The incorporation of computer-assisted instruction (CAI) into the aural skills curriculum has been a part of college music theory classes for some time (Eddis, 1981). Typically, the CAI component supplements the usual dictation and sight-singing drill of the classroom. As a patient provider of interval, rhythmic, harmonic, and melodic drill, the computer program can provide an "assistant" to play multiple and varied patterns for the student to practice.

Students vary in their appreciation of the strengths of the CAI components of aural-skills courses. Many students are not only positive about those components, but also enjoy the challenges they present. Other students complain about the lack of humanity in the computer; some are frustrated by the computer skills necessary to participate in CAI components; and some are unconvinced that CAI lab work has any major positive impact on the overall aural skills necessary for adequate performance in a music degree plan. Students who are successful in the CAI lab sometimes feel frustrated when the CAI units are competency-based, requiring them to complete a certain number of problems in a limited amount of time (Hofstetter, 1981; Killam, 1984).

College music theory programs include CAI components based on the assumption that the CAI drill will improve students' dictation and sightsinging skills. Anecdotal evidence and implicit theory has been used to justify the inclusion of these components. Many early studies during the 1970's and early 80's documented the positive effect of CAI on aural skills outcomes (Diehl, 1971; Herold, 1974; Gross, Griffin, 1982). More recent studies have measured the effect of computer assisted drill on beginning solfege (Ozeas, 1992), have compared computer-generated materials to pre-taped exercises (Garton, 1982), and have compared kinesthetic and aural reinforcement in CAI instruction (Polot, 1992). Substituting an entirely computer-based training program for a standard classroom curriculum supplemented with CAI was the subject of another recent study (Hess,1997). However, despite the advancement of computer technology and the increasingly common computer skills of current college students, the author could find no recent statistical study relating outcomes in aural skills exams to CAI supplemental instruction.

At the University of Georgia, the aural skills portion of music theory has been supplemented by a CAI program called GUIDO (Graded Units for Interactive Dictation Operations). Each music theory course has an assigned set of units to be completed successfully (Table 1). If the required units for a particular course are not completed, the student's final grade is lowered by one letter. No credit is given for completing a portion of the required units.

Seventeen freshman/sophomore music theory classes at the University of Georgia School of Music between 1994 and 1999 were surveyed to ascertain if the student's completion of required computer units had any effect on their final exam grades in aural skills and their final averages in written theory. Grades were compiled from 1247 students, all taught by the author (Table 2). These students comprised over half of all students enrolled in freshman/sophomore music theory courses at the University of Georgia during this period. Theory courses taught by other faculty during this time maintained similar standards and performances.

All data was analyzed using the SAS statistical package. An initial set of analyses was done in which correlation was computed between completion rates of GUIDO and final aural skills exam grades (sightsinging and dictation) and written music theory averages.

In the first set of analyses, some courses showed a strong correlation between satisfactory GUIDO completion and final exam grades, while in other courses, the correlation was considerably smaller Ö and, in some instances, negligible (Table 3).

In all aural skills courses, there are many factors that impact a studentís ultimate degree of accomplishment. Prior musical experience and training influences aural discrimination skills (Rogers, Watkins, 1986). Measures of musical aptitude and academic ability impact sightsinging and eartraining skills (Harrison, 1991; Butterworth, Fogarty, Rorke, 1993). Motivation also effects aural skills results. Thus, the actual benefit of experiences offered by GUIDO can be masked by differences among students on those factors. This would be the case even for something as seemingly straightforward as students' performances on the dictation or sightsinging final. Ideally, if one could adjust the estimate of the correlation between GUIDO completion and performance on aural skills final exams for such factors, one could obtain a far more sensitive assessment of that relationship.

Overall performance on the written theory component of the course was available in the compiled data. This should also be strongly affected by prior musical experience, student ability, and motivation, but it reflects a different set of skills than those that should be affected directly by student interaction with GUIDO. Using written theory performance as a proxy for those other student characteristics that cloud the relationship between GUIDO completion and performance on the dictation final gives a clearer relationship between GUIDO completion and higher grades on the aural skills final exams.

Accordingly, a set of multiple regression analyses was carried out relating performance on the dictation final with both theory performance and GUIDO completion.

In the second set of analyses (Table 4), using a student's music theory average as a regression coefficient made the impact of GUIDO completion on students' final grades much clearer. Nine of the seventeen estimated regression coefficients relating GUIDO completion to performance on the dictation final controlling for theory performance were statistically significant. Even more importantly, all but one of the estimated regression coefficients was positive. Thus the overall evidence for GUIDO completion leading to better performance on the dictation final seems quite clear. Furthermore, the regression coefficients themselves can be interpreted as the unique contribution of GUIDO completion to points earned on the dictation final. Across all seventeen classes this averages an increase of 9.4% in student's grades.

*Statistical Significance at Alpha 2.05

As might be expected, the influence of GUIDO completion on sightsinging exam finals is less compelling, since the skills necessary for GUIDO completion are more directly related to dictation skills. However, another set of multiple regression analyses was carried out relating performance on the sightsinging final with theory performance and GUIDO completion. The analysis (Table 5) shows a statistically significant increase of scores in seven of the seventeen classes, and an average increase of 5% in the students' grades.

This study showed that successful GUIDO completion translated into higher scores overall in sightsinging and dictation final exams when controlled by written theory averages. The dictation final was affected by 9%, which is close to a letter grade. The increase in sightsinging grades, at 5%, is still significant. The overall success in GUIDO completion definitely translates into higher scores on the aural skills final exam. This confirms intuitive beliefs of teachers that CAI-based requirements used to supplement the aural skills courses taught at the university level actually have a significant positive benefit.

Additional studies should be done to see if significant improvement is made using different software programs. While one might assume that the improvements in outcomes would be similar for all aural skills software packages, different approaches from different software companies might significantly affect the improvement in grades.

References: