Who goes to school during their holiday? Nerds. Says the girl typing this blog post from her dorm room while supposedly being on holiday. It's been just a week and a half since my official last day as a Partner & Director at InWithForward. I stepped down from my day-to-day role there and am continuing to stay connected to the IWF work and superstar team, now as Senior Advisor. After a whirlwind 2 years with IWF, I am taking this month for myself... and taking a course in social research methods at the London School of Economics to kick it off. This is the beginning of processing my whole incredible experience with IWF and of figuring out what will be next for me - I have some ideas ;)
The LSE program has been awesome so far.
At InWithForward, the lab approach blends social science rigour with strategic design methods. Some of my favourite parts of this approach have been the social sciency bits, especially ethnographic research and all that goes with it. So when I found this program, I knew I had to apply.
So here I am, nerding out on qualitative research methods with 40 other practitioners and scholars from around the world - including a doctor from Ireland, lawyers from Spain and from Kenya, business university directors from India, an energy advisor from Saudi Arabia, researchers from Croatia and UK and Norway and Germany and Indonesia, an educator from Belgium (from Antwerp to be exact!), development workers from UK and Kyrgyzstan... and it goes on. Needless to say I'm the least interesting person in the room and I'm totally loving it.
One thing I learned from being part of developing Kudoz at InWithForward, is that: humans learn from reflecting on an experience, not from the experience on it's own. That means: for me to get the most from this course, I need to spend time reflecting on what I'm learning.
Below are some of my reflections... they're still rough but they're there - done is better than perfect.
From the book "Qualitative research for development: A Guide for Practitioners" by Morten Skoval & Flora Cornish, which was highly recommended by the prof.
TAKEAWAYS thus far
1. Opinions are being formed and reformed all the time.
A constructivist point of view starts from the premise that our opinions are fluid rather than fixed. That our opinions are being formed and reformed all the time based on what we come in contact with in the world, who we're speaking with, what we're reflecting on - basically that humans are messy complex creatures. That we don’t have a set opinion, that things can change. Makes sense to me.
What's interesting for social research methods is that it means different kinds of information (data) will be unearthed in interviews versus in focus groups. That's because group conversations are a place where opinions and stances are formed, shifted, clustered, negotiated and debated. Where as in interviews, it's a two way exchange and the interview subject may present a particular side of themselves to you.
And, the Opinions Are Fluid perspective becomes interesting when thinking about the most common criticism of focus groups: that people's opinions will be swayed or silenced by whoever is the more dominant voice in the group.
BUT A) if we accept that opinions are constantly shifting, then this is just part of what group discussion is about, so what we want to watch for are the dynamics and social norms about the topics are. At what point are people being silenced, how and why might that be happening, what can that tell us about attitutes? AND B) group composition plays a role - not having obvious power clashes or situations where some members of the group are intimidated to speak up.
2. Focus groups give us a window into what people really think - the noble, the shameful, the gossip.
Focus groups are good at gathering opinions or understanding how opinions shift or are negotiated or debated in groups. They can tell us about what are the complexities of issues - which likely won't come out in an interview, for ex. because you may not want to press on about a sensitive topic but a peer may be able to.
It seems counter-intuitive, but people may be more likely to say more taboo things in a group, if the group composition is right and if the group feels safe. For ex. in an article about attitudes about HIV/AIDS, a focus group members revealed racist opinions about where aids comes from. In a one-on-one interview, the same person may not feel the support of the group in order to be able to say waht they really think, for fear of judgement.
Focus groups are also a great way to learn about a group's shared reference points, slang or jargon, the latest gossip. For ex. if you were studuing 16 year old girls attitudes towards social media, they are going to have a very different conversation with you, an adult, than with their gal pals. Because teenagers relate differently among their peers than they do with adults.
So focus groups can be a particularly powerful tool in getting useful and interesting data.
However, focus groups are not great at eliciting people's personal stories. Because telling your own story in a group setting isn’t so helpful for conversation flow. People sit there politely waiting for you to finish. They can’t debate it bcause it's your story. But if it’s an opinion, they can jump on it and make jokes about it or refute it or whatever else.
