Advances in music technology: The effect of multimedia

on musical learning and musicological investigation*

Scott D. Lipscomb

University of California, Los Angeles

This paper provides a brief discussion of Bloom’s (1956) "taxonomy of educational objectives" in the cognitive domain, and suggests the relevance of his framework in relation to several types of computer-assisted instruction (CAI). A review of experimental research confirms the efficacy of computer-instruction in improving upon (or at least equaling) traditional teaching methods. These results are discussed in terms of student achievement, attitude toward instruction, and time required of the instructor.

When utilizing the multimedia personal computer (MPC) as a tutor, involving the student in the learning process interactively is of utmost importance. Computers are not constrained to the linearity evidenced in textbooks or earlier instructional software. The use of hypertext and hyperlinking allows the student to determine his/her own presentation sequence, within limits established by the instructor and/or programmer. Uses of the computer in experimental investigations will also be discussed.

Instructional software developed by the author and representing several types of CAI and two examples of research software will be presented to illustrate possible uses of this new technology. Concluding remarks emphasize the importance of student interaction and hyperlink capability as critical elements of multimedia instructional software and suggest utilization of the MPC to its greatest potential in musicological investigation.

Recent technological advances have provided academic institutions with a powerful tool for the enhancement of both musical learning and understanding of the knowledge-acquisition process itself. The multimedia personal computer (MPC) provides a means of individual instruction in which a student may proceed at his/her own pace without feeling the pressure often experienced in classroom situations. Furthermore, the MPC offers potential improvement of the very methods by which researchers study these learning processes, providing a single instrument capable of presenting auditory, visual, or audio-visual stimuli as well as storing subject responses for later statistical analysis.

The Learning Process

In an effort to explicate the specific intentions of our educational system, Benjamin S. Bloom (1956) and his colleagues published a "taxonomy of educational objectives" in the cognitive domain. According to Bloom, the taxonomy "is designed to be a classification of the student behaviors which represent the intended outcomes of the educational process" (p. 12). His taxonomy consists of six major classes and their associated subclasses (see Table 1). These classes are arranged in hierarchical order from simple to complex. The most basic level, knowledge, is exemplified by the simple recall of information (e.g. specific facts, universals, methods, etc.). According to Bloom, this process "involves little more than bringing to mind the appropriate material" (p. 201). At this level, the taxonomy refers only to the knowledge itself, not the utilization or application of this knowledge.

Table 1

The other levels in the taxonomy are distinguished from the first level as "intellectual abilities and skills." In other words, levels 2.00 to 6.20 require "organized modes of operation and generalized techniques for dealing with materials and problems" (Bloom, 1956, p. 204). Such abilities and skills involve the mental processes of organization and reorganization in order to accomplish an intended goal. Comprehension is the lowest level of intellectual ability and requires only that the student knows what is being communicated. With this fundamental understanding, the student is able to translate or rearrange the information without distorting its original meaning. In order to attain the next level, the student must be able to apply the appropriate abstraction (i.e. theory, principle, idea, or method) without being prompted. In order to correctly solve a problem of this nature, the student must be able to identify familiar elements in an initially unfamiliar context, using these elements as a guide in restructuring the problem within a familiar context (see Bloom, 1956, p. 121). Analysis implies the ability of a student to breakdown information into its constituent elements and to explicate the relationships between the various ideas expressed. This process is divided into three parts: analysis of elements, analysis of relationships, and analysis of organizational principles. In contrast to analysis, synthesis involves the process of putting together parts in order to form a whole, i.e. creating a novel pattern or structure. At this level, the student moves into the role of a "producer" (Jones, 1990, p. 268). The highest level within the cognitive domain, evaluation, requires that the student make both quantitative and qualitative judgments concerning the extent to which criteria are satisfied by certain materials or methods. Such evaluations are made on the basis of internal evidence (i.e. logical accuracy and consistency) or in terms of external criteria (i.e. a comparative process).

Computer-assisted Instruction

How can students be tested on the ability to function at the highest levels of Bloom’s classificatory system using the MPC? In answering this question, delineation of several of the most common types of computer-assisted instruction (CAI) and their relation to the taxonomy will prove useful.

