Transcript from one-on-one interview
Tuesday, May 14, 2019
10:23 AM – 11:28 AM
Location: Research office of interlocutor (Nairobi, Kenya)
Participant: Kenyan woman who resides in Nairobi and has been working at this research organization since mid-2012.
Discussion conducted entirely in English. Audio recording was uploaded to Otter.ai which did the initial rough transcription. Research Assistant did initial cleaning. Angela Okune then listened through and edited for accuracy before redacting according to interloutor's requests.
Total discussion time: 1hr 05 mins
Angela Okune 0:01
Sawa (Swahili: Okay). Today is May 14th. At 10:23 AM. Okay, thanks again, for agreeing to be interviewed. I really appreciate your time. I'll try to keep it short.
Angela Okune 0:19
So we just looked at this artifact here on my computer that I had uploaded, based on the the materials you had shared with me earlier. Emm, and so I just want to ask you a couple of questions. Because I think looking at it here on this platform allows us to think a bit more about worries and also opportunities about a data sharing platform like this.
Angela Okune 0:45
So based on what you've seen here, would you feel comfortable sharing this data?
Umm, yes and no. Yeah? Yes, I would...I would ahh, like to see a platform where we share...the people who are doing qualitative, qualitative research where we share methodologies, findings. Yeah? But then raw data, I know you've, you've not put names and everything. But, I would, I would just avoid raw data and have maybe something processed like something, a finding or something from a study that can be very useful to somebody else who is trying to design studies of the same nature. That would be more helpful because I'm thinking...not so many people would have the time to go through the transcript and pick on, very important things that they want to know about Tanzania, but if we can have a summary, like a paragraph, talking about this particular project and the methodology maybe, how it was used and how effective it was. That would add a lot of value to us who are doing qualitative research. Yeah.
Angela Okune 2:14
So maybe this could be together with the instrument. It could be together with the final report, like having a packet of all the different materials.
And even abstracts.
Angela Okune 2:25
And abstract and a quick summary?
Yeah, quick summary with ah...main ah...if it is an evaluation, main evaluation questions, the methodology because not always we'll use focus group discussions, we might use other ways of getting data. And what is the most significant thing that came out of that study? Yeah.
Angela Okune 2:47
How do you feel...because sometimes maybe let's say I'm a researcher who studies cotton.
Angela Okune 2:51
So I somehow found my way to this transcript...Maybe I would find because I'm focused on cotton, I could be super interested in actually seeing this. But you see if you do like the findings or summary, maybe yourself you were interested in maize or sorghum. So maybe you don't even mention cotton. Kumbe (Swahili: actually) there was actually some discussion of cotton. So do you think there's value in also having this...raw, if you will...maybe it's not what most people would look at, but at least it's an option in case somebody is interested.
Yeah, like, for example, now we were doing, we were trying to find out on the neglected crops in Tanzania, specific parts of Tanzania where the client wanted to do an intervention and support that particular community. So this one, they would list all the crops that they grow, but then there are those that they would call, the main crops or the traditional, the indigenous crops and the rest. So you...as a cotton, as interested ...if you're interested in cotton growing...you might really not get the information that you want eh? Because first of all, this is one...one region, one district, you might not get a lot of information about cotton growing in Tanzania, but there are other types of research that they might do something focusing on cotton in Tanzania, and it will give a lot of information about cotton growing in a summary form, the results that they got from that particular study. For somebody who is looking at cotton in Tanzania, this, this transcript might not be very useful for them, because this is just one transcript, for one FGD and maybe out of six participants that we had, maybe only four grew cotton. First of all, it's not representative. And then it might not say a lot about the cotton because the...the questions are not focused on cotton growing or how they do the...but we just trying to establish the kind of...so depending on the type of research that has been done...
Angela Okune 5:10
...if we document the abstract or the executive summary well, somebody might be able to get the kind of information that they're looking for. And if there could be some linkages, if I'm interested in this particular kind of data, I may need to contact this organization, or I might need to contact this particular person, so that I can get the raw data and see how I can work with it. That that would be very helpful, because I wanted to avoid lengthy documents in the... in the web, and then somebody might not even take interest in using all of that. Yeah. That's my opinion.
Angela Okune 5:57
Yeah. And have you ever tried to find other data for your own projects like let's say as part of the lit review or anything like that? Have you...
