The interviews were analyzed separately by two researchers (one of which is the author of this paper), using an iterative constant-comparison method (Strauss and Corbin, 1998). First, the data were analyzed separately by the two researchers according to conventional qualitative methods (Mason, 1996 and Silverman, 1993), looking for patterns and relevant themes, which were then coded to allow for further analysis. The researchers then met to discuss the coding scheme that each of them had devised in order to assess how prior expectations or theoretical inclinations influenced the schemes and to discuss the hypotheses by using negative case analysis. Through this discussion an agreed-upon coding scheme was devised. A second reading of all the recorded materials enabled a systematic reduction in the number of categories by combining similar terms and eliminating redundant ones. In case of conflict, agreement between the coders was reached after extensive discussions (Glassner & Loughlin, 1987).
The analysis had two foci in accordance with the foci of the interviews. One was students’ perceptions on the usefulness of the Internet for schoolwork. The students often replied to the relevant question by comparing the Internet and computers with books. Therefore, the results were analyzed accordingly and refer also to books. A list of the themes and the frequencies of their appearance in the interviews are presented in Table 1).