Logo Programming and Problem Solving

In the world of educational computing, programming is a major activ-ity, occupying several million precollege students a year in this country alone. As yet very little is known about what kinds of cognitive activities computer programming requires and whether, in the classroom contexts that are representative of microcomputer use in schools today, children are capable of making substantial progress in learning to program. In the cyclical program-development process of problem understanding, program design and planning, programming code composition, debugging, and comprehension, what gains do children make on the many developmental fronts represented in the complex of mental activities required by programming? Do conceptual limitations impede their understanding of any of the central program-ming concepts, such as flow of control structures, variables, proce-durality, and the like? We have begun to address aspects of these questions in our developmental research on children learning to do Logo programming. I would like to make five points which will be explicated in the re-mainder of 'this paper: 1. Systematic developmental research documenting what children are learning as they learn to program is necessary, rather than existing anecdotes. Our studies focus on Logo because it is a pro-gramming environment that is exciting to many educators, it has great potential for introducing children to many of the central concepts involved in programming and problem solving, and because gran