Subhotosh Khan, Research Assistant, Mechanical Engineering and Mechanics
Russell K. Dean, Assist. Prof., Mechanical Engineering and Mechanics
West Virginia University
1980 ASEE Annual Conference Proceedings
As professional engineers, our students are expected to perform as expert problem solvers after graduation. To attain such a status, one must acquire: (1) an ability to teach himself an idea or theory which was never presented to him before and (2) the discipline and self-confidence required to solve the problem after such self-study.
There have been several methods of individualized instruction (Keller Plan, Guided Design, etc.) which have been devised to allow students to achieve these capabilities. Not only do these methods help students develop as engineers, they are designed to have the flexibility to accommodate various rearming styles. Elastically Structured Teaching (EST) is one such system which has been developed at West Virginia University. This study looks at various ways that EST has been employed at West Virginia University.
Different studies conducted at West Virginia University 1 show that sections studying under programmed instruction (PI) in Elastically Structured Teaching (EST), perform significantly better on the departmental final than the sections studying under lecture instruction. It is also a well known fact that writing and reproducing programmed instructions is a costly and time consuming affair. Study guides, administered in EST, cost much less. The problem that will be studied in this paper is the question of whether study guides are as effective as PI. If the study guides are less effective, then the question is whether or not they are more effective than lecture.
With this goal in mind, study guides for an introductory statics course were written by Professor Russell K. Dean. The effectiveness test will be made by comparing the common final examination score of Program-EST, Study guide-EST and lecture sections. To maintain uniformity, the common final is given on the same night to all the sections and is prepared by instructors who are not teaching the course. One instructor grades one problem for all the sections. The one-tail t-test will be used as the method of testing the hypothesis. The data will be tested for normality by the Chi-Square goodness-of-fit test. The GPA of the students in the various sections will be used to verify the homogeneity of the data.
EST is an instructor paced system in which
there is a minimum rate at which the student muse progress
through the course. On the first day of the class, a handout is
distributed in the class which describes how the course works and
it also includes a schedule. The schedule informs the student
when a quiz will be administered in the class. The student must
take the quiz on or before the date that the quiz is administered
in the class.
The total number of quizzes a student must take is about 30. Students are allowed to take any five quizzes late, without any explanation. Any more than five late quizzes receive a grade zero as its first attempt. In the event that a student scores less than 70 percent on a quiz, he is encouraged to retake a different version of the quiz on the same material. The two grades are averaged to determine the grade for the unit.
In a typical day in the classroom, the first 15-20 minutes are spent on administering the quiz. After collecting the papers, the instructor solves the problem for the class. The instructor stresses the interpretation, diagnosis, and strategy of solving the problem and tries to invoke responses from the class while going through the problem solving procedure. The rest of the time is used for answering specific questions from the student.
On some days (about one out of four class days), practice problem sessions are held. On these days, problems which require a combined knowledge of the previous units for their solution are brought to the class. Students are encouraged to work in groups and the instructor acts like a consultant instead of a tutor. Near the end of the class period, the students put the solutions to the problems on the blackboard. As an incentive, credit points are occasionally given to the group putting up the solution.
Students are given three one-hour tests at approximately equal intervals. At he end of the semester, they have a chance to retake one of the teats. The highest grade obtained on these two attempts is considered as the grade for the particular test. Students are required to take the common departmental final. The final grade of the students is computed by summing 1/3 of his quiz average, 1/3 of his hour test average, and 1/3 of his final examination.
The EST system places heavy emphasis on rapid
feedback to the student. With the student's permission, quiz
grades are posted on a bulletin board in the SELF-Study
Laboratory. There are files in the SELF-Study Laboratory. one for
each student, where his graded and corrected quizzes are kept for
SELF-Study Laboratory is a room where the students primarily retake quizzes. There are usually two graders in the laboratory when it is open. The graders are responsible for administering the retake quizzes, late quizzes and grading the quizzes. They sometimes also act as tutors.
Students are encouraged to come to the SELF Study Laboratory and review their graded quizzes. They are also encouraged to consult a faculty or grader to straighten out any difficulties they are having with the course. However, a student must make an attempt at the quiz before he can receive any help on that unit. This is done to discourage his dependence on the instructor and encourage his own learning discipline.
The EST system emphasizes written instructional material and it uses proctors to grade quizzes in a manner similar to the PSI system (Keller Plan). Unlike PSI system, the EST system does not require a mastery level and it is instructor paced as opposed to being self-paced. There is no motivational lecture as in PSI. Instead, there is instructor modeling in almost all the class meetings.
As an alternative to PI, study guides have been developed to be used in conjunction with the textbooks 3 used in the lecture sections of the introductory statics course. Each study guide is approximately eight pages long and contains the following:
The study guides follow approximately the same order of presentation as the PI.
A validation scheme was generated which was used to determine if a study guide taught the material which it was intended to teach. A study guide was accepted as viable when at least 85% of the students in the class obtained a grade of 85% or higher on the quiz. Otherwise the study guide was rejected. The rejected study guides were reviewed to find the difficulties and were then rewritten.
The average time taken to write a study guide was about three hours. Master preparation and proofreading took about one hour. One more hour was expended for duplication and collation, requiring five hours total to complete a study guide for a unit.
The other media which is used in the EST system is the programmed instruction book. 4 Here active participation on the student's part is demanded throughout the learning process. All the units are self-con aired. That is, no other book is required.
