Obtaining the data you really need: on conducting qualitative interviews

I have been planning a new qualitative research project, and reading draft proposals and draft methodology chapters for students I am coaching, so I have been thinking about qualitative data lately; particularly how to get the right kinds of data from participants when we are conducting interviews.

There are three main forms of interview that are discussed in methodology texts and guides: structured, semi-structured and unstructured. Generally, when embarking on a qualitative project, students tend to opt for semi-structured interviews. This enables them to have a set series of questions, hopefully well connected to their theory and literature, but also to create space for participants, or interviewees, to add views and insights that may not strictly follow the questions. It’s sort of a best-of-both-worlds scenario. Unstructured interviews are hard to manage, especially for postgraduate students, many of who are doing this form of fieldwork for the first time. And fully structured interviews can veer into not allowing any space for additional, unpredicted insights and information – they can, in other words, limit the conversation, and also the kind of data collected. *Caveat: you do need to choose the right tool for your project, regardless of whether it is hard to do or not.

But, semi-structured interviews, while they seem to be a dominant preference for many students doing qualitative research via interviews, are not easy to do well. A key issue I am thinking about in relation to my new project, and that I assist other researchers with, is this: how do I conduct the actual interview so that I get the data I really need? What tools or techniques so I need to be aware of? I have listed a few points here that have helped me, and that are also the product of my own mistakes and learning in conducting qualitative interviews.

interview 1

Photo by nappy from Pexels

  • The point of the first interview is to get a second interview. 

I didn’t know this during my own PhD, which was also my first major research project. I did one set of interviews, and I do recall, on reading the transcripts, wishing that I could go back and ask some of the questions differently or find out more about specific issues. But I didn’t really know I could (and I didn’t plan for this so I ran out of time). But, now it makes complete sense. Unless you are interviewing someone you already know well, chances are you and the interviewee will be strangers to one another. Establishing a rapport, and trust, can take time – certainly more time than one interview allows for. So, you need to make the goal of the first interview establishing a connection that will enable you to go back, and talk to that person again. You may get the basic data you need in the first interview, but chances are that you will need more that interview #1 can yield. You will want richer, or deeper responses to some of the questions, or will see that certain questions could be asked differently, to yield slightly better, or more relevant, responses considering your research questions and aims.

Plan enough time for at least two interviews, and ensure that you check with the interviewee that you could approach them again if needed. Listen to the recording of the interview as soon as you can, then, make notes, and think about what the data is saying in relation to those research questions. Then work out whether you have enough of the rights kinds of data to include in your analysis, where any gaps are, and plan interview #2 accordingly.

  • Record your interviews and take notes.

You always want to audio-record the interviews, so that your transcription, analysis and reporting will be accurate, and rich. But, you also want to take notes – not so many that you are unable to maintain eye contact and engagement in the interview, but have a notebook and pen to record things that the audio may not capture. Perhaps body language is a factor to consider in terms of the kind of data you are generating – if participants appear nervous, or distracted, this may be something you want to consider in the analysis alongside their words. Perhaps what their words say, and what their body language says are two different things, and this could be important as a possible finding. Perhaps they mention names of other people you could consider interviewing as well, or names of websites, documents and other sources of information you need to follow up on. All of this can be captured in a few short notes, and can add to the richness of your overall findings and analysis later on.

  • Plan for a pilot interview.

This has been a big point of learning for me. I tend to get over-eager in interviews, and I just plain talk too much. I find myself listening to the audio and cringing, and wishing I could just be quiet, and let the interviewee talk! There are two benefits, here, to a pilot interview: the first is that you can practise being an interviewer. You can work out how to record and take notes, how to talk less and listen more, how to direct and redirect the conversation as needed, and how to manage your own facial expressions and body language, to remain as neutral as possible so as not to bias or shift the interview in unhelpful directions (like looking shocked or disapproving at something an interviewee says to you). The second is that you can figure out whether the style you are using and questions you have as a guide are yielding the answers you need. Are there too many questions? Are some of the questions too long and ‘wordy’? Do any of the questions yield answers that seem a bit off topic, or that are repetitive? Make notes and adjust the interview plan before the real interviews begin. You need to do a pilot with someone who represents the interviewee demographic, and preferably with enough time to make changes and adjustments before you start meeting with the study’s group of participants.

interview 2

Photo by nappy from Pexels

  • Plan for a conversation, rather than a Q&A session.

