The main instrument of the research was a questionnaire I developed to investigate language teachers’ use of CooSpace, supplemented with semi-structured interviews. The choice of a mixed methods paradigm was motivated by my intention to examine different aspects of the problem, as well as to provide an opportunity for triangulation. The design was sequential, in which the findings of one method are expanded using another method (Creswell, 2003). Thus, data yielded by the retrospective interviews was hoped to complement and elaborate on the questionnaire data, providing underlying reasons for participants’ answers. Finally, the validity of research can be improved by using mixed methods and corresponding evidence can also increase the transferability of the results (Dörnyei, 2007). In the following section first the characteristics of the participants will be described. Then the detailed description of the questionnaire and the interview schedule along with their development will be provided, as well as the procedures of their administration and data analysis.
5.2.1 Participants
The participants for the questionnaire were 44 language teachers of the college, 31 teachers from Faculty 1 (F1) and 13 teachers from Faculty 2 (F2). The original idea was to include all three faculties of the college in the research and to compare the teachers’ use of CooSpace. However, when teachers of Faculty 3 were approached, it was discovered that neither CooSpace, nor any other virtual learning environment is used there. Consequently, only the other two faculties could be investigated. Nevertheless, proper comparison of these two faculties was neither possible, due to the low return rate of the questionnaire (26%) at F2, which prevented collecting representative data for that faculty. The high return rate of the questionnaire at F1 (84%) can probably be attributed to the fact that the researcher also works there. Table 16 shows details about the participants for the questionnaire.
107 Table 16
Participants for the questionnaire in Phase 2
F1 F2 Total
N % N % N %
Gender Male 4 12.9% 1 7.7% 5 11.4%
Female 27 87.1% 12 92.3% 39 88.6%
Age 25-35 1 3.2% 1 7.7% 2 4.5%
36-45 11 35.5% 5 38.5% 16 36.4%
46-55 9 29% 2 15.3% 11 25%
Above 55 10 32.3% 5 38.5% 15 34.1%
Major English 14 45.2% 5 38.5% 19 43.2%
German 9 29% 6 46.2% 15 34.1%
French 3 9.7% 0 0% 3 6.8%
Italian 2 6.4% 2 15.3% 4 9.1%
Spanish 3 9.7% 0 0% 3 6.8%
As only two teachers volunteered for an interview in the questionnaire, two interviews were conducted. The reason for the very low number of volunteers could be that the time of the research was the busiest period of the term at the college. Details about the two participants can be seen in Table 17. In order to ensure anonymity, pseudonyms are used for the participants.
Table 17
Participants for the interview in Phase 2
Sheila Joan
Age 44 58
Teaching experience 22 36
CooSpace use no yes
5.2.2 Instruments
The instruments of the current research were a questionnaire with 43 questions and a semi-structured interview schedule that I have developed. The questionnaire contained 40 questions, starting with 10 questions about the participants’ background and the use of CooSpace in general. The second part contained 14 Likert-scale questions about the frequency of use of 14 functions, while the third part 14 Likert-scale questions about their usefulness, where participants had to indicate on a five-point scale how often they use a function and how
108
useful they find it. In the second part two open-ended questions were asked about additional and problematic functions. As the participants were teachers of different languages, the questions were in Hungarian to make sure that all teachers understood them. The choice of language excluded one native English teacher, who does not speak Hungarian. However, as the language of CooSpace is also Hungarian, he is not able to use it anyway. I developed the questionnaire and it was peer-checked by an expert and five fellow-students at each stage of the process. As a next step it was piloted by two teachers of the college (a male and a female), who were asked to think aloud while filling in the questionnaire. Problematic items, which included some questions, instructions and scale labels, were reworded. The order of the questions was also modified to make it more logical. The paper-based version of the questionnaire and its English translation can be seen in Appendix C and D.
In the second part of the research project semi-structured interviews were conducted with two participants. As both of them are Hungarian, the language of the interviews was Hungarian. The interview questions were grouped around five topics. First, questions the teachers’ professional background and language teaching, as well as language learning experience were asked about to establish rapport. Next, teachers’ answers to the questionnaire were looked at and they were asked to clarify any ambiguous answers, as well as to provide reasons for their answers. In this part questions about their feelings about using CooSpace were also included. Then, differences between students and teaching methods in the past and today were discussed in order to introduce the topic of e-learning. In the last two parts teachers’ opinions about computers and the internet in general and their suitability for teaching including language teaching were in the focus of the questions. Similarly to the questionnaire, the interview schedule was peer-checked continuously during its development.
The interview schedule and its English translation can be seen in Appendix E and F.
109 5.2.3 Data collection
The link to a web-based electronic questionnaire with 43 Hungarian questions was sent to the participants by email. The reason for online administration was that it can reach participants more easily and it is convenient to fill it in and send it back. Moreover, the possibility of skipping questions on the basis of previous answers makes it more user-friendly than a paper-based questionnaire. Additionally, it can save time for the researcher, because the coding and recording of the answers are automatic (Dörnyei, 2010). Nevertheless, the option of filling in the paper-based version was also offered to the participants and one teacher asked for it. An issue that had to be considered due to the mixed-mode administration was anonymity. While the web-based questionnaire was truly anonymous, the participant who chose the paper-based version could easily be identified. However, when I discussed the problem with her, she gave her consent to being identified and even volunteered for an interview.
The two interviews were conducted in an office at the college which was out of use at that time, thus an undisturbed recording process was guaranteed. The interviews were recorded with the help of a mobile telephone after obtaining consent from the participants and lasted between 35-45 minutes. Although less data was yielded by the two interviews than originally expected, they can still supplement the findings of the questionnaires, especially as the two teachers represent the two ends of the spectrum: one uses CooSpace very frequently, while the other never uses it.
5.2.4 Data analysis
All the questionnaires were computer coded and SPSS (Statistical Package for Social Sciences) 17.0 was used to analyse the results with the significance level set for p<.05. First, mean scores were calculated for the frequency and usefulness scores of each function. Then paired samples T-tests were computed to compare the mean values. To find out if any of the
110
individual characteristics which were asked about in the first part of the questionnaire have an influence on the teachers’ use of CooSpace, independent samples t-tests for questions with two possible answers and one-way ANOVA tests for questions with more than two choices were carried out. The interviews were conducted at the college in Hungarian, then transcribed and analysed following Maykut and Morehouse’s (1994) qualitative data analysis principles.
As a first step meaning units were identified in the transcripts, then they were categorized.
After that an attempt was made to find emerging themes and patterns. Finally, the categories were analysed to see if the questions yielded meaningful results.