Reprinted with permission of copyright holder

COMPUTER-MANAGED INSTRUCTION

Alessi, S.M. and Trollip, S.R. (1991). Computer-based Instruction: Methods and Development. Second Edition, Chapter 1. Prentice Hall, Englewood Cliffs, NJ. Chapter 13.

INTRODUCTION

This chapter describes computer-managed instruction (CMI) systems. These systems provide administrative support to instructors for managing instructional materials and activities. We begin with a review of the history and theory of computer-managed instruction. We then give our own description of the 'state of the art" of this technology. Lastly, we show and describe a prototype CMI system for microcomputers.

Computer-managed instruction (CMI) was among the first successful uses of the computer in instruction. In the early days of mainframe computing, before facilities were available for students to work individually at computer terminals, several university and military projects developed programs to aid in the management of classroom instruction, and to keep track of student progress on tests and instructional modules. These programs were developed to deal with the data processing needs of individualized instruction and mastery learning programs. Such programs (Bloom, 1976; Keller, 1968) arose as alternatives to traditional group-based instruction.

Individualized instruction and mastery learning require more frequent testing of students, keeping records on the educational progress and activities of individual students, and reporting of information. Instructors involved in such programs find themselves overburdened with scoring tests, filling out performance and activity sheets for each student, and analyzing student information to produce class reports. Little instructor time is left for working with individual students on things for which they need help-the whole idea of individualized instruction programs.

In answer to these problems, computer programs were developed that included a database of course structure and objectives, and test items and instructional activities linked to each objective. After a student takes a test covering relevant objectives, the computer scores the test and provides a prescription of activities that teach the objectives not yet mastered. Additionally, the computer stores the information about the student's past progress and current activities.

The instructor periodically receives reports showing individual student and total class progress.

EARLY CMI SYSTEMS

Among the first CMI systems was the Program for Learning in Accordance with Needs (PLAN), begun in about 1966 (Flanagan, 1969,1970; Flanagan, Shanner, Brudner, & Marker, 1975). This CMI system was oriented towards elementary and secondary school curriculum management. Developed by the American Institutes for Research and later marketed commercially by Westinghouse Learning Corporation, PLAN provided service to many school districts across the country. In its early implementations, information on student achievement tests was entered into a computer terminal via punch cards and processed at a central mainframe computer. The computer scored the achievement tests and on a daily basis provided printouts showing individual student achievement by curricular objectives, instructional materials recommended for each student, and various summary reports across time and across students.

Soon, several school districts, universities, and businesses began developing CMI systems. Early systems typically required students to take progress tests on optical scanning sheets (the same kind of computer-scoreable sheets that are used for most national exams). These were either mailed to a center having the necessary computer hardware or were entered into a local scoring machine and transmitted to a remote computer. The remote computer scored the tests, stored information from the new student data, and produced reports that were either mailed back or transmitted electronically to the school. Reports of different types were available: such as for individual students at any point in time (showing progress on objectives and current activities), summary information for a student over a period of time, and summary information for groups of students.

As telecommunication networks became more widely available, CMI activities became easier and more interactive. The U.S. Navy CMI system (Hansen, Ross, Bowman, &Thurmond, 1975; Mayo, 1974), begun in 1973, allowed students to take a test using an optical sense answer form, feed the form themselves into an optical form reader, and get a printout almost immediately showing their results, progress, and next assignment. The analysis and data storage was still on a remote computer connected by telephone to the scoring machine and printer.

The Computer-Assisted Instruction Study Management System (CAISMS) begun in 1974 (Anderson et al., 1974) permitted students to take a test interactively on a computer terminal and receive immediate results and study assignments on the same terminal. Little instructor intervention was needed for day-to-day student activity. This management system also provided instructors with programs to facilitate course grading, report generation for individuals or groups of students, communications with students, and scheduling of discussion sections arid other resources.

In November 1974 a conference on CMI projects (Mitzel, 1974) reported twenty-four operational CMI systems. Most systems were operating on mainframe computers, but a few existed on minicomputers. Microcomputers did not yet exist. Great progress was predicted for the future of CMI, and many experts felt that it would prove more successful and cost-effective than computer delivery of instruction (CBI). Research evaluating the effectiveness of instructional computer programs (reviewed in Bangert-Drowns, Kulik, & Kulik, 1985) provided evidence that CMI was more effective than CBI for older students. Baker (1978) provided the first comprehensive discussion, both theoretical and practical, on the design and implementation of CMI systems. Although Baker deals with mainframe CMI systems, his discussion of theoretical, social, and management issues is still pertinent to microcomputer-based systems.

