Online Assignments in Economics: A Test of Their Effectiveness

Abstract

This article compares the effectiveness of online and paper-based assignments and tutorials using summative assessment results. All of the students in a large managerial economics course at National University of Ireland, Galway were asked to do six assignments online using Aplia and to do two on paper. The authors examined whether a student's performance on a particular section of the exam is affected (1) by how he or she performed on the corresponding assignment and (2) by whether the student completed the corresponding assignment on paper or online. Our results provide little evidence that a student's performance on an assignment helps him or her perform better on the corresponding section of the exam. We also found little evidence that the way in which one completes an assignment—on paper or online—has an effect on how one performs on a particular section of the exam.

Keywords:

• assessment
• learning
• online assignments
• paper assignments

JEL codes:

• A20
• A22

Acknowledgments

The authors received generous support from the National Academy for Integration of Research & Teaching & Learning (NAIRTL) at University College Cork, from the Aplia Corporation, and from the Cairnes School of Business and Economics at NUI Galway. The authors thank three anonymous referees for very helpful comments on this article. Drafts of the article were presented at the Irish Economics Association annual meeting in Cork, at the annual meeting of the Eastern Economic Association in Philadelphia, at a workshop on Teaching and Learning in Economics at NUI Galway, at the Development in Economics Education conference in Cardiff, at the annual conference of NAIRTL in Dublin, and at seminars in NUI Galway. The authors thank the participants at these events for helpful suggestions.

Notes

1. The eight sections of the exam covered the following topics: Individual and Market Demand; Applications of Consumer Choice Theory; Production; Costs; Perfect Competition; Monopoly and Pricing Strategies; Monopolistic Competition and Oligopoly; and Game Theory and the Economics of Information.

2. The main differences between universities in Ireland and in the United States are (a) that students in Ireland apply for particular degree programs before they enter university rather than declaring a major after they have been in university for a certain period of time and (b) that students in Ireland progress from one year of a program to another rather than accumulate credits. Thus, students must pass all of the courses they take in first year before they can proceed to the second year of their program.

3. Formal midterm examinations are rarely given in Irish universities. Semesters at NUI Galway run for 12 weeks.

4. A student's overall assignment mark was based only on the assignments that corresponded to the sections of the exam that each student attempted.

5. Students must take six subjects in this exam. English, mathematics, and an Irish language course are compulsory for all students. The optional subjects include science, business, continental languages, and history. The leaving certificate exam is much more like the GCSE or A-level examinations in the UK than the SAT examination in the United States.

6. The detailed results of these regressions are available on request.

7. The fixed-effects regression also was run on the larger sample of students who took the course, but the results were very similar to the results presented here on the reduced sample.

8. The detailed results of these regressions are available on request.

9. One limitation of the fixed-effects model is that it forces us to exclude variables that do not vary within the individual such as our variables on CAO points and the student's first-year results. We also estimated a random-effects model that allowed us to include student characteristics such as CAO points and first-year results. The random-effects model assumes that none of the explanatory variables are correlated with the random effects. This is a strong assumption in our model because it is quite possible that a variable such as assignment score will be affected by an unobserved characteristic such as whether a student has cheated. The random-effects model as a whole is rejected by the Hausman test, so we have not reported the results.