Wright and Forcier (1985, p. 96) define CAI as "a learning environment characterized by instructional interaction between computer and student.... [The teacher] sets up the learning environment, ensures that each student has the necessary skills to engage in a particular activity, and adjusts the learning activities according to the students’ needs." Of the many classifications of CAI available, five specific types seem to be most often utilized for educational purposes (see Table 2). Drill and Practice instructional programs simply assist the student in remembering and utilizing information that the teacher has already presented, reinforcing previous learning through repetition. Tutorials are designed to introduce unfamiliar subject matter. The format of a computer tutorial often emulates a dialogue between the computer and the student, i.e. information is presented, questions are asked of the student, and, on the basis of the response given, a decision is made to move on to new material or review what has already been presented. These first two CAI types are most successful at improving the Knowledge and Comprehension levels of Bloom’s taxonomy (1.00-2.30, as listed in Table 1). Instructional Games present course content in a competitive and entertaining manner, in an effort to maintain a high level of student interest. Though most frequently used to reinforce factual knowledge at the lower levels of the taxonomy, it is quite possible to create instructional games that demand skills from all levels.

Table 2

In contrast to the first three types of CAI, simulations require the student to apply acquired knowledge to a novel situation. As a result, the student must analyze a presented scenario, make decisions based on the information given, and determine a course of action. The simulated environment must change based on the course of action taken, presenting a significant challenge to the programmer. Successful performance relies on skills up to Bloom’s level of Analysis (4.30). Similarly, problem-solving software requires the student to use high level cognitive abilities in the process of considering the problem at hand, analyzing the problem situation and its various solutions, predicting respective outcomes, determining which specific plan to attempt, and enacting the appropriate action(s). Well-designed software that fits this classification, may require abilities from all levels of the taxonomy. However, perhaps the best way to have a student use abilities of synthesis is to have him/her create a novel hypertext system (see below for explanation; Jones, 1990, p. 270). In this case, the student would be forced to identify relationships and evaluate all aspects of the chosen set of course materials. Evaluative ability can be tested (and improved) throughout programs representing any of these five types of CAI by prompting the student at significant times during the session and providing appropriate feedback or explanation.

In addition to the delineated types of CAI, it is also possible to provide a discovery-environment (Kendall, 1987, p. 192; Shute, 1993) within which the student is given a high level of freedom in determining the specific information presented during each session, as well as the order of presentation.

A major advantage of CAI is that, by necessity, it requires the student to be an active participant in the learning process. It is not only possible, but necessary for the student to interact with the computer or else nothing will happen (Chabay & Sherwood, 1992, p. 154). In order to progress from one screen of information to the next, in most cases, the student must respond using the computer’s peripheral hardware (e.g. keyboard, mouse, joystick, or specially-designed devices). As a result, it is impossible for the student to assume the role of a mere observer (Lockard et al., 1987, p. 144). With appropriate monitoring and assessment, it is possible to ensure that the required material is covered and the desired level of competency is attained.

Within the context of instructional software, one of the most useful programming techniques to evolve in recent years is hypertext. Its basic concepts were defined in a pioneering article by Vannevar Bush published in 1945 (cited in Bui, 1989, p. 14). Simply stated, hypertext involves the linking of one piece of information to another. Practically speaking, if a student comes across an unfamiliar word or concept within a presentation--specifically, a word or concept to which the instructor has chosen to assign a hyperlink--the student may click on the word with a mouse pointer in order to immediately display the instructor’s definition or additional explanation on the screen. Of course, this is only one extremely simplistic application of a hyperlink. Hypermedia (the multimedia generalization of hypertext) utilizes the additional capabilities of the MPC, so that sounds, animations, visual displays, MIDI files, etc. may likewise be linked. For example, when clicking on a picture of Beethoven, the opening motif of his Fifth Symphony might sound or, when the term "simple harmonic motion" is clicked, an animation could begin that illustrates the molecular motion produced by a sine wave. The basic building blocks of hypertext are nodes (individual pieces of information), collections (sets of nodes) and links (references between nodes; Mühlhäuser, 1990).

