Individualized Instruction: Tutorials, Learning Agents, and Other Tools
Mr. Johnson was responsible for a new training course. The problem was that mastery of the material required a background in a number of subareas. In the past he would have included each of these topics in his training course. Although it would have bored some students who already knew those topics, at least he would be sure to cover the material for all of the students. He would rationalize that good review for those who already know the material can't hurt anyway. But now with computer assisted individualized instruction, he didn't need to worry about that. Each topic was written as a module. If students answered the questions correctly at the beginning of a module, they skipped over it. If they didn't, the program continued through the module until they got all of the questions correct. Thus, each student might go through a different set of modules depending on their prior knowledge. While it was a lot of work to write the programs for CAI, it had been worth it.
But the problem in the present case was that the training course had to be done in a group in a classroom setting. There would be ten 75 minute sessions over two weeks. There were eight modules that needed to be covered. But each module required knowledge in one or more subareas. Mr. Johnson identified 12 subareas that might need remedial help. Using an authoring system for computer tutorials, he programmed these modules. On the first day he would use a test to identify which students needed to work through which tutorials. These would then be scheduled so that work would be as evenly distributed over the two weeks as possible and so that each subarea would be mastered prior to the class session that required that information. Mr. Johnson wondered if this should be called "just-in-time" learning or "adjusted-in-time" time. Which ever, in the follow-up course evaluation, the students reported that they had never felt so well prepared for and competent for the training sessions due to the out of class tutorials that they completed prior to the classes.
Programmed learning, computer-based instruction (CBI), and computer-assisted instruction (CAI) have been around for a number years. These systems have the distinct advantage of being able to accommodate for the individual needs of the student both in terms of the material covered and the time that it takes for the student to learn the material. This chapter will not attempt to review work but will instead relate computer-based individualized instruction to overall learning objectives in the switched-on classroom. The question is how to link CAI to the needs of the student in the classroom in a seamless and integrated manner.
New learning systems have been developed in the last few years that combine artificial intelligence agents with instructional materials and goals. Intelligent tutors move beyond the rote drill and practice of the programmed learning by using expert systems to diagnose the learner's errors and generate steps to correct the learner's knowledge base and problem solving strategy. In a classroom setting, such tools may be used not only for tutorials but also in conjunction with regular studying and test taking activities.
While these approaches provide individualized instruction in one sense, they do not necessarily foster independent thinking, creativity, or self-management of learning. Most forms of CAI create a dependence of the student on the system. While this may be beneficial during the early stages of the educational process, eventually, the student must be weaned from dependence on the system to acquire a role of mastery over both the learning process and the instructional system. It is at this point that the electronic educational environment must provide learning tools that are similar in many ways to productivity tools in the work place. The student needs tools for personal planning, scheduling, and self-assessment of learning activities.
Tutorials and individualized tools in support of classroom education will be discussed under these three types: (a) self-contained modules for learning pre-requisite or ancillary material as illustrated in the opening scenario, (b) intelligent tutors that assist the student in learning problem solving skills in addition to the course material, and (c) tools for self-management of learning course materials and generic tools for developing study and work skills.
Remedial and Ancillary Study
Nearly every teacher has been frustrated by the student who in the middle of a lecture asks a question about some background material that everyone in the class should already know. Other students, knowing that they should know the information, don't ask for clarification and are left never understanding the material. How can we make sure that all of the students are up to speed for either the lecture or the discussion? We continue to encourage students to read the assigned material prior to coming to class; and we may suggest that students who are weak in some area should do some extra reading. But it never fails that many students come into a class with a deficit and continue to fall behind.
Generally what is needed is direct intervention. The most threatening intervention is the pop quiz at the beginning of class to see who on top of the material and who is not. Quick 3-5 minute quizzes can be included in HyperCourseware prior to lectures and group activities. These may consist of 5 multiple choice or fill-in-the-blank questions. The results can be immediately viewed by the instructor, fed back to the students privately, and/or viewed by the class as a whole in summary form. If the instructor views the results, he or she may suggest additional readings to particular students. If the student views the results, he or she may decide to study additional material. If the group as whole views the answers, it may be used as a time to discuss questions that most students got wrong. The use of exams and quizzes will be discussed at greater length in the next chapter. At this point, we will assume that we have some indicator of the need for tutorials.
Tutorials should in general be conducted outside of class time. As in the scenario opening this chapter, they should be as evenly distributed over the course of the semester as possible. As a part of HyperCourseware, tutorials and computer assisted instruction modules may be listed as a part of the Course Materials Module. Each module might be listed in an index with scope and sequence information. In addition, tutorials may be scheduled on the syllabus as readings in the Reading Module or as assignments using the Assignments Module. Since the scheduling of tutorials and the results of the tutorials would be stored on the network in the Student Space, the instructor would have access to the number of students working on tutorials and their progress. This information should prove useful in knowing the level of knowledge of the class prior to class sessions.
