168 – Collecting Data for Community Research

Kate: [00:00:00] Welcome to the IDEMS podcast. I’m Kate Fleming, a director of IDEMS, and I’m here with Lucie Hazelgrove Planel, who I just gave a new title to, which we were just discussing, which is Community Research Methods Lead, because that is really what you do, although you’re also technically a Postdoctoral Impact Activation Fellow. 

Lucie: No, no, no. I haven’t been that for a while. 

Kate: So then what would your title have been? 

Lucie: Social Impact Scientist. 

Kate: Social Impact Scientist. You are definitely, yeah.

Lucie: Ironically, a lot of my work is working with statisticians, which it is community research, but yeah, it’s not how a lot of people think of community research, I think also.

Kate: Well, okay, so this brings us, I think this topic relates to what our topic for the day is, which is really the issue of why communities don’t necessarily share their best data, or why it’s hard to get at that data. And [00:01:00] this came out for me, I was giving a presentation to a bunch of mathematicians and technologists, and we were talking about the Global Parenting Initiative work.

And the results they found, and there was a data scientist in the audience who was like, how did you get that? That data can’t be right. How could you even get that? Because it was about things like how are you engaging with your children? Qualitative research, but quantitative research too that is quite personal and that people could lie about. So how do you actually gather that data?

And my answer was, you’re not depending on technology, like these are very much about how do you have people who work in the community who have trusted relationships? How are you building programs that are getting people to share things in a context where they trust? And that takes years to build those relationships. The idea that you’re just going to deploy a software that’s going to suddenly collect valuable data, I think was what this data scientist was reacting to, [00:02:00] that that is his experience of how you gather data.

Lucie: Yeah, that’s really interesting. So they were surprised that it could, they didn’t trust the data because they were aware that it’s sensitive data and people aren’t usually comfortable sharing that sort of information.

Kate: Yes. And it’s like any study that’s about something that’s personal, I think you build that into the results. How much of this is, people might exaggerate or lie, how do you control for those things?

Lucie: We often see issues with that sort of sensitive data collection with regards to a lot of the researchers we work with in West Africa. They ask members their income, like their annual income or something, which, just on a first point of view, a lot of the smallholder farmers are not aware of their annual income. They don’t have a bank which tells them and gives them summaries. But then also secondly, who’s gonna share that information?

Kate: And it is the sensitive information piece. I think this is such an interesting issue in the context of sex [00:03:00] work where one of the issues in sex work, this is what I came from working on, you can imagine that sex workers and clients actually have a lot of knowledge about where something bad has happened or where something is sketchy.

Probably there are word of mouth networks and in fact, not even probably there are word of mouth networks where people will share information about abusive clients or, in the UK that’s been structured in National Ugly Mugs, which is an organization that tries to systematize that and have that as a resource.

But you can understand that because of a fraught history with law enforcement and even nonprofit organisations that often come in with a rescue mentality, sex workers are not interested in sharing that data with just anyone.

And often even things like if they’ve experienced violence, they don’t want to share that because that gets weaponized as this is why we need to just get rid of sex work. And you’re saying, well, that’s not what I’m [00:04:00] asking for. I’m just asking for workplace protections. I just wanna make sure that , you know, this thing that other people can take for granted in their work, I would like to be able to feel that I have in my work too. 

Lucie: And that’s a really interesting example. But normally, sharing that information should be of direct benefit to them. The police would be able to have somebody on their record that this is a dangerous person. And it should be able to have a therefore positive impact on the sex workers’ lives. But it’s being used as you said, in a different way and in the way which the sex workers can’t trust it.

Kate: Yes. So, I think what’s interesting, one is how much data can be used for ideological purposes. I mean, I think we see this all the time where some study comes out and someone has drawn these conclusions, and if you bring a scientific mind, you’re often like, huh, that seems, those, that I don’t quite follow that logic, that seems, yeah, it doesn’t seem to quite add up or like, it’s not a logical [00:05:00] conclusion from the data that they seem to have gathered necessarily. And so I think there’s a lot of taking data and shaping it to serve one’s purpose, which I don’t think is incredibly difficult to do.

