10 tips for semi-structured qualitative interviewing

Many qualitative researchers spend a lot of time interviewing participants, so here are some quick tips to make interviews go as smooth as possible: before, during and after!

 

1. Let your participants choose the location

If you want your interviewees to be comfortable in sharing sometimes personal or sensitive information, make sure they can do it in a comfortable location. For some people, this might be their own house, or a neutral territory like a local cafe. Giving them the choice can help build trust, and gives the right impression: that you are accomodating them. However, make sure you make it clear that you need a relatively quiet location free from interruptions: a pub that plays loud music will not only stop you hearing each other, but usually makes recordings unusable!

 

2. Remember that they are helping you

Be polite and curtious, and be grateful to them for sharing their time and experiences. This always gets interviews off on the right foot. Also, try and think about participants motivations for taking part. Do they want the research to help others? Are they looking for a theraputic discussion? Do they just like a chat? Understanding this will help you guide the interview, and make sure you meet their expectations.

 

3. A conversation, not an interregation!

Interviews work best when they are a friendly dialogue: don't be afraid to start with some small talk, even when the tape is running. It turns a weird situation into a much more normal human experience, and starting with some easy 'starter for 10' questions helps people open up. Even a chatty "How did you hear about the project?" can gives you useful information.

 

4. Memorise the topic guide, but keep it to hand

Knowing all the questions in the topic guide can really help, so group them thematically, and memorise them as much as you can. It will really help the flow of information if you can segue seamlessly from one question to another relevant one. However, it's always useful to keep a print-out in front of you, not just for if you forget something, but also to make you seem more human, with a specific role. Joking about remembering all the questions is a great icebreaker, and it gives you something to look at other than the participant, to stop the session turning into a staring match!

 

5. Use open body language and encouraging cues

Face the participant in a friendly way, and nod or look sympathetic at the right times. Sometimes it's tempting for the interviewer to keep quiet during the responses, and not put in any normal encouraging noises like "Yeah", "Hmm" or "Right" knowing how odd these read in a transcript. But these are important cues that people use to know when to keep talking, so if you are going to drop them, make sure you make positive eye contact, and nod at the right times instead!

 

Quirkos - simple qualitative analysis software

 

6. Write notes, even if you don't use them

It always helps me to scribble down some one-word notes on the topic guide when you are doing an interview: first of all it helps focus my thoughts, and remind me about interesting things that the participant mentioned that I want to go back to. But it also helps show you are listening, and makes sure if the recording goes wrong, there is something to fall back on.

 

7. Write-up the interivew as soon as you finish

Just take 15 minutes after each interview to reflect: the main points that came up, how open the respondent was, any context or distractions that might have impared the flow. This helps you think about things to do better in the next interview, and will help you later to remember each interview.

 

8. Return to difficult issues

If a particular topic is clearly a difficult question (either emotionally, or just because someone can't remember) don't be afraid to leave the topic and come back to it later, asking in a different way. It can really help recall to have a break talking about something easier, and then approach the issue sideways later on.

 

9. Ask stupid questions

Don't assume you know anything. In these kinds of interviews, it's usually not about getting the right answer, but getting the respondent's view or opinion. Asking 'What do you mean by family?' is really useful if you discover someone has adopted children, step-sisters and a beloved family dog that all share the house. Don't make any assumptions, let people tell you what they mean. Even if you have to ask something that makes you sound ignorant on a specialist subject, you could discover that someone didn't know the difference between their chemotherapy and radiotherapy.

 

10. Say thank you

And follow up: send a nice card after the interview, don't be like a date they never hear from again! Also, try and make sure they get a summary of the findings of the study they took part in. It's not just about being nice, but to make sure people have a good experience as a research subject, and will want to be involved in the next project that comes along, which might be yours or mine!

 

I hope these tips have been hopeful, don't forget Qurikos makes your transcribed interviews easy to analyse, as well as a visual and engaging process. Find out more and download a free trial from our website. Our blog is updated with articles like this every week, and you can hear about it first by following our Twitter feed @quirkossoftware.

 

 

Quirkos is coming...

key

 

Quirkos is intended to be a big step forward for qualitative research. The central idea is to make text analysis so easy, that anyone can do it.

That includes people who don't know what qualitative analysis is, or that it could help them to better understand their world. This could be a council or hospital trust wanting to better understand the needs of people that use their services, or a team developing a new product, wanting feedback from users and consumers.

