In this paper, we provide evidence against the common idea that worked examples should be designed to convey problem categories and category-specific solution procedures. Instead we propose that instructional examples should be designed in a way that supports the understanding of relations between structural problem features and individual solution steps, i.e. relations that hold below the category level. We illustrate in the domain of probability word problems how category-avoiding instructional examples can be constructed. In two experiments we provide evidence that category-avoiding examples reduce cognitive load during learning and that they foster subsequent problem-solving performance. Problem-Type Schemata and Skill Acquisition It has often been argued that one important prerequisite for skilled problem solving in well-structured and knowledgerich domains (e.g., physics, mathematics or programming) is the availability of problem-type schemata (Gick & Holyoak, 1983; Reed, 1993), i.e., representations of problem categories together with category-specific solution procedures. Once a problem has been identified as belonging to a known problem category the relevant schema is retrieved from memory, is instantiated with the information that is specific to the to-be-solved problem, and finally the categoryspecific solution procedure attached to the schema is executed in order to produce a solution to the problem (cf. Derry, 1989). Schema-based problem solving is considered to be very efficient and therefore often seen as a marking feature of experts’ problem solving (VanLehn, 1996). Accordingly, a substantial amount of research has focused on the question of how such schemata can be acquired. A ubiquitous answer to this question is that studying concrete instances of problem categories (i.e., examples) is necessary for schema acquisition (Atkinson, Derry, Renkl, & Wortham, 2000; Sweller, van Merrienboer, & Paas, 1998). Schema Acquisition from Worked Examples In particular worked examples (i.e., example problems together with a step-by-step solution) play an important role in schema acquisition (cf. Atkinson et al., 2000). However, the mere availability of instructional examples seems not to be sufficient to guarantee an adequate representation of problem categories and an understanding of category-specific solution procedures. Rather, a profitable processing of worked examples has to be ensured. Such processing is likely to include example comparisons and example elaborations as the most important activities. Many approaches to improve the instructional design of worked examples subscribe to the general doctrine of using worked examples as a means of conveying problem categories and their associated solution procedures by fostering these activities. Example comparisons: Providing multiple examples allows a learner to compare examples within and among problem categories with regard to their differences and similarities. These comparisons might enable learners to identify the defining features of problem categories and to avoid confusions by examples' surface features (Cummins, 1992; Quilici & Mayer, 1996). According to Bernardo (1994, p. 379) there is "a consensus that problem-type schemata are acquired through some inductive or generalization process involving comparisons among similar or analogous problems of one type." Without these comparison processes learners might tend to categorize test problems according to their surface features and in turn to apply inappropriate solution procedures to them. Example elaborations: A commonly found problem is that learners “tend to form solution procedures that consist of a long series of steps – which are frequently tied to incidental features of the problems – rather than more meaningful representations that would enable them to successfully tackle new problems” (Catrambone, 1998, p. 355). To overcome these shallow representations of solutions, learners have to draw inferences concerning the structure of example solutions, the rationale behind solution procedures, and the goals that are accomplished by individual solution steps. In order to foster an understanding of category-specific solution procedures several methods have been suggested. One is that solution steps can be grouped according to their subgoals (Catrambone, 1998), which is thought to provide