a minor second inversion was followed by a diminished second inversion.
Pitch
Pitch discrimination items contain two successive musical notes. Students had to decide which one was higher, first or second. Tasks employed the diatonic scale system; pitches could differ by between one and four semitones. We used a piano timbre from the Sibelius 7 sound library for a clear sound.
Melody
The melody discrimination task measures how students memorize, group, and compare musical excerpts. In these items, two melodies follow each other; these could be the same or there could be a small difference between them. To obtain comparable items, we defined nine musical parameters: (1) As for the style of the excerpts we chose, we employed folk songs whose sound set is familiar to the participants, even though they are not able to recognize the song. (2) We defined the length of the melodies in eighth notes. (3) To avoid validity problems with melodies with complex rhythmic patterns, we used excerpts which contain quarters and eighths. (4) The tempo of each melody was 110 BPM. (5) As for the melody contour, items do not have repeated bars. (6) The musical notes in the melody ranged from D4 to D5 to correspond to the vocal range of the sample. (7) Using a range of melodies, we can differentiate two main categories: perfect fifth-range and perfect octave-range items. (8) In those items where the first and second melodies are different, the change can happen at the first, the middle (4th or 5th), or the last note. (9) The difference between the two melodies was that the one note was higher or lower than in the previous one by a second or third. The changed note matches the tonality and melody contour of the original melody well (Figure 3).
Melody discrimination task items.
Visual connection
The connection between visual and auditory modality is related to the development of reading and writing music. To examine the main characteristics of the ability, participants have to choose the one picture of four that corresponds to the melody contour that they hear (Figure 4).
Visual connection task. Visual representation (above) and melody contour (the notes below only represent the melody played) of the visual connection task.
Dynamics
Dynamics can be defined as the volume change during a given time period. We created this task to exclude special musical terms; students simply had to decide the volume characteristics of monophonic and polyphonic musical excerpts (strengthened/faded/stayed the same). We designed musical items with piano, string quartet, and orchestral timbres for the task. The average volume difference between the first and second parts of the excerpt was 0dB, 10dB, or 15dB.
Timbre
Timbre is a qualitative component of music; its perception includes the sound spectrum and envelope. We used excerpts of polyphonic music involving different instruments. To avoid knowledge-related aspects, the same melody was heard twice and participants had to decide which melody involved more instruments. The number of instruments was manipulated with the Sibelius software (Figure 5).
Timbre discrimination task (Joseph Haydn: Divertimento HOB. XIV. 9. Menuet (1–8.) in Editio Musica, Budapest).
Questionnaire
The musical test was followed by an online questionnaire asking students about their school marks, school- and subject-related attitudes, leisure-time activities, and music-related habits as well as education and socio-economic background variables.
During the data analysis, the focus was on the main characteristics of the measurement tool, the characteristics of musical perception skills and the differences between subsamples. We applied confirmatory factor analysis to confirm the measurement model of the test.
Results
Psychometric characteristics of the tests
The coefficient of internal consistency, Cronbach’s α, for the whole test was .867. The subtest coefficient ranged from a low of .263 (timbre) to a high of .74 (pitch) (Table 2). From the viewpoint of different grades, we found a growing tendency from .67 (Grade 1) to .9 (Grade 10) (Table 3). Some of the low reliability coefficients may be attributed to the low number of items in the subtests. These data also indicate that the tests in their current form are more suitable for measuring the older cohorts.
Internal consistency values of the test.
Internal consistency values by grade.
The validity of the measurement instrument is supported by the fact that test achievement significantly related to the variables of students’ music education. Students who are learning or have learnt how to play a musical instrument (the variable was number of years, r = .34, p < .001) and those who attend schools with a special music curriculum (as a dichotomous variable, r = .44, p < .001) performed better on the test.
Developmental tendencies in musical abilities
Test achievement shows an increasing tendency during the school years under examination (Figure 6). However, the differences between age groups are not significant between any of the successive grades. There were no significant differences between Grades 1 and 2, Grades 3 and 4, Grades 5 and 6, Grades 6 and 7, and Grades 9 and 10. Typically, there are major differences between the different school program types (between the lower and higher grades in primary education and between the higher primary grades and secondary schools).
