Analysing unstructured qualitative data in various formats turns out to be a real challenge. Immersing myself in auto-ethnographic research in online social networks, microblogging and the wider blogosphere at this early stage of preparation for my Master thesis is going to produce rich descriptions, large amounts of questions, notes, comments, ideas to pursue, concepts to challenge, theories to be considered….I need some powerful helpers. Categorising, tagging, sorting, re-grouping and transcribing data seem to be among the most time-consuming tasks. It’s daunting. And the hunt for affordable, reliable and useful can-do software is something I don’t want to leave to the last minute. A current debate on the AoIR list made me think it is a good moment to look into the range of options in a somewhat more systematic manner.
First of all the market seems to restrict choices as there is the Mac / Linux / Microsoft question. Second, money is an issue. So, applications that offer at least a free trial are my preferred choice. Third, qualitative data is meant to be worked with, otherwise the layers of meaning may not start to unfold, so I want the software to do part of the work for me, but please, do not outsmarten me. Finally, I am not dogmatic: handwriting notes, typing them or recording them on a dictaphone, mp3 player or other devices is not a question of ideology – but to a significant extent an issue of practicalities. Also, I find that various methods of notetaking and recording help me keep my questioning mind fresh and make it less tiring to rework through the material, which in turn may feed into methodological triangulation and help with participant validation (either point needs more in-depth work, no doubt). They also help me being as efficient as possible: working full-time in a different industry means I can listen to some material while commuting. Not least, I am happy to save the time on typing and spend it rather on analysing and reporting.
NVivo imports, sorts and analyses straight from transcripts as well as audio files, videos, digital photos, PDF, rich text and plain text documents – it’s been designed for Microsoft. It allows to code data that contains tables and images and offers graphic presentation of project information. There is a query function, merging and sharing of files (which includes tracking of reviews) – also with non-NVivo users – is available by help of HTML pages. The 30 days trial version is free and a student version comes at GBP130, for 12 months only, though. For the PhD project, the full licence at GBP330 might be an investment worthwhile. The computer will need to fulfill a range of specifications that might have an impact on the available budget: NVivo system requirements.
Transcription and Qualitative Analysis Software packages HyperTRANSCRIBE 2.8 + HyperRESEARCH 1.0 offer a powerful combination of tools with a number of graphic options that help to build the coding, construct hyopothesis, theory and reports. While the full version costs about USD400, a FREE version for Mac, Linux and Windows is available for smaller projects with the following restrictions:
– the Master Code List is limited to 75 codes
– the study is limited to 7 cases
– each case has no more than 50 code instances.
Express Scribe Transcription Playback Software is FREE software that assists with transcription of audio recordings. It’s been designed for Microsoft, Mac and Linux and it is able to recognise a range of audio files from various sources. So my recorded comments and ideas on the SanDisk SANSA mp3 player can get transformed into text. The same applies to some podcasts which I think will provide quotes for reports. A pair of pedals could help to make the most of my limited time by delegating rewinding and fast-forwarding to the feet. ExpressScribe works with speech recognition software such as Dragon Naturally Speaking to automatically convert speech to text.
Dragon NaturallySpeaking 10 is available for about USD200 and helps with transcribing in-depth qualitative interviews. It does not require special script reading and offers to even surf the web by voice plus dictate and edit in most Windows-based applications. It claims to be 99% accurate and three times faster than typing but I suppose I will need train to it well, speak accent-free and verbalise punctuation. Washington State University provide a site with helpful screenshots on how to record with Dragon Naturally Speaking.
I came across two further tools which I found convincing, they are currently only available for Mac though, but hopefully soon also a Windows version will hit the markets: Tinderbox is a tool to organise, map and visualise complex fieldnotes and comments – and it even comes with a public file exchange option. TAMSAnalayzer is a FREE text analysis markup system works with Mac and Linux.
You can subscribe to my site’s updates using this link to my RSS feed.
I recently came across Mariana Goya Martinez’s and Aimee Hope Morrison’s work. The latter is a researcher in Canada and mommy-blogger, her presentation raised a number of interesting questions. Morrison argued that the diary-style blog, as sub-category of the blogosphere, constitutes a hybrid as it links the private and the public mainly by comments. These two areas are also interdependent as they nurture each other – comments provided on each other’s blogposts help to maintain and build networks and relationships. The theoretical framework was very much rooted in Goffman’s notion of the frontstage and the presentation of the self in the everyday. An everyday I would also define as a hybrid of the public and private, online and offline. The extent to which the bloggers’ narrative self-disclosure was part of an everyday life lived offline, though, was not subject to analysis.
Guilty pleasure as a concept sprang to mind as it was argued that many of those [female] bloggers ‘under investigation’ avoided tagging and categorising by purpose in order to prevent identification. It seemed, underlying the lifecycle of those blogs were also strongly correlated events and stages in offline life which forced the researchers to be either in a similar situation in order to comment and engage with the authors or to remain in an undisclosed role. Participant observation, textual analysis and auto-ethnography constituted the methodological framework.
Researching blogs without revealing one’s own identity, motives and activities might be very tempting and legitimate in ethical terms as the blogs are publicly available. Yet, when it comes to evaluation of findings the lack of participant validation as one strategy to ensure triangulation strikes me. Part of the problem may be the fragility of a blog’s lifecycle which is certainly subject to periodical crisis. Revaluation and redesign seem to be part of a blog’s coming of age with comments as currency generally providing a more or less constant flow of stimulation, positively or negatively. This may even result in its death with the role of flame wars in this context certainly not to be underestimated. Without interviewing bloggers at a more in-depth level, though, the complexity of events, complex decision-making processes and the impact of the unconscious upon all this must remain behind the public presentation and largely obscure – or subject to speculation at the researcher’s end.
This is the nexus where it becomes clear that participant observation and textual analysis are no doubt valuable methods but they may need to be reviewed as bloggers become increasingly aware of being subject to scholarly scrutiny with them being subjectified to observation and their products to textual analysis. The politics of ethnography would benefit greatly from a more sophisticated understanding of an unequal distribution of power in the blogosphere and what strategies bloggers resort to in order to prevent their blog becoming a casualty of the politics of ethnography.