An Instructional Theory for Introductory Computer Science for Non-Computer Science Students
Spring, 1997
Table of Contents
Visual representation of the theory
Figure 1: Triangulating student skills and schemata to understand concepts.
This instructional theory addresses computing concepts and principles for non-computer science students at the university level. It is intended to address an entire course and the structure of the individual lessons. The lessons are designed to help students construct mental models of how computers and software work to enhance retention and facilitate transfer to learning new software to solve new problems. Rather than concentrating on the recall of facts, the lessons focus on the process of solving problems in collaborative learning groups. In this way, students help each other learn the material, and in the process must explicitly reflect upon their own understanding. A key component of this theory is the assessment, a modified mastery model in which students must demonstrate their ability to solve problems from a range of disciplines using a variety of computer software. The assessment is non-competitive to encourage students to participate fully in the collaborative learning model. This theory is being operationalized in a new computer science course at Michigan State University: CPS 101.
Prior to designing this course and theory, we conducted a series of 67 interviews with the chairs of the various client departments whose students could be the audience for a new introductory course. We gathered data on the types of computing competencies and concepts the chairs saw their students needing, both for their subsequent course work and in their prospective careers. Among the most interesting outcomes:
- Across disciplines, employers expect graduates not only to have knowledge in their domains and a basic set of computing skills, but they also expect graduates to act as computing mentors for their other employees.
- While many disciplines expect students to use computers for sophisticated purposes, none of the chairs wanted students to learn programming. In many disciplines, tasks such as data analysis which once required programming knowledge are now done with spreadsheets and specialized analysis tools.
- Across disciplines, the chairs acknowledged that their students needed to be able to adapt to changing computing environments and learn to use new software in new ways to solve problems. They did not merely want training on particular software.
These results helped us define the purpose of CPS 101. We wanted a course that would help students from a wide variety of disciplines learn a core set of computing concepts and competencies to help them succeed in a wide range of disciplines. We needed to do so without teaching programming.
Even though computer science (CS) is a rapidly evolving discipline, much of the underlying cannon – and hence the purpose and desired outcomes of introductory computer science courses for CS majors – has remained unchanged for years. Some of the conditions – computing platforms and languages – have evolved rapidly and many of the instructional methods have changed as a result. However, when considering computer science instruction for non-computer science majors, the purpose, conditions, and desired outcomes have changed even more radically than those for computer science majors.
There is general agreement among computer science professionals as to what should constitute the curriculum for a degree in computer science . The purpose of an introductory CS-1 course usually includes an introduction to the field of computer science, problem solving, algorithms and programming to prepare students for subsequent courses in the curriculum . The conditions of instruction vary by institution, but generally the students in CS-1 courses have relatively homogeneous math and science backgrounds and are motivated to learn the material in order to pursue further courses in computer science. The desired outcomes are for students to have the background they need to move into a CS-2 course and to be prepared for the rest of the computer science curriculum. There are a number of instructional methods that are appropriate for meeting these outcomes, but in most CS-1 courses students spend a substantial amount of time designing and writing computer programs as exemplars of problem solving and algorithm development. This curriculum is intended to provide students with a deep structural understanding of computing concepts and principles.
Introductory Computer Science for Non-majors
With the advent of microcomputers, introductory computer science courses specifically intended for non-computer science and non-science/engineering majors flourished at most colleges and universities. The purpose of these courses was still to introduce students to computer programming, since programming was still necessary to use microcomputers. Microcomputers changed a number of the conditions for these courses, by allowing stratification of the curriculum. While computer science students learned Pascal with an emphasis on algorithms and data structures, engineering and science students continued to learn FORTRAN, with an emphasis on solving computational problems. At the same time, students who had not previously taken computer science courses began to take introductory courses that focused on programming in BASIC . However, the assumption of these courses was still that understanding computing concepts required understanding how to program.
As microcomputers continued to increase in power, the availability of a wide range of applications for solving a variety of problems without the need for programming resulted in another bifurcation in the non-computer science curriculum. For the first time, programming was not the focus of some introductory computer science courses. However, these courses often concentrated on procedural skills, such as learning to use specific software packages, at the expense of concepts and principles. The result was that students often were unable to transfer their knowledge to the use of other software.
This instructional theory must accommodate a range of students: from no prior computing experience to some who are likely to have a great deal of incoming knowledge of computing as the course has no prerequisites. Moreover these students are much more heterogeneous than CS majors. They come from 67 different disciplines and have a wide range of interests and motivations. Analysis of the SAT scores of the students enrolled in MSU’s introductory CS-1 course compared with students enrolled in the non-major courses during the 1995-96 academic year showed that the non-majors math and verbal scores averaged almost 100 points lower than the scores of the CS majors.
