| A great deal of instruction focuses upon procedures for solving small, often arbitrary problems. These procedures often become manipulations of notational systems, such as mathematical expressions or graphs, without meaning for the students. This sort of rote procedure may be sufficient to allow students to achieve success at school tasks. However, students often have great difficulty applying these procedures to real world problems.;Solving real world problems requires not only applying a solution procedure, but also identifying the problem itself from the surrounding information. This in turn requires students to apply their knowledge about real world agents and objects to identify the problem and appropriate factors to include in a solution. Then students must choose a relevant solution procedure and correctly apply it. Students may also need to use their real world knowledge to support their procedural problem solving, by helping them identify errors and recognize success. Of course, the notation may lead students to identify misconceptions held about the real world, as well. Thus, learning new domains sufficiently well to allow real world problem solving requires the interaction of procedural knowledge that operates upon notational systems and also general background knowledge of the world and its situations.;The three papers that form the body of my dissertation look at different aspects of this issue. The first chapter develops my theory of learning environment design and further explains the difficulties of learning from problem solving. The second chapter describes the reasoning students need to do to master a new domain, and provides empirical support for the argument in the domain of LISP programming. This chapter reports two experimental comparisons of novices in different learning environments. |