...I have tried. For some projects, the client might give you a very brief... uh...I mean the information, brief summary of what they want to do without a clear background on that particular area where they want to assess, or the...some things are not very clear. Like, for example, we were supposed to do a survey for REDACTED ORG NAME in Tanzania. And Tanzania is bigger than Kenya. And so we wanted to have data that could be comparable between the two countries. For Kenya, we...we were able to select a few locations that are known to be doing very well in agriculture. And so we were not...we were not so much concerned about the kind of crop that they grow. But now for Tanzania, given the regions that they have, there are so many regions and the area... geographical coverage is bigger than Kenya. So we had to make a decision on how we do the sampling. Like if we want to do sampling, first of all, which areas do we select, which are known to have smallholder farmers? Which kind of crops are they growing? Yeah? And how do we come up with a sample that is similar to what we have done in Kenya for the results to be comparable. So we had to review so many documents, just online documents, trying to get [REDACTED ORG NAME] reports, ummm, [REDACTED ORG NAME], what they had done because [REDACTED ORG NAME], their project [REDACTED PROJECT NAME], more on digital inclusion of farmers, financial inclusion, digital inclusion of farmers. So we had to look for different data sets just to get a feel of where would be appropriate for us to go do the sampling. And maybe what kind of farmers do we talk to, because talking blindly to all the smallholder farmers was not going to be easy for Tanzania. So yes, at times with the kind of survey that you want to do, you might be forced to get more information from other sources. And it's so unfortunate that most clients would not provide these sources where you can get this information. You have to look for them. Yeah.
Angela Okune 8:44
So if like in that example, you were looking like at [REDACTED ORG NAME]'s reports or also the [REDACTED ORG NAME], the statistics... because I know [REDACTED ORG NAME] has some data, in terms of numerical data, isn't it?
That was, uhhh. how do we call it? There's some survey done by [REDACTED ORG NAME], I think several, on the same, just the same topic as [REDACTED ORG NAME]. I'm forgetting this, it's not MasterCard...ahhh... if I remember the name of the organization, I'll let you know, so they have different data sets on...actually, the design of the survey was very similar to what they had done. And they had done this in Tanzania and other Asian countries and compared the results across those countries. So we tried to get hold of that data. And we even talked to the client for [REDACTED ORG NAME], because they were having some interactions. It's a [REDACTED ORG NAME] thing, but I don't remember the [REDACTED ORG NAME] group that was doing that. So they had some interaction and they shared... ummm... some slides, what they had done in Tanzania, two waves I think, and data, raw data. We could also access.. they call it, is it mega data or something...Yeah.
Angela Okune 10:04
They were numbers? Statistics?
Yes, yes, yes. They sent to us. And we tried to do our secondary analyses and tried to use that information for something. And we came up with three crops that would ...base our sampling on.
Angela Okune 10:23
But it wasn't like, umm... qualitative data in this kind of way, was it?
No, no no.
Angela Okune 10:29
Interviews and that kind of...
That was purely quantitative, but then we were supposed to do interviews, and then FDG with farmers and key informant interviews with key players in the agricultural sector, for the crops that we had selected.
Angela Okune 10:49
So you used the statistical data to figure out where you would be and what you would focus on?
Angela Okune 10:54
And then you generated qualitative...
Angela Okune 10:57
And that data then went to the client.
Yeah of course yeah.
Angela Okune 11:00
But they didn't...like...how did they...did they analyze it? Or did you guys analyze it?
Umm... so the analysis was done at two levels. One, we did the basic analysis, like, we did the transcripts. And we did summaries, yeah? Like presentation, triangulating data from quantitative and qualitative. So like there are some questions in the quantitative that needed support from the qualitative, like this is what the participants were saying about supporting a second finding in the quantitative. And then the other part of analyses went to [REDACTED ORG NAME]. Because they are working with them as as their strategy partner. So that other part, we had our data, our reports sent to the Dalberg team and they did another analysis. They said ecosystem something. Yeah.
Angela Okune 12:03
So in terms of project, let's say, in your opinion, do you think sharing, let's say, some of those qualitative materials afterwards... what kind of researcher do you think might be interested in reading such things?
Well, ummm, you see, like for development partners who are funding this project. They might need to have that information at hand. Even implementing partners, say like [REDACTED ORG NAME], if they could have access to that kind of information, before we even designed the project for them or the research project for them. It would have been very useful because they would say we are interested in 1,2,3. Yeah? So let's not focus on this this and this and that. We'll just be very focused to a certain...I mean to the issues that they need to address issues. Issues that have not been addressed before. But ahh...and that would have made their work easier and cheaper. Yeah? Because now, if we completely lack information, yet all that information has been collected, then it's like, every time somebody goes in with the same...who needs the same kind of information...they have to spend a lot of time, a lot of money to get similar information that has...somebody else has, maybe they are not using at the moment. Yeah.
