Reiser, B. J., Copen, W. A., Ranney, M., Hamid, A. & Kimberg, D. Y. (1994)
Technical Report #54, The Institute for the Learning Sciences, Northwestern University

A central controversy in the design of instruction concerns the amount of freedom or guidance that should be provided to students. We examined the cognitive and the motivational consequences of guidance and freedom in a learning environment used by students learning introductory programming. Students worked with one of three interactive learning environments that varied in the amount of freedom to explore or guidance provided.

The Guided group solved a set of assigned problems worked with a tutoring system (GIL) that interrupts and provides explanatory feedback whenever students make mistakes. An Exploratory group worked on the same problems using an exploration- oriented version of the system that does not interrupt upon errors, but instead enables students to test ideas and receive feedback at their request. A Free group used the same exploratory learning system but was free to generate their own tasks.

The tutorial guidance helped students solve the assigned problems more quickly, although there were no differences in the subjects’ understanding of programming constructs or their ability to construct programs. However, both discovery learning groups were more effective in detecting bugs in programs, presumably due to their experiences finding and repairing their own errors.

For students required to solve assigned problems, the positive or negative nature of the learning environment’s motivational consequences appeared to depend upon the relative ability of the student. The learning environments differentially affected the attitudes of high and low ability students toward the domain and their assessments of the success of their performance. Low ability students in the discovery learning situation tended to exhibit more negative judgements about their performance and the computer’s assistance than comparable students who received tutoring. In contrast, high ability students tended to exhibit somewhat more positive opinions of their performance in the discovery learning environment than comparable students in the tutoring situation.

We argue that discovery learning creates more opportunities for students to assess how well they can overcome obstacles, and their resulting attitudes toward their past and future success in the domain relies heavily on this type of attribution. The positive or negative nature of that attribution will depend on their relative success in achieving their goals.