Effects of innovation versus efficiency tasks on collaboration and learning.

摘要

What makes for a naturally productive collaborative task? Some researchers have suggested that optimal tasks for productive collaboration are ill-structured and allow for exploration and construction of multiple possible solutions (e.g. Cohen, 1994). Others have suggested that tasks should have one solution and be well-defined such that everyone can agree on their answers (e.g. Steiner, 1972). In a search for a way to reconcile this dilemma and begin to answer the driving question, two dimensions, innovation and efficiency, were examined for their effects on collaboration and learning in two experiments with university students.; Innovation involves the use of prior knowledge to construct solutions to unfamiliar problems. The goal is to prepare students to perceive and appreciate how an expert solution works when they receive instruction on it. Efficiency involves being given the canonical solution and then having an opportunity to practice it. The goal is to promote speed and accuracy in applying the expert solution.; These dimensions were recently found to be informative to the field of transfer. Transfer is the generalization of learning from one situation to another. Schwartz, Bransford, and Sears (2005) suggested that optimal instruction for promoting transfer involves cycles of innovation and efficiency, rather than just one or the other approach. Thus, they described them as complementary components for promoting thorough understanding.; For the two experiments reported here, it was hypothesized that tasks with an innovation component would yield more productive interactions and learning than tasks with strictly efficiency components. The first experiment involved concept-mapping of the relationship between cholesterol and the circulatory system. The second involved learning the chi-square formula.; Results of the first study indicated that Innovation dyads collaborated more than Efficiency dyads. Results of the second revealed that the Innovation condition scored significantly higher on transfer problems, and Innovation dyads showed the greatest performance on a novel measure of preparation for future learning (Bransford & Schwartz, 1999). The strongest indicator that tasks with innovation components might naturally support collaborative learning came from the finding that Innovation dyads exceeded nominal dyads (mathematically modeled dyads based on individuals' scores) on the novel measure.