Supporting Online Collaborative Learning in Mathematics

Most school math problems do not require multiple cycles of designing, testing and refining (Lesh & Doerr, 2003), and therefore, do not elicit the collaboration between people with different repertoires of knowledge that most authentic math problems elicit (Nason & Woodruff, 2004). Another factor that limits the potential of most school math problems for eliciting knowledgebuilding discourse is that the answers generated from school math problems do not provide students with much worth discussing (Bereiter, 2002a). Another factor that has prevented most students from engaging in ongoing discourse and other mathematical knowledge-building activity within CSCL environments is the limitations inherent in most computerbased mathematical representational tools (Nason et al., 1996). Most of these tools are unable to carry out the crucial knowledge-building functions of: 1) generating multiple representations of mathematical concepts, 2) linking the different representations, and 3) transmitting meaning, sense and understanding.