Two Measures of Self-paced Instruction

Wallace S. Venable - Instructor, Theoretical and Applied Mechanics

Helen L. Plants - Professor, Theoretical and Applied Mechanics

American Society for Engineering Education - 1972 Annual Conference


An experimental comparison of self-paced instruction and group-paced instruction was made in which students under both methods used identical study materials. Both groups used the same programmed instruction texts, took similar quizzes and examinations, and had the same instructors. Differences between the methods in examination performance were insignificant. A comparison of the results of this study with earlier tests of the programmed materials indicates that self-paced instruction is an effective teaching method but it is not uniquely effective.


The relative performance of educational systems is of particular importance in any situation in which a decision has to be made regarding the adoption of a different method of instruction. Such performance may be assessed on the basis of cost, achievement, student or faculty attitude and many other factors. Assessments of any of these must be as objective as possible since objective measurements make it feasible to establish relationships which lead in turn to ability to predict effects of smaller changes.

Since it is impossible for every institution and individual to try every possible method, comparisons must be undertaken in such a way that the results obtained are objective rather than subjective in nature and are therefore applicable to a variety of situations.

It was in this spirit of establishing valid basic comparisons that the study reported herein was undertaken.

It reports the results of an experimental trial of Self-Pace Instruction (SPI). This study has involved both mathematical and non-mathematical evaluation. The relative performance of the classes conducted in the self-paced format was compared with the performance of classes-conducted in a lecture mode and with classes which used programmed instruction in a group-paced system.

The study was directly concerned with achievement only. The data gathered support the contention that the use of individually paced instruction will result in higher overall achievement than the use of the ''traditional lecture" mode of instruction. This study does not support any claims that SPI results in uniquely high achievement, but rather it is concluded that self-paced instruction will produce achievement comparable to that produced by several other methods of organization which are also based on behavioral psychology.

Concern with relative performance is refuted to use of educational materials as consumers. Educational designers must also be concerned with internal relationships since knowledge of the relationships between variables within classes may eventually permit educational specialists to anticipate the effect of relatively minor changes in organization.

Two aspects of the mastery concept, that is the practice of requiring students to repeat mastery or formative tests until a pre-set level of performance is reached, were examined in this study. Both sets of results lend support to the prediction that repetitive quizzing results in greater long term achievement.

The Experiment

This study of SPI is not the first experiment with innovative educational techniques which the engineering mechanics faculty at West Virginia University has attempted. For the past ten years the department has experimented with original text books, programmed instruction, instructional administration and demonstrations. From these experiments two formats have evolved. Both formats are popular with individual students and faculty members and both are offered concurrently. One is the traditional "lecture-homework-examination" format using popular texts. The other format is based on the use of programmed instruction, formative tests and interactive class sessions.

We call the second format Elastically Structured Teaching (EST). Elastically Structured Teaching is a system that provides a great deal of elasticity for the individual within a rather rigidly structured course. Under this system students are required to demonstrate their level of accomplishment each time they complete a two to four hour study unit. Each study unit is a programmed text section which integrates new information, instruction, and practice in solving textbook problems. During class periods the student is generally expected to be tested on the new material and will then have an opportunity to discuss the test and the material with his classmates and instructor. The instructor may use any remaining time to show the class his personal tricks and applications for the concept. Table top demonstrations are often used as discussion starters.

Grades are based in large part on traditional style examinations, but the frequent graded quizzes and discussions keep each student informed of his mastery of each concept so that he may review knowledgeably for the more critical examinations,

While classes and quizzes are held on a regularly scheduled basis, attendance is not required. Those students who fall behind may take a makeup quiz in a Self Study center. Students are expected to complete the material before the end of the course, and nearly all of them do. Those who need extra help are encouraged to visit the instructor in his office.

The instructors accept any excuse for a temporary delay in completing a study unit, but no-one is excused from completing all work. The individual student seems to respond to this by setting up his own standard of permissible excuses for postponing an assignment.

Our search for improvement in our teaching methods leads us to consider a third alternative, self-paced instruction. The idea of individually paced instruction with the teacher playing the role of consultant had great appeal to us. Consequently, we set up a system conducted much like the Keller System of PSI. 1 We included hour exams and a final examination in addition to mastery quizzes, so that the final model tested most closely approximates that set up by Bloom.2

The SPI classes were conducted using the same programmed instruction as the EST classes. The same number and type of post-tests and examinations were used with questions drawn from the same pool of test items.

At the beginning of the semester the students in the SPI classes were given written descriptions of the rules of conduct for the course and the schedule which had been followed by classes conducted in the standard method. The SPI students were told that the written schedule was an example of one workable plan for completing the course within a semester but that each would be permitted to work at his own pace, whether faster or slower than the published version.

Students were told that they were not required to attend classes. If they chose to attend class, the instructor would have copies of post-tests available for them to take and would discuss any of the material with them. If a student wished to take a post-test outside of the class hour, he could report to a room near the instructor's office where a staff member would give him the desired post-test.

Students were required to complete the units in the order given in the schedule. Post-tests were graded on a mastery basis and a unit was considered to be complete when an appropriate post-test was completed without a serious mistake. Students were required to be present while their post-tests were graded. Usually the instructor or an undergraduate assistant graded the students' work within five or ten minutes of the time he completed it.

Students were required to take hour examinations at four points in the course. These points were specified in terms of units to be completed before a student was to be permitted to take an exam rather than in terms of an examination date. Examinations generally were administered to students individually, although examinations were also administered to small groups in class on the dates listed on the suggested schedule.

Since all the students were required to attain "perfect" post-test scores to complete the course, post-tests were required for passing the course but were not "averaged" in determining the course grade. Course grade was based on examination scores alone.

