Solo, together, apart: evaluating modes of CSCL for learning a problem solving task

ABSTRACT Many researchers are investigating systems that aim to exploit the benefits of collaborative computing technology in learning. There is, however, a surprising lack of empirical evidence on the benefits and costs associated with different modes of computer support for collaborative learning. This paper describes an experiment that compares how well 50 children (aged ten and eleven) learned to solve a small puzzle in three different learning conditions. In the first ‘solo’ condition, the children learned the puzzle on their own. In the second ‘contention’ condition, single gender pairs of children shared access to the same mouse and display. In the third ‘groupware’ condition, single gender pairs of children used a synchronous groupware system with a WYSIWIS (What You See Is What I See) display to mediate their collaboration around the puzzle. After learning the puzzle in one of these three ‘training’ conditions, all of the children solved the puzzle alone in a ‘testing’ condition. Results show significant reductions in solution time, number of moves, and latency between moves between the training and testing stages, assuring us that learning did take place. There was not a significant difference between the three training conditions (overall the children learnt a similar amount when trained in the three conditions). There were, however, significant interactions indicating that collaboration made solving the puzzle more difficult during training, and that girls found it harder than boys to solve the puzzle when collaborating.