Economics Network IREE Virtual Edition

Comparing Student Performance Using Cooperative Learning

Daniel R. Marburger
International Review of Economics Education, volume 4, issue 1 (2005), pp. 46-57
DOI: 10.1016/S1477-3880(15)30138-9 (Note that this link takes you to the Elsevier version of this paper)

Up: HomeLecturer ResourcesIREEVolume 4 Issue 1

JEL Classification: A22


The traditional lecture is the predominant means through which 'principles of economics' classes are taught (Siegfried, et al., 1996). Despite the popularity of lectures, alternative teaching pedagogies that employ active learning have received increasing attention in economics education in recent years. In contrast to passive learning pedagogies such as lectures, active learning requires the student to be actively engaged in the learning process. 'Active learning' is a fairly broad concept and might include in-class exercises or experiments, writing assignments, or case studies.(note 1)

A subset of active learning is co-operative learning. With co-operative learning, students work on exercises in small groups. The exercises may be brief ('Think, Pair, Share') or the students may be called upon to resolve a fairly complicated exercise. The common bond among the variants is that the students uncover knowledge through small-group interaction rather than by passively listening to lectures.

Curiously, despite the attention paid to co-operative learning, relatively little research has been conducted in economics education to measure its impact on learning. Maier and Keenan (1994) and Watts and Moore (1998) concentrated primarily on inferring student satisfaction. Johnston et al. (2000) found that students who participated in collaborative learning tutorials did not perform better in multiple choice questions and an evaluative essay question than those who did not. Finally, Frank (1997) found that students who had participated in an in-class experiment on the tragedy of the commons were more likely to perform better in multiple choice questions on the subject than students who had been exposed to the material through lectures.

Regarding the impact of co-operative learning on student performance, one possible explanation may lie in the work of Marton and Saljo (1976). They distinguished between surface learning and deep learning. Surface learning refers to students who commit assorted unrelated facts to their short-term memory. They, therefore, are less able to apply theoretical concepts to new contexts. Deep learning implies that students acquire a level of understanding sufficient to apply concepts to different situations. Surface learning may, of course, reflect the demands of the learning task. If the assessment instrument requires surface learning, the students may respond accordingly. In this vein, one may wonder if the dependent variable in the empirical studies measures the impact of co-operative learning on surface learning or deep learning.

The purpose of this paper is to investigate empirically student performance in principles of microeconomics classes taught via co-operative learning versus the traditional lecture. In a fall semester, I taught one cohort of micro principles students as a traditional lecture, while presenting the course content to the other cohort via co-operative learning. A major distinction between this study and previous empirical works is that co-operative learning did not serve as a supplement to the traditional lecture. Rather, co-operative learning exercises essentially replaced the traditional lecture. The evidence reveals that whereas performance on multiple choice exams was fairly comparable, students who were enrolled in the co-operative learning class were better able to apply theory on a project that required a higher level of economic reasoning than those who learned the course content through the lecture.


My variation of co-operative learning included handouts or exercises that unveiled the material I would normally introduce through a lecture. An example appears in Appendix I. In this sample exercise, the students were given information that required them to construct a linear production possibilities curve. Next, they determined the production of goods that would maximize utility. Some of the groups determined the optimal allocation for the Country X whereas the others worked on the solution for Country Y. Having determined the optimal production of goods as self-sufficient countries, the two countries looked for trades that would benefit them under the proviso that they must not agree to a trade that makes them worse off. Their findings set the stage for a subsequent discussion on comparative advantage and gains from trade.

Importantly, exercises such as this one did not follow the lecture. They were substituted for the lecture. This distinguishes the study from Johnston, et al. (2000) and Frank (1997), in which co-operative learning exercises merely supplemented the traditional lecture. In my class, the formal 'lecture' simply consisted of a short (five minutes or less) debriefing. In addition to the debriefing, I circulated around the room as the students worked on an exercise, posing questions to individual students within groups or asking them to explain what they had discovered. Given the number of students in the co-operative learning class, my ability to give each student individual attention was quite limited: most interpersonal interaction during class time was between students.

Although the use of class time clearly differed in style between the two cohorts (i.e. lecture versus group exercises), efforts were made to make the classes as comparable as possible. The duration of each class meeting was the same. Each class meeting concentrated on one or two critical concepts from the corresponding chapter. For all practical purposes, lecture notes from one cohort were simply 'translated' into co-operative learning exercises for the other. Consequently, the students in the co-operative learning cohort gained exposure to the same concepts, examples, and numbers that were presented to the other cohort in lecture form. Moreover, the students in both classes had access to the same ancillaries (i.e. study guide, tutorial software, instructor web page) and both cohorts had identical homework assignments.

In addition to comparing performance on an identical comprehensive multiple choice final exam, I also sought to compare performance on a project that required the students to assimilate and apply their knowledge. As a graded homework assignment, the students were given two one-page magazine articles regarding the passage of Proposition 103 in California in the late 1980s. Proposition 103 mandated a 20% rollback on auto insurance premiums.(note 2)

The readings provided specific examples of auto insurance premium rates that Californians were paying and mentioned that over 400 firms sold auto insurance in the state. The articles also summarised the proponents' and opponents' positions regarding the Proposition. The Proponents alleged that the auto insurers were guilty of price-fixing and 'disguising unconscionable profits'. In contrast, the opponents' asserted that the premiums were driven by rising costs in the industry, and they provided specific examples of percentage increases in various costs of offering insurance.

