Extreme Testing Could Be the Future Of Education

"Taking a test." Most people, at the sound of those three words, imagine sitting in a room with a pencil and paper, scrawling essays or dutifully filling in Scantron sheets. The experience is commonly imagined to be equal parts boring and stressful. Increasingly, tests determine economic and social standing ... and there seem to more of them every year, for all age groups and professions. In 2008, one count has it that 50 million tests were administered to k-12 students.

However, the practice of administering tests may be changing in response to criticisms that tests are biased, encourage an ineffective type of rote learning, and encourage cheating by both students and teachers. For the latter, I'm shocked at how often a new scandal emerges. Atlanta is the most recent cheating teacher incident, in case you're keeping track.

A new paradigm of testing is emerging along with new educational philosophies that encourage collaboration, problem-solving, and practical application of abstract knowledge. 

Anecdotally, I think this type of testing has a lot to recommend for it. As a college sophomore, I remember that my economics class involved several simulated experiments. In one, students were pooled at computers. Some were told to sell a product for as much as they could get and others were told to buy as much of the product that they could for a given allotment of money. The interesting thing was that a market for the good had not been established, so that no one knew how much to sell for or to buy for. But over the course of just a few minutes, the constraints operating on the various sellers and buyers resulted in a market price shaking out. I sat at my computer and marveled at how quickly it had all happened. I found that I simply could not sell my goods for any higher than a certain target price. The market has spoken.

The orientation toward games is only present in college and it's not just about economics.

Some educational professionals are already seriously debating the wisdom of making learning whether biology or reading into a game, complete with rewards, opportunities for creativity, and all the rest. This would be a kind of PolicyMic for education, and it's already being practiced like places such as Quest To Learn, a New York City charter school that schedules classroom activities as quests. In that program, they ask students to make maps, plan expeditions, and calculate expenditures for their budget. To complete these missions, quests, or whatever you want to call them, the student must master the relevant skills (if you're interested in hearing more about the philosophy of gamification, see this post).

The possibilities for this new type of testing are extreme. Take, for example, the test that UCLA ecology professor Peter Nonacs recently gave to his class on the biological applications for game theory. He told them that the test would be hard, but that they could use whatever nefarious tactics they could come up with to succeed ... even talking during the test, using the Internet, and copying the answers of other students. His idea was that it's a jungle out there, and why not communicate that fact by turning the classroom into one as well?

I think the idea could work, and it shows how much more fun it can be to play games rather than to merely memorize information. Nonacs reported that his class did better than average and that their scheming leading up to the actual test day helped them appreciate strategic action and strategic thinking by forcing them to live it.

There are provisos, though, and lessons for what could be done in the future.

First, allowing students to "live" a certain subject area may not be the best way to learn about it, and that's because knowing how to do something is very different than knowing that something is the case. For instance, insects behave strategically in certain hive organizations they "live" game theory every day but that doesn't make them experts about game theory. Of course, the students didn't just execute game theoretical strategies; one presumes that they reflected at least a little bit on what they were doing, and so learned something that way. 

The other thing to keep in mind is that unleashing the creativity of a college class on getting a good grade mirrors the way that cooperation can be good for a species or community, but in education, we care very much about the attainment of individual students. In other words, the fact that the class came up with a good answer to the test question (there was only one big question that Nonacs asked) says very little about what each student learned. What about the kid in the back who is charming and uses his social capital to earn him the right to copy answers? Or what about the person who searches the Internet to answer questions as the dialogue goes on, but knows very little about how those answers link together?

To remedy these flaws, one could imagine introducing multiple variants of the test that was administered at UCLA. For instance, what if the professor randomly selected five students from the lecture hall to take a second pencil and paper test after the first test was turned in? Free-riders would be put at risk by such a strategy, introducing a whole new strategic dynamic into the process. Or what if you were unknowingly paired with another student in the room, and his grade was yours and your grade was his? Again, the incentive would be on making sure that even the marginal students "got it." Other variations could be even wilder: what if there were no absolute grades, only class ranks? This would probably create a vicious kind of competition, in which students might have reason to lie to each other about the answers and to introduce distractions.

These variants would not be boring pencil and paper affairs, but some of them might be even more stressful than the rules we have now: one person, one test.