This has been a really big point of learning for me. In a qualitative study, regardless of the topic, you need rich, detailed data. Qualitative research is about depth, rather than breadth, in simple terms. Thus, you need depth in the data, to achieve depth in your analysis, findings and conclusions. Thus, what you want to do with conducting the interview is create a conversational space, rather than a stiff, Q&A session where you ask a long list of questions, and the interviewee responds. The best way, I find, is to start off by asking the interviewees to tell me a story, related to the research I am doing and my interview questions. For example, if I am doing a project on postgraduate supervision, and my interviewees are students and supervisors, I don’t want to start off with something like: “Do you enjoy supervising students, and why or why not?” (Typical kind of question you find in interview schedules). This will not help me to get a richer sense of what the supervisor does as a supervisor, and what aspects they do and do not enjoy. So, what I could rather do is ask the question like this: “Can you start off by telling me a bit about your current supervision situation?” And then just listen. In that response, then, I listen for what aspects they seem to like, or not like, and in the follow up, I could ask: “You seem to find supervision a struggle, rather than a joy, right now. Would you say that is accurate? Could you say a bit more about that?” And then listen. And so on. So my questions become a guide, but not a determinant to the structure of the interview.

The main thing in doing qualitative interviews, I am learning, is for an interviewer to have empathy. Trying to be in the moment, and create a sense for the interviewee that their stories are valuable, and worth sharing and hearing, enables the creation of a rapport that can lead to follow-on interviews, and that can encourage interviewees to see me as an ally, and someone who will share their stories responsibly, ethically and with care. This is crucial when you are working with people who trust you enough to give you truthful stories, experiences and accounts. Ultimately, you need to listen closely, and record accurately, so that what you share in your study is able to meaningfully shape knowledge, research and practice in your field.

Developing well-constructed data gathering tools, or methods, for your study

I spent the better part of last week working with emerging researchers who are at the stage of their PhD work where they are either working out what data they will need and how to get it, or sitting with all their data and working out how to make sense of it. So, we are talking theory, literature, methodology, analysis, meaning making, and also planning. In this post I want to focus on planning your data gathering phase, specifically developing ‘instruments’, such as questionnaires, interview schedules and so on.

tools

Whether your proposed study is quantitative, qualitative or mixed methods,  you will need some kind of data to base your thesis argument on. Examples may include data gathered from documents in the media, in archives, or from official sources; interviews and/or focus groups; statistical datasets; or surveys. Whatever data your research question tells you to generate, so as to find an answer, you need to think very carefully about how your theory and literature can be drawn into developing the instruments you will use to generate or gather this data.

In a lot of the postgraduate writing I have read and given feedback on, there are two main trends I have noticed in the development of research methods. The first is what I considered ‘too much theory’, and the other ‘not quite enough’. In the first instance, this is seen in researchers putting technical or conceptual terms into their interview questions, and actually asking the research questions in the survey form or interview schedule. For example: ‘Do you think that X political party believes in principle of non-racialism?’ Firstly, this was the overall research question, more or less. Secondly, this researcher wanted to interview students on campus, and needed to seriously think about whether this question would yield any useful data  – would her participants know what she meant by ‘the principle of non-racialism’ as she understood it theoretically, or even have the relevant contextual knowledge? Let’s unpack this a bit, before moving on to trend #2.

The first issue here is that you are not a reporter, you are a researcher. This means you are theorising and abstracting from your data to find an answer that has significance beyond your case study or examples. Your research questions are thus developed out of a deep engagement with relevant research and theory in your field that enables you to see both the ‘bigger picture’ as well as your specific piece of it. If you ask people to answer your research question, without a shared understanding of the technical/conceptual/theoretical terms and their meanings, you may well end up conflating their versions of these with your own, reporting on what they say as being a kind of ‘truth’, rather than trying to elicit, through theorising, valid, robust and substantiated answers to your research questions, using their input.