And then I guess the other piece that’s like if it’s not ideologically driven, often it’s profit driven. So it’s if I’m gonna share that with you, and I think you probably see this with farmers in Africa, and I think this has come up in a number of contexts where they recognize that when they’re participating in an outside technology and they’re not actually the keepers of their data, that data is going to be somehow used to monetize them. I use my little quotation marks around that. It’s going to be something that is taken and then used to come in and sell them something or change the system in some way that doesn’t necessarily serve them.

Lucie: Well, it doesn’t even need to go that far. I think a lot of people are just research fatigued.

Kate: Yeah.

Lucie: So I’ve seen both in the Pacific [00:06:00] Islands that people get interviewed a lot about climate change, which can have a lot of impacts if someone keeps on asking you every year sort of how worried are you about climate change? And then in West Africa they get lots of researchers coming and asking them about their different agricultural methods? And then they see no results. The smallholder farmers, they spend an hour or two answering a really long, fairly boring survey, usually very boring, I’ve got to say. And what do they get out of it?

In a way, at best, they get a suggestion of the monetization you’ve just said. But usually they don’t even get that.

Kate: No, and I guess that comes back to the sex worker example, there’s some history possibly of even sharing data, but then seeing no results. So it’s that that’s probably a big driver where it’s like, why would I bother to participate? The worst case is that new harms emerge, but like there’s probably a big middle ground where it’s just that nothing happens.

Lucie: So a really small example of this, is that in [00:07:00] our West Africa team, they give and they’re giving more and more trainings within the region to different student groups mostly at the moment. And we always give a feedback survey at the end of it. And just even in that the number of people who answer a feedback survey can be interesting. So in the last one, which has just ended yesterday, nine out of 10 people answered and they answered every question and gave, proper answers, which is really interesting.

I’m wondering if it comes back to this idea of how engaged they were in the training. I’m wondering if it comes back to what you were saying about the relationship of trust.

Kate: I think that was what I was going to ask you because you are doing this work a lot. It’s easy to say why don’t communities share their best data? Here’s why. But then the question becomes how do you get them to share their best data? Who even decides what that best data is?

If you’re asking a systemic climate change question, that’s not particularly good data for them. So I guess that would be my question for you, what does that look like in your [00:08:00] experience and what’s your hypothesis for why people did have such a engaged reaction to this survey that you just did?

Lucie: If I can go back to the Parenting for Lifelong Health example you started with, there, and correct me if I’m wrong, it is a program to help parents, as part of that program, they’re also gathering data. So the parents are being supported, they are involved in a program and they see the benefits of that program, so are happy to share information to advance the program and therefore advance themselves as well. 

Kate: Yes.

Lucie: And I think that’s the same situation with feedback survey. I think the facilitators who were in the room, they were highlighting that it’s to help us improve the next time. And it’s a lot, it’s really what we’re seeing in West Africa with the different research programs that we’re involved in.

The farmer organisations, they are really pushing for not just surveys, they collect [00:09:00] data from hundreds of people and give lots of information to researchers, but surveys which directly give information back to the farmer, so to see no data collection as just a process of extraction, but to see data collection always as a conversation.

So I’ve seen some interesting, no, I’ve seen some potential, let’s say, I’ve seen potential for some interesting conversations about food nutrition for example. In fact, the researchers just wanted to do a survey asking families, to try and get the idea of how many different nutrients they get in a week. So asking about how many leafy greens, how many orange foods and things like that.

But in the initial version of the survey, there was no explanation to the households why they were asking these questions and what is the importance of the green vegetables, the orange vegetables. So restructured in a conversation where you were exchanging information, already [00:10:00] it’s a different view of what research is. It’s a different view of how data is used and what the role of data is even. 

Kate: It’s interesting as you’re talking, ’cause I’m thinking about the logic of like wellness culture, which is the quantified self and the way you get people addicted to technology is you’re like constantly giving them this feedback loop of self-improvement or how many steps did you take or what did you eat? Like how do you get people to do these behaviours? And in that context, I think I have problems with it because often it’s trying to addict you to the technology even as it is purporting, and probably is hopefully having some good offline effects. It’s not entirely focused on, it’s definitely focused on having you do things in the app that you wanna pay for.