Developmental tendency of musical ability (overall test achievement).
The structure of musical abilities
To explore the structure and dimensions of the musical abilities under examination, we carried out confirmatory factor analysis with MPlus software (Figure 7). As the internal consistency of the test was relatively low in the lower grades in primary school, we analyzed the structure from 5th to 11th grades (n = 2213). A reduction of the sample resulted in reliable fit indexes (CFI = .984; RMSEA = .049).
Structural model of musical perception skills.
The analysis confirmed the hypothetic factor structure. The rhythm and tempo discrimination task related to the same factor. Rhythm and tempo are determined by the relative length of notes and their absolute length, respectively; thus, the basis for their perception is temporal processing. Melody discrimination and visual connection tasks measure contour processing, and they form another latent factor. As the accurate procession of small frequency changes is essential in pitch and harmony discrimination, they are also related to a common factor. The interrelation between the factors of contour and frequency differences is medium strength. A common feature of dynamics and timbre processing is that they play a determining role in musical expressivity. They are also important in non-musical contexts, and both of them are multiple, complex phenomena, with qualitative processing as a common latent factor.
Differences between students in music schools and regular schools
We examined the differences in the test performances between students in music and non-music schools by independent-samples t-tests (Table 4). No significant difference was found between the test results of 1st- and 3rd-grade students, but in later school years the difference between the two groups shows an increasing trend and the developmental tendency between grades is more dynamic among music school students.
Means and standard deviations of subsamples (%).
Performance on musical subtests was also examined in every grade (Figure 8). We found that only the visual connection task shows a significant difference (p < .05) between music and non-music students in first grade. While the majority of the abilities demonstrate significant differences in the higher grades, visual connection and pitch discrimination are the only tasks to show significant differences (p < .001) in every age group after the first year. In secondary school, all the tasks are significantly different.
Performance differences between pitch discrimination and visual connection tasks in music and non-music schools (R: regular schools; M: schools with special musical curricula).
While both tasks show a ceiling effect in the 9th–11th grades in schools with special musical curricula, the highest means are much lower in the highest grades in average schools (Pitch M = 82.4; Visual Connection M = 69.31).
Relationships between students’ achievement on the musical test, school achievement, and social background variables
As there are large differences between students if we consider the entire age range explored in this study, we have to choose a narrower sample for structural analyses. Taking into account the characteristics of the Hungarian school system, we have found Grades 5–8 to be most appropriate. These are the lower secondary grades, taught by teachers specialized in particular subjects instead of class teachers who cover all the subjects. From the perspective of music education, teachers trained for music education are responsible for students’ musical development.
A multiple regression was performed to identify the variables that predict performance on the musical ability tests. Socio-economic background variables were examined in our questionnaire with questions about the number of rooms, bathrooms, cars, computers, phones, televisions, and books at home. The common predictive value of these variables is also relatively low, R2 = .099, F(7, 2188) = 33.708, p < .0001.
We found a more significant predictive value in the context of musical aptitude and school marks, R2 = .144, F(12, 1600) = 22.396, p < .0001. Parents’ educational attainment also has a predictive value on musical aptitude, R2 = .114, F(2, 2449) = 154.229, p < .0001. While family’s financial background has a lower impact on the level of students’ musical perception skills (9.9%), parents’ educational attainment (11.4%) and students’ school marks (16.9%) have a greater predictive value.
We also explored gender differences for the whole test and the subtests, and at various ages, but we found no large and significant differences in any of these comparisons. The largest differences were found in the case of the harmony subtest, where boys slightly outperformed girls, but the significance level in this case was only .05. These results are in line with previous evaluations of musical perception tests among adults (Janurik, 2010; Ullén et al., 2014; Wallentin et al., 2010). Although a comparative study carried out among pre-school children (Pollatou, Karadimou, & Gerodimos, 2004) found that girls outperformed boys in rhythm performance, we found no statistical differences in the perceptual form of the rhythm task even in the lower grades.