Due to institutional constraints, we must deliver this course with the same resources devoted to the courses that it will replace. We anticipate demand from 1700 to 2000 students per semester, so we need to design instruction that can be delivered with relatively few faculty resources and a large number of Teaching Assistants.
Our desired outcome is for students to learn the underlying computing concepts and principles so they can use computers to solve problems from a variety of disciplines. Furthermore, students should have sufficient ability and confidence to be able to learn to use new software to solve new problems on their own. This means that the instruction and assessment must focus on genuine problem-solving tasks , rather than static measures of component facts, such as those measured by multiple choice exams.
In addition, concern for student motivation is important when designing instruction from which we expect a high degree of retention and transfer, as specified in our desired outcome. Keller notes four dimensions of motivation: (1) interest, whether the learners’ curiosity is aroused; (2) relevance, whether the learner perceives the instruction meets personal needs or goals; (3) expectancy, the degree to which the learner’s perceived likelihood of success is under his or her control; and (4) satisfaction, the learner’s intrinsic motivations and reactions to extrinsic rewards. Similarly, Yelon notes that students are best motivated to learn something new when they see it as relevant .
In courses for computer science majors, we often presume a high degree of interest, relevance and sense of control. However, for the highly diverse non-computer science majors, it is more difficult to stimulate interest among the large numbers of students who do not understand the relevance of what many perceive to be a "requirement." Many non-computer science majors have enormous anxiety about using computers and fear that this is an area in which they have little control . When these motivational factors are combined with traditional assessment measures such as quizzes and multiple choice exams, students primary motivation often becomes a quest for the extrinsic rewards of "points" in an effort to "beat the curve." Under these conditions, there is little retention and even less transfer to solving new problems.
This theory addresses these motivational concerns. First, the course will offer several different "tracks," with each track having focal problems from a variety of domains to pique student curiosity (see Table 1). Thus, students who are interested in fields in which data analysis is crucial may take the track that concentrates on problems requiring the collection and analysis of data. Students interested in disciplines that place greater emphasis on research and writing may take a track that concentrates on searching databases and preparing reports and presentations. Students can select the track most appropriate to their interests or major, thus increasing their perceptions of the relevance of the materials. Regardless of the focal problems in the track, the purpose of each track will be to help students learn the underlying computing concepts and principles that are common to all computers and software, such as data representation and manipulation.
Second, the collaborative learning model of instruction specifically addresses student expectancy of success. Exercises are designed to help students help each other succeed and engender feelings of competence, rather than to stratify and categorize students in a competitive manner . Finally, the assessments are designed to be authentic tasks that require the students to apply their knowledge to solve problems similar to those they will encounter in courses in their majors or in the workplace .
In most introductory computer science courses for CS majors, students learn about computing from the "bottom up" by learning to design and write computer programs. In contrast, this course content is organized to introduce students to computing concepts by having them solve a series of problems that epitomize classes of problems for which various computing skills are the solutions . Yet, for students to be able to adapt to new computing systems and solve new problems, they must have a deeper understanding than is generally acquired by being trained to use specific software packages. The goal is for the students to construct mental models or schemata of the computing systems they are using without learning the minutiae of computer programming. To accomplish this goal, the course will follow a spiral curriculum in which students are presented with increasingly more challenging problems to solve using a variety of software packages. In the process of doing so, they will work downward from the procedural skill towards a conceptual understanding, rather than trying to first learn decontextualized concepts and then attempting to use those concepts to solve problems.
As Tennyson and Cocchiarella note, people learn concepts "as contextual entities (correlational structures), with common attributes that are the most typical, or average, members of a class" . As students grapple with apparently unrelated problems, the instructor must tie them together, showing how they are each examples of particular concepts or principles. As students are presented with subsequent problems that have these concepts and principles behind them, the instruction must relate the new problems to the previously learned concepts and principles. Students can thereby "triangulate" on these concepts and principles, refining their schemata as they solve successively more abstract problems. This pedagogy is consistent with a constructivist perspective on how novice knowledge evolves into expert knowledge .