Angela Okune 13:45
And so do you think there would be benefit beyond just let's say the final report, which sometimes is published in the public, sometimes it's not, do you think there is maybe additional information that can be learned when someone has more of this supplementary additional materials?
Yeah. If, for example, we have access to this, this platform, we're not just doing research for one organization. We're doing research for different organizations, different subjects. So if I'm able to get to say, different organizations who have done something, say in Kenya, about a given crop, or maybe health survey...they have some findings maybe varying findings, but something that I can connect and design my own study in that line, then that would be very helpful for us who are using that platform. Because first of all, we go with a clear focus on what we want in the field. And we know where we're going to get it from because others have done it here. Have done it here. And this is what they were able to get it...to get. So we can go there and get more information maybe based on what our needs are. That would mean saving us a lot of time, again. And ahhh...yeah, and we'd have a clear focus on what we really want to get. If we have different reports or different executive summaries that we can go through and say okay, here we are able to get this information. Maybe we can request Angela send this particular data set. We are interested in this one, and they can anonymize it and share with us and yeah, that that can really help us in getting very clear information, rather than having to use tools with so many questions to get into one particular question that has not been answered correctly or appropriately. Yeah.
Angela Okune 16:06
So it sounds like basically, you're saying that this could particularly be helpful for research organizations like this one and maybe others. What do you think maybe something like this for policymakers or students or are there other types of groups you think might benefit?
Ah, policymakers? Yes. That would be helpful. Policymakers, development partners, other research organizations. Students? Ummm, I think I would, I would think of the level of the students...are they studying masters? PhD? Because masters, PhD is usually directed into a certain field ah? Rather than undergraduate... because undergraduates they might not see the importance of this. But when you're now specializing in a certain sector, this might be very useful, because I believe it will have information for different sectors as it is available. So if I'm from the energy sector, for example, and I'm able to get a few things from here, this, maybe combined with other sources, that would help me a lot in also designing my research for my studies.
Angela Okune 17:32
And if you were to share this, like, let's say this one specifically, just to think about, what would you put any conditions, let's say, on someone else using it, like would you want them to acknowledge the company, would that you want them to acknowledge you? Would you want them to pay? Like what kinds of conditions could you think of that maybe to share this...?
I think payment would be the appropriate thing. Yeah? Pay. But then if there is specific information that you want, then you need to be assured that you can get that information. It needs to be restricted so that the information in the platform is not misused for some reasons. Because if we make it free, and there are no restrictions on login and everything, people might...might misuse it, like we've seen in other platforms, how they do it. And that might not be very useful. But if you say it's for payment, a small amount of money, and there are some restrictions also there's some information that maybe you have to get through the admin or somebody. That would even make it a more serious thing and most of the serious organizations would be interested in the platform. Yeah.
Angela Okune 19:10
And if you let's say you contributed this, would you want to get the money to you or would you be okay if it goes towards let's say supporting the server what what what. Like to contribute, would you want to get money?
No no, okay for me, I wouldn't look at the money coming back because the money should be supporting this so that it's sustainable.
Angela Okune 19:34
Yeah, and it's a small amount of money. Yeah?
Angela Okune 19:38
So you're imagining like a small amount of money that guys contribute to be part of this here and then they can also use...
Maybe what what...something else you can think about...like because even [REDACTED ORG NAME], we get the data for free. They just filter out what they don't want the public to see. And yeah...I think just a small...a small fee for the maintenance of the platform. Yeah, that that's it. Yeah.
Angela Okune 20:09
And you said misuse. So like, how do you...how could this type of information be misused? In your mind?
Misusing the information. Now, if we make it free, and anybody can access this, nowadays people are very clever. They can even think of coming up with their own platforms and getting information from this platform and making their own like, you know. So for you, or for the organizations that are involved, they're really working hard to make sure that this thing is coming into the platform and others are benefiting from it. Yet there is somebody else now who is able to get this information so easily. Yeah? They don't have to pay for it, they don't have to...I mean, there are no restrictions, they can have access to it, and they can even sell it to others, because it will be a lot of information that would be there. My imagination. Yeah, so to me that is misuse. Because they are getting it for free, and they can do as they want with it.
Angela Okune 21:21
So maybe putting in some sort of clause so that if someone is using, they also need to contribute something back.
Of course, of course. Yeah. Yeah. And all the requests as they come, they need to be directed to a central point, somebody to address those requests. Like, I don't have to send a request directly to somebody saying Kantar, I need this and this. No. Maybe either through the admin or something, a system where one person or a group of people can be leading the others in getting that information. Yeah.
Angela Okune 22:05
And how do you think this might work...let's say with the different market research companies because I know there are many. And I know that often they'll, let's say all of them will bid for one project, isn't it?