Discussion
The study demonstrates that the online assessment system and the eDia platform may be used for testing musical abilities. The tests, which were delivered through the system, have been available through the schools’ local infrastructure and have proved to be easy to use. In contrast to most other musical test batteries, the multimodality of the current measurement tool provides reliable overall information about the main abilities of musical perception through measurements of timbre, tempo, dynamics, and pitch, among other factors. As an ability measurement tool, the musical perception test is suited to providing reliable information about students’ music aptitude in average school settings.
The test was originally designed for assessment in the primary grades, and the achievement and the current version is optimal for around Grade 3; after this age, the discriminating power of the test decreases. One improvement of the test would be to create and add more difficult items. Another direction for the development of the test will be creation of other item formats.
Some subtests have a relatively low consistency value, which can be attributed to the small number of items in them. Longer tests with more items should be prepared for a more precise assessment of the specific skills, and other item types with student-created responses may also be explored. The correlation between test achievement and the level of musical training may be considered to be an indicator of validity, and the effects of taking lessons in music or a musical instrument are demonstrated in nearly every grade. Results are comparable at the class, school, region, and country levels, thus suggesting the presence of conscious pedagogic development and realistic self-assessment.
Continuous development was found during the period under examination in the case of the whole test and subtests. Based on a confirmatory factor analysis, four latent factors of musical perception were identified. The demonstration of the consistency between the measurement and the hypothetical model also confirms the test validity. Results confirm that musical training has a slower but long-term effect, as we found no significant differences between music and non-music classes during the first school years, although later we can see growing differences. Our study also confirmed that music training is significantly associated with better musical perception skills in line with previous research in the field (e.g. Mosing, Madison, Pedersen, Kuja-Halkola, & Ullén, 2014). It is not at all common for low-achieving children to drop out of elementary schools with special musical curricula. Only secondary music schools have special selection criteria related to musical abilities. The differences between the subsamples grow in the elementary grades. These results also indicate that musical abilities are amenable and an increased number of hours devoted to music education at school would have a measurable impact.
The family’s financial conditions have a small impact on test achievement, while school marks and parents’ educational attainment are more decisive. As with previous research, gender differences were not found in our study.
Limitations of the present study
The purpose of devising the current form of the test was to construct an instrument that is short enough for administration during a single class and at the same time provides general information on the developmental level of the students’ musical perception. A detailed examination of the component skills related to the subtests requires further research.
More developmental work is needed to adapt the test to the ability level of the participants. More attention is required in particular in lower grades, where the reliability of the current measurement tool is weak.
Further innovations of the online tool will consist of the recording of rhythm and melody reproduction. While the former can be resolved with the use of a keyboard or mouse, technology-based measurement of melody reproduction raises new technical questions. Although excellent solutions are available for this method (e.g. EarMaster software), using it in a classroom environment will pose problems.
As the data collection was cross-sectional, we can only make estimations concerning the real developmental curves. Longitudinal assessment is required to confirm our results and outline the development more precisely.
Conclusions and further prospects for studying the development of musical abilities
This study has shown the potential of online assessment in a classroom situation. On the other hand, it has also indicated where the shortcomings of the present test are and how they could be overcome. Some subtests need more items to improve their internal consistency. More difficult items should also be created to increase the discriminating power of the subtests for older age groups.
As the testing time may be limited in school contexts, adaptive testing or multistage testing may be a solution. Larger item banks have to be prepared for this purpose. Our study has shown the main directions for building such item banks.
A further developed musical ability test may be an ideal tool for measuring more students in class in average school settings. The online technology-based musical ability test could be a useful supplement to high-quality music pedagogy, as diagnostic information can facilitate a focused student development process. Further research is required to evaluate the predictive and diagnostic value of the assessment of musical abilities and to establish a complex musical development program. The online measurement tool provides an opportunity for transfer research with other cognitive and affective domains and for international comparative studies and longitudinal projects. In an ongoing study, the tool is currently being adapted for the kindergarten age group.
Footnotes
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study has been supported by the MTA-SZTE Research Group on the Development of Competencies and the Center for Research on Learning and Instruction, University of Szeged.