FUNCTIONS OF CMI SYSTEMS

As Baker described CMI at the time, the typical functions of a system were the following: input and storage of student data (primarily test scores and activities completed), input and storage of curricular data (generally objectives, test items linked to objectives, and instructional materials or activities linked to objectives), retrieval and analysis of the data relating student scores and activity to the curriculum data, and generation of various reports showing individual or group progress and current status. The primary purpose of the analysis and reporting was to provide diagnosis and prescription of student learning problems or needs and appropriate instructional activities to remedy them.

Baker categorized CMI systems into small-scale (managing a single course at a single institution), medium-scale (managing multiple courses at a single institution), and large-scale (managing multiple courses at multiple institutions).

Specific CMI systems added features beyond the basic ones Baker discussed. On mainframe systems that could provide computer-based instruction as well as CMI (for example, PLATO), the CMI system provided routing of students directly to the appropriate computer programs and back to the management program after their completion.

Some added generalized roster or gradebook features, allowing instructors to enter and use information they had always kept about students, such as names, addresses, grades on classroom tests and homework, attendance, and so on.

Some added an "attention needed" feature, allowing the instructor to query the computer as to which students most needed attention and why. The computer would analyze the status of all students and produce a display or printout. One student might be identified as not having taken any quizzes in three weeks. Another as having received a very low score on the most recent test. Another as having completed all instructional units for the current semester and having nothing to do.

Systems with many terminals provided administration of exams on-line. This has advantages and disadvantages. On-line exams may be scored immediately and the test data accurately transferred to the CMI database. This reduces the time the instructor spends scoring tests and entering data, as well as reducing data entry errors. However, on-line examinations, if they are to be scored automatically, limit tests to objective items.

CMI systems integrated with CBI also provided a common base for the CBI component. That is, the CMI system takes care of asking and storing names, teaching about use of the keyboard, and providing instruction about general lesson procedures, and so on. This not only is efficient, eliminating redundant directions and activities at the start of every CBI lesson, but it also encourages standardization of procedures across lessons, making them easier for students.

Similarly, some CMI systems provided information for students, which was in contrast to Baker's point of view, that the CMI system's primary purpose was providing information to instructors. This typically is a subset of that information available to the instructor, providing the students with their own scores, status, and assigned activities.

Some CMI systems provided scheduling of scarce resources. Scarce resources are anything less numerous than the students. Textbooks, desks, paper, and pencils are generally not scarce resources. There are enough to go around and do not require scheduling. In contrast, reference books, computers, instructors, laboratory equipment, films, seminar rooms, and instructor office hours are all scarce resources. A CMI system can provide a calendar and scheduling program for such resources, allowing students to sign up and reserve in advance the use of a computer, an appointment with an instructor, or use of a videotape. Such scheduling enables efficient use of limited resources as well as fair access to them by all students. This feature can also alert the instructor when more of a particular scarce resource is needed.

The final CMI system feature we discuss, which was added on large-scale systems, is communications features for instructors and students. Because CMI systems are used in and partially automate individualized instruction systems, they increase the tendency for students to work alone and have less interaction with other students and instructors. There is a tendency for classes to meet less as a group with the instructor. As a result, the opportunities for instructors to make announcements to all students are restricted, and the ability of individual students to ask for help, make an appointment, and so on, are limited. Communication features allow instructors to easily leave announcements that all students will see next time they use the system, and allow students and instructors to exchange messages. If most students and the instructor use the CMI system on a daily basis, on-line communication can be more efficient and effective than mail or the telephone.

These additional features of some CMI systems are not the essential ones, according to Baker. On some large-scale CBI systems such as PLATO, many of these features are considered a part of the general system (or operating system) while the CMI component is only that part which coordinates tests, objectives, and assignments. However, since these activities are all aspects of the management of instruction, we believe they should be considered in the design and evaluation of CMI systems. Similarly we regard any system which includes these features to be a type of CMI system.

FACTORS INHIBITING CMI IN SCHOOLS

The same year that Baker's book on CMI was published, 1978, saw the introduction of commercially available microcomputers such as the Apple II, the Radio Shack TRS-80, and the Commodore Pet. As a result, elementary and secondary schools, and to a lesser extent universities, began switching from the use of mainframe and minicomputers for educational purposes to using microcomputers. The introduction of microcomputers in schools was a great setback for the use of CMI for two reasons. First, the effectiveness of a CMI system assumes and requires a shared database of all students' academic data. Independent microcomputers do

Second, with the development (for microcomputers) of many CBI programs and other software useful in classrooms, also came copy protection. Almost all software, from $30 math drills to $300 word processing programs were copy protected. Copy protection of programs prevents their management via CMI programs in several ways. It is usually impossible to use the software on a network or shared mass storage device. More importantly, copy-protected software often must be run by restarting the microcomputer. As a result, a coordinating CMI system cannot activate a CBI lesson or instructor utility, let alone return to the CMI system after that program's completion.