In each unit, the student learns a problem solving procedure which is broken into several parts. Thus, students learn the whole procedure by learning each step individually. Then in combination with each previously learned step, a step is repeated several times to aid reinforcement. Throughout the programmed instruction, the students are asked to respond to different questions which form the steps to the solution of the part of the problem. The answers to each step are given on the next page. The step size is small enough that the student is successful about 90% of the time. The questions forming a step can be a simple problem, a multiple choice question, a true-false question, etc. Each unit contains about three complete solved problems and several parts of the problem.
At different stages throughout a unit, the student is referred to the notebook. It acts as a reference book for the program and is partially completed by the students. The theoretical parts (derivation of formulae, etc.) of the subject are discussed in the notebook in a comprehensive and concise manner. Students are also asked to solve several complete problems in the notebook. The answers and key steps to the solution of the problems are given in the main textbook.
In programmed instruction, the students are taught to solve a complex problem piece by piece. However, in study guides, only complete problems are solved. Each strategic decision and diagnostic decision is discussed explicitly when the problem solving procedure is explained to the student. In the study guides, the student is not asked to react actively in the learning process through response and feedback. Only at the time of quizzes and tests is the student in study guide sections asked to respond actively.
The primary features of EST which are common factors among PI and study guide sections and which differentiate it from most common lecture courses include the following:
Similarly, there are two principal elements which are common to the media of the study guides and PI. These include:
As indicated earlier, the effectiveness of different media for teaching was tested through the performance of the students on the common final examination. The data were collected for two semesters. In order to incorporate both semesters' data, the second semester (spring) scores were transformed through the z-score, before being lumped with the first semester (fall) scores. The transformation was done in the following way:
This transformation assumes that the two testing procedure (two final examinations) and the population were inherently similar. The data are tabulated in Table I.
Table 1. Lumped Results of the Departmental Final Examinations.
|Method||Group Size||Score||Standard Deviation|
Analysis of this data was carried out by using a one-tail t-test to compare various groups two at a time (i.e., SG and lecture, SG and PI and PI and lecture). It was observed that the null hypothesis of no significant difference among the means of the groups could be rejected at 95% confidence level for all the groups. Thus, it was shown that the sections with programmed instruction faired significantly better on the final examination than the sections with either study guides or lecture instruction. The sections with study guide instructions faired significantly better than the sections with lecture instruction.
A Chi-Square goodness-of-fit test was run on the final examination scores of each group to insure that the data was normally distributed. In each case, the distributions were determined to be normal at an sigma level of .05.
Although the complete data was not available at the time of this writing, initial calculations indicated that there was no significant difference in the GPA's of the three groups when they were entering the statics course. Since there appears to be good correlation between GPA and final examination scores for each group, the GPA is a good predictor for homogeneity of the data. Thus, it will probably be found that there was no significant difference in the populations before they took the statics course. Therefore, it can be concluded that any differences which appeared in the final examination scores is probably a result of the different instructional methods.
The mayor conclusion of this experiment is that the study guides are not able to teach the basic status material as well as PI; however, both are able to teach it better than traditional lecture. This conclusion would indicate that study guides used in an EST format are a viable alternative to the traditional lecture approach. Although PI performed even better than studyguides, it must be remembered that the relative advantages of study guides are that they take less time to write and they are less expensive to reproduce than PI.
There is a secondary conclusion that can be drawn from this study. Venable did an experiment 6 in which he taught the same statics course with PI using two different methods. One was EST and the other was PSI. The results of his experiment showed that there was no significant difference between these two groups on final examination scores. Both groups performed significantly better than the lecture group. Combining these results with the present experiment indicate that instructional media is probably the most important variable in determining the success of an instructional scheme.
There is also an observation that was made in the affective domain. Students' continents indicated that they like the study guide EST more than PI EST. This preference appears to be based on their desire to have a "real" textbook which they can use as a reference later in their professional lives. This may very well be a reasonable thought.
The results of this experiment with study guide EST have been sufficiently satisfactory that the Mechanical Engineering and Mechanics Department at West Virginia University is presently developing study guide EST for a thermodynamics course. It is anticipated that this course will be as successful as the other EST courses which are presently offered in our Department.
Mr. Subhotosh Khan is presently pursuing a Ph.D. in Mechanical Engineering and Mechanics at West Virginia University. He received his B. Tech. degree in Metallurgical Engineering from Indian Institute of Technology, Kharagpur, India, and his M.S. degree in Materials Sciences from Duke University.
He has worked as a practicing mechanical engineer. While attending graduate school, he has been a research assistant associated with different research projects. His basic research interests are in the fields of fracture mechanics and failure analysis. He has also taught basic mechanics courses as a teaching fellow at West Virginia University. and has had extensive experience wit the Elastically Structured Teaching system at West Virginia University, and is presently-involved with extending the system to teaching the thermodynamics service course.
RUSSELL K. DEAN
Russell K. Dean is Assistant Professor of Mechanical Engineering and Mechanics at West Virginia University. His duties consist of teaching graduate and undergraduate mechanics courses and carrying out research in engineering education. His present interests include studying whether method or media is a more significant factor when designing problem- solving courses.
Professor Dean is a member of IEEE, ASEE, Eta Kappa Nu, Tau Beta Pi and Phi Kappa Phi, as well as an associate member of ASME. His education consists of a BSEE (1974), MSME (1976), and Ph.D. (expected summer-l980). All these degrees have been earned at West Virginia University.