Multimedia personal computers (MPCs) have provided educational institutions with a tool of inestimable value for both teaching and research. The following pages will delineate ways in which past research has shown computer-assisted instruction (CAI) to be particularly effective, describe methods for using the computer within music research contexts, and introduce a Windows-based software package developed at UCLA for the purpose of carrying out musicological investigations.

During the past decade, computer use within our schools has become ever more prevalent. A Market Data Retrieval Survey compiled in October, 1993 revealed that, in the five years between 1988 and 1993, there was a significant increase in the number of computers available for students from kindergarten to the 12th grade. In 1988, the student-to-computer ratio was 30-to-1. By 1993, the ratio had fallen to 16-to-1, meaning that the number of computers had almost doubled in that five year period (Austin, 1994). Predictions suggest that the K-12 mulitimedia industry (currently $1.7 billion annually) may reach $3.5 billion by 1996 (Fliegel, 1993). It is imperative, therefore, that teachers at all levels become "computer literate" in order to assist students-of-the-future in their academic endeavors.

Apart from administrative usage, applications of the computer in music educational institutions falls into two broad categories. First, the MPC can serve an educational purpose within the music curriculum. Second, it can assist in the process of advancing our knowledge of musical learning through systematic research. In addition to the basic computer, monitor, and keyboard, the MPC also incorporates a sound card with both digital sampling and MIDI capabilities, a CD-ROM player, and external speakers connected via an amplifier for sound reproduction.

David Sebald (1994) suggested that multimedia may serve four functions as a part of music education. Using the mnemonic "TIME," he delineates these functions as follows: to Train, to Inform, to Motivate, and to Entertain. The value of the final item might be questioned by some. However, if students find lessons entertaining–and the multimedia capabilities of the computer certainly can entertain–then it is likely that they will spend more time with their studies.

A recent literature review (Lipscomb, 1994) suggests that CAI has been an effective addition to the learning process. Achievement of students using computers for instruction equals or surpasses learning by traditional methods. Students exhibit a positive attitude toward using the computer. The time necessary for students to learn material is sometimes significantly reduced by incorporating computer instruction. Finally, the amount of time required of the instructor for simple drill-and-practice instruction (i.e., knowledge-level learning, as defined by Bloom’s taxonomy) may be reduced, so that more energy can be devoted to higher level activities (e.g., analysis, synthesis, and evaluation).

Computers are often utilized in university music departments to assist with the following areas of study: Ear Training & Sight Singing, Theory & Analysis, History & Literature, Composition, Music Performance, Music Perception, and even Music Technology itself. Each of these uses will be discussed briefly in the following paragraphs, though this review is not intended to be exhaustive.

The ear training lab was one of the first places in the music department where computer-aided instruction found a home. Weaver (1994) explains the benefit of giving students the ability to "initiate, direct, and immediately monitor their own practice and training" (p. 169). Lorek and Pembrook (1989) introduced a program that performs a fundamental frequency analysis of notes sung by a student to assess sight-singing accuracy.

The MPC can to be a perfect companion to the instructor of Music Theory or Formal Analysis. Using its sonic capabilities to the fullest (e.g., CD-ROM, digital sound files, and MIDI), along with impressive graphics, Beethoven’s symphonies can be brought to life for the music major and for the nonmajor alike. Robert Winter’s Voyager series (including Beethoven’s Symphony No. 9, Mozart’s Dissonant String Quartet, Stravinsky’s Rite of Spring, and Strauss’ Tone Poems), David Sonnenschein’s (1994) "Anatomy of Music" series (the latest additions include a collection of Mozart’s late piano concerti) and Dilthey’s (1994) "Horizontal Graph Analysis" are excellent examples of this type of instructional material. The student is lead through the formal design of a piece with the software author’s commentary. It is also possible for the student to navigate at will through the piece simply by clicking the mouse on the appropriate text or icon. These software packages will be useful to the general Music Appreciation class as well as core Music Theory courses within the music curriculum.

When considering multimedia software, the line between what is useful to the Theory/Analysis instructor versus what is useful to the Music History/Literature instructor is decidedly blurred. For instance, both Robert Winter’s and David Sonnenschein’s software discussed above contain historical information to provide a background against which to consider the music. In fact, Winter provides a great deal of historical data and even incorporates a game to test the students’ comprehension of both the formal analysis of the piece and the historical information about the composer. In addition, there is software available that focuses on historical facts. Compton’s Jazz: A Multimedia History provides a detailed account of all of the major figures in the evolution of this American artform. Microsoft’s CD-ROM entitled Musical Instruments provides a marvelous collection of instruments from around the world along with information about their design, developmental history, and classification. In the coming years, there will be many more of these "electronic textbooks" as the MPC becomes commonplace within our educational institutions.