Separate from the issue of copy protection were several other factors preventing good CMI on microcomputers. The lack of adequate networks until fairly recently prevented maintenance of centralized databases and the incompatibility of software impeded management of microcomputer courseware. Most microcomputers allow and encourage developers to design totally different user inter-faces, data storage techniques, and peripheral devices. It is very difficult, perhaps impossible, to design a good CMI system that can control many different lessons which use different methods of entry, exit, data storage, and so forth.

The result is that medium- and large-scale CMI has grown very little, and perhaps even decreased, in schools. In contrast, CMI has grown more in industry and military training, where mainframes are still used and much courseware is internally produced and does not contain copy protection. There are some large school districts throughout the country that use CMI systems, but they are few and far between.

Separate from medium- and large-scale CMI systems are small, or perhaps even mini-CMI systems. That is to say, some specific CBI lessons or curricula have CMI components designed into them. For example, in the Milliken Math Sequences (Milliken, 1980) the CMI component stores student scores by name in a centralized database. It determines what drills the student must do next based on scores, and allows the instructor to manage the database, such as producing reports and removing old data to begin new classes. However, these mini-CMI systems are unique to a particular software series and cannot deal with off-line materials. If several such software packages are used, the instructor must work separately with the different CMI options for each package.

RECENT FACTORS ENCOURAGING CMI IN SCHOOLS

Very recently, we have become more optimistic about the future of CMI. First, the technology of microcomputer networking is rapidly progressing. Second, mass storage devices have increased in capacity while decreasing in cost. The result is to enable microcomputer networks to work with a single, very large database. Third, there has been a considerable decrease in the use of copy protection and hence an ability to run software stored on a networked database. And fourth, especially on Macintosh microcomputers, there is growing agreement on interfaces, data storage formats, and data interchange techniques.

As a result of the above changes, it is possible to design a CMI system that runs on a microcomputer network and manages a database stored on a mass storage device (for example, a large hard disk drive) available to all microcomputers on the network. The database may contain many things: CBI lessons, test-item files, objective and other curricular files, assignment (both on-line and off-line) files, and student data files. When lessons are not copy protected and contain more compatible data and entry/exit formats, the CMI system may control routing of students between lessons, interpret data generated by lessons, and pass useful information to lessons (such as student and instructor names, dates and time, previous lesson progress, or pretest scores). The CMI system can easily provide all the functions discussed above, routing, data storage, analysis, reporting, on-line examinations, general information, scheduling, and communications.

CMI AS A BASIS FOR CBI IMPLEMENTATION

It is our contention that CMI can and should be the basis of any medium- or large-scale CBI implementation. In the remainder of this chapter we explain why this is so and describe the phases of implementing a CMI system. We end with an illustration of a prototype CMI system.

By medium-scale CBI implementation we mean an entire CBI curriculum, such as high-school algebra or chemistry. The CBI materials might be any combination of tutorials, drills, simulations, or other methodologies supporting the curriculum, and would certainly include noncomputer activities such as textbooks, laboratory sessions, lectures, and discussions. Large-scale CBI refers to the management of several curricula within a school or across several schools. Again, the CBI materials would be in addition to other instructional activities and media.

Small-scale CBI refers to a situation in which individual instructors choose to use or experiment with one or a few microcomputers to supplement current teaching in their own classes. We do not believe that small-scale CBI implementation requires initial implementation of CMI.

What are the reasons that CMI is a good prelude to and basis for CBI? There are several reasons:

CMI implementation can reduce instructor work loads by automating tedious and routine functions such as grading, scheduling, and keeping track of resources. In contrast, beginning with CBI typically increases an instructor's workload. Thus, prior implementation of CMI can offset the extra learning and work involved in CBI and make it smoother.

CMI can be introduced in phases and adapted to meet the needs and concerns of any particular instructor or group of instructors. Initial implementation might be to provide gradebook facilities. Later phases might include test scoring, test administration, and student routing. Thus, implementation can accommodate an instructor's comfort level and expertise with computer technology. Furthermore, it can begin by replacing those tasks instructors do not typically like, such as grading, rather than replacing those activities instructors typically do like, such as working individually with their students.

CMI can be done with just one computer. This is a situation that instructors typically face-having one computer in their classroom. Although there are useful teaching functions that can be accomplished with one computer, they are few in number. In contrast, the initial phases of CMI are easily accomplished on a single computer used by the instructor. As one progresses to more sophisticated phases of CMI, additional computers are added.

CMI provides an excellent way to organize and integrate both CBI and traditional (off-line) instructional materials. The system permits sequencing, avoiding redundancy, and communicating with students as to when each resource should be utilized. It also helps instructors keep track of what work students have done both on and off the computer.

