ABSTRACT
In this chapter, the author begins with a synopsis of the basic paradigm that has been used in his experiments and continues with a description of some elementary working-memory concepts. He provides some experiments on probability judgment that converged on retrieval failure as the major source of reasoning errors. The author discusses a simple stochastic model that guided an earlier series of experiments. The aim of this first model was to provide independent estimates of two types of working-memory failures in children’s mental arithmetic, namely, breakdowns on the input side and breakdowns in transforming stored traces into numerical outputs. The author presents an experimental results. He summarizes findings from previous studies in which the simplified model was applied to protocols of kindergarten and first-grade children. The author also provides some additional experiments in which the expanded model was applied to protocols of children from the same age range.