This connects to the second issue: it is your job to answer your research question, and it is your participants’ job to tell you what they know about relevant or related issues that reference your research question. For example, if you want to know what kinds of knowledge need to be part of an inclusive curriculum, you don’t ask this exact question in interviews with lecturers. Rather, you need to try and find out the answer by asking them to share their curriculum design process with you, talk you through how they decide what to include and exclude, ask them about their views on student learning, and university culture, and the role of the curriculum, and knowledge, in education. This rich data will give you far more with which to find an answer to that question than asking it right out could. You ask around your research questions, using theory and literature to help you devise sensible, accessible and research-relevant questions. This also goes for criteria for selecting and collating documents to research, should you be doing a study that does not involve people directly.

analysis of data

Photo by Startup Stock Photos from Pexels

The second trend is ‘not enough theory’. This tends to take the form of having theory that indicates a certain approach to generating data, yet not using or evidencing this theory in your research instruments.  For example structuralist theories would require you finding out what kinds of structures lie beneath the surface of everyday life and events, and also perhaps how they shape people, events and so on. An example of disconnected interview questions could be asking people whether they enjoy working in their university, and whether there are any issues they feel could be addressed and why, and what their ideal job conditions would be, etc., rather than using the theoretical insights to focus, for example, on how they experience doing research and teaching, and what kinds of support they get from their department, and what kinds of support they feel they need and where that does and should come from, etc. You need to come back to using the theory to make sense of your data, through analysis, so you need to ensure that you use the theory to help you create clear, unambiguous, focused questions that will get your participants, or documents, talking to you about what matters to your study. Disconnecting the research instruments from your theory, and from the point of the research, may lead to a frustrating analysis process where the data will be too ‘thin’ or off point to really enable a rich analysis.

Data gathering tools, or methods for getting the data you need to answer your research questions, is a crucial part of a postgraduate research study. Our data gives us a slice of the bigger research body we are connecting our study to, and enables us to say something about a larger phenomenon or set of meanings that can push collective knowledge forward, or challenge existing knowledge. This is where we make a significant part of our overall contribution to knowledge, so it is really important to see these instruments, or methods, not as technical or arbitrary requirements for some ethics committee. Rather, we need to conceptualise them as tools for putting our methodology into action, informed and guided by both the literature our study is situated within as well as what counts as our theoretical or principled knowledge. Taking the time to do this step well will ensure that your golden thread is more clearly pulled through the earlier sections of your argument, through your data and into your analysis and findings.

 

 

PhD workout: warming up your writing muscles

So, I am writing a book. I have been sort-of-kind-of writing a book for a long time now. We have an on and off relationship, my book and I. But, a proposal is being reviewed, and the hope is that the feedback will be a green light, so I have to get writing. And soon. But, I am a bit out of practice. I wrote a fair bit last year – 3 book chapters (a few drafts each) as well as part of a paper with colleagues. But this is a different beast altogether – as long and as complex as a PhD thesis. I am finding I am out of shape here.

This is not an unfamiliar feeling. I wrote here and here about moving from one year of PhD or post-doc into the next, after having a break and getting a bit flabby around the writing middle, so to speak. I know, therefore, that I have felt unfit before, and have made myself fitter and gotten the writing work done. But, this is – like actual fitness – hard work and requires a level of emotional and psychic energy that can be hard to find sometimes. I have decided, therefore, that I am going to start with gentle warm-ups, rather than jumping straight into the whole thing (Thank you, Roger Federer :-)).

rfed warm up

The first thing I am doing is starting with something manageable, that I could want to do every day – or at least 4-5 times a week. If I want to do it, and it feels manageable, it is very likely I will actually do it (and enjoy the experience). Instead of doing what I too often do, and writing ‘Chapter 1 draft’ on one day of my calendar, I am writing ‘one pomodoro’ every other day. I can do this. It’s 30 minutes of writing. I can then tick this off, and actually add days as a I go, or keep it every other day and work up to 2 pomodoros at least. If I can do it, I won’t fail, and if I don’t fail, I can keep enjoying this writing time and make it productive. Too often I set myself overly lofty goals, in life and writing, and set myself up to fail rather than succeed. Last week I wrote my first blog post in over 4 months, scheduled this post, and also managed about 1000 words on my book. HUGE success I say. All in these little manageable chunks.