CMI helps instructors and curriculum planners determine curriculum needs. By collecting achievement data from progress tests and CBI lessons and by cross-referencing those data with the intended objectives, it becomes apparent which objectives are being met and which are not. Thus, rather than developing redundant materials or changing parts of the curriculum that are already working well, energy may be channeled into producing materials for those objectives that require attention. New development may be on a computer or any other medium. Subsequently, the CMI system's data collection and analysis can assess the success of newly created or changed components.

CMI may be used for either individualized or group instruction. Although CMI systems were created to meet the data organization needs of individualized and mastery learning programs (Baker, 1978), we do not believe that such a distinction remains necessary. First, we endorse a broadening of the definition of CMI to include not only the individualized testing and assignment components of instructional systems, but group testing, resource scheduling, grading, and other

instructional management functions, which are equally useful in group instruction, and may be even more useful for large group instruction such as college lecture-oriented courses.

Second, new technologies are blurring the distinction between individualized and group instruction. Whereas it used to be the case that a course might be all group or all individualized instruction, today we see many cases where some individualized CBI is included in a course that is otherwise largely group oriented. At all levels, from elementary school to college, instruction is becoming more eclectic with group-oriented components (such as lectures, discussions, and team projects) and individual-oriented components (such as reading textbooks, CBI, and term papers). Even if a course is largely group oriented, whenever there are individualized components, CMI can ease data management for the instructor and guide the student through the activities.

CMI provides a basis for evaluation of both students and instruction. Especially when achievement testing is automated, a CMI system can assist in grading and can assess (by virtue of their association with particular objectives and achievement measures) which instructional materials are teaching successfully. The former alleviates the tedious job of grading, while the latter permits more frequent course improvement.

CMI can solve some problems posed by incompatible software and hardware. Educators wishing to use CBI today are faced with an array of different brands and models of microcomputers and software intended for particular microcomputers. In addition, even software packages designed for the same model microcomputers are typically incompatible in terms of student access and data collection. Students and instructors must deal with the frustration of knowing a number of different access methods. One program requires restarting the computer. Another requires a special startup disk. A third requires using the computer manufacturer's special DOS disk first, and then typing in commands to run the program. Although these different techniques also present problems for good automation, a CMI system can deal with them better than students and instructors, thus easing the burden. Even if the instructional software does not allow the CMI system to route the student to and from it directly and to transfer data automatically, the CMI system may include directions to students and instructors and provide facilities for data to be transferred in a manual but guided fashion.

INCREMENTAL IMPLEMENTATION OF CMI AND CBI

In the preceding section we made the case that it often makes sense to begin with CMI rather than with CBI. However, that is not to say that one should start with a fully fledged large- or even medium-scale CMI system. Rather, CMI and subsequently CBI should be phased in. What we mean by this is as follows.

Take the example of an instructor in a school that has recently begun installation of computer facilities. The school has one or more central lab facilities as well as mobile computer carts for use in classroom. A reasonable way to proceed is to start using the computer technology in a simple way and then gradually extend its use-what we call an incremental implementation.

First, the instructor should learn to use the computer via simple class management and material preparation programs. A word processing program may be used to develop handouts, exercises, assignments, and lecture notes. A gradebook or spreadsheet program can be used to score tests and compute term grades. A database or test item banking program should be used for producing and printing copies of tests. A graphics program may be used for producing handouts and in small groups may be used as an electronic blackboard. A time-line program may be used for planning classroom schedules. The database program may also be used to inventory classroom materials, library books, and parent mailing lists. The spreadsheet program can also be used as an electronic blackboard program in teaching some subjects, such as math, science, or business.

When comfortable with basic computer usage and having used it to automate some regular teaching functions, the instructor progresses to a second phase of CMI implementation. This and subsequent phases require multiple computers, either in the classroom or in a central lab facility. The main feature of the second phase is networking the microcomputers. This is essential for good organization. Multiple, unconnected microcomputers only complicate organization of data and activities, and get in the way of automating them. Connecting the microcomputers so they share databases (such as student data and CBI programs) and resources (such as printers and telecommunication devices) enables a high degree of organization and automation. With the microcomputers networked, CMI components may be implemented for student signons, data collection, and routing to individual CBI lessons. Routing to lessons should be simple at this point, limited to students choosing from a hierarchy of menus with some capability for the instructor to turn parts of the hierarchy on or off.

In the third phase, assuming sufficient microcomputers are available, CMI components can schedule resources and provide communications and online progress testing. On-line testing is the most important component and ultimately facilitates the major advantages of CMI, namely, automatic assignment of appropriate instructional materials and evaluation of both students and the curriculum. Some testing systems, such as The Examiner (Trollip & Brown, 1989) which was described in Chapter 6, can be activated automatically by a CMI system.

The fourth phase adds more sophisticated routing to CBI lessons. Instructors may make assignments for any student or group of students. For lessons which permit it, data such as the student's name and previous activity may be passed from the CMI system and other data returned to the CMI system such as performance in the lesson.

