Collecting Qualitative Data

I spent the day on Thursday with the 16th Regiment of the Royal Artillery of the British Army.

My task was to have conversations with the different people in the army surrounding the service and support of equipment.

Having not done this for awhile, it took about an hour to warm up (a bit rusty) but once I got going, some of my old tactics came back to me. Later, I asked Guo Lei (my PhD student) to recount what she has learnt from the session. Guo Lei is undergoing training with me as she will be 'going native' for her data collection in understanding language and relationships and will be doing qualitative data collection. While it's fresh on my mind, I thought I'll write some if it down.

1. The objective of the conversation is not for them to answer the question. In qualitative research (particularly the grounded theory type), you do not really know what questions you want answered. So you just construct some questions as a framework and the true objective is to get GOOD QUALITY DATA. What is good quality data? Words that they MEAN, and not words that they think should be said. That would mean that you need to get data on 3 levels:

(A) The facts (objective facts that are verifiable to be the truth)
(B) Their opinion on the facts
(C) Their true feelings on the facts (which would give you a true picture of the reality)

(A) is easy. Ask the question, and they will answer. It's a fact finding exercise and gives the background context to the study.

(B) is a little harder. Instead of asking what is going on, you have to ask them what they THINK is going on. Junior researchers have difficulty with this. They think they know the answer and they don't ask obvious questions (examples below).

(C) is the hardest. Usually, (C) cannot be obtained instantly - you would need to go back and read the transcript, between the lines and understand the patterns and code them... HOWEVER, to get (C), you would need QUALITY DATA as QUALITY DATA allows you to get all (A), (B) and (C). That means you need to dig deep.

Here are some of the tactics I use:

1. Tactics for them to forget that they are being recorded.

This means no looking at a paper with questions. No looking away from them. Get the participant to lose himself or herself into the conversation. I use body language a lot. I lean back and relax, and they tend to mimic that action when they start to relax. Or I lean forward and listen intently like a student, as though they are 'teaching' me something. Or I look away and think about his/her point. I laugh, nod, frown, tease, in short - turn on every bit of charm, cajoling and engagement tactic I have. When people see you frown, they feel the need to clarify. When they see you smile, they focus on you. So body language of the researcher is the fastest way to get a respondent to forget his/her environment.

2. Tactics to get their feelings about something.

I start with simple questions. If I want them to divulge their feelings, I start with simple 'non-threatening' and 'non-revealing' feelings e.g. why don't you like your coffee with sugar? and they tell you why and how they feel about sugar etc. Once they start getting used to telling you how they feel, you can push the boundary further and ask about how they feel about other things.

There was a point in a conversation where the respondent said 'and they all wear the company corporate uniform' and I said 'uniform'? why? Now, I know perfectly why but I needed to hear it from them because when I hear it from them, they will use words that show if they agree or disagree to wear uniforms. Sometimes you don't need to ask a question outright about their feelings. Just asking why and getting their explanations will give you a good idea whether the 'reason' is something they agree or disagree with. Their words give positive or negative connotations.

3. Tactics to get answers but not ask questions and to confirm their feelings and perspectives

When a respondent says something, it is useful to echo back what s/he says. Reflecting back what was said allows the respondent to 'back-peddle', clarify, in short, shows what s/he truly feels. In many cases, I volunteer my own feelings. By saying 'I think that's terrible', they can say 'yes it is isn't it?' or 'not really, because...' when you volunteer your feelings, they volunteer their emotions. So you don't have to ask any questions, and they still give you the answer.

Another tactic is to answer a question yourself deliberately wrongly so that they can correct you and clarify why and how they feel about it.

Still another tactic is to use their terms and jargon, as though I'm in their 'inner circle'. For this project, I use it a lot, because the military has so much jargon. I use the acronyms as though I have been using this all my life and this helps with making them feel that they share a bond and reveal more.

I have put down just some of my tactics but they really come with a lot of practice and immersing yourself into a situation with the technique is crucial. To me, it's a bit like the flow and movement of martial arts. It's used without the respondent knowing it, and all tactics flow out during the interaction. EVERYTHING I do has relevance - a twitch, a smile, crossing hands in front of me - EVERYTHING. Human beings react to body language faster than through spoken language and it's very instinctive. Combining body language and my words, I get them to reveal what they feel. Their words ebb and flow differently when they respond to you and you have to be very very sharp to immediately grasp at a point, twist it around and change the direction of the conversation to go where you want it to go (which is tough because remember you don't know what you want, just that they have to reveal as much as possible). Whenever I do qualitative data collection, I always feel like I'm an instrument (this is also taught in class although I think many junior researchers don't understand what it really means). Over the years, I've just been sharpening 'me' as an instrument.

Guo Lei said that it's like a theatrical performance. She was all amazed with the technique. Funny, I've been doing it so long I've forgotten how it's like seen from the outside but she was sitting next to me watching the entire day. I guess seeing it first hand being used in real life rather than learning it theoretically in class can be quite amazing.

All I can say is that I was very drained at the end of it - doing this from 10 to 4pm was the longest I have done in a long time. It's even harder than going native. Going native is more natural and the respondent is often unaware of being observed or recorded. This is a conversation where the person knows he's being recorded and yet you have to use every tactic in the book to get the person to reveal as much as possible. You can tell them that it's all anonymous and their rights protected at all times but human beings are often paranoid when they are being recorded.

Yet, I am happy to know that I've still got it. I got loads of very good data (contradictions, revelations, emotions etc.) - but I still need to do much more.

The army officer I was interviewing was on to me I think. He gave me the best set of data I had for the day, revealing some very interesting details. Later, when the interview was over and as we had coffee, he asked - 'where did you learn your interrogation skills from? The CIA?' I was flattered, if not a little alarmed. The thing about this type of methodology is that no respondent should know how well the methodology is employed. It is indeed ironic that the best practitioners of this methodology should always be unknown.