The electronic music studio and its computer-driven technology have provided a significant enhancement for instruction in compositional techniques. Most obvious, perhaps, is the ease of hearing a composition during the creative process. Decisions concerning voicings or instrumental combinations can be made by trying several alternatives as "performed" by the electronic instruments controlled by a computer sequencer. Falling prices for this advanced technology have made it possible for students to acquire their own reasonably high quality MIDI home studio capable of this same type of trial-and-error compositional process. An interesting analytical study of composition style (Cope, 1991) suggests that the computer may be used to analyze the style of various composers and create novel music "in the style" of any given composer–what Cope refers to as "recombinant music." Analyzing occurrences of pitch classes, note durations, and specific pitch and rhythmic patterns, the computer utilizes similar patterns in the process of "creating" a new composition. In an attempt to further understand the creative process, some investigators have used the computer to study the manner in which young children approach musical composition (Upitis, 1989; Webster, 1989).

Uses of computer technology in performance instruction is widely varied. Learning improvisational skills has been greatly enhanced by the introduction of auto-rhythm accompaniment and recorded MIDI tracks (Stampfli, 1994). A student can practice improvising over a series of chord changes for hours without tiring a rhythm section. Researchers at McGill University are currently developing a computer system that can analyze incoming MIDI data played by live instrumentalists and improvise a solo in real time (Pennycook, Stammen, & Reynolds, 1993). Noting the inadequacy of traditional Western notation in transcribing musical sound in differentiating one skilled performance from another (Kendall & Carterette, 1990; Palmer, 1989; Bengtsson & Gabrielsson, 1980), some investigators have attempted to use an analysis-by-synthesis approach to determine performance rules in order to create an expressively acceptable computer performance (Sundberg, 1988).

The computer can even be used as an instructional guide to Music Technology. For instance, Gottschalk and Feng (1994) introduced a computer-assisted composition program that actually taught students how to utilize the facilities in their music laboratory, as they created a composition. Psychology of Music and Musical Acoustics courses can benefit greatly from the MPC. My own teaching experience has shown that the process of building complex signal shapes, illustrating the associated molecular motion using animation, and hearing the resulting musical timbre serve to clarify the meaning of some of these complicated topics far better than mere words alone.

With all of these potential uses currently within the music curriculum, the MPC has already become an integral part of music education. In the years ahead, there will certainly be an increase in the number of institutions utilizing CAI, resulting in the much-needed development of methods for using the computer in training students to make use of high-level educational skills in addition to simple drill-and-practice instruction.

The MPC has already made a significant contribution to musicological research. Digital sampling has greatly increased the ecological validity of stimulus materials (i.e., sampled trumpet tones resemble natural trumpet tones to a greater degree than synthesized brass sounds). Secondly, use of the computer increases measurement precision. Computer presentation of experimental stimuli allows the presentation order to be randomized for each subject individually, eliminating the possibility of an order effect. Finally, the MPC provides a central workstation from which all aspects of research can be carried out at a single, convenient location. The Music Experiment Design System 4.0 (a.k.a., MEDS; created by Roger A. Kendall at UCLA) takes full advantage of the multimedia capabilities of the computer in the process of designing an experiment, selecting and/or creating stimuli, collecting subject responses, performing statistical analyses, and printing 2- and 3-dimensional graphs of sound files. The remainder of this monograph will focus on specific aspects of MEDS, a Windows-based program that has been integrated into the current research activities at UCLA’s Music Perception and Cognition Laboratory.

The ultimate goal of MEDS is to provide a flexible and user-friendly environment for research design, data collection, and data analysis. In this sense, it truly meets the qualifications of a "central workstation," as outlined above. When the program begins, the user is confronted with the MEDS Control Center, where menu items and Quick Access buttons allow the user to select which operation is to be performed. These operations include designing & running an experiment, editing experimental data, creating lists (e.g. sound files & animations, verbal response scales, or graphic images) for incorporation into an experiment, synthesizing stimulus materials, performing FFT and RMS analyses of sound files, and exporting data to a statistical software package for quantitative analysis. It is impossible to discuss all of the capabilities of this program within the scope of the present paper. However, some of the most important functions will be delineated below.

