Tests enhance learning—Compared to what?

Paclitaxel is a chemotherapy drug that has been used to treat cancer since the 1960s. It is relatively inexpensive because there is a generic version. In recent years, more expensive drugs have been developed to replace paclitaxel. After successful clinical trials, both Abraxane and Ixempra have been approved by the FDA. For their makers, this is the good news. The bad news is that, according to a study released in June 2012, neither of the newer drugs works as well as paclitaxel (Berkrot, 2012). New drugs are initially compared to placebos. But in medicine, being more effective than the placebo is not enough. To become a recommended treatment, a new drug should be better than the current treatment. The same is true in education. Roediger and Pyc (2012) make this point clear: “The gold standard of educational innovation for any kind of new educational technique should be a strong research base showing that the new method produces positive results relative to standard practice (Whitehurst, 2010).” Roediger and Pyc’s (2012) article ably describes three principles that enhance learning: distributed practice, explanatory questioning, and the one we focus on here, test-enhanced learning. In the literature to date, the value of testing has been demonstrated by comparing tests to two other activities: restudying the same information in the same way as before, or not restudying at all (Roediger & Pyc). Tests appear to be better than restudying. But is restudying an effective standard practice, like paclitaxel, or is it more like a placebo? In this article, we consider the value of tests as compared