Active participation in comprehensive activity may support not only advanced conceptual understanding, but also the emergency of new metacognitive beliefs about knowing, and particularly about the importance of understanding Hatano & Inagaki (1992); as cited by Lehtinen, et al., (2004). The typical classroom does not have the resources needed for successful collaboration because there are not enough available anchor points at which action and attention can be coordinated, "however computers and the accompanying capabilities of computers present an opportunity for mediation tools that assist students in focusing their attention on objects that are mutually shared." Jarvela, Bonk, Lehtinen; as cited by Lehtinen et al., (2004).
Effects of CSCL on learning and achievement have been studied through a large meta-analyses specifically on the effectiveness of computers with results stating that the use of technology has markedly improved learning outcomes e.g. Khaili & Shgashaani, (1994), Kulik, Kulik & Cohen, (1998); as cited by Lehtinen, et al. (2004). The conclusions that CSCL-based learning frameworks enable higher levels of learning both for in-class and distance taught courses (Macdonald, 2003) further underscores the need for an online learning management system that is tailorable to support both instruction strategies. The concepts of Khan (2003) and others (Bannon, 1989) (Kanfer, 2001) highlight the need for an agile yet foundationally solid learning network...
Of all methodologies that could potentially be used for analyzing the influence of CSCL-based teaching strategies that include customized scaffolding strategies for students, the test/re-test methodology shows the greatest potential level of reliability and validity over time according to Anderson (1998). Taking into account the concentration on using argumentative teaching strategies in conjunction with CSCL-based platforms and scaffolding yields a complex set of independent variables that are best measured in a test/re-test methodological approach. The combined influences of all variables together can be isolated by specific variable through the use of pair-wise t-tests and correlation analysis once the data is accumulated. Using statistical approaches that introduce any of the given independent variables into the analysis during the learning process introduce too much of a potential for sampling error, in addition to spurious results given the influence of the independent variable's introduction rather than its actual effect of the variable's value itself.