Examples of training modules using computer based instruction abound. However, we have not yet come to a point where an instructor looking for a particular training module can with good probability find one to meet his or her objectives. The only recourse is to write one's own. This, however, is not an easy task. While a number of good authoring programs exist, the amount of effort is still perhaps greater than writing a traditional textbook on the subject. Deciding on and specifying the branching, authoring the multimedia materials, and programming simulations, games, and other interactive exercises are monumental tasks. At present few educators and publishers are willing to take on challenge and the investment.
Figure 9.1 shows two simple frames from a short tutorial on formal problem solving theory hosted in HyperCourseware. As in most systems, study material is presented in a frame; in the next frame the student must answer a simple question about the material, and so on. If the questions are correct, the system marches through the material in a fairly linear manner. If the student misses questions, the system branches to repeat study material or possibly to additional study material and repeats the same or similar questions until the student has mastered the topic. Most systems try to use encouraging feedback.
Many tutorial systems allow the author to build and see a map of the branching structure. The learner is generally in the dark, however, as to the length and branching of the tutorial. The prototype in HyperCourseware allows the student, as well as the author, to see the structure of the lesson. This is shown at the right of each screen. In this case, squares represent lesson material and circles test or practice material. Filled or checked nodes indicate that the material has been covered.
Computer Tutorials Instead of Textbooks
In our initial efforts to incorporate computer based instruction and tutorials, we started with traditional study materials (e.g., the textbook, selected readings, etc.), assume that the student has read and studied the materials prior to class, and then test that knowledge in class. This approach, however, suffers from its inception by using traditional study materials. Instead of using a passive set of study materials, the electronic education al environment should begin with interactive materials that combine multimedia material with individualized instruction and an assessment of the student's knowledge of the information prior to class rather than during class. Hard copy textbooks should be replaced by programmed learning materials or at a minimum supplemented by assessment measures. If the materials are on-line, the instructor may at least know if the student has "cracked" the textbook, or additionally, how long the student studied the material, or ideally view a profile of strengths and weaknesses revealed during the study session.
The problem with tutorials is not their lack of educational potential. The problem is authoring systems of tutorials that are integrated with the course objectives and sufficiently extensive to cover the scope and sequence of the course. The problem is that tutorials have been perceived as being useful for remediation and for supplementary materials rather than as a complete solution to replace textbooks and other materials. On the hand, tutorials have been used to replace the human instructor than to augment the role of teacher. Instead computer tutorials and other computer based instruction should be seen as replacing the textbook, allowing the instructor to work at a higher plane in the classroom, knowing that the students have mastered the prerequisite material prior to the class.
Intelligent Tutors for Course Materials
Finally, it should be noted that programmed learning in the form of computer based instruction has been quite successful for material that is primarily factual, textbook information. However, it has been less successful for procedural learning involving the application of problem solving steps. It is one thing to memorize the equations for the distance and rate of travel and the definition of solving simultaneous equations by substitution. It is another to successfully solve word problems that require the application of these facts. This is where intelligent tutors come in. Computer programs that present problems, monitor the students attempts to solve the problems, and provide corrective guidance to the student.
Computer based instruction has a number of great advantages over textbooks and other hard copy materials. In addition to branching and repetition based on mastery of the material, computer tutorials can use multimedia materials and be truly interactive. They can allow the student to see molecules rotate, to see machine parts in operation, and to hear the sounds of animals as well as there pictures. Now with virtual reality, it is also possible to experience the materials in a three dimensional world through which the student can navigate spatially. More significantly, interactive systems allow the student to explore by experimentation. They can learn by directly manipulating the systems whether ecological, economical, or biological. Construction programs allow the students to learn by doing. They can follow procedures, design objects, build them, and subject them to tests.
Self-Management of the Learning Process
Computer assisted instruction and intelligent tutors take on the task of objectively evaluating the student's performance based on mastery of the subject. However, an important part of education is self-assessment. How do we know if we have learned something? How do we know when we need to learn something for future use? Too often this is not the function of the student but of the educator who plans the "scope and sequence" of the material and programs it into the computer. In order to learn B you must first know A. While this is true, classroom education cannot guarantee that a particular student will have learned Subject A before the class moves on to subject B.
Students may be tested for mastery of the topic and if they pass they move on to the next level. If they fail, they repeat the subject until they master. However, things are not always so cut and dried. Tests are not totally reliable. Subjects are not monolithic. Understanding may be superficial.