Lucie: Yeah, there’s few examples of that where it’s towards subscribing more.

Kate: Yes. But fundamentally there is a lesson and an equivalence there, the thing you’re tapping into is people [00:11:00] want things that serve them. And I think you can, I think that probably the toxicity of the quantified self is that it’s so self-oriented where a lot of the things we’re talking about are quite community oriented. 

Lucie: Oh, that’s so interesting.

Kate: Yeah. And I’ve read about successful places where, like there’s some town in the US where the mayor was overweight and he set in motion this whole community thing. And I think there was an app where it’s we as a town are going to lose all this weight. And apparently it was really successful because that community, it was actually quite social, it was doing a lot of offline things because it wasn’t about mediating human behaviors through technology, which is very much about you don’t actually want people to connect directly, you want the app to always be the thing that defines how people interact with themselves with each other.

Whereas something that just emerges from a community is actually often very focused on the benefits of the whole, the benefits for everyone, I want my neighbours also to [00:12:00] benefit from this. It’s not just about I’m gonna be the best and this kind of like ruthless, self-serving kind of thing.

So I guess I was hearing in what you’re saying there’s like an interesting interplay there of self, of community, of what are the incentives of the system you’re designing. I was also thinking while you were talking about the example that we point to a lot, which is the Fuma Gaskiya example, which was the farmer cooperative that really transformed research when they gained ownership of the technology, the systems, how data was collected and stored.

That when you change the power dynamics, where before it had been outside researchers deciding what research looked like and getting funding for that. And then there was a shift to what does it look like when the farmers decide? And it really changed because it was very focused on not what are the effects of climate change? Or, what should we be planting in this region? It was about literally what do the people in our community need? We all [00:13:00] win when these most vulnerable people in our community, we’re developing things that serve them.

So there’s a really different focus. Yeah, just what data you’re collecting, what the purpose of that data is. 

Lucie: And as part of that sort of filtering out then of you know what data you’re collecting, I think a lot of researchers tend to think more data is better. Whereas, well, if it’s a farmer, if it’s perhaps a sex worker who’s thinking about what data they want to give, then they’re gonna think more directly to what interests them. And they’re not gonna want to waste their time on all of these other questions, which can be helpful for a researcher.

Kate: But I would argue actually that if you step your way, that if you start with the big picture, they’re definitely not going to be interested. I think one of the theories I had with sex workers is if you give people ownership and control, if they can drive data, like they can hold their system, I think they’re just as interested in finding human trafficking and identifying that as anyone else.

It’s that [00:14:00] they can now do it through the lens of oh, it’s not that other people are going to lump me in with trafficking victims, or decide that like the treatment we decide for trafficking victims is the same as for consensual sex workers. And actually they see that solving that problem will make their lives easier because it means that the whole anti-trafficking movement isn’t coming for sex work.

So I think you can get people interested in systemic problems. There’s just a really different pathway to doing that and the sort of like self-serving outsider view that just wants to use one group to serve its interest in one place, that doesn’t work. 

Lucie: So what I was trying to say was that I think a researcher about sex work would probably ask all sorts of questions which aren’t that relevant directly to the sex workers and to their specific problem. I completely agree that there’s ways of getting people engaged in really detailed, really specific things. But often it takes, it’s a journey.

So [00:15:00] if you do want to get into that, you start at one level perhaps, and then as the community gets more engaged, then you can start. For example, Fuma Gaskiya that you mentioned, it’s what they’re doing. They started off with a small level, a light touch, let’s say, of data collection. And then as the farmers are getting more engaged, they’re exploring different things that they want to share data on, or like the different aspects that farmers want to research, basically.

Kate: I suppose also underlying that is some hierarchy of needs where it’s like you’ve met their most basic needs. So once they solved, okay, we figured out this affordable fertilizer that works for, you know, the poorest people, which was human urine, which I find such a fascinating study. So you’ve solved that, just acute pain and then you can start to step toward now I’m more curious about these other things that are a little bit more playful. Maybe they’ll fail, maybe they won’t.