The second thing I am going to do is keep it steady. Rather than having a good week, and thinking I can now escalate to high levels of writing productivity, I am going to keep going at this pace for now. Probably, realistically, this will be the pace for the year, with bursts of higher productivity around deadlines and when I have excess time and energy. As one of my writing students said to me last year: ‘Eat the elephant one bite at a time’. Apologies to elephant lovers – I am one too – but this is a good metaphor for taking it steady with life and writing. One task, one pomodoro, one idea at a time. This way, things actually do get done as opposed to being menacing, un-ticked-off tasks on your to-do list.

Finally, for now anyway, I am going to get me some writing buddies. Face-to-face if I can, but virtually if not. I am always thinking I should join a Twitter shut-up-and-write group, or create my own writing group. And then work, and kids, and life, and my writing gets pushed down (with me attached) to the bottom of my list. My writing time is also time for me – it’s personal as well as professional. So, I have to actually value it, and myself. As a working mother I am too often too far down my list. And so is my writing. I am hopeful, that with positive peer encouragement, we can collectively make our writing more present each week in the to-do lists, and make appreciable progress on our projects.

group yoga

Warming up these tired writing muscles to fuller strength will take some time – what do people say?If it’s too easy you’re not doing it right? Maybe so. I don’t think writing should always be hard, but good writing should take effort and time. Maybe you are in this spot too, coming back to work and PhD and research writing, and working out how to begin your “elephant meal”. Hopefully some of these steps to warming up your writing muscles will help you, too.

If you have other ideas, please share in the comments. All the best for 2019!

Having theory and ‘theorising’: reclaiming the verb

Jeepers, I have been away a while! But, rather than dwelling on all the Things I let get in the way of my writing and creativity last year, I will look to a new year, and new ideas for the blog instead. In a conference workshop in November, I had a really interesting discussion with colleagues about theory – where theory comes from, and whether its origins are relevant to its current use (for a later post). Part of that discussion focused on how we use theory. It is this I would like to focus on in this post, to kick off the new year.

Theory is key part of research – particularly at postgraduate and doctoral level. In the social sciences we tend to really spell out the theory we use – critical social theory, behavioural theory, post-structural theory and so on, whereas is the natural sciences there is less obvious mention of theories although they are most certainly there (I am told by scientist friends it is usually some version of positivism, but I stand to be corrected). In essence, theory is always present when we are trying to make sense of a smaller part of the bigger picture – connecting the specifics of our study with a more general or wider phenomenon. But, the way we use theory to do this is often inadequately talked about or grappled with in postgraduate spaces. Do we ‘have theory’ or do we ‘theorise’? 

question mark

Patrick White, in an excellent book on research writing, argues that to be called by such a name, theories need to be abstract, explanatory and testable. In other words, they need to be able to apply across more than one field; they need to explain, rather than just name, the part of the world we are researching; and they need to be able to be tested, to see whether and how they apply to the problem we are researching. But, in many projects, especially those where the theory needs to shape choices made in the research design, methods, and analysis of data, theory can be under-utilised in the act of making meaning and creating new knowledge. In other words, I am arguing that there are projects that ‘have theory’, but this is not adequately used in the act of ‘theorising’ new understandings and explanations of the problems we research. 

So what, then, do I mean in my distinction between ‘having theory’ and ‘theorising’?