The fifth phase adds procedures for automatic assignment and routing based on progress tests. This requires valid and reliable progress tests and good cross-referencing of test items, curricular objectives, and instructional materials.

The sixth phase, and a continuing one, is to provide features that the instructor desires. These are likely to include analysis and reporting functions that aid in day-to-day as well as semester-to-semester duties. Graphs that show a student's activity and progress might be useful for parent/instructor meetings A summary listing of test scores and other information might be useful for final grading. A summary of lesson popularity and effectiveness might be useful for planning subsequent semesters. Lastly, the weaknesses in the curriculum pointed out by progress test scores would lead to requests for added instructional materials to complement the instructional system.

The point of a phased implementation as we have just described is to progress to each successive phase as instructors become comfortable with the activities involved in the preceding phase. Failure can easily result from trying to implement a full-fledged CMI system right from the start. For each of the phases described, instructors may require a semester or even a full academic year to become competent and comfortable with the activities. Then they are ready to move on to something more complex.

A PROTOTYPE CMI SYSTEM

We end this chapter by describing and illustrating a hypothetical CMI system. The design of CMI depends on many factors. Chief among them are the instructor and student activities to be automated, the type of instructional materials being managed, the educational level of the students and the amount of control they are given over their learning activities, and the number of courses and instructors involved.

For our example we choose a somewhat complex case, namely the management of instruction across several courses and departments at the university level. We begin by assuming that the fictitious North Woods University has several microcomputer instructional classrooms on campus that are interconnected by a network and share a single large database of student data and courseware. From any microcomputer all faculty members can retrieve information on their advisees or students enrolled in their classes. Similarly, from any microcomputer, students can sign on to the network and access instructional materials and advice for all university courses in which they are enrolled. Access to personal records is restricted by virtue of both a person's signon (whether student or faculty member) and passwords. A design decision of the North Woods University (NWU) instructional system is that every student and every instructor has only one signon, from which all appropriate programs and data can be accessed. This provides for good integration of all relevant data across the system. For example, by unifying the instructional system with the university's academic data processing database, advisors can see not only a student's current course activity, but grade history and any other information an advisor requires. It is even possible, given this level of integration, that the system could be used for course registration.

A name is insufficient to distinguish students reliably, for students may have identical names and should not be forced to use a modified name just to suit the system.

Figure 13-1 shows the initial signon page for the network. In Figure 13-2 a student, William Long, has typed in his name and is now prompted to enter his identification number. A name is insufficient to distinguish students reliable, for students may have identical names and should not be forced to use a modified name just to suit the system.

In Figure 13-3 William has entered his ID number as well. This, and not his name, uniquely identifies him on the system. He is now prompted for a pass-word, which prevents other persons from using the computer system under his name. This is essential, because the system includes personal data, tests, grades, and other sensitive information.

In Figure 13-4 the system has informed William that he entered his password incorrectly and must try again. A well-designed system must allow the student to change the password at will to maintain privacy.

When William successfully signs on (Figure 13-5), he sees a page which consists of his personal instructional activities (the options listed vertically) and those management activities to which all students have access (the options listed horizontally at the bottom of the screen). Using a mouse or other pointing device, William can point to and choose individual CBI lessons (the first two options on the vertical list), courses (Calculus II through Intermediate German I), or the general options to read or write electronic mail, see schedules, see information about his academic status, get help on using the system, or exit and sign off.

Figure 13-6 shows a sample schedule. Having chosen the schedule option on Figure 13-5 and then choosing the Math Computer Lab schedule, William can obtain information about computer availability and, by pointing, choose a day and time to reserve a computer. Because all computers are networked, the appropriate computer can prevent any other student from signing on at the time William has reserved it. Again there are options along the bottom of the screen which can be pointed at for other actions. For example, if the microcomputer has an attached printer, William can obtain a printout of his reservation.

The most important options listed on Figure 13-5 are the courses, Calculus II, Engineering II, Electricity and Magnetism, Freshman Rhetoric, and Intermediate German I. These names are identical to the university courses in which William is enrolled. Through the use of a networked university-wide CMI system, students have access from any computer to instructional materials and guidance for all their courses. A course option, as we will see next, provides the student with on-line instruction (CBI), progress tests for instructional decision making, and advice and assignments for activities not on the computer such as reading the textbook, discussion and review classes, or films.

In Figure 13-7 William picks the Calculus II course and receives the information shown. Based on the work he has done on the computer, he is ready to take a unit posttest (similar to the typtcal untversity hourly exam) assuming he has also completed his off-line work The computer, of course, cannot tell whether he has done noncomputer work. While it is possible to allow the student to take a test basedjust on CBI progress, here it is deemed better to allow instructors some control. To take the test, William must use the mail option to send an electronic letter to the course instructor, in which he indicates he has completed all the off-line work. The instructor, as we will see later, may choose to permit the test, sending a response back to William to that effect.