MEDS’ Design Form is divided into three windows. The area filled with icons at the top of the form allows a user to create an experimental procedure by simply clicking on the appropriate icon so that it is copied into the first open square in the Experiment window. In the Experiment window, double-clicking on an icon opens a form in which specific parameters of that item may be set (e.g., lists of sound files, random/nonrandom presentation, etc.). A single mouse click on an icon in the Experiment window causes the current parameter values for that item to be displayed in the Information window on the right-hand side of the form. The structure of the experiment incorporates repeat signs to indicate when a series of operations is to be performed more than once. This particular experimental design represents a basic semantic differential task where listeners are presented with several musical examples and are then asked to respond on a number of verbal scales. The icons in the Experiment Window represent–from left top to bottom right–User Input (subject name), Message Box (instructions), Open Repeat (for Play List object), Play List (sound files to be presented), Play A (presentation of a selected item from Play List–either sequentially or randomly selected), nested Open Repeat (for Scales object), Scales object (list of verbal scales for subject responses–also either sequentially or randomly presented), nested Close Repeat (cycles through all items of the preceding verbal response scales before playing the next item from the play list), Close Repeat (signaling a return to the Play List item for selecting the next sound file), and Message Box (thanking the subject for participating). After an experiment has been designed, MEDS will automatically check the design for logical consistency, before allowing any subjects to be run.

Subjects taking part in the experiment described above will first be presented with an input box requesting their name. Then a set of instructions (provided by the researcher) appears, explaining the task. Next the subject will hear a musical example, followed by a series of verbal scales with bipolar adjectives anchoring either end of a scroll bar. The subject responds by moving the button on the scroll bar to a location that best represents an appropriate response. Each item of the Play List will be presented with every verbal scale. After responding to all of the sound files, a message box will appear thanking the subject for taking part in the experiment. At a fundamental level, that is all that is required to design and run experiments in MEDS. Due to space constraints in the present volume, it is not feasible to elaborate on the more complex possibilities of experimental design in MEDS. However, such information is readily available in MEDS’ on-line Help file. Alternative experimental tasks include similarity judgment, categorization, response time measurement, matching musical sound to visual images, tapping keys on the computer keyboard to register a subjects "perceived pulse" when listening to musical sound or computer animations, and other experimental tasks, limited only by the investigator’s imagination.

Of course, Control Center and Design Form are not the only modules available. For instance, in order to use the multimedia capabilities of the computer within an experimental context, lists of sound files and graphic images must be compiled. MEDS’ Play List form is used for this purpose. A Play List can consist of any combination of sound files (WAVE, MIDI, or MicroSound SF files), laser disk excerpts, CD audio, FLI Animation files, Video for Windows AVI files, and raw MIDI data compiled in MEDS. Once a list has been created and saved, it is easy to incorporate it into an experiment by double-clicking on a Play List icon in the Experiment window and selecting the appropriate file name. Building word lists for verbal responses and lists of pictures for incorporation into an experiment is also easily accomplished in a similar manner.

Once an experiment has been completed, MEDS provides a method for accessing the experimental data and importing that information into one of two popular statistical software packages: SYSTAT for Windows or SPSS for Windows. MEDS’ Data Editor provides two windows for examining data. On the left side of the form is a text box listing the complete contents of the data file. On the right side is a spreadsheet, allowing the user to examine portions of the data file (e.g., eliminating warm-up sessions, subject name, etc.) and manipulate the number of rows and columns. Once the data in the spreadsheet is arranged appropriately, it can be transferred for statistical analysis by selecting either SPSS or SYSTAT from the Transfer menu.

MEDS also provides tools for examining physical characteristics of the sound files used in an experimental investigation. A Fast Fourier Transform (FFT) can be calculated for any WAVE file, providing a graphic representation of how the spectral characteristics of the sound file change over time. MEDS offers both 2- & 3-dimensional graphs of the resulting FFT as well as the capability of observing the signal graph (i.e., waveform). Root-mean square (RMS) amplitude analysis is also available so that the investigator can observe how the amplitude of a signal changes over time. The RMS module provides the option of calculating attack time (operationally defined) and decay time (user-defined) for a sound file. Finally, MEDS’ Synthesis Module allows the user to create WAVE files using additive synthesis. By determining the number of sine components and their relative amplitudes, complex signals of user-defined durations can be created for incorporation into a Play List.

MEDS has been an extremely productive tool for our research in UCLA’s Music Perception and Cognition Laboratory. We have found it to be reliable and easy to use in the process of designing experiments, collecting data, and analyzing results. This type of central workstation approach provides the necessary flexibility within a user-friendly environment that can facilitate incorporation of convergent methods (Kendall & Carterette, 1992) into the process of answering a research question with a few simple clicks of the mouse.