Visual representation of the theory
Figure 1 represents the process. Students learn Skill 1, represented by the smaller circle in the figure, and build a schema representing a generalization of this particular skill, represented by the larger circle, Schema 1. This schema may consist of accurate representations of the computing concepts along with misconceptions. Students then learn Skill 2 and build Schema 2, representing the generalization of that skill. This schema may also contain some accurate and inaccurate concepts. As the students learn additional skills and build additional schemata (represented by Skill 3 and Schema 3) the intersections of these schemata triangulate on more accurate representations of the concept that is common to the set of skills. Moreover, understanding the concept underlying the increasing number of skills reduces the cognitive load required to represent the knowledge compared with the requirements of storing an increasing number of discrete, unrelated skills . Eventually, the students’ conceptual understanding will become rich enough to support independent problem solving beyond that which is possible with only procedural skill. Although this approach entails additional instructional expense over a more behavioral model , the goals of retention and transfer justify the additional effort,
To operationalize these goals, we have incorporated two important instructional methods in the design of CPS 101:
- active, collaborative learning as a key instructional component and
- a modified mastery model for assessment.
Active, Collaborative Learning
Virtually all of the educational psychology literature on learning in recent years acknowledges that learners construct their knowledge by interacting with their environment and other people. There are a number of constructivist schools of thought. Some focus primarily on the individual learner . Others focus primarily on the social nature of knowledge construction . In either case, the consensus is that education is not the mere transmission of knowledge from the teacher to the student but requires that students be active. Furthermore, collaborative learning is becoming a key component in many college classrooms with several benefits :
- Improved achievement
- Enhanced critical thinking competencies
- Improved attitudes towards the subject area
- Reduced student anxiety
Students will spend the majority of class time in CPS 101 solving focal problems that require them to learn skills that are applications of the concepts from the readings. The role of the Teaching Assistant (TA) is not lecturer, but rather facilitator. Thus, in addition to designing the required classroom activities, a primary role for the faculty in this model is the ongoing training of the Teaching Assistant staff in this new role.
Each class period consists of a series of focal problems, the solution of which requires students to learn and practice new skills. Each problem should build on previous skills and concepts, extending the range of the students’ capabilities. Generally, students will spend some time discussing possible solutions to the problem in small groups and then attempt to solve the problem in pairs or individually. They will then come back to their small groups to compare their results and discuss problems they encountered and solutions they discovered. These small group discussions should then be followed by a whole class debriefing where each group reports on their solutions and problems. The TA facilitates this discussion and helps the students reflect on their schemata and how their schemata relate to the underlying concepts.
Assessments must be consistent with the instructional objectives and require students to demonstrate competence in a variety of situations . Because of a lifelong experience in school testing environments, students are often preoccupied with the extrinsic reward of the grade. We must consider both of these factors when designing instruction. Mastery learning is one successful strategy for coupling assessment with desired outcomes. In traditional mastery learning, students continue to work on the course materials until they demonstrate mastery of specified materials at the desired level. They do not take a fixed set of examinations in order to receive a grade on the basis of single-attempt assessments . Instead of the instructor setting the pace, mastery learning can accommodate individual student variation, . However, in a large university curriculum, students are expected to complete courses within a single semester, so there is usually little opportunity to use true mastery learning.
To accommodate individual student differences, keep assessments consistent with our goals of encouraging student problem-solving, and work within the institutional constraints of a fixed-credit semester, this instructional theory uses a modified mastery assessment model. In this model, the curriculum progresses at the pace specified in the instructional design. At regularly scheduled intervals, students take a "bridge" task that requires them to synthesize the concepts and competencies to that point in the course. The students must use their homework, in-class assignments and materials provided to them as part of the bridge task to solve the problem.
A key factor in the bridge tasks is that each one will contain one or more "extension tasks." These are tasks that require students to apply the concepts and principles they have learned to solve new problems they have not previously encountered. Thus, while learning a set of skills for using particular software may help the students complete routine tasks, they must also understand the underlying concepts or principles to complete these extension tasks.
Bridge tasks are evaluated on a mastery level pass/fail basis. The bridge tasks are individualized by defining a number of dimensions that encompass the concepts and competencies for that task. The specifics for each students’ bridge task are selected from among a number of possibilities in each dimension. The pass/fail criteria for each bridge task are determined by combining the pass/fail evaluations for the separate dimensions of the bridge task. Some dimensions may be required and others may be optional. For example, a bridge task may have six dimensions, two of which are required, that is the student must pass both of them. The other four dimensions may be optional and require that the student pass any three of the four.
If a student demonstrates sufficient mastery on the first bridge task, he or she will receive a grade of 1.0 in the course. If a student fails a bridge task, he or she continues in the class. However, the student must repeat the failed bridge task until he or she has successfully passed it before being allowed to take subsequent bridge tasks. For each subsequent bridge task passed, the student’s course grade is incremented by 0.5 until she or he has passed the 3.0 bridge task. Once the student passes the 3.0 bridge, he or she completes an integrative semester project that may increase the course grade to 3.5 or 4.0.
