As Creswell (2002, 257) explains, qualitative researchers first collect data and then prepare it for data analysis. I prepared the data so that both during and after each action research cycle I organized the data related to the specific cycle into its own database. This organizing was facilitated by the fact that most of the data was in digital form, e.g. logs of e-mail and computer conferencing messages. I didn’t tran-scribe the hand-written field notes as their number was not that high, and, there-fore, I felt that I was able to deal with the notes by just resorting to my notebook. I also had all data available in printed form as I felt more comfortable with dealing with paper than the screen. At an early stage, I had been tempted to try out some computer programs for analyzing qualitative data, but after experimenting with some, I gave up the idea as the programs seemed quite complicated, and I therefore was not convinced whether the time spent to learn to use them would bring added value to the analysis from the perspective of my particular study.
Creswell (2002, 257) specifies that after the data is organized, its analysis ini-tially consists of first developing a sense of the data, and then coding description and themes about the central phenomena. The process involved is thereby primarily inductive in form, moving from the detailed data to developing general themes. In my analysis of the data, I largely employed qualitative content analysis that Bry-man (2001, 381) regards as probably the most prevalent approach to the qualitative analysis of documents. He also emphasizes that such an analysis comprises a searching-out of underlying themes in the materials.
For the purpose of conducting qualitative content analysis, I used coding as my data analysis strategy. As explained by Miles and Huberman (1994, 56), codes are “tags or labels for assigning units of meaning to the descriptive or inferential information compiled during a study.” They further explain that codes are usually attached to “chunks” of varying size, e.g. words, sentences or paragraphs. Coding helps to retrieve and organize the chunks.
When creating the codes, I resorted to my three chosen research areas and the research questions related to them. The first research question was related to identi-fying issues and problems arising in connection with introducing ICTs as an edu-cational innovation into higher education communication teaching. The data re-garding this particular question consisted of the actual events and procedures in-volved during the project. Consequently, the ‘chunks’ to be coded were mainly
‘events’, not just words or paragraphs. In educational action research, in order for the teacher-researcher to be able to analyze such events, a narrative account of the entire action research project, a reliable ‘story’ so to say, needs to be written.
Writing a detailed and vivid account is of the utmost importance because the ac-count is expected to enable others to see how the researcher has arrived at his/her conclusions regarding the specific research question. Consequently, writing the account is a challenging task. Although the teacher-researcher has personally en-countered the issues and problems arising in connection with the action research project, it still does not mean that he/she could be able to identify them or interpret their significance. Indeed, it is not until through attempting to write a coherent and accurate report about the events that the researcher can better understand the vari-ous aspects involved.
For writing my account I utilized various types of the collected data. The logs of the e-mail message exchanges between myself and my partner teacher or other collaborators in the project were particularly useful as they most often dealt with various types of issues that had come up or problems that needed to be solved. Re-reading the logs of the students’ messages in the discussion forums was also help-ful in the sense that it helped me to follow the order of the events and also in that some of the messages confirmed my choices for what to include in the narrative account. As those choices are crucial from the point of view of the comprehen-siveness of the account, I used the structure proposed by Cohen and Manion (1989) for writing my action research account. I was carrying out an analysis of the types of issues and problems already while writing my account and I then completed the analysis after the account was completed.
The other research questions were related to the affordances of the media and the pedagogical issues of social presence, community, and teacher and learner roles. My codes included directly research question-related codes such as ‘teacher role’, ‘learner role’, ‘presence’, ‘community’, but I also adopted new codes if they seemed to emerge from the data. One of my original codes included ‘features of the
media’ as it was what I was originally looking for in one of my research questions.
It turned out during the coding process that I had difficulty in finding chunks re-lated to that theme, whereas instances of what the media enabled the participants to do seemed more evident. Consequently, this led to reformulating the research question for which I adopted the code ‘affordances’.
I did my coding by hand, marking the codes with a pen in the margins of the printed material. I then copied the coded chunks from the database and organized the chunks with the same code into a file and named the file according to the code, e.g. ‘teacher role’. The coding process also called for reading the materials over and over again. As pointed out by Creswell (2002, 258), qualitative researchers analyze their data by reading through it several times and by conducting an analy-sis each time. Some chunks to be coded seemed easier to find than others depend-ing on how explicitly or implicitly they appeared in the documents. For instance, some chunks to be coded seemed to be more explicit and easier to identify, whereas coding social presence, for instance, seemed to be more subtle and there-fore more complicated.
As is commonly known, qualitative data collecting often results in the accu-mulation of very large amounts of data. One of the problems with such material can be that some of the data collected may remain superfluous. So was the case with the data collected for the current study as well. For instance, during the analy-sis of the data, I did not end up utilizing the videotaped recordings of the videocon-ferences conducted during Cycles 2 and 3. First of all, to be of use, the sessions should have been transcribed the way my first course assistant had transcribed the recordings of the videoconferences during the first cycle. In the final analysis, even those transcripts did not provide as much specific content for the purposes of the current study that I had initially thought. Although I had read through the three transcripts several times and re-watched the tapes, it turned out that the more suit-able method for analyzing them would have been discourse analysis as the content was not, by and large, really geared to accommodate the pedagogy-related research areas that I had in mind. Using a different method of analysis such as discourse analysis could have given information, for instance, on the students’ communica-tion strategies during videoconferences, which research area I had eliminated at an early stage from the sphere of the current study in order to limit the focus of the study.
Another issue that could be considered somewhat problematic is that the proc-ess of content analysis may involve interpretative problems, which could be said to be the case with all qualitative data analysis. As the teacher-researcher I was re-sponsible for the coding and the interpretation of the themes arising. At the initial stages of the study I had hoped that in the final analysis I could use an inter-coder, but things did not work out that way, mainly because of the extended length of time I had spent on carrying out the analysis; as a result, the potential inter-coders were
not available any more. On the other hand, the use of inter-coders is not without problems either as the issue of inter-coder reliability may be an area of major con-cern. What compensated for the lack of an inter-coder at least to some extent was the fact that I had analyzed the process of the action research project extensively through discussions with my partner teacher from the other participating university during and after each action research cycle. Consequently, it was not only the mat-ter of how I myself inmat-terpreted the events, but the mutual discussions seemed to help me validate many of my interpretations.
8 Results and Interpretations
“High quality teaching can take place anywhere and any time and it’s not to be connected only with a high-tech environment.”
(A comment in a student’s course report.
Action Research Cycle 1)
This chapter will deal with the current study’s main results and their interpretations in the order of the research questions presented in Sections 1.2 and 1.3. The first research question related to the research area focusing on ICTs as an educational innovation will be addressed through rendering an account of the actual action re-search project. Next, the second rere-search area of affordances of educational ICTs followed by the third area centering on the pedagogical aspects of social presence and the roles of the teacher and learner in network-based learning environments will be dealt with.