The Effect of Computer-Based Lessons on StudentsÕ Perception in Rehearsal
Rick Dammers
University of Illinois/ Ladue School District
January 28, 2003
The nature of studentsÕ musical perception is an important issue for music educators. Bamberger (1991) states that ÒÉthe goal of music development is to have access to multiple dimensions of musical structure, to be able to co-ordinate these dimensions, and most important, to be able to choose selectively among them, to change focus at will.Ó (p. 4). For instrumental music educators, the question becomes how does this complex act of listening interact with the similarly complex act of playing an instrument? Does performing help focus studentsÕ perception on important features of a piece of music, or does it draw the studentsÕ cognitive resources away from attending to the expressive qualities of a piece? Assuming that performance might both enhance and detract from perception of salient aspects of a composition, it seems reasonable to provide multiple experiences, including performance as well as listening and analysis, in order to develop studentsÕ musical perception most effectively. This study will serve as a preliminary investigation into this issue by examining the effect of a computer-based lesson on studentsÕ abilities to identify melody and accompaniment roles in a rehearsal setting.
Research Questions
The research questions for this study are grouped into three areas of inquiry: the relationship between performance tasks and student perception, the effect of the computer based lesson, and studentsÕ attitudes toward the computer-based lesson.
Performance and
Perception
Effect of Lesson
Student Response
The participants for this study were ninth (n=38), tenth (n=29), and eleventh (n=29) grade woodwind and brass students from a suburban high school in the Midwest. A stratified, random sample of the students (n=40) was selected to receive four computer-based lessons on melody and accompaniment. The subjects were stratified into four groups based on the role and tessitura of their instruments. Group 1 consisted of the high (soprano) instruments that tend to have melody most frequently (flute, oboe, clarinet and trumpet); group 2 were the alto instruments (alto saxophone, horn), group 3 the tenor instruments (tenor saxophone, trombone, and baritone), and group 4 the bass instruments (baritone saxophone, bassoon, and tuba). Each of the four groupsÕ representation in the experimental group was based on the overall instrumentation of the band. The percussion section was rehearsing separately during the experimental period and did not participate in the study.
Pieces
Four grade 3 pieces that were being rehearsed by the band during the experimental period were selected for use in this study. The pieces used in the study were The Time Warp (arr. Wallace), Overture on a Shaker Tune (arr. Higgins), Birdland (Zaniwul, arr. Wallace), and Prelude Hymn on Ode to Joy (arr. Curnow). Seventeen sections of these pieces were selected from these four pieces. Each of these sections consisted of a complete statement of melodic material by an instrument or group of instruments. These sections were reviewed by three experienced middle school band directors for clarity of melody and accompaniment roles.
At the beginning of the two-week experimental period, all of the subjects (N=85) took a listening pre-test during rehearsal. At the seventeen pre-designated points, the conductor stopped the group and the investigator asked one of the following questions: ÒIn the section that we just played, which instruments played the melody?Ó or ÒIn the section that we just played, which instruments played the accompaniment?Ó The students responded by placing checks next to the appropriate instruments on a response sheet. For the purposes of this study, melody was defined as the lead line of greatest importance. Accompaniment was defined as all parts that are not the melody. The response sheets were scored based upon the answer key agreed upon by the judging panel. Any instrument that was incorrectly identified or that was omitted by the subject was marked as incorrect. Each of the twelve instrument types in the band was scored as one point, giving each item a total possible score of twelve.
Following the pre-test, the experimental group received four 20 minute computer lessons on alternate days. The lessons were administered in two groups of twenty using FreeStyle 2.01, a sequencing program running on iBook laptop computers with headphones. Since none of the tasks in the lessons required the use of synthesizer or keyboard as an input device, the researcher decided to use QuickTime Musical instruments (a software synthesizer) as the sound source. The lessons took place in a small room adjacent to the room where the band was rehearsing. While the close proximity made it easier for students to leave and return to rehearsal, sound leakage between the rooms was a negative factor.
Portions of two pieces (Overture on a Shaker Tune and Birdland) were pre-recorded into FreeStyle for use in listening and analysis activities. In the first two lessons the students were presented with sections of Birdland and Variations on a Shaker Tune and asked to turn off all of the accompaniment parts. At the beginning of the first lesson, the students were given a brief introduction to the playback and instrument control panels necessary to complete the task. The students were encouraged to use multiple listenings as they identified the melodic lines. In additional to the audio feedback, the instrument parts were represented in graphic notation. This notation allowed the students to view the full score and understand large portions of a piece without prior score reading experience. All of the students worked for the full twenty minutes at their own pace. None of the students completed all of the prepared sections. At the end of each session students were asked to save their completed files.