There are several advantages to this assessment model.
- The assessment provides greater opportunity for learning than traditional multiple choice examinations. The student’s extrinsic motivation shifts from accumulating points, to mastering the concepts so they can complete tasks similar to those they will encounter in their subsequent courses and after graduation.
- The course grade actually indicates that the student has mastered the concepts and competencies up to that point in the course. By contrast, a grade of 2.0 often means that a student has accumulated enough points from a variety of exams and homework assignments to receive a 2.0. It does not indicate what knowledge the student has. In this model the course grade will now indicate what concepts and competencies a student actually has mastered.
- While students complete the bridge tasks individually, they are not evaluated on a competitive basis; students' grades are not dependent on doing better or worse than their peers. This is an absolute requirement of a collaborative learning model that fosters cooperation and peer-teaching .
The observational checklist may be used to verify that the instruction is consistent with the theory.
This theory covers both an entire course and the individual lessons within the course. The following examples show first the structure of the entire course and then an example structure for a single lesson.
Table 1 shows the topics by class day for each of the tracks. Note that the first 17 days of each track represent a set of common skills and competencies identified by all client departments as essential for their students. These include basic knowledge of operating systems, networks, files, electronic mail, searching databases and word processing.
After students have successfully completed the 2.0 bridge task on day 17, the tracks diverge to concentrate on different types of software that are appropriate for different majors. For example, in the General track, students learn additional word processing operations, some basic spreadsheet skills, and presentation software. In the Data Analysis track, students focus in greater depth on spreadsheets, with a focus on data collection and analysis. In the Financial Analysis track, students also learn more about spreadsheets but the focus is on financial modeling rather than data analysis.
While the particular skills vary from track to track, the underlying computing concepts are the same. Hence, while the particular software may be different, abstractions such as data representation and manipulation are at the core of the more advanced topics in each track. In word processing, these are exemplified by text formatting and style templates while in spreadsheets functions and graphing have data representation as underlying concepts.
In each track, after students complete the 3.0 bridge task, they may select a project from a range that represent authentic problems in a variety of disciplines. These projects will be provided by faculty in various client departments as representative of the problems that students will encounter in their major courses. While each project may come from a different domain, they will each require that the students apply their skills with several software packages and also demonstrate their abilities to learn new software to solve a particular problem. For example, in the data analysis track, students will have to collect data, use spreadsheets to transform and consolidate the data and then use statistical software to perform analyses they have not previously learned to perform. Finally, they will have to merge the results from these packages into a report using word processing software.
Each class is held in a microcomputer lab and lasts 110 minutes. Most of the class time is devoted to a series of exercises that students complete in small groups, pairs, or alone using the computers. Table 2 is a sample lesson plan from class day 9 in which the students learn to search a variety of bibliographic and reference databases using a client program called WinSpirs. The readings prior to class introduce the concepts of databases and Boolean operators for constructing searches.
Each day’s lesson plan consists of four columns and several rows. Each row in the lesson plan is a single exercise. Generally, exercises run between 10 and 30 minutes, with most averaging 20 minutes. The first column is a brief description of each exercise the students will do. The second column contains notes for the TA to help her/him organize and complete the activity successfully. Generally, the primary role of the TA is to outline each exercise and then move among the student groups while they are completing the exercise, facilitating their activities. The third column is the time allocated for the exercise, indicated as both running class time and time for each exercise. The fourth column contains notes for the Assistant TA. The assistant TA is an apprentice who is learning to facilitate the exercises in preparation for becoming a TA. In addition to this apprenticeship, the primary role of the assistant TA is to monitor class activities and keep the activities on schedule, record the results of each exercise on an evaluation form, and assist the TA with student group facilitation.
Each exercise focuses on solving a constrained problem. Generally, the new skills required for each exercise are logical extensions of previous skills. By having students working in small groups, they can help each other with the procedural aspects of learning the skills which is often their immediate concern. As the TA moves among the groups, s/he helps the students reflect on how these particular skills triangulate on the underlying concepts and asks questions to help the students remain aware of their metacognition while trying to solve the problems. What does this problem have in common with other problems they have seen? How is it different?
As students from a variety of disciplines prepare to be the "information workers" of tomorrow, they must be able to use a variety of rapidly changing computing systems and tools to solve an ever-expanding range of problems across disciplines. Rapid advances in application software has eliminated the need to learn computer programming in order to use computers. However, literate computer uses must still have a deep understanding of computing concepts and principles if they are to adapt to constantly changing computing systems and use them to solve new problems in new ways. Instruction designed for this purpose must motivate the students’ interest, and engage them in authentic tasks so that they may construct the rich schemata of experts. By using a spiral curriculum of focal problems, this instructional theory helps students triangulate their schemata on the underlying computing concepts and principles without learning computer programming.
