One of the key tools that we have used for our qualitative analysis is Maxqda (http://www.maxqda.com/ ), a software package that has much in common with NVivo (http://www.qsrinternational.com/products_nvivo.aspx). We chose Maxqda because of its potential flexibility and versatility, particularly when working longitudinally, and working with mixed methods.
In order to undertake analysis using Maxqda, all text needs to be in an electronic format. Therefore the MOP scripts that we had sampled needed to be copied into a digital pdf format then transcribed into Word. When undertaking the task of transcription we asked our transcribers to be as faithful as possible to the idiosyncrasies of each piece of writing, particularly where a script had been handwritten. We asked transcribers to keep spelling and grammatical errors, crossings out, abbreviations and notes in the margins; and keep to the same line and page length. When analysing the texts we have also looked at the digital copies of the original scripts, using these alongside the transcribed documents that have been imported into the software. Our view is that the texts themselves are artefacts, and may have their own stories to tell.
Each individual script that we have imported has been given a unique reference, which identifies the MO writer, the directive they are responding to, the year that they are writing in, and their age when they were writing that particular script. This referencing will allow us to follow writers across the years that they have written for Mass Observation, and to track changes to individual life-courses, volunteering attitudes and behaviours, and views on voluntarism and the state.
The software allows us to identify how each script answers our various research questions, and to identify emerging themes.
As we read through each script, we highlight pieces of text and place them into individually labelled ‘folders’, for example, if a writer says “I disagree with the concept of volunteering, and therefore I don’t volunteer”, we might highlight this and place into three different ‘folders’:
- One labelled ‘Says doesn’t volunteer’
- Another labelled ‘Negative attitude to volunteering’
This is known as Coding.
Some codes were set up prior to analysis (deductive coding) and other codes have emerged (inductive coding) as we have begun to analyse the scripts.
Analysis of our coding will allow us to:
- Explore how views and behaviours change or continue over time. For example, we can look at how a 50 year old woman writing in 1983 envisaged what her life might be like when she retires in ten to fifteen years’ time. Does she anticipate volunteering in her retirement? By following her writing throughout her life-course we can map what she has written about preparing for her retirement, the process of retiring, how she feels about retirement since she gave up work, and how and why she has volunteered during retirement. We can compare this to her earlier thoughts on what retirement might be like. (This process of following a writer through time is a diachronic, or longitudinal perspective)
- Compare views from people of a similar age range, for example we can capture and compare all coded references to retirement written by women in our sample who have written about retirement at the age of 50. (An aggregate perspective)
- Compare how all of our MOP writers write about retirement at a particular time point, for example, during the economic crisis in 2008. (This is a synchronic or cross-sectional perspective)
There are many other variations in ways in which we can analyse coded scripts.
Maxqda software also allows us to collate demographic information about the MOP writers. Underpinning the software is a spreadsheet containing information on each document. Here we can gradually record information relating to each document and its author, such as: the author’s unique MOP number (e.g. S496), age when s/he wrote the document, gender, where the author was living at the time of writing, if s/he is in a long-term relationship, occupation at the time of writing. This spreadsheet will enable us to compare some of the key characteristics of the MOP writers with those of the individuals who have taken part in the surveys that we are using for the quantitative side of this mixed method project.
The language that writers employ can be interesting, informative, surprising and thought-provoking. Sometimes very close analysis of the language used in portions of scripts, or in whole scripts, can be as rich as looking at several scripts more superficially.
We have identified individual scripts, and portions of individual scripts, which we have scrutinised closely, analysing the language and discourses used. This type of analysis has provided fresh insights, particularly in relation to our understanding of individual attitudes towards volunteering
MAXQDA is helping us to follow individuals through time, and to organise and structure themes that are emerging out of MOP writing. However we are keen not to lose sight of the individual writers who make up our writers’ sample. We have, therefore, built up portraits of individual writers’ lives across the time in which they have been writing.
Portraits pull together, and chart the information that writers have provided in their writing on contemporary and recalled key life events. These vary according to each writer, but encompass attending school, gaining or not gaining qualifications, work histories, engagement in volunteering, births of children, and sadly, deaths of partners. One of our aims, in putting these biographical portraits together, has been to examine the difference between what we as researchers think of as an important life event, or useful piece of information, compared to the emphases that an individual writer places on these events.
We are aware that in 2008 MOP writers were asked to write for a directive entitled ‘Your Lifeline’, in which they were invited to discuss their key life events. We have wondered whether we should have chosen to sample, transcribe and analyse responses to this directive, rather than to collate this information through reading scripts written over 30 years. However, one of the advantages of longitudinal analysis is that things that are important to a writer at one point in time can hold less importance at other points in time. Building biographical portraits across time has allowed us to build up a cumulative picture of key life events, whilst also enabling us to focus on the substantive research aims of this project.
Two interesting findings have emerged from our various analyses. Firstly, we were unsure whether writers would write in the same way throughout their life-course. We found that some individuals vary the style, tone and voices that they employ in their writing. We are in the process of enquiring why this this variation has occurred. Does it relate to the passing of time, the subject matter, or other influences?
Secondly, throughout the process of analysing MOP writing, we have been aware of our position as researchers. The process of analysis can create a situation of looking at, gazing at, or ‘judging’ the writers – a hierarchical situation. We wondered what would happen if we simply allowed our responders’ writing to speak for them, if we as researchers were to abandon our hierarchical position, and to effectively ‘walk alongside’[i] our writers (Neale, 2012)?
We have experimented with I-Voice analysis (a technique that Ros Edwards, who is part of the team, has written about with Susie Weller[ii]). This involved looking at an individual writer’s use of personal pronouns (and sometimes personal adjectives) and the verbs that accompany these, in isolation from the rest of the text. The aim was to take a position of ‘walking alongside’ the writer, to let the text speak for itself rather than to contextualise it. The results were interesting and our view is that I-Voice analysis can provide additional insight into, and new ways of looking at, an individual writer’s identity and personality.
CONCLUSIONS ON QUALITATIVE ANALYTICAL METHODS
This is a very complex project. It explores time, the life-course, voluntarism and volunteering, using longitudinal and mixed data sources. In order to get to grips with these sources, concepts and substantive subject areas, we have had to draw on a variety of different methodological techniques, and be reflective and responsive when identifying the best methodological fit for the project.
[i] Neale, B, Henwood K, and Holland J (2012) ‘An introduction to the Timescapes approach: researching lives through time.’ in Qualitative Research 12: 4-15.
[ii] Edwards S and Weller S, 2012, ‘Shifting analytic ontology: using I-poems in qualitative longitudinal research’ in Qualitative Research 12 (2) pp.202-217