Angela Okune 22:16
And so often there's a feeling of a bit of competition, because everyone is trying to get the same, the same bids.
Angela Okune 22:23
How do you think a platform like data sharing now with possible competitors, how would that...how would that play into things?
Yeah, because now what we are sharing is, is information for projects that are complete, isn't it? Not projects that are bidding for. So what we've done and in sharing that information, it all depends on the person who is reading this information. You might have something very important, I mean a very important finding from that particular data in your own way. But there's something else I can see and feel it's more important for me than what you see as being important. So for me, I wouldn't see the competition part of it, I would, I would see the knowledge, information, part of it, information sharing. I'm only interested in that, because it's not all about business, I think. It's all ...it's all about how knowledgeable...what are you able to cover? What do you know about this sector? Because looking at the kinds of businesses that we have in Kenya at the moment...first of all, it depends on who do you know to give you this assignment. The other thing is the budget, not necessarily the methodology, because the methodology and other things, you can come and beat them once you come into an agreement, and they say, okay, we want them to be done like this. From experience, who do you know, how much is your budget? That is what...
Angela Okune 24:22
Yeah. It's not what you know about...most of the time.
Angela Okune 24:27
So you don't think that would be a barrier for...?
I don't see it as a barrier because we are sharing information about projects that are complete, isn't it? Yeah.
Angela Okune 24:37
And for yourself, let's say you're coming to this fresh, you didn't... You had never heard of this project. And you were just seeing this. Would you use this information? Let's say you have a new project...
Related? Yeah. So like I've told you, if I get a summary and abstract, because we usually...sometimes as I had mentioned, we usually sometimes go to the net, try to get for information I'm doing something in agriculture in Tanzania. So if I'm able to get something on agriculture, which is of interest to me, I can just read the abstract and see, okay, this one covers what I'm supposed to do. Maybe I can request for more information from this.
Angela Okune 25:22
What if you are sent this, this transcript, would you? How would you use it?
[Laughs] I'll read it and maybe when I'm very desparate [both laugh]. But I will just have a look at it and say ahhh no.
Angela Okune 25:40
It's too long?
Angela Okune 25:41
Too many words?
Yeah. Yeah. Yeah.
Angela Okune 25:44
I see. Interesting. And what other information then? So... the abstract, the executive summary, would you trust this? Like, would you trust the findings? How would you know that this wasn't just someone making something up? Like why would you trust that you could...that this was real?
Okay. I would trust it because I'm part of it now. Yeah? You're getting information from me. So I know, I've contributed something. So, definitely somebody else has contributed something. And if they are willing to provide some information like this, they're giving it in good faith. So whatever they are giving is to the best of their knowledge. This is what...I can play with the data and come up with other findings, but for them, because those findings were of interest for their client or for themselves during that particular time, they have summarized it based on their objectives. I can get that data and come up with something else based on my findings, I mean, what I'm interested in. Yeah? If they have the specific questions that I need to to answer, I can play with that data and get something that might be meaningful to me. Yeah. But whatever they provide there is a summary, is an abstract, is what was of importance to them then. Yeah. So if I get the data I'll know the specific questions that were asked, I can...I can single out the ones I'm interested in and get the information that I need. Yeah. For my own use.
Angela Okune 27:26
Because I think this, you see this as the data...So let's say once you decided what your product is, nini nini (Swahili: what what). Instead of even going and trying to run this FGD, I don't know...could you use this data as your own original data? Like could you do your analysis on this data that someone else collected, and now write your own report and maybe go to the field with this, but not maybe as much as you would have?
Angela Okune 27:54
You can because..ummm...there's some projects that we use secondary data, yeah? Like the reports that I've just mentioned, we used several reports, and we asked for the raw data. And we were able to come up with tables that enabled us to do the sampling for Tanzania. So you can use secondary information...data that has been collected by somebody else...to do your secondary analysis, and then try to establish the gaps and do your own primary data collection. Yeah. "This is missing from this data. So I need to go and collect it from this location." Yeah.
Angela Okune 28:40
Okay, I think it makes sense. How do you think so I think what we've said is that this, in and of itself cannot be by itself. It needs some sort of abstract, executive summary.
Angela Okune 28:52
Maybe more information about the study itself, maybe the instruments as a separate document.
Yeah. And even the, the institution or the client that we worked for. If they don't have a problem with that, ah? You can even put it there. Because now if we're thinking about sharing the platform with the partner organizations, umm, it might be easy for them to contact those organizations if they need of some information. Because we might do our own analysis our own way. But they also do their own analysis their own way and donors they understand each other and they can collaborate. Yeah, so that...