Let’s start with theorising. This is an act – something we have to do, usually in an iterative manner. This is the act of bringing theory and data into conversation with one another, in the process seeking meanings that will help us both explain the nature of the problem we are researching, and test the viability of the theory we have chosen in making sense of that problem. We are also locating the problem, through theory, literature and methodology, within the bigger picture it connects with. For example, using critical social theory, such as Margaret Archer’s morphogenetic cycles to critique and understand processes of change or stasis in one university’s leadership environment (understood within the broader context of neoliberalism and marketisation of higher education). To theorise, we need to make the theory do work. The work of making sense of the world – through a chosen and always partial lens – but making sense nonetheless.

When we use theory to theorise – to seek, debate and construct meanings so as to create and add to knowledge about the part of the world we are researching – we need to acknowledge that the theory we have chosen is but one of probably a number of theories that could offer an explanation. Research is about looking for truth – a form of truth that enables us to see, know and act with greater knowledge and insight. Thus, the theory we choose, following White, must be able to offer the kinds of explanations that enable this. Your research question – the problem you want to solve – must guide the choices you make here: what theoretical tools, frameworks, explanations will enable you to solve this problem in the clearest, most useful way at this point in time? We choose, and build, theoretical frameworks, or theoryologies as I have called them, based on the problems we are trying to solve, or the truths we seek. They are not just found, fully formed, ready to be placed into the right part of the thesis. 

constructing theory

Photo by David McBee from Pexels

This brings me to dissertation work that has theory, but does not fully theorise. There may be a chapter entitled ‘Theoretical framework’ or some version of that, and the theory is laid out and explained. (In the sciences the theory may be implied in the earlier parts of the thesis rather than explained in its own chapter.) Yet, when the analysis and data sections are reached, the theory is oddly silent, or under-utilised. Rather, you may find a thematically organised discussion of the data that recounts what the researcher has found, and makes suppositions and suggestions of what it could mean without fully engaging the powerful theory to really make meanings that can, following White again, show that the theory has been tested, and is able to explain the data in ways that enable the researcher to create knowledge that adds to what we already know. These findings can then be built on by other researchers as part of the abstracted explanations they may offer for why different parts of the truth, and new solutions to current problems still need to be sought, and found. 

Theory is powerful. Or, it can be when used to actively theorise – to make meanings that are new, or additional to those we already have access to. This process is not simple, or linear. It requires immersion in your data, it requires you to suspend some of your assumptions or beliefs about what is truth, and what your data is saying, so you can allow the theory to act as a lens with which to look at your data with more ‘naïve’ or unassuming eyes. Theorising is an iterative, at times frustrating process, that is both intellectually and personally challenging. But, skimping this process to get the thesis done belies a misunderstanding of what the doctorate – or research – is. It is not (just) the thesis, or the paper. It is the process of engaging with current truths and meanings, findings under-explored problems and questions, and working to make meanings that will add to our knowledge about the world, and how we live in it (and that will transform your own understandings and knowledge).

In these uncertain, and challenging times in which we all live, we all need to embrace the difficult act of theorising – we need to reclaim the verb, over the noun. We need to embrace the research process and the learning therein – both personal and professional. The product that emerges in the form of that thesis or paper – and its author and readers – will be the richer for it. 

 

On acronyms in academic writing

I am not a huge fan of acronyms. I feel I should start with this disclaimer. I know that they serve a purpose in academic writing, and I do use them. But with caution, and only when needed. I think acronyms are, essentially, un-reader-friendly, and should be used judiciously to create and communicate meaning.

Photo by Alex Andrews from Pexels

Let’s start with what is useful about acronyms. Firstly, they can save you space and typing time. If you have a long term you need to use, such as “Southern African Development Community”, and you’ll be writing this several times in your paper or thesis, you can shorten it to SADC. This will reduce your overall word count, and also type 4 letters every time you use it, instead of 4 longer words. 