In the third and fourth lessons, the students were presented with the same sections of these pieces with the volume level for each instrument turned down to zero. The students were then asked to set the volume level for each instrument as if they were the band director. At the end of the experimental period, a listening post-test, identical to the pre-test, was administered to the entire ensemble (N=82 ). Due to student absence, some experimental mortality occurred. By the end of the study, the control group stood at (n=43) and the experimental group (n=35). This attrition was not unexpected and the size of both groups still allowed for accurate use of parametric statistical procedures.
Following the completion of the post-test, the experimental group was asked to respond to a questionnaire (n=32) regarding the computer lessons. Further attrition (three surveys) occurred due to end of the year schedule changes in the school. This questionnaire included eight close-ended questions with a six point likert scale. Three of the items were directed to student attitudes about the lessons, two were directed to changes in understanding of the pieces used in the lessons, two were directed at more general changes of listening in rehearsal while one question inquired about whether the lessons improved student performance or not. To avoid bias, each of the questions used in the survey was asked both in a positive and negative direction. An open-ended section for student comments was also included at the end of the survey.
The mean score for the listening pretest (N=85) was 8.62 out of 12, indicating a 72% accuracy rate. The scores followed a normal distribution with narrow range of scores (7.19-10, SD of .526). Considering that one instrument type accounts for almost 10% of each section, this distribution indicates a strong across the board ability to identify melody and accompaniment on this test.
Since some of the instrument groups were small, the non-parametric Kruskal-Wallis test was used to examine for differences between the each groupÕs pre-test scores. No significant differences were found (H(3)= 2.806, p=.423). The same procedure was used to test for differences between the three grade levels (ninth, tenth, and eleventh grades). Again, no significant differences were revealed (H(2)= .904, p=.636). A paired t-test was used to compare the pre and post test scores of the control group (n=43) to examine for the influence of familiarity of the piece. This also did not reveal any significant differences (t(42) =-.826, p=.414).
SPSS statistical software was used to compute an ANCOVA in order to compare the mean scores of the experimental and control groups. The pretest scores were used as the covariate in order to control for each student's initial ability to identify melody and harmony. No statistically significant variances between the mean scores were found between the mean scores for the two pieces used in lessons or the two pieces not used in the lessons.
Student Response
The experimental subjectsÕ responses to the questionnaire were tallied and combined into four indexes. Each index had a possible score of six, with six being a positive response. The Attitude Index mean score was 4.82, (SD=. 931). This indicates that students generally enjoyed the lessons and would welcome them as a regular part of their band rehearsals. The Understanding Index mean score was 4.52, (SD=1.0). The General Listening Index mean score was 4.58 (SD=1.04), while the Performance question received a mean score of 4.03 (SD=.999).
Discussion
The relatively
high score (72% accuracy) indicates that these subjects possess a
well-developed ability to identify melody and accompaniment that was not
impeded by these performance demands.
The narrow range of distribution, coupled with the lack of significant
differences for any of the variables included in this study, may indicate the
presence of a ceiling effect. It is possible that, in this population, the
ability to identify melody and accompaniment had been secured at an earlier age
and that by high school this perceptual skill may be in a secure, static state. It is also important to note that the
performance demands of these grade 3 pieces are lower than those found in the
grade 4 and 5 pieces typically performed by the subjects at other points in the
school year. Given these factors, no conclusions can be drawn about the
effectiveness of the computer-based lessons. The positive responses received on the questionnaire
indicate a positive student attitude toward these lessons. The positive
response on the Understanding and General Listening questions may indicate
positive effects that were not measured by the listening tests.
Implications for Practice
While the experimental results of this study are inconclusive, the implementation of these lessons provided to concrete implications for practice. The successful use of laptops in these lessons underscored the benefits of laptop computers. The use of laptops in this study solved space and access issues that would have otherwise been prohibitive. These lessons also illustrated how MIDI keyboards are not always essential for using music software. The design and delivery of these lessons also demonstrated that computer based lessons can extend the studentsÕ rehearsal experiences by allowing them to work with the whole score in way that encourages higher order musical thought and expressive decision making.