The students were given a series of questions to assist them in using economic analysis to analyse the case. The questions not only forced them to read the articles carefully, but also to gather the information necessary to apply economic theory. A copy of the questions appears in Appendix II. Each student was required to hand in the answers to the questions as well as a written analysis of the Proposition.

Empirical Results

Prior to comparing exam and assignment performances, the composition of the cohorts was tested to assure comparability. Neither the mean Grade Point Average (GPA) nor the class composition (in terms of the percentage of freshman, sophomores, etc., differed significantly between the two cohorts.(note 3)

The comparisons between the two cohorts are summarized in Table 1. As the table shows, the mean scores were fairly close. The lecture class scored an average of 19.76 out of 30 questions on the test, or 65.9%, whereas the co-operative learning section posted a mean score of 20.76, or 69.2%. Using a proportions test, the differences were not found to be significant at any conventional level of confidence.(note 4)

Some significant differences did exist, however, with regard to the students' analyses of Proposition 103. Although both classes overwhelmingly identified California's auto insurance industry as either perfectly competitive or monopolistically competitive, they differed in their perceptions of the allegations made by the Proposition's proponents. Specifically, 62.5% of the students in the lecture class thought that the 400-plus firms in the industry were price-fixing, compared with only 37% in the co-operative learning class. Similarly, 67% of the students in the lecture class believed these firms were earning 'unconscionably high profits' as compared with 42% in the co-operative learning class. The differences in both of these cases were statistically significant. In the lecture class, 44% of the students supported the Proposition, whereas 30% of the students in the co-operative learning class favoured the law. This difference, however, was not significant.

The empirical evidence may cast some new light on the impact of co-operative learning on deep learning. To the extent that multiple choice exams 'decompose' course content into smaller chunks, one may wonder whether previous empirical tests adequately distinguish between surface learning and deep learning.(note 5) In this case, whereas no difference in multiple choice exam performance was discovered, co-operative learning did appear to affect the students' ability to analyse a case and apply their knowledge.


Co-operative learning is an attractive alternative or supplement to the traditional lecture. To the extent that multiple choice examinations are used to infer student knowledge, co-operative learning did not have an appreciable effect on students' test scores in this study. However, most instructors aspire to have their students 'think like economists'. Deep learning requires students to integrate their knowledge and apply it to comprehensive 'real-world' situations. If the ability to analyse and apply economic theory in the Proposition 103 case study is consistent with deep learning and 'thinking like economists', the results in this study provide evidence that co-operative learning may improve those skills.

TABLE 1: Comparisons Between the Lecture and Co-operative Learning Classes

Lecture Co-operative Significant Learning Difference?
Number of Students: 31 54
Mean GPA: 2.86 3.05 No
Percentage of Juniors and Seniors: 76% 70% No
Multiple Choice Final Exam:
Percentage of correct answers on final exam: 65.9% 69.2% No
Proposition 103 Assignment:
Percentage of class identifying industry as perfectly competitive or monopolistically competitive: 83% 87% No
Percentage of class who believed the firms were price-fixing: 62.5% 37% Yes**
Percentage of class who believed the firms were earning 'unconscionably high profits': 67% 42% Yes*
Percentage of class supporting Proposition 103: 44% 30% No

**Significant at 5% level
* Significant at 10% level

Appendix I

Appendix II


[1] For an excellent list of references on active learning pedagogies, see Becker and Watts, 1995.

[2] Amongst the many references to Proposition 103 available on the web,, provides a useful account. None of the students indicated having any prior awareness of the proposition.

[3] One may not expect differences in the percentages of freshman and sophomores to influence student outcomes. However, junior and senior students may be more intellectually mature than younger students or more savvy in their ability to apply their knowledge.

[4] Multiple choice examinations may be more common in the US than in other countries. Similarly, within the US, they may be more commonplace in large universities than in smaller liberal arts colleges. An advantage in using the multiple choice exam in this study is the lack of subjectivity in comparing student performances across cohorts.

[5] In fact, given that multiple choice examinations constituted the largest proportion of the students'grade, one may wonder, in the context of Marton and Saljo (1976), if multiple choice assessments push the students toward surface learning.


Becker, W. E. and Watts, M. (1995) 'Teaching Tools: Teaching Methods in Undergraduate Economics', Economic Inquiry, October, pp. 692-700.

Frank, B. (1997) 'The Impact of Classroom Experiments on the Learning of Economics: An Empirical Investigation', Economic Inquiry, October, pp. 763-69.

Johnston, C. G., James, R. H., Lye, J. N. and McDonald, I. M. (2000) 'An Evaluation of the Introduction of Collaborative Problem-Solving for Learning Economics' Journal of Economic Education, vol. 31, no.1, pp. 13-29.

Maier, M. H. and Keenan, D. (1994) 'Co-operative Learning in Economics' Economic Inquiry, April, pp. 358-61.

Marton, F. and Saljo, R. (1976) 'On Qualitative Differences in Learning: Outcomes and Process' British Journal of Educational Psychology vol. 46, pp. 4-11.

Watts, M. and Moore, R. L. (1998) 'Teaching Introductory Economics with a Collaborative Learning Lab Component' Journal of Economic Education, vol. 29, no. 4, pp. 321-330.

Siegfried, J. J., Saunders, P., Stinar, E. and Zhang, H. (1996) 'Teaching Tools: How is Introductory Economics Taught in America'? Economic Inquiry, January, pp.182-92.

Contact Details

Daniel R. Marburger
Arkansas State University, Box 4181, State University, AR 72467
Tel: (870) 972-3416

Top | IREE Home | Economics Network | Share this page