3. Grounded Theory
InWithForward uses grounded theory as a research design approach, and I've heard a lot about Grounded Theory from Sarah and Daniela. So I was particularly interested in how they explained this topic. Grounded Theory emerged in response to the prevailing rigid research methods of the time - researchers felt they needed to get back to the data and be led by the data. The main point about Grounded Theory is its heavy focus on being inductive and working from the data without preconceptions. That means you don't start with a fixed research questions, rather your question gets more and more refined as you go along, and you keep circling back to get more indepth data based on what you're hearing.
This is consistent with how Kudoz emerged: first from a study about social isolation, then as more was learned, the research focus was refined to studying poverty of experiences among adults living with a disability. It's really neat to be learning, through this course, about the methodological decisions that Sarah has been making in her research design / it is making the course material come alive.
summary slide about Grounded from the course
4. Reading about a study's Research Methods turns out to be super interesting!
Qualitative Researchers spell out, in great detail, exactly what they did in their study.
This is because rigour is determined by transparency and soundness of the methodology rather than by replicability. It would be impossible to have the exact same interview twice - people won't say the same thing, things will have changed in their lives, and anyway that's not the point with qualitative research. By spelling it out, the reader can decide for themselves whether the research design and execution is sound.
And, we also have to keep in mind that methodological decisions are made around constraints and practicalities. So one must read the research methods with practical considerations in mind. For ex. in a report about infertility among women in Milawi, the researchers recruited their research sample by asking nurses to recommend women who had infertility problems in the past. The researchers would then approach the women, let them know that the nurse recommended them, then ask them if they indeed had fertility problems. If the woman agreed, they could be part of the research sample. This recruitment approach is problematic because a) there is pressure to say yes, since the nurse said so, and b) they may not actually think of themselves as infertile but then are asked to confirm this label - and there is tremendous stigma related to infertility among this community, so it could be distressing to be labelled as such if you did not think fo yourself as such. However, if we consider the context and constraints, going through the nurses may have been the only practical way to recruit research participants.
Moral of the story: qualitative research is murky and complex. As the Researcher, the best you can do is be as ethical as possible (including prior approval by an ethics board), lay out all the details of what you did, and let the reader decide whether it is sound.
5. There are power imbalances between researcher and research subject - act responsibly.
With design, the user is king. The user has the opportunity to reject the thing that was designed. If the user thinks it doesn’t work, doesn't like it, doesn’t choose it, then who cares. It is worthless.
However, for many researchers, the intent is not to create a product or a service that will be used by the people being studied. Rather, the intent is usually to further knowledge around a certain phenomenon.
So what happens in practice is that: the researcher goes into the field, extracts data from the field, analyses the data, writes it up, and uses it for their own purposes - be it teaching, presentations, in journals, to share with the social science community. So the researcher has power over what happens with the data and the purpose is not necessarily to create value for the people they interviewed - and the language is about "not doing harm" rather than "doing good for". Pretty different stances.
This difference is really intersting to me, as it creeped into assumptions about methodological considerations. Still stewing on this one.
6. Some questions that remain about interviewing...
To what extent is the researcher guiding the interview and facilitating, and to what extent do you let it go naturally in the direction it wants to go - like a conversation. Obviously you have goals of the types of things you want to know about. But you are there to listen.
How direct can you be with your questoins, especially difficult questions - what can be learned from the field of journalism about probing into sensitive or controversial topics? And what ethical considerations are needed so that you don't cause harm to the interviewee by being too intense.
.... more reflections coming soon!
TWEETS FROM SATSUKO
I wrote about how reinforcement learning and multi-armed bandits power the recommendation engines we know and love… https://t.co/sKbEn4Ah2b12 hours ago
“Questions to ask before starting user research” by @mmargolis https://t.co/hkXNMcK6E613 hours ago
“Never Ask What They Want — 3 Better Questions to Ask in User Interviews” by @chuckjliu https://t.co/WtOjDAYlBs13 hours ago