Digital Games as Assessment?

by Jennifer Kotler, Ph.D.
June 1, 2011

In January, I attended a workshop dedicated to games, assessment and learning hosted by the MacArthur and Gates Foundations and the USC Game Innovation Lab. The workshop brought together game designers, educators, and researchers to work together on designing games around various curricula topics that would be engaging, educational, and contain features to allow for the collection and feedback about how players were faring when engaged in the game. The conversation went beyond what players could learn from games: We also focused on the valuable information we can gather from patterns of game play, such as where players might make errors and the kind of errors players might be making so that either the game or another knowledgeable player can help provide the necessary support to improve game play and therefore, learning.

This kind of thinking always reminds me of math class tests where we were asked to “show our work” so that the teachers could see how we went about solving a particular problem. A wrong answer to a division problem that had more to do with a simple subtraction error is very different from getting the wrong answer because of a fundamental lack of understanding of how to approach the problem. Patterns of responses can provide much more specific information than whether children get the answer right or wrong (as many standardized assessments generally report). Game play data may indeed provide another valuable way to assess patterns of children’s understanding in a less threatening way than common testing conditions.

Not only might such games be useful in formal learning situations for assessment, but they might also encourage parents to become more engaged in children’s learning. As part of some recent research around Prankster Planet on The Electric Company website that Mindy Brooks wrote about in last month’s blog post, we asked parents (about 40 of them) to fill out a survey. The survey included questions about parents’ interest in receiving feedback about how well their children were doing on the math and literacy activities within Prankster Planet.

I assumed that perhaps only a third of parents would be interested in receiving information on how well children were doing on the game. Surprisingly, the vast majority of parents (over 70%) said would be very likely to use information about how well their children were doing on the games. Furthermore, even more said they wanted specific feedback as to how to support the activities that the children were doing in the games even more so than general suggestions how to work on math and reading skills with their children. Parents said they would be most receptive to receiving this information through an email (rather than a text message or in a password protected site). This might be a particularly interesting opportunity to engage more parents and provide very specific information about how to extend children’s learning based on children’s individual game play patterns.

Before we rally for more widespread use of games as assessment tools, we likely need more investigation as to whether scores, errors, and successes in games are indeed highly correlated with the very same things that success on standardized or classroom tests are supposed to predict. Clearly, this assumes that standardized measures or classroom tests are the “gold standard” for information about what children “know” and that, of course, is the topic of much debate. Still, at this point in time, children are often classified or assigned to particular learning interventions based upon standardized assessments.

Games might provide a less “frightening” testing environment. Perhaps games might indeed reduce what Dr. Claude Steele termed “stereotype vulnerability.” Girls and children of minority status might do better under conditions that don’t seem test-like because they have been unfortunately conditioned to believe that children like them do not do as well as others on academic tests. Games might provide a neutral playing ground as well as reduce test anxiety.

On the other hand, perhaps children take more risks in games that they would not do if they were being tested, which may in fact be what educators encourage, but might interfere with their scores. Furthermore, I have seen situations where some children may actually choose wrong answers every so often just because the wrong answer feedback was funny, or perhaps they were just interested in seeing what would happen with a wrong answer choice.

Nevertheless, games provide a very efficient and engaging way to collect valuable information about performance. To be most useful as an assessment tool, however, game designers should work with educators and experts in assessment to ensure that information is captured in meaningful ways. Using the data in ways to support further learning is critical. Providing additional opportunities to practice skills and expand learning through additional gaming or materials for parents can only help make gaming experiences richer for children.