A little later, in Figure 13-8, we see that the instructor has enabled William's test. When William chooses the Calculus II option on his main menu again, he sees this page. He has several options (along the bottom) such as reviewing CBI lessons before taking the exam.

Later (Figure 13-9) William signs on in a proctored lab to take the test. Tests must be taken in a proctored lab for security because even though students sign on with a name and secret password, there is no way for a computer to know that the person working is really the person whose name is typed. Nor can the computer tell if the student has a friend sitting next to him or her to help on the exam, or if the student has a textbook or notes handy. In a proctored room, a proctor may check photo identification to be sure the correct person takes the exam and ensure that no cheating occurs. The proctor may also be of assistance to students concerning exam procedures or provide help in the event of a computer malfunction. In Figure 13-9 William is on the third item of his test. In keeping with the principles of Chapter 6, he can move around freely among items, mark items for later review, and change answers.

When he eventually clicks on the done option he will receive test feedback (not shown) followed by his next Calculus II assignment (Figure 10). This display tells William that he has successfully passed Unit 1 and is ready for the next unit. He is informed about the CBI lessons and the off-line assignments for the next unit. Along the bottom of the display are the usual options.

The next day William signs on and, from his main menu (Figure 13-5), chooses the see status option. He sees the display in Figure 13-11 which summarizes all his current on-line and off-line assignments and his status within the on-line assignments.

Returning to the menu, William chooses the Calculus II option and receives the display shown in Figure 13-12. It should be obvious by now that every time a course option is picked, such as Calculus II, the student may receive a different activity. Depending on the course and its structure, the student may have a choice of lessons or may be directed, as in the case of CalculusII to a specific lesson. William is now told he is ready for a particular lesson. Even so, the options along the bottom of the screen include ones that allow him to review previously studied CBI lessons for this course (by pointing at the review-map option).

Two days later, having done some of his calculus lesson, William chooses the review-map option and sees the display in Figure 13-13. This is an abbreviated map showing the layout of the entire on-line portion of the Calculus II curriculum. The map shows that the course begins with a pretest (which William took at some time in the past) to assess whether the student is ready for the course, or which may direct the student to study some or all of the course's units. There are six units in Calculus II, each ending with a posttest that determines if the student is ready for the next unit. Posttests may also be used for grading, if the instructor so desires. At this time William has completed all the lessons in Unit 1 and has just begun Unit 2. The completed lessons are shown in dimmed (non-bold) type and William can point at any completed lesson to review it. A lesson in review mode is frequently abbreviated and does not take as long as the original lesson.

Summarizing the activities and options for our imaginary student William Long, the CMI system administers diagnostic tests based on which he is given study assignments, both on and off the computer. The computer automatically routes him to on-line assignments and keeps track of his progress, but an instructor must intervene to tell the system when off-line assignments are considered done. The system accommodates all of William's courses, each of which has its own unique structure. Not all have on-line tests or CBI lessons. Some may be strictly sequenced like Calculus II, while others may provide a menu of options for the student to choose in any order desired. Even when a course is sequenced in a strict fashion, the system may provide the student with the option to review. At almost any time, students may see their status and assignments for all courses, may see the map for a course to ascertain their progress in the course, may write and read electronic mail to communicate with instructors, and may see schedules and make reservations for scarce resources.

The Instructor's Perspective

We now look at how the CMI system is used by Janet Brown, the professor who teaches the Calculus II course. She signs on the network in the same way as any student (as in Figures 13-1 through 13-4). However, the system recognizes that she is an instructor and displays a menu of instructor options, Figure 13-14. Janet may do many things. She may see a roster of students in any course she teaches (or to which she has authorized access) or a roster of students whom she advises. She may see the catalog of all NWU courses and CBI lessons or may inspect and try out CBI lessons or tests. She may read and reply to electronic mail from students or other faculty. Like a student she may see schedules and make reservations, but as an instructor she may also be able to set reservation schedules, determining, for example, the number of students that may enroll in an exam review session, the time and location where computerized exams may be taken, or the hours that the Math Department Computer Lab is open.

In Figure 13-15 we see that Janet Brown has chosen to see the course roster for Calculus II. Since this is a large enrollment course (taken by all freshmen science and engineering students), there are multiple sections to facilitate instructor management. Sections might also correspond to teaching assistant assignments, so that Janet can assign each of her teaching assistants a different section for student help and grading. She is now looking at Section C, in which William Long is enrolled. Student names are shown with their unique ID numbers, as is necessary in the case of the two John Smiths. Janet does not need to type long ID numbers however, or even names to access student records. She may pick a student by pointing and clicking with the mouse. Along the bottom of the screen are many management options for a course roster. They include adding and deleting students from the roster, seeing the progress data for any individual, printing the roster, reading or writing electronic mail, and so on.