That is part of what the scientist is doing. They’re getting to have fun because it’s all just theoretical. These are interesting problems, I wanna explore this. And so it’s all seen as [00:16:00] malleable pieces and like things you can experiment with. Whereas when you’re living the day to day and you just feel acute scarcity or precarity or, you’re really close to the edge of survival, it’s very hard to spend much time feeling yeah, your fun little experiments about climate change, yay.

I mean, I suppose that’s part of trust, like if I feel you’re a researcher who’s just coming in with this view toward your own ends, I’m not gonna trust you.

Lucie: No.

Kate: Whereas if I feel you’re a researcher who’s coming in with real respect for the community is helping us, helping us gain your skills and apply those skills in ways that serve us and humbling yourself in that context, that probably starts to create that stepping stone to way more interesting studies and unlocking, and exactly as you’re describing with your survey that you had such a great response rate.

Lucie: But what’s interesting with that sort of thing, if you’ve [00:17:00] got a research group that’s doing research about human urine, in this case, and then they start getting interested in some other aspects of a small holder farm, then in terms of research, you’re quite interested to get that continued year on year data to see the evolution, to see is human urine continuing to have a good effect or was it only the first three years? So there’s an interesting question there about keeping people motivated, even when they think that they have solved their problem. 

Kate: Wouldn’t that be, and this is something we think a lot about, and it’s your work, a lot of that is training somebody locally to have the research mindset. So it’s not about the outside researcher keeping that alive, it would be that you have somebody locally who’s aware of, oh, we need to be keeping an eye on this because it might have had some short term spike for reasons we actually, we’ve misdiagnosed why this works so well.

Lucie: It’s perhaps also, I’m just thinking, at a later stage, where you’re monitoring if there’s continued [00:18:00] impact, there’s also perhaps an evolution in what indicators you’re gonna start using. So can you reduce the number of, literally, data collector points in the sense of, rather than collecting loads of data on all different aspects of how human urine is used, can you just look at the yield, and can you just look at the quantity of human urine used on that particular farm? 

Kate: And I guess the question in that would be who is designing that research? Does it really take the local, does it mean that every community needs to be developing their local data scientist or the person who’s quite aware of how research works? Or researcher, not even data scientist. You need the data scientist to be another person. But is the goal to develop grassroots community level skills there? I mean, is that part of the problem that you just don’t have people who are skilled enough and then you become dependent on outside researchers who just easily lose the spirit and [00:19:00] value and the needs of the community?

Lucie: Yeah, it’s an interesting question. Because it depends on the sort of data, there’s a level where it’s farmer household data and so just one person. If the households can’t collect the data themselves, then it needs to be a second or third person helping them with that, which can be at the community level, absolutely. 

Kate: I suppose the answer is there isn’t always a clear answer, it’s the common IDEMS theme that it’s options by context. There are going to be contexts… so the thing about Fuma Gaskiya is it’s a collective, so it’s a lot of farmers, it has scale. And so the idea that at scale you can start to have your own research or your own data scientist, you could see how that would be sustainable, how that’s an option.

Lucie: I think actually, there’s people who are marketing the human urine too. It’s not only that they’ve identified that it is helpful on the farms, it’s also a business now for some people because other people want to buy it rather than producing it themselves. 

Kate: That’s actually a great [00:20:00] example of when data is used to create a sustainable local system. Then there is something there that is, there’s like a very different relationship to the data and its value. And that gets back to if you have a researcher who’s coming in and doing sort of systemic research, there’s nothing sustainable in that for the community often.

There might be some little like intervention that comes their way, here, this is what we’ve found in small communities serves people, but it’s not localized. It’s like they’ve flattened the data to focus on what would be global expert research, where this is your field, you wanna understand it.

I think that probably if we’re thinking about what are the pillars of how you get communities to share their data, it’s that the data actually serves them, that they’re getting the value of it. I mean, that’s what we see in tech right now is for the most part, small scale communities are not seeing the benefit of that data. The financial rewards of data are accruing to tech [00:21:00] companies and outside stakeholders, so, increasingly, if you’re paying attention, you’d be inclined to lock down your data.

