Should I Provide Incentives?

Don't be afraid to conduct a classroom experiment because of the expense of paying students to participate - classroom experiments don't usually involve money. Often the desire to "do well" and rivalry with other students to see who can earn the most hypothetical money is all that is required to make students try to do their best.

On the other hand, there are reasons to explore rewarding students for performance in economics classroom experiments. One is that students are typically given grade rewards (which are usually performance based) for completing any class assignment. Giving students points for participating or doing well in a classroom experiment signals that the classroom experiment is an important part of the course and needs to be taken seriously. The second is that research experiments in economics typically follow Vernon Smith's Induced Value Theory so that participants in experiments are rewarded for the decisions that they make. Including some sort of performance incentive more closely aligns the classroom experiment with research experiments.

Arguments Against Performance Based Rewards

• In many experiments players have different characteristics so not all students have the same chance to earn points. This has fairness implications.
• "Doing well in classroom experiments" is not a learning objective for the course, so it might not be appropriate to base grades on this sort of performance. Put another way, it is possible for a student to master the material related to the experiment and still perform poorly in the experiment itself.
  
• Students can be given "participation credit" if they participate in the experiment.
• In large classes, recording point earnings and computing grades may not be a good use of the instructor's time.
  

Arguments For Performance Based Rewards

• Some students enjoy earning class points this way.
• It makes the games more interesting for some students.
  
• Frequently students will deliberately make poor decisions to "mess up" the game. They enjoy watching the reaction of the rest of the class to an obviously "wrong" decision - a grade reward counteracts this incentive. For example, in an auction game a single seller who sells at a price of zero can move the trading price and quantity away from the equilibrium. This affects the learning experience for the entire class. As a result it is appropriate for there to be something at stake.
  

Strategies

• Use a conversion rate that converts game points into dollars and randomly select and pay one student for their performance. Note: Don't pay the person who earns the most because this payment strategy makes students act in more risk-seeking ways than they otherwise would do.
• In a class with a lot of experiments, students can contribute a few dollars to a "prize fund" at the beginning of the semester from which a random winner can be selected for each experiment. Make sure to consult your university's policy on "lab fees" before choosing this strategy.
• Assign grade points for for participation as a fraction of the maximum a student could have earned, given their role or private information in the game. This is more complex to calculate but does solve the problem of some students being disadvantaged by their role or private information in the game.
  
• Assign most of the students' grade based on whether or not they participated, and only a small part based on performance success.
  

References

Smith, Vernon L, 1976. "Experimental Economics: Induced Value Theory," American Economic Review, American Economic Association, vol. 66(2), pages 274-79, May.