The decision to study material is also based on one's self assessment. The student may feel that he or she knows the material. This is all too often an incorrect assumption. Students might feel that they know a topic based on what they remember. Unfortunately they are not in tune with what they do not remember. They are subject to the "availability bias."
According to many educators, one of the major problems that students have is one of scheduling. The problem is knowing how much time one needs to study and allocating the time to do it. This is difficult enough for one course, but compounding it with three or four other courses can result in a complex problem. Many students do poorly on exams and other courseware simply because they do not understand the scheduling problem or cannot solve it.
The problem of scheduling has a long history in business and industry and many computerized tools have bee written to increase productivity and ensure that complex projects are completed on time. Yet these tools have not been applied to individual learning in a serious way. Personal productivity tools need to be developed learning activities particularly for allocating study time and the competition of assignments given the due dates imposed by the instructor. Such tools should take into consideration individual differences such as reading speed, typing speed, information retrieval speed, parameters of the learning curve, and other parameters involved in project completion. Robust algorithms could be used to generate time needed to complete tasks and a scheduler could be used to distribute required times over the course of the semester.
Figure 9.2 shows a prototype screen in HyperCourseware for scheduling (a) studying for two exams, (b) completion of three writing assignments, and (c) a term project. Individual parameters were assessed on other screens and are factored into the projections on the scheduling screen.
This screen starts from the date of an exam or due date of an assignment, estimates the amount of time needed for study or completion of the assignment, projects that time backward and allocates study and work time from the time resources available to the student. If the amount of time required exceeds that available, the student is alerted and must make a decision to either allocate more time by forfeiting discretionary time (e.g., time for social engagements and entertainment), forfeiting time allocated to other courses, or decide to jeopardize the grade on the exam or assignment. In addition, to allocating time the scheduling may be used to help order the steps required to complete an assignment or study for an exam. For example, it may suggest the following steps for writing a term paper: (a) decide on the topic and its scope, (b) do a literature review, (c) outline the paper, (d) write sections, and (e) finally, proof and format the paper.
A number of educators are advocating the use of journals by students. The journals may be used to keep class notes, records of experiences, personal reflections on the course material, and a log of activities. The purpose of the journal is help the student focus and reflect on the topic and to process the material at a deeper and more personal level. In traditional classrooms if the instructor wants to review the journals of the students, they must be collected, read, and finally, returned to the students. In the electronic classroom, journals are not only easier to manage but they can be enhanced to include multimedia materials and links to other sources.
Figure 9.3 shows a sample page of an electronic journal taken by one student. Since it is electronic and available from the student workspace, the instructor can review it anytime and add comments to it.
As educators in the new electronic environment, we can and must now more than ever take into consideration in the learning process the individual differences that students bring to the classroom. Students differ in a number of strengths and weaknesses, from prior learning and mastery of material to cognitive abilities and motivation. A number of these differences can be compensated for by individualized instruction using the computer. Prerequisite knowledge and skills can be ensured prior to class sessions by remedial computer-based tutorials designed and scheduled to bring students up to speed. Differences in cognitive abilities can be factored into the learning process allowing students make the most of their strengths.
In courses with fairly well-defined content material and learning objectives, it possible to write, with some effort, computerized tutorials that manage the learning process by drill and practice. In courses requiring students to develop problem solving skills and engage in creative projects, intelligent tutors may be designed to assess the steps and strategies used by the students, diagnose errors and to guide them to better solutions. At the highest level of learning, courses may require students to engage in uncharted exploration, develop skills in self-assessment and self-management, and develop personal character. While neither programmed instruction or intelligent tutors can be written for this level of learning, other electronic tools can help the student. These include personal productivity tools, spreadsheets, schedulers, and on-line journals. At this point, it will be up to the students to take control of and to expand the electronic educational environment themselves. This issue will be revisited in the last chapter.
Exercises and Projects
1. Decide on a simple topic and write a computer tutorial or computer based training program for it. How difficult is it to generate such a module? What advantages does it have over a simple instruction manual?
2. What sort of information would you like to collect on the individualized instructional material from students prior to holding a class session? How would you use this information to advance the level of the class during the lecture or discussion?
3. Start an electronic journal. What thinks would you record in the journal that would help improve your abilities as a student?
Alessi, S. M., & Trollip, S. R. (1991). Computer-Based Instruction: Methods and Development, (2nd. ed.), Boston: Allyn & Bacon.
Gagné, R. M., Briggs, L. J., & Wagner, W. W. (1988). Principles of instructional design (3rd. ed.), New York: Holt, Rinehart, and Winston.
Soulier, J. S. (1988). The design and development of computer-based instruction. Boston: Allyn & Bacon.
[Table of Contents] [Chapter 8] [Chapter 10]