Students were told that if they had completed more than half of the course work before the last week of the semester they could be granted a grade of "Incomplete" and permitted additional time to complete the material, if they requested it. Students were informed that their rate of completion of the material would not be considered as a factor in establishing grades as long as they fulfilled the minimum requirements. As in the EST classes, all examinations, grades, and post-tests standards were criterion referenced.

The similarities and differences between the EST and SPI classes are summarized in Table 1.

Table 1. Summary of Differences Between Methods of Instruction

Feature Elastically Structured Teaching


Self-Paced Instruction


Course pace Teacher specified Student evolved
Examination dates Pre-set by teacher When student ready
Class attendance Limited absences No requirements
When given Scheduled date or on request On request
Sequence Not specified In listed order only
Grade format Percent correct "Mastery" or "Not completed"
Conditions of grading In batch in office Singly with student
Weight Counted toward better grade Required to pass
Grade announced Within two or three days Within 15 minutes
Classroom activity    
Post-Tests Yes Yes
Teacher instigated discussion Yes No
Student initiated discussion Yes Yes
Limited to scheduled topic Yes No
"Incomplete" grade Special circumstances 50% completion plus request
Examination grades Criterion referenced Criterion referenced

Results and Conclusions

Several hypotheses about the behavior of students in the self-paced classes were tested, as well as hypotheses about the relative performance of SPI and EST. The questions about behavior within the self-paced classes will be discussed first.

One basic assumption of mastery based instructional systems is that early or repeated failure does not indicate that a student is unable to master a particular piece of material but only measures his rate of learning. A test of this assumption was made in this study by computing the correlation coefficient between the number of attempts which students made on mastery quizzes and their examination scores. We found a highly significant correlation on a short term basis (in Statics r = .464, n = 22, in Dynamics r = .535, n = 22). Quick learners had higher hour examination scores. However, on a longer range basis, as measured by the correlation of quiz attempts with final examination score, there was no significant relationship (in Statics r = .246, in Dynamics r = .225). We did not find that all students did well, merely that their rate of learning was not significantly related to final achievement.

Another basic assumption in the use of mastery quizzes is that students will learn between retests and/or that they will learn as a result of the testing. This assumption was tested by comparing the percentage of quizzes which students in the SPI group passed with the percentage on which the control groups had scored eighty seven percent or higher. These comparisons are summarized in Table 2. We found that the SPI group required significantly fewer quiz attempts than would have been predicted from the EST group behavior. We therefore concluded that the mastery quiz strategy is an effective way of increasing short range learning.

Table 2. Confidence Interval Predictions of Number of Quizzes Passed and Corresponding Observations

Course CI95 CI99 Observed Value Sample Size
Statics 580-588 521-597 660 917
Dynamics 582-644 654-672 690 1021

On the basis of the above results we have concluded that self-paced instruction is an effective teaching system and that many of the operating principles work in the manner which educational psychologists have predicted.

We next turn our attention to the relative performance of the classes taught with Self-Paced Instruction and Elastically Structured Teaching.

The first comparative finding of this study is that there was no significant difference in examination scores between the methods.

Not only were the averages of the examination scores essentially the same within each of the courses, but the two methods also had similar ranges and variances. These differences are summarized in Table 3.

Table 3. Summary of Differences Between Methods in Examination Scores

Course Method Mean Standard Deviation Number of Students
Statics EST 81.6 7.5 71
  SPI 80.6 7.4 21
Dynamics EST 79.8 9.5 49
  SPI 76.7 9.2 21

We therefore conclude that although some students may benefit greatly from self pacing, overall class performance will not be greatly affected by this factor. It further seems that the repetition of quizzes until the "mastery" level is reached does not result in significantly greater command of the material than can be achieved with lower standards plus early knowledge of results of all attempts.

The second comparative finding is that the percentage of students successfully completing the EST courses was significantly higher than the percentage of successful completions in the SPI classes (Table 4). This is true even after including those students who took up to one year to complete their work. Many of the students who did not successfully complete their course in the self-paced sections had been doing work of acceptable quality before they decided to withdraw from the course. The larger numbers of students withdrawing from the self-paced classes may have been due to their reactions to the instructional method used, however the effect also may be related to changes in college policy on the approval of withdrawals which have been made over the past few years.

Table 4. Summary of Differences Between Methods in Completion Rate

Course Method % Completion

One Semester

% Completion

Two Semesters

Statics EST 81 81
  SPI 52 74
Dynamics EST 81 81
  SPI 50 66

We found the self-paced classes harder to conduct than our method of teaching with the programs. The instructors and the undergraduate assistants all felt that they were under considerably greater pressure with SPI than with EST. It is our impression that we received much less favorable student response from our self-paced classes. We therefore have returned to the use of Elastically Structured Teaching in our courses.

However, we have found on the basis of departmental examinations, that cognitive performance of students in EST classes is higher than that of students in standard lecture classes as shown in Figure 1. Since we found no significant difference between the cognitive learning of students in the EST classes and that of those in the SP! classes we conclude that both methods are superior to the' traditional approach insofar as results may be measured by final examinations.

Figure 1. Mean scores of classes on departmental final examination in Dynamics.

Summary Conclusion

We have found that SPI is an effective educational method, but we conclude that it is not demons/ratably more effective than several others, including EST.

Because we found two methods which used the same material equally effective we feel that the superior performance of both may be due more to the quality of the materials than to the qualities of the systems.


1. Keller, F. S., "Goodbye, Teacher...," J. Applied Behavior Analysis v. I, n. 1, 1968, pp. 79-89.

2. Bloom, B.S., "Mastery Learning," in Block, Ed. Mastery Learning Theory and Practice, Holt, Rinehart and Winston, New York, 1971, 152 p.