How a lawyer (and PhD candidate) uses NVivo to transcribe, classify and visualize data
I am a PhD candidate, researching how Australian lawyers who teach practical legal skills engage with scholarship of teaching and learning. I was one of those lawyers before starting this research. Before that, I was a lawyer in a regional firm. Before becoming a lawyer, I performed in theatre, film and television for about 20 years. I am half way through my 3-year candidature.
I was born with a moderate-severe hearing impairment; without hearing aids and other technologies it would be very difficult for me to work or study. Technology enables me, but it must ‘earn its keep.’ I prefer robust utility to bells and whistles.
I’m please to say that NVivo 10 is an excellent ‘enabler’. I use it for my literature reviews, and to store, organise, analyse and visualise data. I use NVivo to produce materials (memos, charts, models), which are easy to incorporate when I ‘write up’ my qualitative research.
Lawyers who teach
My research involves theory and practice in law, policy, education and training in practical legal training (“PLT”). PLT is an eligibility requirement for admission to the Australian legal profession, and is usually undertaken after completion of a law degree.
I am studying how lawyers who teach PLT (“PLT practitioners”) engage with scholarly activities around their teaching and learning work. This involves asking questions about PLT practitioners’ motivations and capacities to engage with such activities, and the resources and symbolic support their employers provide.
Data – documents and recordings
My data sources include documents and recorded interviews. The documents are legislation, subordinate instruments, commentary, scholarly articles, reports, mission statements, profiles, usually in PDF format. The semi-structured interviews were recorded on digital audio-visual media. I found it easy to import the documents and recordings into NVivo.
I transcribed the interviews in NVivo, using the built-in media player, to create time-stamped transcript rows for the recordings. I used Dragon speech-to-text software to dictate both sides of the interview into the transcript rows. It took a little while to improve my workflow, but by the time I finished the transcriptions I was able to transcribe 4-5 times faster than my touch-typing.
Classifying the data
The draft transcripts were sent to interviewees for verification and comment. In the meantime, I set up classification sheets and attributes for the sources. For example, attributes for the documents included ‘type of instrument’, ‘jurisdiction’, and ‘date proclaimed’. Attributes for interview recordings include ‘interviewee’ (a pseudonym), ‘date’, ‘format’, ‘location’ and ‘organisation.’ I created a node called ‘interviewees’, with attributes such as ‘jurisdiction’, ‘affiliation’, ‘gender’, ‘role’ and ‘name’ (a pseudonym).
Comparing perspectives on paramount duty
The attributes help me to explore the data where certain nodes cluster around certain attributes. For example, I asked interviewees, “does a lawyer’s paramount duty to the Court intersect with their teaching and assessment work in PLT?” I set up a parent node for the responses, and created child nodes for different themes emerging from the responses. I could then compare child nodes with attributes, such as ‘jurisdiction’, ‘affiliation’ or ‘role’.
The interviewees’ responses identified other individuals, organisations, and ‘field forces’ relevant to my “paramount duty” question. With NVivo I could create models representing those agents and forces. This model proved to be an excellent catalyst for discussion during a recent conference presentation.
If only I had known!
I am still getting to know NVivo but I am already persuaded that it is a powerful tool for my research. I wish I had known about it when I was a practising lawyer! It would have been very useful for analysing documents and evidence in my legal work.