ACM/IEEE-CS Joint Curriculum Task Force. (1991). Computing Curricula : ACM/IEEE.
Barrett, M. (1996). Emphasizing design in CS1. SIGCSE Bulletin, 28(1), 315-318.
Beilin, H. (1985). Dispensible and nondispensible elements in Piaget's theory. In J. Montangero (Ed.), Genetic epistemology: Yesterday and today. (pp. 107-125). NY: City University of New York.
Berliner, D. C. (1992, August, 1992). The Science of Psychology and the Practice of Schooling: The One Hundred Year Journey of Educational Psychology from Interest, to Disdain, to respect for Practice. Paper presented at the American Psychological Association, Washington, D.C.
Block, J. H., Efthim, H. E., & Burns, R. B. (1989). Building effective mastery learning schools. New York: Longman.
Bruner, J. S. (1960). The process of education. New York: Vintage Books.
Cahan, E. D. (1992). John Dewey and human development. Developmental Psychology, 28(2), 205-214.
Chandler, P., & Sweller, J. (1991). Cognitive load theory and the format of instruction. Cognition and Instruction, 8(4), 292-332.
Foshay, R. (1991). Sharpen up your schemata. Data Trainig(May), 18-25.
Gergen, K. J. (1994). Social construction and the educational process. In L. P. Steffe & J. Gale (Eds.), Constructivism in Education (pp. 17-39). Hillsdale, NJ: Lawrence Erlbaum Associates.
Greeno, J. G., Collins, A. M., & Resnick, L. B. (1996). Cognition and Learning. In D. Berliner & R. Calfee (Eds.), Handbook of educational psychology (pp. 15-46). New York: Macmillan.
Johnson, D. W., Johnson, R. T., & Smith, K. A. (1991). Active learning: Cooperation in the college classroom. Edina, MN: Interaction Book Company.
Keller, J. M. (1983). Motivational design of instruction. In C. M. Reigeluth (Ed.), Instructional-design theories and models: An overview of their Current Status (pp. 383-434). Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Lee, J. F., & Pruitt, K. W. (1984). Providing for individual differences in student learning: a mastery learning approach. Springfield, Ill.: C. C. Thomas.
Levine, D. U. (1985). Improving student achievement through mastery learning programs. San Francisco: Jossey-Bass.
McInerney, V., McInerney, D. M., & Sinclair, K. E. (1994). Student teachers, computer anxiety and computer experience. Journal of Educational Computer Research, 11(1), 27-50.
Merrill, M. D. (1983). Component display theory. In C. M. Reigeluth (Ed.), Instructional-design theories and models: An overview of their Current Status (pp. 279-333). Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Reigeluth, C. M., & Stein, F. S. (1983). The elaboration theory of instruction. In C. M. Reigeluth (Ed.), Instructional-design theories and models: An overview of their Current Status (pp. 338-381). Hillsdale, New Jersey: Lawrence Erlbaum Associates.
Rogoff, B. (1994). Developing understanding of the idea of communities of learners. Mind, Culture, and Activity, 1(4), 209-229.
Smith, J. P., III, diSessa, A. A., & Roschelle, J. (1993). Misconceptions reconceived: A constructivist analysis of knowledge in transition. The Journal of the Learning Sciences, 3(2), 115-163.
St. Julien, J. A. (1994, April 4-8, 1994). Social constructivism: Vygotskian activity and connectionist representation. Paper presented at the Annual meeting of the American Educational Research Association, New Orleans.
Tennyson, R. D., & Cocchiarella, M. J. (1986). An empirically based instructional design theory for teaching concepts. Review of Educational Research, 56(1), 40-71.
Vygotsky, L. (1978). Mind in Society: The development of higher psychological processes. Cambridge, MA: Harvard University Press.
Weinshank, D. J., Urban-Lurain, M., & Danieli, T. (1990). Introduction to computing: Telecourse with Waterloo BASIC. (2nd ed.). Dubuque, Iowa: Kendall/Hunt Publishing Company.
Weinshank, D. J., Urban-Lurain, M., Danieli, T., & McCuaig, G. (1995). Integrated Introduction to Computing. (updated first edition, revised and enlarged ed.). Dubuque, Iowa: Kendall/Hunt Publishing Company.
Yelon, S. L. (1996). Powerful Principles of Instruction. White Plains, NY: Longman.