Imagine that you encounter three events A, B and C that tend to occur together. What are the relationships between A, B, and C? One possibility is that A causes B and B causes C; another possibility is that A separately causes both B and C. Research on what is termed causal structure learning examines how we go about figuring out the structure of the relationships between events. When we learn about causal structure, we are essentially learning about how the world works, thus this type of learning is fundamental to the development of knowledge itself. This project aims to examine the cues that children and adults use to figure out causal structure, and more specifically to explore whether there are changes with age in the accuracy of such learning and in the ways in which different cues are used. In particular, it will examine the ways in which cues about the timing of events are used, and whether children can learn about the relationship between events through acting or intervening on them. The proposed project will test competing theories about the benefits of intervention and also provide a systematic exploration of developmental changes in the ability to learn through intervention.

This project involved testing of children and adults using a series of tasks specifically designed for this project. Participants were tested individually in their schools or in a psychology laboratory. Seven studies were conducted in total. Studies 1-3 involved 188 children aged 4-8 years using variations of a short cognitive task in which children observed a mechanical system operate and then made judgments about its causal structure. Children were also asked to make judgments about what would happen if the experimenter fixed or altered one of the components. Study 4 involved 227 children aged 5-8 years and 74 adults. Participants again completed a short cognitive task that required them making causal structure judgments, with the cues for these judgments varied across three conditions. Studies 5 & 6 involved 229 children aged 5-8 years and 107 adults. Participants completed a similar cognitive task as in the previous studies, but the cues that were available for making their causal judgments were different than in the previous studies. Children watched as the experimenter acted on components in the system. In Study 5, there were three conditions which again varied the specific nature of the cues; in Study 6 there was one condition, but the cues were richer than in Study 5. Study 7 involved 235 children aged 5-8 assigned to one of three conditions in which children either acted freely on the system, observed the experimenter act, or acted themselves according to the experimenter's instructions. Data from each one of these studies can be made available as an SPSS file, although each will require an accompanying document explaining the nature of the data. Thus, there would be 14 files in total.