Teaching with worked examples – Why the selection of problems for exemplification is critical

Abstract

Educational research has explored many aspects of the design of worked examples for teaching problem-solving skills but has yet to consider the influence of the problem itself that is chosen for exemplification. It is argued that the selection of problems to exemplify influences (meta-)comprehension since some problems lead to worked-out solutions that could be more or less ambiguous. Contrasting theories and findings are discussed which suggest both negative as well as positive effects of ambiguous examples on learning, at least when students are sufficiently supported to deal with that ambiguity. Study 1 (50 university students) confirmed that the selection of problems to exemplify impacts the ambiguity of examples. Studies 2–4 (90 university students; 100 university students; 142 high school students) showed that the groups receiving the problem that led to the more ambiguous example performed worse on the learning test, more often exhibited misconceptions, and overestimated their test performance. These effects were mostly large, consistent across learning topics (designating a factorial design and calculating the h-index), and even held on when learners received instructional support in the form of verbal explanations, which could have helped to disentangle the ambiguity. Amongst others, the results emphasize the need to pay more attention to the selection of problems for exemplification and consider an example’s ambiguity as an important feature for its effectiveness. In addition, they illustrate the preference of learners to follow examples over explanations, how misconceptions may arise, and the importance of guiding learners in early skill-acquisition phases.