Angela Okune 29:40
So including the funder. And do you think you would need to ask the funders' permission before you share something like this?
Yeah, definitely. Definitely we have to. Yeah.
Angela Okune 29:50
And who else? Let's say before you can make this public at a certain level, who else would you need to get permission from? Like, let's say you are...you have those transcripts. Before you put them all there...Within the organization, you would need to confirm with...?
[REDACTED PERSON NAME]. Both [REDACTED PERSON NAMES]. I might need to confirm with them if it's okay.
Angela Okune 30:12
And even when I'm doing the analysis, they have to see what I've done. And maybe see if it makes sense to have that in the platform. Or if there's any information that they would want to, to remove.
Angela Okune 30:29
And then the funder. What about maybe the people who were part of the FGD?
The participants? Usually with the FDG participants, what we have, they usually sign a consent form, eh? And for most studies, they ask would you like...first of all there's a consent for pictures...if you want to take pictures...and then the consent for sharing or publishing that information. So if they say no, they don't want that information to be published. Or, the information can be published, but don't mention our names. Yeah, that's what we usually do. So we have consent forms that we usually...
Angela Okune 31:12
So you won't need to go back again?
No no, no no. So for me now, I'll have known I have consent for publication, but no mention of names, then I can come up with my own names like we usually have those stories that we usually do, yeah.
Angela Okune 31:28
Ummm. How would you change maybe the title of this thing?
So now agriculture, I would put agriculture first of all, as a broader capture. And then I can try to have, ummm, that would be a lot of work, but I can try to have agriculture say, Kenya, Tanzania, maybe East African countries, Uganda, like that. And then under Kenya, I can have different subtitles. Coffee farming in Kenya, maybe smallholder farmers in Kenya, something like that. So that when somebody comes to click on smallholder farmers in Kenya, they can get all the reports, all the abstracts on smallholder farmers in Kenya. Yeah, that would make work easy. Yeah.
Angela Okune 32:32
Yeah. I think part of the challenge is always how to make something...You see, maybe yourself, you know, this project, so you know it is a certain way.
Angela Okune 32:40
But someone who is a stranger to the project, how can they know what it's about quickly?
Yes. Yeah, yeah, exactly.
Angela Okune 32:48
Okay. And when you're looking for data, what kinds of data did you find were so easy to, to access and which ones maybe were hard to find?
First of all, with quantitative data, it's usually very difficult to understand it if you don't have the questionnaire. Yeah? And if it's already coded, and you don't have a codebook...that that's another thing. So, for for the quantitative data it is usually not easy. I have not tried to look for qualitative data, the kind of qualitative data maybe I can, I can look for on the net is ah, just Google a thematic area and see what I can get. And maybe just read that and summarize it like that without necessarily having the questions. Or maybe you can ummm, do Google sample qualitative report on this subject and see if you can get a report attached with a guide and everything. So that you can see the questions. But that I've not done so many times, but with the quantitative data I can, I can assure you, when you get quantitative data, and you are not part of the data collection team, it's not always easy. It's very difficult, unless you get the questionnaire, the code book and even ask a few questions to the person who was responsible for that data. Because there are so many other things that may not, for example, they might do weighted calculations. You have to follow up on that and understand how they calculated the weights.
Angela Okune 34:40
So in that way, do you think maybe sharing qualitative data like this is a bit easier because you don't need to necessarily...the aim is not necessarily comparative, it's about understanding the story and what people are saying in this moment. So you don't need to...and you can read it like that. What do you think?
Maybe what would be helpful, the findings, important findings and the guides. Yeah? Yeah. [pause] Because for the responses to be very helpful, I need to see transcripts for all the focus group discussions.
Angela Okune 35:21
Okay. So you would need to see everything?
Angela Okune 35:24
Mmmh. Even just one is not enough?
Angela Okune 35:27
Because this you cannot generate information that would be...represent everybody. I'm very sure there were others, which are very different from this one, based on the location, based on the village, yeah. And the participants themselves. Even, comparing women focus group discussions and male focus group discussion, you would get different information.
Angela Okune 35:56
But let's say as the user if you're going to look for this. Maybe there were 10 different transcripts. You see, you're seeing this one and you're like, Oh, so many words. So if you had 10 of them...
Angela Okune 36:07
How do you think the user would be able to go through all of that information?
For me? If I have to go through all that information, I will just focus on the questions that I'm interested in and how they were answered. Yeah? So like the way they have moderator, blah, blah, blah, I would be able to just single out and see how the responses are coming out from...for those particular questions I'm interested in.
Angela Okune 36:39
That's important, because you see, unless you scroll through this, there's no way for you to see the question.