With the arrival of hypertext and hypermedia, the discovery-environment mentioned previously became a viable alternative to the textbook-like sequence of most early instructional software. The result is increased freedom for the user, but the consequence is a potential loss of the sequential thread (i.e. intended structural ordering) provided by the instructor. A system of this type supplies the user with the ability of non-linear navigation through instructional materials, provides the capabilities of all media devices present in the computer, and allows determination of the specific material presented in any session to be based on the needs of the individual student (Stubenrauch, 1989, pp. 541-42). Contrasted with traditional methods, Spiro & Jehng (1990, p. 165) suggest that this may provide a more appropriate method of CAI at advanced levels of knowledge acquisition, as content becomes more complex and relationships across cases to which similar knowledge must be applied become more irregular.

The added flexibility and freedom provided by a hypermedia system result in a new set of problems (Hammond, 1989). Because of the size of the knowledge base, the user may get "lost" among the variety of choices available. Secondly, it may be difficult for some students to see the relationship between the various parts of the presentation, resulting in a failure to gain an overview of the material. The student may also have a hard time finding specific information even if it is known to be present. In addition, users not accustomed to such an exploratory environment may wander aimlessly from one topic to another in a highly inefficient manner. Lastly, the process of learning to use the interface and the various capabilities of the system may get in the way of learning the materials.

With all of these potential difficulties, is it worth the effort involved to solve the problems surrounding the creation of such a learning environment? Since computers have been utilized for educational purposes for approximately 30 years, results of recent research relating to CAI should provide a reasonably reliable metric for the efficacy of this instructional method. Relevant criteria in this determination are the observable effects on student achievement and student attitude. In addition, due to decreasing availability of funds for education, it is crucial to consider the amount of time required of the instructor in the learning process.

Kulik and his colleagues carried out two comprehensive studies, integrating findings from a large number of independent investigations evaluating the use of CAI. Using Glass’ (1976) meta-analysis technique, the authors examined 51 studies utilizing students in grades 6-12 (Kulik, Bangert, and Williams, 1983) and 59 studies using college students (Kulik, Kulik, & Cohen, 1980). Their results illustrated that generally CAI made "small but significant contributions" to student achievement, also producing positive effects on student attitude toward instruction and toward the subject matter. Studies conducted with elementary school students over a 4-year period by the Educational Testing Service showed that students who used a computer for only 10 minutes per day to improve mathematical skills scored significantly higher than those students who had no such access to a computer (cited in Bracey, 1982). Results of another study (Willett & Netusil, 1989) observed that students drilled by a computer using G. David Peters’ (1983) Clef Notes software scored significantly higher than students in a classroom drill when learning bass clef notes . A questionnaire designed by the investigators provided evidence of positive student attitudes toward using the computer in this process. Dalby (1992) was able to improve his students’ ability to make harmonic intonation judgments by using specifically-designed computer software. A questionnaire filled out by the students assigned to CAI indicated that most of them felt positively toward this method of instruction. When comparing a computer-assisted program to a traditional method designed to teach pitch and rhythm error detection skills at the university level (Ramsey, 1978), results of a study by Deal (1985) indicated that both groups experienced significant gains in the ability to detect errors, but found no significant difference between groups.

Economic realities of the present demand that educators consider efficient and economic alternatives to traditional teaching practices. Even if CAI performs only at levels comparable with normal classroom methods (though the trends in the research cited previously were always toward improved performance), it certainly remains worthy of consideration. If microcomputers can be used effectively for the development of basic skills, for example, then the instructor would be free to devote more time to those aspects of the educational process that are not capable of being taught through a process of simple repetition and assessment. Learning to play a musical instrument--more specifically, learning the expressive capabilities inherent in the process--is an excellent example of an ability that is passed, one-on-one, from teacher to pupil. It is questionable whether these abilities (beyond learning to read notation and memorizing fingerings, that is) will ever be taught by a computer.