One of the very powerful options a good CMI system may provide is the attention option In Figure 13-16 we see that Janet Brown has chosen this option. The computer scans all records and displays the names of those students who need attention soon. From this section four students are listed together with reasons they need attention. This feature allows the instructor to check who needs help quickly, which encourages doing so often. The instructor needs not look through each student's records, test scores, or file folders, nor wait for students to come asking for help.

In Figure 13-17 we see that Janet Brown has chosen to see data for William Long (choosing that option from Figure 13-15 or from 13-16 if William had been on the attention needed list). She can see a summary for all of William's on-line assignments Among the many available options, she can choose to see detailed data for her own course (Calculus II), including the map (Figure 13-13) that visually shows William 5 locations within the course activities.

As seen in Figure 13-18, she can also choose to modify William Long's on-line assignment. In this display she is inspecting the catalog of Math Department courses and can add or drop William from them. She could also add or delete individual CBI lessons which are not associated with particular university courses.

In Figures 13-19 and 13-2 Janet Brown uses the system's electronic mail facilities. In Figure 13-19 she reads William Long's letter requesting permission for the next test. In Figure 13-20 she replies to him, indicating that permission has been granted.

Finally in Figure 13-21, another faculty member, Dennis Harrison, has signed on the system and chosen to see information about one of his advisees, William Long. The advisor sees a summary of current assignments very similar to that shown to students (Figure 13-11) and course instructors (Figure 13-17). The advisor cannot access details of courses unless he actually teaches them, but can access other information about the student such as grades and academic history. If Dennis Harrison were to note a problem in a course, such as Calculus II, he could send electronic mail to Janet Brown to discuss the problem. Although not shown, the faculty advisor may access a roster of all advisees (similar to the course roster in Figure 13-15) and a list of advisees needing attention (similar to Figure13-16).

CONCLUSION

Computer-managed instruction was among the first uses of the computer in education but faded almost into obscurity due to the shift from mainframe computers to microcomputers. It now has the potential for a comeback due to advances in mass-storage technology and availability of networking. We believe instructors will benefit by using computers to assist in instructional management before using them for delivery of instruction. Even small CMI systems provide the basis for course improvement by reducing the instructor's time spent with management functions and freeing them for teaching and helping functions. A large-scale, integrated CMI system, such as that illustrated in this chapter, should be the long-term aim of any educational institution wishing to utilize computer technology fully to improve instruction Such a system can integrate all teaching, learning, and administrative functions in a way that enables all members of the educational community to teach and learn more efficiently.

REFERENCES AND BIBLIOGRAPHY

ANDERSON D.O., & GLOWINSKI D. J. (1982). Effectiveness of a computer-managed instruction program. Journal of Learning Disabilities, 15(9), 555-556.

ANDERSON T. H., ANDERSON R. C. ALESSI, S. M. DALGAARD B. R., PADEN, D., BIDDLE, B., SURER, J. R., & SMOCK H.R. (1975). A multifaceted computer based course management system. In 0. Lecarme & R. Lewis (Eds.), Proceedings of the Second World Conference on Computers in Education. Amsterdam: North Holland Publishing Co.

ANDERSON T. H., ANDERSON R. C., DALGAARD B. R., PADEN, D. W., BIDDLE, W. B., SURBER, J. R., & ALESSI, S. M. (1975). An experimental evaluation of a computer based study management system. Educational Psychologist, 11, 184-190.

ANDERSON, T. H., ANDERSON, R. C., DALGAARD, B. R., WIETECHA, E. J., BIDDLE, W. B., PADEN, D. W., SMOCK, H. R., ALESSI S. M., SURBER, J. R., & KLEMT, L. L. (1974). A computer based study management system. Educational Psychologist, 11, 36-45.

BAKER, F. B. (1971). Computer-based instructional management systems: A first look Review of Educational Research, 41(1), 51-70.

BAKER, F. B. (1978). Computer-Managed Instruction: Theory and Practice. Cliffs, NJ: EducationalTechnology Publications

BAKER F. B. (1981). Computer-managed instruction: A context for. computer-based instruction. In H. F. O'Neil, (Ed.). Computer-Based Instruction: A State-of-the-Art Assessment New York: Academic Press.

BANGERT-DROWNS, R. L., KULIK, J. A.A, & KULIK, C-L.C. (1985). Effectiveness of computer-based education in secondary schools. Journal of Computer-Based Instruction 12(3), 59-68.

BLOOM, B. S. (1976). Human Characteristics and School Learning. New York: McGraw-Hill.

BLUHM, H. P. (1987). Computer-managed instruction: A useful tool for educators? Educational Technology, 27(1), 7-13.