Angela Okune 36:46
There needs to be a better way to actually see...maybe even as part of the tags or somewhere to see now what is the question that is being answered.
Angela Okune 36:57
Hmmm. Yeah, that's important... Okay. [pause] Interesting, because I've been wondering, so this idea of curation is this idea that you just take, let's say, like a sample, a small subset or like...it's not comprehensive, the intent is not to be everything. Because if you put a lot of information it can also be overwhelming. So it's about picking out maybe what you think is like the core important aspects? Or maybe even just for that taste?
Angela Okune 37:29
And putting it there. So I've been struggling...you see, I also just put one, cuz I was like, I'm not going to do all of them yet.
Angela Okune 37:38
Because you also need to have a good way of, of organizing information.
Angela Okune 37:43
You can't just have everything "pa" like that.
Angela Okune 37:46
So I've been trying to figure out what's the best way to organize it... and thinking about that. But for you, let's say as a researcher, you would need to see everything? All of the transcripts? Or you would be okay, maybe with that small sample, and then if you want more, you could also reach out...
Yeah, I think I would, I would first want to see just a small piece of...maybe, and see this if this is an interesting study for me. And then I can follow up...what kinds of questions did you ask? And then if I'm interested in all the transcripts then maybe they can be sent to me, but then I'll just select a few questions that I'm interested in. Of course, I might not be interested in all the information that has been provided. Somebody might just be interested in nutrition and not the crops that they grow. So how, how do they for example, from the women focus group discussions, we were asking them breastfeeding...How does it take? What do you feed your children at this time, that time. So I'm only interested in nutrition, I just check female focus group discussion, this set of questions and I'm able to come up with something that is of interest for me. If I'm interested in knowing the kind of crops that they grow and maybe the patterns, which crops are associated with male and which crops are associated with female, then I would just go to the section on crops for all the FDGs and see and do a comparison women are mainly growing this and men this...
Angela Okune 39:21
You've just outlined two very nice studies. [Laughs] Two ways you could reuse this thing. Yeah! I think that's for me why data is interesting is because it allows...you see when you're reading a final report, you can't do such like re-analysis, isn't it? Like you can't...within the report, someone has already baked in their, their findings...very in, in, in the way they're reporting, but here, someone can go in and actually read now the responses.
Angela Okune 39:50
...and reinterpret them in their own way.
...in their own way, yeah.
Angela Okune 39:53
For me, that's why data and not just open access of the final reports is exciting because I think it allows for that re-analysis or reinterpretation.
Mmmhhh, yeah. But now presentation of the data...
Angela Okune 40:07
Yeah, yeah, definitely. Okay, let me think. Just in terms of this platform, other aspects that you think might be important? Let me just read for you what I had brainstormed... restricting access to authorized individuals. I think you had said that that was important.
Angela Okune 40:29
And then being able to cite the data set. So sometimes, like even if you're writing a paper, or maybe a report to be able to point, you know, the way you can cite, like an article or a journal [yeah] to be able to cite this data sets, [mmmhmmm], do you think that would be important?
Yeah it is.
Angela Okune 40:46
Yeah? Okay. And then, oh, here ensuring someone can access all the materials together so that they don't just take one transcript out of context.
Angela Okune 41:00
I think that's what we were just saying.
Yeah, that's what we've talked about.
Angela Okune 41:02
Yeah. For you being able to see usage statistics, to see how many people have accessed this data.
Angela Okune 41:09
Very important, why?
We need to know, how many users, who are these users also might also be important.
Angela Okune 41:18
Okay. Yeah, that's the other one, being able to gather information about the people. So like, who are they?
Angela Okune 41:23
Okay. Anything else you can think of? That you would want to know?
Mmmh...mm mm. For now, yeah.
Angela Okune 41:35
Beyond usage, are there other measurements that you would want to apply to your data? So let's say you put this up. Is there any other information let's say after three months, you would want to know? You'd say, Angela, I put that data set. So what would you want to know about? After you shared?
How many people are interested in using that data? Ummm. an how they want to use that data? Has the data been useful to the users, in what way? Yeah.
Angela Okune 42:17
And actually one question I had, do you think that someone similar to yourself, with similar expertise, if they just encountered this here, would they be able to understand and properly use the data? Based on the way it is now?
Angela Okune 42:35
We need questions. We need a bit of a background.
Speaker 1 42:40
For them to understand, yeah. Not so many people would know what is Lubanga maybe for example. It's a male focus group of six participants. Were you only doing male groups, or mixed or how was it yeah? Is this all what you've done? I think that would be important information for the, for the user. Yeah. And the questions I think.
Angela Okune 43:11
And do you think any of the participants would be interested in this material?
