Using the Book in Large Classes

Here are some suggestions for larger classes:

• Run the experiments in recitations: In classes that have a large lecture as well as several smaller recitation sections, these experiments can be run in the recitation sessions. Experiments in the recitations give students a stronger incentive to attend these sessions and allows students to interact in relatively small groups. Recitation based experiments are also an attractive way of making effective use of teaching assistants who vary in skill and experience.

• Break the class into smaller groups and run the experiments simultaneously: You can divide your class into appropriate sized groups and run the experiments simultaneously. You need to make sure that the groups have enough separation, so that subjects in one session don't inadvertently wander into another session. Thus, you should designate widely separated ``landmarks'' for each group to congregate around, and some territorial boundaries that are not to be crossed. You should be able to handle all of the preliminaries for each experiment (discussion of basic procedures, warm-up exercises, and the answering of any questions) in front of the entire class before splitting into groups. Each group will need a market manager. TAs are an obvious choice, but students from the class will also work---you may want to talk to them ahead of time, but in reality the skills required are minimal (hand out personal information sheets, keep track of incoming forms, and post prices). To keep the groups separate and minimize confusion, you may want to print up the materials for each group on separate colors of paper. Finally, when ever you are managing a group of this size remember to do the following: tell them not to move until you give complete instructions, then tell them where to go and how to assemble once they get there, and finally what to do when they are done, finally, tell them to move.

• Run separate laboratory sections: This is a variant on the previous two ideas, but we thought it might be useful to mention it explicitly. You can always set up smaller laboratory sessions that meet in separate rooms, and then require the students to attend these sessions when appropriate. These sessions can either meet at the same time as the class (and therefore substitute for a class meeting), or at various times throughout the week. This idea is similar to what happens in introductory courses in the physical and biological sciences. In these courses, experiments are conducted in separate laboratory sections that are run concurrently with the main introductory course. (In fact, if you think back to your own experience in these types of courses, it is likely that the laboratory sessions provide some of your most vivid memories.) The pace of the laboratory course is tied to that of the regular introductory courses (but some variance is acceptable).

• Assign teams to manage the activity of each subject: If you have, say, 180 students, you can always assign three people per personal information sheet, and have them act as a team. Often times, teams have the advantage in that they can discuss various ways to play the game among themselves, and in so doing, gain some useful insights about the problem. Make sure the each team understands that it must act as a unit, and therefore it can not do anything that would violate the rules for an individual player (for example, make two deals when only one is allowed, etc.). You can either have the whole team get up and participate in each session, or you may want to have only one representative from each team come down into the trading area for each session. If you do this, make sure that the team members rotate who is sent into the trading area.

• Have the larger class watch a sample group: You can always randomly pick a subgroup of students (50 or so) and have them run the experiment in front of the rest of the class. For many of the experiments, the process of watching the market is quite informative (especially if you give the observers some key things to look for (and think about) before the market gets underway). Obviously, those students who are not in the experiment, lose some of the impact of direct participation---this can be mitigated by having the students who participated in the experiment trade-off with those who have only been observing. For example, at the end of each session, have each of the students in the experiment hand their Personal Information Sheet to someone who has not yet participated. Note that this approach with trade-off is very similar to the team approach discussed previously (especially if you are having the teams rotate), and the team approach is probably better as it keeps the students more invested in what is happening on the trading floor.

• Conduct the experiments in the large session: We have developed techniques to run some of the market experiments in large classes (there is nothing like watching 350 people erupt in a flurry of trading activity). In this scheme, students don't need to come down to a trading area, but rather are allowed to negotiate with their neighbors. Since this occurs simultaneously, the process is quick even for large classes. (You might think that the lack of information about sales prices would hamper this process, but in fact, there is a lot of information about prices floating around in terms of the bids and offers that are being overheard by the students.) You will need to develop some system for collecting and reporting the signed contracts. A few teaching assistants or students from the class can be assigned to stay in the aisles and do this job (each should be given some Sales Contracts ahead of time), you will also need to designate a ``runner'' or two to go up and down the aisles collecting completed contracts from the TAs or designated students and bringing them down front for you to post. Given the size of this market, it would be a great burden to the students if they had to complete the lab reports with all of the actual transactions (since they need to calculate profit for each trade, graph all transactions, etc.). Therefore, make the following modifications. Rather than report all of the data, you should create a sample market of a more convenient (under 60) size, say, 50 participants. To do this, report buyer values and seller costs as if you only handed out personal information sheets to 50 students. Then, report every n/50th (where n is the actual class size at the experiment) transaction. Thus, if you had 250 students, use every 5th transaction. (If n/50 is not close to an integer value, then adjust the number of participants until you get a value relatively close, for example, if you have 230 participants, you could have a market of say, 46 buyers and sellers, sampling every 5th transaction.) This method should work well on experiments 1, 3, 5, and 6. You can also run sessions 1 and 3 of experiment 7 (monopolies) by breaking the class into monopolistic firms and having each firm decide on the price and quantity, but not running the actual market at these prices, and making sessions 2 and 4 into thought experiments. On most of the other experiments, the students will still benefit from reading the discussion and doing the homework (just tell them in advance to avoid any question that requires data from the experiment---these are fairly obvious in any event).

When creating your course, you might find these page references useful.