BORTON, W. M. (1988). 988). The effects of computer-managed aged mastery learning on mathematics test scores in the elementary school. Journal of Computer-Based Instruction, 15(3). 95-98.

BROWN, N P. (1982). CAMEO: Computer-assisted management of educational objectives Exceptional Children, 49(2), 151-153.

BRUNO, (1987). Admissible probability measures in instructional management. Journal of Computer-Based Instruction, 14(1), 23-30.

FERRICO, P-A. (1982). Individual differences in cognitive characteristics and computer-managed mastery learning. Journal of Computer-Based Instruction, 9(1). 10-18.

FEDERICO P-A. (1983). Changes in the cognitive components of achievement as students proceed through computer-managed instruction. Journal of Computer-Based Instruction, 9(4), 156-168.

FLANAGAN, J. C. (1969). Program for learning in accordance with needs. Psychology in the Schools, 6(2), 133-136.

FLANAGAN J. C. (1970). The role of the computer in PLAN Journal of Educational Data Processing, 7(1), 7-17.

FLANAGAN, J. C., SHANNER, W. M., BRUDNER, H. J., & MARKER, R. W. (1975). An individualized instructional system: PLAN. In H. Talmadge, (Ed.), Systems of Individualized Education. Berkeley: McCutchan.

GOODSON, B. (1984). Software report: Are computer-managed instruction programs worth the trouble? Electronic Learning, 4(1), 8.

HANSEN, D. N., ROSS, S. M., BOWMAN, H. L., & THURMOND, P. (1975). Navy Computer-Managed Instruction: Past, Present, and Future. Memphis, TN: Memphis State University, Tennessee Bureau of Educational Research and Services. (ERIC Document Reproduction Service No. ED 114051)

HOUGHTON MIFFLIN. (1986). Houghton Mifflin Computer Management System. [Computer Program]. Boston: Houghton Mifflin.

KELLER, F. S. (1968). "Good-bye, Teacher dance...Journal of Applied Behavior Analysis,1, 79-89.

LECHOWICZ, & WISE, F. H.. (1976). Tagging course objectives for the computer. Educational Technology, 16(4), 40-41.

LEIBLUM, M. D. (1982). Computer-managed instruction: An explanation and overview. AEDS Journal, 15(3), 126-142.

LILLIE, D. K., HANNUM, W. H.., & STUCK, G. B. (1989). Computers and Effective Instruction: Using Computers and Software in the Classroom. White Plains, NY: Longman.

LOCKARD, J., ABRAMS, P. D., & MANY, W. A. (1987). Microcomputers for Educators. Boston: Little, Brown and Company.

MAYO, G. D. (1974). Computer based instructional systems. Journal of Educational Technology Systems, 2(3), 191-200.

MIDDLETON, M. G., PAPETI, C. J., & MICHELL, G. S. (1974). Computer-Managed Instruction in Navy Training. (TAEG Tech. Rep. No.14). Orlando), FL: Training Analysis and Evaluation Group.

Milliken PUBLISHING CO. (1980). Milliken Math Sequences. [Computer Program.] St. Louis: Milliken Publishing Co.

MITZEL, H. E. (Ed.). (1974). An Examination of the Short-Range Potential of Computer-Managed Instruction. (Conference proceedings for Grant No. NIE-C-74-0091). National Institute of Education.

ROBINSON, C. A., TOMBLIN, E. A., & HOUSTON A. (1981). Computer-Managed Instruction in Navy Technical Training:: An Attitudinal Survey. (Tech. Rep. NO). NPRDC-TR-82-19. San Diego: Navy Personnel Research and Development Center. (ERIC Document Reproduction Service No. ED 212290)

SWOPE, W. M., COREY, J. M., EVANS, R. M., & MORRIS, C. L. (1982). Analysis of Factors Affecting the Performance of the Navy's Computer-Managed Instructional System. (Tech. Rep. NO). TAEG-TR-19). Orlando), FL: Naval Training Analysis and Evaluation Group. (ERIC Document Reproduction Service NO). ED 223 240)

TELEM, M. (1982). CMI and the MIS-An integration needed. AEDS Journal, 16(1), 48-55.

TROLLIP, S. R., & BROWN, G. B. (1989). Examiner, V 2.I. [Computer Program]. Mendota Heights, MN: Media Computer Enterprises.

VAN MATRE, N. (1980). Computer-Managed Instruction in the Navy: Research Background and Status. (Tech. Rep. NO). NPRDC-SR-80-33). San Diego): Navy Personnel Research and Development Center. (ERIC Document Reproduction Service NO). ED 196411)

WAGER, W. (1985). Computer-managed instruction-How teachers and principals can improve learning. NASSP Bulletin, 69(478), 22-27.