Other research has suggested that the time required to learn material actually decreases when students use CAI. Kulik et al. (1983) noted a substantial reduction in the amount of time needed for instruction among their subjects (grades 6 through 12 and university students) when computers were utilized as a means of instruction. Applying CAI in the training of Army electricians, Blaschke & Sweeney (1977) reported a 10 percent reduction in training time when compared with traditional methods. Lunetta (1972) claimed that high school physics students were able to reduce the 745 minutes of classroom time conventionally devoted to learning the required subject matter to 90 minutes of individual instruction and study using a computer!!

The preceding studies support the theory that, when participating in CAI programs, student achievement equals or surpasses the performance of individuals learning by traditional means, students exhibit a positive attitude about using computers in the educational process, learning time can sometimes be significantly decreased, and time required of the instructor may also be reduced. Therefore, the question is no longer whether or not to use computers as an educational tool, but rather how they can be used most effectively as part of the learning process. The multimedia personal computer (MPC) provides important capabilities to assist in meeting this challenge.

The Multimedia Personal Computer

As described above, the MPC merges a battery of media capabilities with the power already inherent in a typical personal computer (PC). Specific media types include (but are not limited to) CD audio, digital audio tape, digital video, scanner images, MIDI sequencer, WAVE audio, videotape recorder, and laser disc (Microsoft Corporation, 1991, pp. 7-4 & 7-5).

Using the application of Bloom’s taxonomy as a point of reference, several computer programs (created for IBM-compatibles) will now be presented, representing several types of CAI. It is hoped that the following discussion will assist in pointing out strengths and weaknesses of the general methods and specific techniques utilized in each program. Another goal of this presentation is to plant seeds of potential in the fertile minds of musician-programmers who may be sitting in the audience today. All CAI programs were created by the author (including one collaborative effort) in order to fulfill a specific educational purpose. In most cases, there are commercial software packages available that could have substituted as examples.

The first program (EXAMPREP) is a DOS-based example of simple Drill & Practice CAI. It was created as a means of preparing for tests requiring a large amount of fact recall (e.g. dates, composer names, etc.). The computer serves simply as a kind of electronic flashcard, randomly presenting material that the student has entered (Fig. 1). In this particular version (designed for the purpose of testing knowledge of musicological investigations in the literature), the student provides a set of records, each of which includes the following fields: Author Name(s), Date of Publication, Topic Categorization, and Summary of Results. Once this information is entered, it is stored to disk and available for presentation at a later time. Naturally, the student is able to edit the records, providing additional information or identifying newly-discovered relationships between records when appropriate. Compared with its flashcard analog, this computer program provides the student with the flexibility of being quizzed on the contents of any field without entering the information multiple times, e.g. provided the name and summary, the student must provide the date of publication; provided the author name(s) and date, the student must provide the summary of results; etc. The limited capabilities of this program, however, only allow it to test knowledge of facts and do not require multimedia capability.

Fig. 1

Captured Screen from EXAMPREP

The second program (NAMENOTE) is a Windows-based application designed to teach young children to read notes on the musical staff in both the treble and bass clefs. The graphical interface is kept as simple as possible, in an effort to avoid confusion. Appropriate feedback is provided after each response (Fig. 2). If the incorrect answer is given, one of several encouraging remarks is made and the correct response is provided so the child may learn from the experience. Such feedback keeps the student aware of his/her level of performance, maintaining a positive attitude about the learning process and method of instruction. In addition to the immediate feedback given after each note or interval, following every tenth example, the percentage correct for the most recent ten problems is provided. When exiting the program, the student receives a composite percentage for the entire session.

Fig. 2

Positive feedback provided for a correct response in NAMENOTE

NAMENOTE has recently been modified to include practice on interval identification for more advanced music students (Fig. 3). With this addition, it could almost be considered a Tutorial. A help screen is provided that lists all of the acceptable responses, their meaning, and the number of semitones in each interval (Fig. 4). The student then translates this information by responding, correctly or incorrectly, to multiple occurrences of the various intervals presented in a randomly-generated sequence. Using the sonic capabilities of the MPC, the computer plays each note as it appears on the screen.

Fig. 3

Interval identification option in NAMENOTE

Help screen provided for the interval identification portion of NAMENOTE

Up to this point, the programs discussed have used only the most fundamental capabilities of the MPC and, in reality, could have just as effectively been utilized on a PC with a built-in speaker and no capacity for multimedia presentation. However, in the next program, the potential of the MPC becomes fully realized.

