The Details: When they arrived in class, we asked each student to write on a piece of paper the kind of car they drove and to submit it anonymously to me. I wrote the list of cars on the board. We then divided the class into four groups, and gave each group a profile that Theresa and I had crafted in advance (e.g. "You are a single person," "You are married with one small child," etc.). Once the groups received their profiles, I showed them how to access the National Highway Traffic Safety Administration's site for car-crash data (see Figures 2 and 3).

We then instructed the students to examine the various ways the NHTSA tests for car safety and to pick which of these measures they believed to be most important, depending upon the profile their group had received. Then, using that measure, we asked the students to arrange the cars driven by the class in order from safest to least safe.

At the end of the class period, we compared the different rankings each group had produced. We observed how their different operationalizations of "safety" produced different evaluations about which cars are the safest.

The Goals: From my standpoint, there were a number of pedagogical goals for this exercise. I wanted students:

• To see that evaluating an object requires defining the terms of that evaluation, a process full of debate

Figure 1:

Figure 2: