Expert example but not negative example standards help learners accurately evaluate the quality of self-generated examples

In acquiring new conceptual knowledge, learners often engage in the generation of examples that illustrate the to-be-learned principles and concepts. Learners are, however, bad at judging the quality of self-generated examples, which can result in suboptimal regulation decisions. A promising means to foster judgment accuracy in this context is providing external standards in form of expert examples after learners have generated own examples. Empirical evidence on this support measure, however, is scarce. Furthermore, it is unclear whether providing learners with poor examples, which include typical wrong illustrations, as negative example standards after they generated own examples would increase judgment accuracy as well. When they generated poor examples themselves, learners might realize similarities between their examples and the negative ones, which could result in more cautious and hence likely more accurate judgments concerning their own examples. Against this background, in a 2 × 2 factorial experiment we prompted N = 128 university students to generate examples that illustrate previously encountered concepts and self-evaluate these examples afterwards. During self-evaluation, we varied whether learners were provided with expert example standards (with vs. without) and negative example standards (with vs. without). In line with previous findings, expert example standards enhanced learners’ judgment accuracy. The newly developed negative example standards showed inconsistent and partly even detrimental effects regarding judgment accuracy. The results substantiate the notion that expert example standards can serve as a promising means to foster accurate self-evaluations in example generation tasks, whereas negative example standards should be treated with caution.