Yes, they would be. The only fear is that...most of them, looking at their location and looking at maybe their literacy levels...they might not be able to log-in and get this information. Or even if they're able to...the information is not in a format where they can read and understand. Maybe some of them might not be able to read in English. But definitely, ummm, this has been an issue, not just in Kenya but many African countries. They don't get the results of what they respond to. So they would be very happy. Not just receiving some report, but at least knowing, I have participated in this. And this has gone this far. Or this is what is happening in relation to this. Yeah.
Angela Okune 44:09
So maybe even as part of this group, whoever ends up managing this kind of platform, part of the mandate could even be that they find ways to interpret or maybe disseminate or share that this materials is available for anyone...
So I think umm, for me, ummm, let's say for this particular project, we had the client, the end client who was funding the project. And we also had an implementing partner in Tanzania, who is involved in this particular project, and it's a three year project. So that was a baseline and then we're going for a mid-line and end-line and the implementing partner is going to be the same. So if this implementing partners can also have access to this one, then they might be able to pass this information to the beneficiaries that they work with, because some of these participants are going to be beneficiaries, and they'll be interacting with them at the village community level. And they can be asking, by the way, we answered this survey questions, what happened? Oh, we have...look [hits hand on table to show putting something down] this is what we got. It will make their work easier, yeah.
Angela Okune 45:25
Actually you raised a good point about baseline, midline and end line. So in terms of when this information would be released, what what do you...what is your thoughts? Because if it's just the baseline, and maybe you're ready to release something, what do you think it should be available immediately or three months or six months or after the project, the end line is completed or what what do you think, when should such data be released?
Yeah, so you see for this particular project, the, the client is also involved in analysis, we might have some basic analysis that we've done, but the main analyses, they're going to do it, because they have to devise ways of working with the implementing partner on the ground. And they'll base most of it on what we have here. So I would prefer that first of all we inform the client we want to do this and that and that. And then some clients are very supportive of such kind of ideas. And they would say, Okay, give us some time, we do our own analysis and then we can share even the abstract from our end which you can put in the platform. For this project, we did it last year. The analysis is not complete until now.
Angela Okune 46:48
Ummm, we are doing it with [REDACTED ORG NAME] and those ones are strict researchers like they have to make sure that they do it the correct way. So last month, they had a few questions on the quantitative data. And we answered that. So that says the analysis is not complete. Even the reporting is not yet complete. Yeah. Yeah, they usually do that, they take a lot of time, but they come up with something. So most likely before we go for the midline, the report would be out. And I think that would be the best to share in the platform. Yeah. So it all depends on them. For other clients, like, I would say, the only clients that we take very short time is like for market research, because they want that information very fast to make some decisions. Yeah? One month, two months, yeah. But I would say once we have finalized with the client and they have given us a go ahead. Yes, we can do this or we need to add this and that to the platform, then I would say that that's the best time to share.
Angela Okune 46:50
And do you think most of the funders, based on your interaction with them, would they be open to sharing at this level? Like, would they...would they agree?
I'm pretty sure some of them would agree like [REDACTED ORG NAME]. The reports that we do for them, they share them in their learning website. They have one, they share with the different organizations. People can access them from there. And for them, it's for free, as long as you work with them, you can have access to the link to the website and get that report. So some of them based on the nature of the work that they do, and knowing that they might also need some kind of information from other organizations they would agree. But I'm very sure there are some that won't agree. For example, if we talk about banks, if we talk about these fast-moving products companies, for some reasons, most of them, they might not accept.
Angela Okune 49:21
And maybe more of the product focused ones would not accept?
They would not accept.
Angela Okune 49:32
And how would you evaluate how good a particular data set is? Or maybe anything...how would you evaluate maybe it's bad or it's suspicious, like what would you look at?
Ah hah. Now. First of all, we have what we call Kenya National Bureau of Statistics. They usually do a lot of work and they usually have a lot of data. Anytime you can access and get good information. We also have, for example, like if I'm doing something on digital financial inclusion, and maybe there is mobile money. I mean MNOs (Mobile Network Operators), they usually provide their own reports, isn't it like penetration, market shares and all that. And I look at a certain data that has been collected by a different organization and then I look at the percentages and the numbers and compare with those ones and I see like, okay, there's a big disparity. Yeah, then I might not be able to use that data set for some reasons. Unless I get some explanations. Maybe there are some weights used somewhere. We don't know. We usually get the polling results. Yeah. And you can even look at them based on where you're located and know, okay, this one, maybe it is really not making a lot of sense. And you might fail to use that information and it might not be very useful. So there are other, say authentic sources, like if you are looking at numbers, say they're doing like weighting of the data based on the national census data and you get some disparity like data is saying this dimension of data is saying this, this percentage is given here. They don't come close to then you might need to put some question marks on that particular data.
