Still reading...

On reading Roschelle (1992), up to the evaluation
The argument being made is that it is possible to construct ‘an integrated approach to collaboration and conceptual change’ (p235). If you accept that the collaboration will occur in, for example, problem solving activities (and that this is not just unconnected conversations), then it would be fair to research the specific ‘collaborative conversational interaction’. Roschelle’s proposal that the ‘crux of collaboration is the problem of convergence’ (p236), seems to assume collaboration as a meeting of joint understanding, whereas this could just as easily be viewed as an agreement to complete or accomplish a given task; however, if we concur with the assumption, the type if research required would indeed necessitate analysis of detailed conversational interactions – ‘conversational analysis’. The research discussions provided are very detailed, with specific notes on the situational activities, such as use of hand gestures, which seems very relevant to setting the context.
The results of the research are presented in a narrative format, with supporting diagram; some use is made of statistical evidence (e.g. p242 ‘…the median point…’, but this is much less so than used in other papers. The evidence ranged from observer comments and analysis, transcriptions of conversations and their associated analysis, as well as post-experiment interviews. ‘Evidence’ is considered to be the conclusions reached by the participants; specifically, it is considered that a ‘non technical’ conclusions (in a language sense), which is equal to approximately the prized ‘correct’ scientific understanding, is acceptable. Is this sufficient ‘evidence’? The researcher also seems to fill in a lot of gaps with assumptions; for example, from p250…’It is plausible that her intention was…’, can hardly be concluded to be evidence?
If the research question is to consider the construction of an integrated approach, then there appears to be much more ‘evidence’ for the collaboration than for the conceptual change. However, the small matter of highlighting the comment ‘…the use of we’, is perhaps a solid indication that the participants were indeed collaborating and did experience conceptual change.
The article used the concept of ‘deep features’ (of situations) as a crucial component of the research framework; however, this definition did not appear to be clarified at the outset and was therefore dependent on personal interpretation. Due to the numerous references to these ‘deep features’ it seems difficult to guarantee that one has grasped the researcher’s meaning, without further reading of one of the cited authors (such as Anzai and Yokohama, 1984); plan to check this out before continuing with the paper and considering Roschelle’s evaluation.