Making it Work
Ideally everyone should participate in an experiment. This is because research shows that actually participating in the experiment is different that just seeing the results. It also shows that experiments are still an effective tool in large classes (for more information on these research findings, see the Why Do Classroom Experiments Work page). Unfortunately, experiments can be a little trickier to manage when classes are large.
There are several strategies that can help make classroom experiments scale up to be manageable and effective for large classes.
- Conduct the experiments themselves in recitation sections rather in the large lecture. In this case, experiments will be most effective if the instructor compiles the data from all of the recitation sections and discusses them in class.
- Have a subset of the class participate in the experiment and let the rest of the class observe. In this case, the experiment becomes more similar to an interactive demonstration.
- Divide students into groups so that every group represents one agent in the experiment. For example, an experiment that is designed to have about 30 participants can easily scale to a class of 120 if you have students work in groups of 4.
- Sometimes it is possible to have students take turns participating in the experiment. For example, if there are three treatments the class could divide into three groups, each of which participates in one treatment.
- A lot of what makes experiments difficult in large classes is the challenge of collecting decisions from individuals and communicating a partner's decision to an individual. This means that Technology can be very useful in conducting experiments in large classes.
Online CoursesExperiments present some technological challenges that can be overcome, often using easily available technology.
- If there are times when all of the students are online at once, consider using polling or survey programs to collect student decisions in experiments when it isn't necessary to know exactly who made what choice. Chat rooms can be used for discussion, however, it would be distracting to try to use them for both discussion and collecting decisions, especially when the class is large or chatty.
- In classes that are entirely asynchronous, consider planning an experiment that takes place over a several day time period. Students can e-mail their responses to the instructor by a certain time. Then there can be a time for posting and reviewing the data from the previous period (and perhaps discussion time) before the next period begins. Actually, this is the procedure used in FCC spectrum auctions!
- Some specialized Technology designed for classroom experiments can also be used to facilitate experiments in online courses.
In a traditional classroom with a small number of students it is easy to conduct a classroom experiments. Technology can help instructors to use classroom experiments in large classes or in distance learning classes. Some resources include:
- Aplia is a publisher that offers both online homework and tutorials as well as classroom experiments for micro principles classes.
- EconPort is a digital library with resources for teaching economics with experiments.
- VeconLab offers online experiments for classes at all levels of economics. It is a complement to Markets, Games & Strategic Behavior by Charles A. Holt, published by Pearson/Addison Wesley.
- In some experiments, like the Prisoner's Dilemma, it is important to be able to tell students what their "partner's" decision was. When this is not the case clickers can be used to collect student responses and run experiments.