The lack of female participation in Information Technology (IT) education and professional work is a challenging and constant issue. This issue is now a well documented research area, but the causes and remedies remain puzzling and complex.
Started in 1995 by Associate Professor Liisa von Hellens and Dr Sue Nielsen, our longitudinal study - called Women in Information Technology, or WinIT project, explores the perceptions and attitudes of women in high school, undergraduate and postgraduate university study and in the work place, and attempts to understand the factors which influence these females to take up and remain in an IT career.
Our research has taken the view that gender and IT are socially constructed. That is, IT is constructed as a domain attractive to certain types of people, primarily as a male domain, in the same way such occupations as child-care and nursing are constructed as female domains. The advantage of this view is that what can be constructed can also be deconstructed and changed. The more women enter professional-level IT education and work, especially in technical areas, the less IT will be viewed as a male domain. Similar changes have been seen in other professional fields, including law and medicine.
To date, over 800 surveys have been collected and approximately 100 interviews have been conducted. Since 2000, NUD*IST has been used to manage the very large data set and to help facilitate communication between members of the project team in Australia, as well as a collaborative project with other researchers in the US.
Our coding scheme is still quite simple and although we intend to elaborate it as our interpretation proceeds, the team has viewed coding as a necessary evil, time consuming and tedious, not as an end in itself. Our interest remains focussed on interpretation. We found that the systematic approach used with NUD*IST is helpful in improving indexing, searching and linking of the WinIT data. We also found that it did support interpretation of the data.
After our initial coding of data at the nodes, we used discourse analysis to identify distinctive patterns of text, and to relate these patterns to the individual interviews and to the whole set of interviews and to previous research in this area. NUD*IST supports the examination of the interview discourse, by allowing us to quickly move from the smaller piece of text to the wider context and back again.
Some unexpected issues arose when the project moved from a single-researcher analysis using the tool to shared analysis but this did not mean that the team were obliged to adopt a single coding of the data. The advantage of using software such as NUD*IST is that the same text (phrase, paragraph or larger section in the interview) may be coded as many ways as the researchers wish. This is made much more manageable with software than when performed manually.
Patton (1990) supports the use of computer assisted coding and suggests that "when data are going to be used by several people, or when data are going to be used over a very long period of time, including additions to the data set over time, such a comprehensive and computerized system can be extremely useful and could actually save time in the long run" (p.384). This statement sums up our experience very well and we are already convinced of the benefits of using NUD*IST to explore new and changing perspectives on the data.