There are accepted acronyms in every field that you and your colleagues and peers will know and use. Think of CHAT (Cultural Historical Activity Theory) in Education, or RAM (Random Access Memory) in Computer Science, or WRT (with respect to) in Mathematics. To use these accepted, known acronyms is to signal your membership of your academic community of knowledge-making and knowers

However, as a writer I think it is useful to put myself in the position of my readers, and read the text, with the acronyms, as they might. Read this:

SG and SD are realised in terms of their relative strength or weakness, and brought together these two organising principles create semantic codes that reveal combinations of stronger and weaker SG and SD together. These codes shift and move over time as SG and SD strengthen and weaken in relation to one another. These movements form what LCT terms a ‘semantic wave’, which can be used to map a teaching and learning event, such as a lecture, part of a lecture or a whole series of lectures (see figure 1). Inverse movements of SG and SD – where SG is stronger at the same time as SD is weaker for example – are potentially important for cumulative knowledge building, as we shall see in the following section. It should be noted, here, though, that SG and SD do not necessarily strengthen and weaken inversely (Maton, 2013), although it is these kinds of waves, for the purpose of illustration and brevity, that will be focused on in this paper.

How do you encounter this text as a reader? You can assume, coming from the middle of a paper, that SG and SD have been defined earlier in the paper. Do you find these easy to make sense of?

Confession, this is a draft of a paper I wrote a couple of years ago. This is how I wrote it. But, when I got the reviews back, the reviewers both pointed out that the use of all of these acronyms had an alienating effect on the reader, especially as these refer to theory, which can already be difficult for readers new to it. I therefore rewrote this paragraph (and subsequent similar paragraphs):

Semantic gravity and semantic density are realised in terms of their relative strength or weakness, and brought together these two organising principles create semantic codes that reveal combinations of stronger and weaker semantic gravity and semantic density together. These codes shift and move over time as semantic gravity and semantic density strengthen and weaken in relation to one another. These movements form what LCT terms a ‘semantic wave’, which can be used to map a teaching and learning event, such as a lecture, part of a lecture or a whole series of lectures (see figure 1). Inverse movements of semantic gravity and semantic density – where SG is stronger at the same time as SD is weaker for example – are potentially important for cumulative knowledge building, as we shall see in the following section. It should be noted, here, though, that semantic gravity and semantic density do not necessarily strengthen and weaken inversely (Maton, 2013), although it is these kinds of waves, for the purpose of illustration and brevity, that will be focused on in this paper.

How does it read now? A little easier to follow? I think so. The thing that concerns me about acronyms, even the accepted ones, is that readers don’t always read out the term in full in their heads. Sometimes, they don’t read ‘SADAC’ or ‘semantic gravity’. Sometimes they read ‘ESS-AYE-DEE-CEE’ or ‘ESS-GEE’. And the more they do the latter, the less readerly the text becomes. Your reader can end up feeling alienated from the meanings you are making, and communicating to them. Readers who have to work too hard to make sense of your text, and remember what all the acronyms stand for, are likely not going to enjoy the reading experience.

I have become, through the process of writing and revising this, and a couple of other papers, more aware of the ‘acronymising’ I do in my writing. I have also become more aware of it in my students’ texts, as I read and offer them feedback. And, my observations and writing practice have led me to this advice:

  1. Try to stick only to the accepted, known acronyms, as far as possible in your text. Try not to create acronyms where there don’t need to be any (like SA instead of South Africa, or HE instead of higher education). 
  2. Put yourself in your reader’s head, and read your text aloud. Do the acronyms work, or does it sound odd, or confusing after a while to have as many as you have included? 
  3. Always define the term you are acronymising first – this is basic, but often something writers forget to do, especially when they know their field well. 
  4. I try to create text-by-text guidelines for myself – if I have a long text, like a thesis, I will use the acronyms carefully, and probably redefine them chapter by chapter to remind my reader what they mean. If I have a shorter text, like a paper, I won’t need to do this. I also try not to include too many acronyms, so I choose the ones that will be most useful and necessary in terms of saving words and typing time, and signalling my knowledge of the field 

I hope this advice helps you to consider your use of acronyms, and focus less on making your job as a writer easier, and a little more on making your text reader-friendly, and your meanings accessible and clear.