Lucie: But this is a question that anthropology often tries to deal with, so this is on the non-technological front, especially PhD students ’cause you usually do field work that lasts for about a year or something. So if you’re going to a small community, generally, you often have to ask permission to do research there. Now again, I’m sorry I’m going back to the Pacific Islands, but a lot of groups there refuse permission, which is really interesting.

This is the equivalent of on an app, refusing permission. They are aware that the researcher is the one who’s going to get something out of that year, basically. They’re the one who’s going to improve their career, they’re the one who’s going to use the data for their own needs as opposed to the local community’s needs. There isn’t really, in terms of anthropology, there’s not a great solution.

The education systems, the academic structures themselves, they’re not set up to [00:22:00] encourage collaborative research, long term collaborative research. I know myself and other people, you know, you go and do your field work you go back to your university to write it all up into a horrible, long book, which no one reads. But then what about the community? They have given you so much time, they have given you, I mean, you know, speaking more in terms of apps, they have given you so much data.

What can you do in terms of that relationship? How can you really… it’s not giving back. I don’t like that idea. But how can you respond to their level of trust and their commitment in your project? 

Kate: I think that is something that actually a lot of researchers feel, and it’s why they often become like real evangelists for wherever the place they work. They want these great people they worked with to be getting more, to be benefiting. But the researcher themselves, they’re often dependent on outside grants and outside funding.

So everyone is just scrambling [00:23:00] to get something. It comes back to that hierarchy of needs. It’s very hard to, if you’re a researcher, to task that researcher you also have to go and lift all boats. Like everyone has to come up with you when you’re probably just getting enough funding to survive yourself.

And the whole system isn’t set up to kind of unlock the levels that are needed and create something that’s sustainable and access the value. There’s so many dead ends, which really do come down to funding models. And even the goals, like the goal of a lot of things is not well, how do we lift the local community? It’s how do we create some thing that we bring home and then create something out of that.

I feel like we always come back to the same themes, but it’s shifting the system of what the expectations are of when you fund research and what successful outcomes are? A successful outcome, I guess it sounds like would start to shift [00:24:00] toward what happened for the community that you embedded in? What were the outcomes for them that improved, that had measurable, sustainable improvement?

Lucie: Goodness me, that would just break down most of academia if you were judging people on that.

Kate: I know. And I actually feel that we often sit at the intersection of research and actual like sustainable commercialization, which is why we’re a social enterprise. I find myself in these places all the time where we’re working with research teams and they have a very particular mindset. You’re trying to get help or funding or advising from people who are very traditional commercial people, and they have such a particular mindset. And so you’re constantly just trying to find your way through this very uncharted territory of like, how do I make these pieces come together? How do I do it? Yeah it’s really hard. It’s really hard.

Lucie: I think I’m very fortunate to be working with the Collaboration for [00:25:00] Resilient Food Systems, which really promotes these farmer organizations, these farmer research networks to be led by farmers too. 

Kate: Yeah, the McKnight Foundation, they were the force for why funding got shifted from outside researchers to the farmers. They had this sense of like, well, they should have more power in this and have more say. That’s probably a big piece of why communities don’t give their best data is we’ll pay them for it. Like the idea that everyone is just existing to be your study and just like waiting for somebody to come along and just study them. That’s not, that’s a very, I think that is stymieing probably a lot of research getting better results or having more impactful outcomes.

Lucie: Yeah. So there’s interesting models that could come out, which are coming out here and there in different places. 

Kate: Yeah. 

Okay. Shall we end it there? It’s an interesting conversation and I mean, no easy answers, but, yeah, good to think about.

Lucie: And there are interesting models coming out of [00:26:00] how researchers are working with communities, how communities are developing their own models for doing research. 

Kate: Let’s make our next thing, we should interview somebody who’s doing this and just talk through how they’re approaching it.

Lucie: Yeah.

Kate: Thanks, Lucie.

Lucie: Thank you, Kate.