wow it's been a hectic two days.
Christ Church Secondary finally gave me their five PowerPoint files, which represent the summative products generated by the students regarding their fieldwork. with these files, i am so relieved to say that all the various bits of data generated by the study are finally in :-) yay! no more loose ends :-)
yesterday, john very kindly set up a meeting with Peter Freebody. he's one of the pioneer batch of staff at the CRPP, but sadly he will be leaving CRPP at the end of the month. all the more reason for me to be grateful to have had an extended two-hour session with him - he's a real statistics guru and he helped both john and myself clarify our thinking on how to represent and process the quantitative data generated from the pilot (yes, the saga of my preparations for my First Year Progress Report lumbers on).
here's what i learned from peter (in no particular order):
- the work of Sinclair and Coulthard, nearly thirty years ago, forms the basis of much of the corpus of research on discourse analysis in the classroom. invaluable for my literature review.
- Ian Hutchby has done some work on discourse analysis with regards telephone conversations (and, to an extent, IRC as well). his book - Conversation and technology: from the telephone to the internet - is a good read but slightly tangential to my needs.
- eventually, the limitations of the study can be turned point-by-point to areas for further exploration. for example, one such would be the influence of particular site attributes on adolescent discourse.
- i raised the question of whether the fact that i had no way of knowing which of the pair-members was doing the texting at any given time would be a major problem in terms of design-critique (eg, during my oral defence). peter said that he did not think so, because the micro-management that that would have entailed would have run counter to the spirit of the naturalistic inquiry. he went on to say that there might hopefully be a levelling-up effect, meaning that the joint performance on the pre- / post-test comparisons for the pairs would be similar to the individual performance of the better of the respective pair-members.
- i also told him that at the present moment, i had little intention of analysing orienteering performance with respect to each of the seven field-sites. he agreed that that would be beyond the scope of the present study (vide supra). but he advised me to address this near the beginning of the report, so as to pre-empt an unpleasant situation during my oral defence should someone query this very issue. hopefully once i've analysed all the data, there will be no significant differences in orienteering performance by either site or school.
- i have learned how to use SPSS to generate what is known as a dendrogram. this dendrogram allows us to group statistically-related variables together, so as to simplify and generate more meaning from the discriminant analyses.
- after several trials, including deciding to combine some of the original discourse modes which were obviously related by definition, john and i have decided to proceed with a three-group discriminant analysis for both the 'leading' and 'following' pairs. not sure whether this may change once the remaining data has been analysed.
whew!
Technorati Tags: discourse, viva