Created by the participants in a seminar within the Department of Ethnomusicology and Systematic Musicology at UCLA (Kendall, Fall, 1991), Computer Instruction Module (CIM) was designed for the purpose of providing a flexible tool for the presentation and review of course materials, using illustrative multimedia examples (Fig. 5). A system of hyperlinking was also developed so that the student could request additional information during the course of a presentation sequence.

Fig. 5

Example frame from CIM

The program distinguishes two modes of operation: Design Mode and Present Mode. The former (password-protected) allows the instructor to enter information in any combination of textual, graphical, and aural formats. The intent of CIM was to provide an end product that retains as much flexibility as possible, so the instructor has total control over the module’s content, appearance, and presentation order. In Design Mode, the "designer" carries out the task of creating "frames" for inclusion in Present Mode. Within each frame, text format and color of each object, as well as size and location of the various elements within the presentation window are completely under the designer’s control. Graphics and sound may be added to some or all frames, causing the appropriate control buttons (e.g. play, pause, stop, rewind, etc.) to automatically become visible and enabled during the presentation at a location specified by the designer.

Incorporation of hyperlink capabilities allows the designer to provide the student (in Present Mode) with additional information upon demand. This assistance is available in three varieties. Firstly, in the text editor, the designer determines words that will have definitions attached via hyperlink. These words are marked by pointed brackets (e.g. "<MPC>"; see Fig. 6). Secondly, certain words or concepts may require more extensive explanation. For these "topical terms," a hyperlink is established which causes the presentation order to pause momentarily while a different frame is displayed providing additional information about the topic of interest. After reading the text, seeing the picture, and/or hearing the aural example, the student is returned to the previous frame for continuation along the chosen presentation sequence. Topical terms are marked in the text editor with square brackets (e.g. "[taxonomy of educational objectives]"). Once defined terms and topical terms have been marked, the designer can select an option from the menu which allows updating of all text input to the frames in the current module. This process identifies the keywords (both defined terms and topical terms) and marks them automatically anywhere they occur on all frames within the module. The third hyperlink option allows the designer to select a portion of the graphical image and provide explanatory text. If the selected portion of the image is clicked during presentation, a box appears supplying the student with the designer’s explanation. For example, if a bitmap image of the physiological structure of the ear is presented, when the student clicks on the portion illustrating the eardrum, the explanatory text might include identification of the anatomical structure, its function, and any other pertinent information.

Fig. 6

Text editor in CIM, illustrating the process of marking terms for hyperlinking

Perhaps the most unique aspect of CIM is its concept of multiple "threads" for presentation of the material within a single module. The term "thread" was suggested by Roger Kendall and effectively expresses the intended flexibility of the sequential arrangement for any given instruction module. The designer is presented with a list of frame labels and simply arranges some or all of them into a sequence that is saved to disk (Fig. 7). In Present Mode, the student may select one of any number of available presentation sequences created by the designer. This provides a means of explicitly illustrating different relationships within a common set of instructional materials simply by changing the presentation order.

Fig. 7

Presentation sequence editor in CIM, where "threads" are created and saved to disk

CIM can be used as a Tutorial, presenting either familiar or novel material to the student along with accompanying visual and aural examples. Beyond this basic function, however, the hyperlink capability provides a means of calling into play higher levels of learning. In the most extreme example, suggested earlier, the student could be required to create a hyperlink system of his/her own using Design Mode. Such a project would require knowledge of the important issues in a field, the ability to express these ideas clearly through verbal, visual, and aural means, analysis and evaluation of the representative theories and their interrelationships, and the capacity to synthesize all of this information into a unique set of materials. Of course, in CIM, it is also possible for the instructor to create an environment in which the student is given complete freedom to explore the module contents. In this case, the thread might consist of only a single frame. On this frame, the text would simply supply a list of topical terms identifying each frame in the module. These topical terms could be linked so that, when clicked during presentation, the appropriate frame is displayed. CIM provides a high level of flexibility without requiring any knowledge of computer programming on the part of the module designer.

