Single Session Simulations: The Effectiveness of Short Congressional Simulations in Introductory American Government Classes

ABSTRACT

Simulations are widely viewed to be effective pedagogical tools; however, evidence concerning their effects on student learning has been largely anecdotal or unsystematic in nature. In addition, most of the literature focuses on fairly complex simulations that require more class time and demand more of students than may be possible in many circumstances, particularly in introductory level classes. In the fall of 2004, I conducted a quasi experiment in order to better understand the potential effectiveness of short and simple legislative simulations in introductory level classes. The conditions under which students in both classes were taught were very similar, with the exception of the simulation, which was only used in one of the classes. Results of a pretest indicated that the classes were not significantly different in either their knowledge of legislative processes or a variety of other potentially relevant characteristics. Analysis of a posttest given to both of the classes after the simulation indicated that there was a difference in knowledge of legislative processes, with the class participating in the simulation performing significantly better than the nonsimulation class. In addition, the simulation class scored significantly higher than the nonsimulation class on the exam following the simulation. These results are important for two primary reasons. First, they provide empirical support for claims regarding the effectiveness of simulations. Second, they indicate that even brief simulations can increase student knowledge.

Keywords:

• congress
• legislative
• simulation
• introductory-level

Notes

*Difference of means significant at .01 level.

*p < .05

**p < .01

***p < .001.

*recent polling data on percentage of your constituents in favor of the Marriage Preservation Act.

**

***year in which you'll be running for reelection.

****

1. On the day the simulation was run, I discussed the next topic on the syllabus (the presidency) with the nonsimulation class. This created a temporary problem, in that it put the simulation class behind the nonsimulation class. By using every available minute of class time in the simulation class (and occasionally a few minutes more), the simulation class had caught up with the nonsimulation class before the next exam, with both classes receiving essentially the same exposure to the material on which they were tested.

2. These are questions 3–13. The first two questions, concerning the number of members in each chamber, were not related to legislative procedures and were included as a simple warm-up for the students. I included several open-ended questions after the true/false and multiple choice questions, but I found that subjectivity in grading student responses made their inclusion in this analysis problematic.

3. The answers of the two classes to these questions were compared using ANOVA and chi-square tests in order to check for any statistically significant differences between the classes in the following areas: number of previous political science classes taken, familiarity with the “I'm Just A Bill” Schoolhouse Rock cartoon, exposure to a high school civics class, grade on the first exam given in their PSC 100 section, interest in politics, work hours (nonschool related) per week, gender, and age. No significant differences between the groups were found in either the pretest or the posttest.

4. A short cartoon that traces the path of a bill through the legislative process.

5. Initially, I planned on including number of previous political science classes taken as a third variable in the ‘prior exposure’ category. However, I chose not to include it in the final model, as almost no students had taken another political science class—75% of the students had taken no previous classes and another 20% had taken only one previous class, almost certainly another introductory course in which they would have been unlikely to be exposed to the legislative process. Runs of preliminary multivariate models suggest that including this variable would not have altered the results in any substantive manner.

6. Several of the students designated as “members” did, in fact, fail to show up for the simulation.