True-False Testing on Trial: Guilty as Charged or Falsely Accused?

Abstract

Although widely used, the true-false test is often regarded as a superficial or even harmful test, one that lacks the pedagogical efficacy of more substantive tests (e.g., cued-recall or short-answer tests). Such charges, however, lack conclusive evidence and may, in some cases, be false. Across four experiments, we investigated how true-false testing of studied passages (e.g., on Yellowstone National Park) might enhance—or be optimized to enhance—performance on subsequent cued-recall tests. In Experiments 1–2, relative to control performance that did not benefit from any additional exposure, we found that (a) the evaluation of true statements enhanced the recall of tested (but not related) content and that (b) the evaluation of false statements enhanced the recall of related (but not tested) content, a differential pattern of benefits that did not depend on the syntactic structure of the test items. Moreover, when competitive clauses were embedded within the true-false items of Experiment 3 (e.g., True or false? Castle Geyser (not Steamboat Geyser) is the tallest geyser), we found that the evaluation of both types of statements enhanced the recall of both types of content. Finally, in Experiment 4, these holistic benefits proved robust to a retention interval of 48 h and were comparable with the benefits of a restudy condition in which learners restudied all of the propositions that could have been retrieved in the evaluation of the true-false items. Accordingly, although it was not uncommon for participants to misremember information as a consequence of true-false practice, our findings broadly indicate that, especially when carefully constructed, true-false tests can elicit beneficial, not superficial, processes that belie their poor reputation.