In addition to music educational applications, utilization of the MPC for musicological investigation is a topic worthy of consideration. The potential impact of this technology upon our understanding of the processes involved in musical learning, music perception, and performance is so great that it would be extremely negligent not to mention at least the most obvious effects on our ability to carry out contemporary research. First of all, the accuracy and reliability of calculations performed by the PC decreases the likelihood of human error in these processes. Second, and most importantly, the MPC provides a single instrument upon which all operations involved in the data-gathering and analysis phases of an experiment may be executed, i.e. presenting aural and/or visual stimuli, collecting responses, and performing statistical analyses on the data. Used effectively, the computer manages the mundane, repetitive, and calculation-intensive tasks, allowing the investigator to focus energy on portions of the investigation demanding human intelligence and integration, e.g. data interpretation and model-building. Two software examples (one of which was created by the author) will illustrate these abilities.

MELCOUNT is a simple DOS-based program written to analyze melodies in terms of pitch class, melodic interval, or note duration. The investigator has the option of inputting a new melody, loading a saved melody, editing or transposing existing melodies, hearing the melody played, or seeing a visual representation of the chosen analysis. For example, two melodies could be compared in terms of the number of pitch classes that occur (i.e. how many Cs, Ds, F#s etc.). Notice how clearly the stylistic differences in pitch class usage between Bach and Schoenberg are illustrated in Figs. 8a & 8b. Once a melody has been input correctly, all available types of analysis can be carried out at any time without fear of misidentification or miscounting in the process.

Fig. 8a

A MELCOUNT pitch class analysis comparison: the first 100 notes of the melody from the first movement of Bach's Brandenburg Concerto, no. 2 and the first 80 notes of the opening theme from Schoenberg's Piano Concerto, op. 42.

A Windows-based system is currently being developed at UCLA by Roger Kendall (Music Experiment Development System, a.k.a. MEDS, version 3.1e) that provides an environment within which an experimental procedure can be designed and implemented, saving subject responses to disk for later statistical analysis. The main window of this program is divided into two parts: the Experiment Design window and the Objects window (Fig. 9). Elements become part of the experiment design by clicking the mouse on one of the buttons in the Objects window. When clicked, the object’s icon is placed in the first available empty box in the Experiment Design window and it becomes part of the experiment. MEDS uses a set of symbols and an organizational structure familiar to the musician. The experiment flows from left to right and top to bottom like a musical score with repeated tasks (e.g. playing a stimulus list, inputting subject responses, etc.) nested between repeat signs. Potential objects include messages displayed to the subject, information input (e.g. subject name, level of musicianship, etc.), creation of lists for presentation (e.g. pictures, sound, or animation), stimulus presentation (in A, AB, or ABX format), and several types of response mechanisms (e.g. semantic differentials, similarity judgment, and categorization). Once an experiment has been designed, the software checks the logic of the newly-created design, informing the user of potential problems if any exist. After the investigator completes the design procedure, the computer will run the entire experiment, saving subject responses in a file on the hard drive.

Using MEDS, an investigator also has the capability of performing signal synthesis, as well as calculating fast Fourier transforms (FFTs) and root mean square (RMS) values for a chosen signal. MEDS utilizes all capabilities of the MPC in a graphical, user-friendly environment. Though the potential of these new technologies are only just beginning to be realized, computer programs like this one will lead us toward the future, expanding the role of the MPC in institutions of musical learning.

Fig. 9

Conclusion

The advent of the MPC has provided a tool to assist in both musical learning and empirical research. However, tools provide only a means for making progress. True progress cannot be made unless the tools are placed in the hands of a skilled craftsperson. Energetic individuals are needed to design innovative software using the most recent technologies and to utilize existing software in new and creative ways. For educational applications, this means testing more than mere rote memorization, involving the student interactively in the learning process, and providing hyperlink capabilities so the student has access to additional information on demand. Within the context of empirical investigation, progress will be exemplified by putting the computer to its greatest possible use in all phases of an experimental procedure.

Our predecessors have provided a powerful tool: the multimedia personal computer. It is our task to develop ways of using it most effectively.

References