And for experienced researchers too, the goal was to create software that helps people engage with their data, rather than being a barrier to it. Over the last decade I've used a variety of approaches to analysing qualitative research, and many collegues and I felt that there had to be a better way.

Quirkos aims to make software to easily manage large projects, search them quickly, and keep them secure. To visualise data on the fly, so findings come alive and are sharable with a team of people. And finally to make powerful tools to sort and understand the connections in the data.

After years of planning, these pieces are finally coming together, and the prototype is already something that I prefer using to any of the other qualitative software packages out there. In the next few weeks, the first version of Quirkos will be sent to intreped researchers around the globe to test in their work. A few months later, we'll be ready to share a polished version with the world, and we're really excited that it will work for everyone: with any level of experience, and on pretty much any computer too.

There are a lot of big firsts in Quirkos, and it's going to be exciting sharing them here over the next few weeks!

An overview of qualitative methods

There are a lot of different ways to collect qualitative data, and this article just provides a brief summary of some of the main methods used in qualitative research. Each one is an art in its own right, with various different techniques, definitions, approaches and proponents.

More on each one will follow in later articles, and it’s worth remembering that these need to be paired with the right questions, sampling, and analysis to get good results.

Interviews

Possibly the richest, and most powerful tool: talking to someone directly. The classic definition is “conversations with a purpose“, the idea being that there is something you are interested in, you ask questions about it, and someone gives useful responses.

There are many different styles for example how structured your questions are (this paper has a wonderful and succinct overview in the introduction). These can range from a rigid script where you ask the same questions every time, or completely open discussion, where the researcher and respondent have freedom to shape the conversation. A common middle ground are semi-structured interviews, which often have a topic guide, listing particualar issues to discuss, but will allow questions for clarification, or to follow up on an interesting tangent.

Participant Observation

Often the remit of ethnography or sociology, participant observation usually involves watching, living or even participating in the daily life of research subjects. However, it can also involve just watching people in a certain setting, such as a work meeting, or using a supermarket.

This is probably the most time intensive and potentially problematic method, as it can involve weeks or even years of placement for a researcher, often on their own. However, it does produce some of the richest data, as well as a level of depth that can really help explain complex issues. This chapter is a fine starting point.

Focus groups

A common method used in market research, where a researcher leads a group discussion on a particular topic. However, it is also a powerful tool for social researchers, especially when looking at group dynamics, or the reactions of particular groups of people. It’s obviously important to consider who is chosen for the group, and how the interactions of people in the group affect the outcome (although this might be what you are looking for).

It’s usually a quicker and cheaper way of gauging many reactions and opinions, but requires some skill in the facilitator to make sure everyone’s voice is being heard, and that people stay on track. Also a headache for any transcribers who have to identify different voices from muffled audio recordings!

Participant Diaries

Getting people to write a diary for a research project is a very useful tool, and is commonly used in looking at taboo behaviours such as drug use or sexuality, not just because researchers don’t have to ask difficult questions face-to-face, but that data can be collected over a long period of time. If you are trying to find out how often a particular behaviour occurs, a daily or weekly record is likely to be more accurate than asking someone in a single interview (as in the studies above).

There are other benefits to the diary method: not least that the participant is in control. They can share as much or as little as they like, and only on topics they wish to. It can also be theraputic for some people, and is more time flexible. Diaries can be paper based, electronic, or even on a voice recorder if there are literacy concerns. However, researchers will probably need to talk to people at the beginning and end of the process, and give regular reminders.

Surveys

Probably one of the most common qualitative methods are the open ended questions on surveys, usually by post, on-line, or ‘guided’ by someone with a clipboard. Common challenges here are

  • Encouraging people to write more than one word, but less than an essay
  • Setting questions carefully so they are clear, but not leading
  • Getting a good response rate and
  • Knowing who has and hasn’t responded

The final challenge is to make sure the responses are useful, and integrating them with the rest of the project, especially quantitative data.

Field notes

Sometimes the most overlooked, but most vaulable source of information can be the notes and field diaries of researchers themselves. These can include not just where and when people did interviews or observations, but crucial context, like the people who refused to take part, and whether a interviewee was nervous. It need not just be for ethnographers doing long field work, it can be very helpful in organising thoughts and work in smaller projects with multiple researchers.

As part of a reflexive method, it might contain comments and thoughts from the researcher, so there can be a risk of autobiographical overindulgence. It is also not easy to integrate ‘data’ from a research diary with other sources of information when writing up a project for a particular output.

 

This is just a whistle-stop introduction, but more on each of these to follow…