Angela Okune 50:15
And what about qualitative data?
Qualitative data is not easy.
Angela Okune 51:57
To judge if it's good or bad?
Not easy. It's not easy unless it's a subject...even if it's a subject that maybe touches on what I know or I'm not representative, and people might have different thoughts different.... uhh...they think differently about different ideas of qualitative, I don't think it would be easy.
Angela Okune 52:24
So maybe you would look at the source...?
You would look at the source, really yes, but maybe the much you can do, ummm, you can get the information and maybe try to set a certain hy...I mean, test a certain hypothesis based on the information that you found and see what you get. Yeah, because most of the qualitative surveys are not done country-wide that is one. Maybe a certain data has been used to generalize some information for women in Kenya, and maybe the data collection was just done, in areas where say women are very submissive. And then I want to do a similar kind of survey and I go to different areas are women are not submissive, we are going to get a variation in the results. So really, I don't think there is a benchmark for qualitative data. I don't think so.
Angela Okune 53:28
And I think we talked about this before, but this one is looking at the idea of responsibilities. And what the responsibilities should be for someone who uploads the data. And also for someone who uses that data, what the difference responsibilities should be...
Uh huh... the person uploading and the person using...So the, the user, they might have access to all the data for a particular subject. Yeah? But they're supposed to request for what they feel is more useful for them. I don't think they need to request for everything at the same time. Honestly, they might not even look at it. Yeah? So if we organize it in terms of thematic areas, subtitles and all that, I might be able to find out a few things that I'm interested in. And maybe talk to the person who is in charge or send a message to the person who is in charge, so that I can get specific data for this particular data sets that I'm interested in. Now, for the person who is doing the upload, that person is going to have a lot of work, yeah? Because now you are a point of contact for all the users of this, the users could be organizations. Could be individuals. Yeah. So you can imagine getting information from all this getting certain information from these particular organizations. That would be a lot of work. But then, yeah, it has to be coordinated that way. And if we're dealing with maybe organizations, one contact person like the IT or somebody should be, like, be able to coordinate with the person uploading this data, so that there's not too many people communicating at the same time.
Angela Okune 55:30
And if you were to let's say, if this were to start to take off, would you want to be the one manually pushing, let's say, the transcripts of the materials that you've collected? Or would you want it to kind of be done maybe automatically, like you would just put it in a folder and it can go...like how would you...would you want to be in control of all of it or would you want it to be automated?
For now, we are trying to have a system where we can send everything to a central storage system. Like the project I do from proposal, contracting and everything, and the invoices also paid, so that it can be somewhere at a central location. So that when I'm sick, or I'm not in the organization, somebody else can read those documents and continue with that work. So, for me, I would say, if organizations have such kind of systems, they would do that. And then they have one central point where now if you want to request for certain information, that person can be given the authority to do that and share to that particular party.
Angela Okune 56:41
Makes sense. Okay. Emmm, do any of your funders require you to share data?
Speaker 1 56:49
Mmmmh, not necessarily with them, I think that's a must. But like, public or to publish the data...?
Usually what they do, we do the data collection. We do the cleaning as per instruction, we send the data to them. Now it's up to them to put...to share the data publicly, or just to use it or share part of it if they want. But ah...
Angela Okune 57:21
They don't expect you to...they don't want you to do it.
They don't want us to do it, yeah.
Angela Okune 57:26
So in their minds, the data is theirs?
Yeah, the data is theirs. And even in terms of contracts and proposals we write. Yeah, the data is theirs.
Angela Okune 57:37
It normally says it explicitly?
Angela Okune 57:39
In the contract?
So if they, if they...yeah, nobody else is supposed to have that access to that data. So if, if they're not able to tell us, okay, some of them might say, just do the analysis for us do this and that and give us the findings, that's it they are not bothered with the data. But majority would want the data and we can store the data...we usually have a close...up to a maximum of two years. After that, now, they should not come back to ask us for the data.
Angela Okune 57:40
...and you normally send it on email?
Yeah. On email.
Angela Okune 58:16
It's not so big, or you send it with... what is it, file transfer?
We transfer... it depends on the size of the data, yeah. And the format that they want, yeah.
Angela Okune 58:30
But for two years you maintain...two years after the project has completed, you maintain...
Actually I have not seen data destroyed two years after...we still have it...but we usually put that phrase like if you need your data we can store it for you up to two years. After that, we're not responsible.
Angela Okune 58:51
Someone might lose their own data and then come back to you and ask for it again.
Angela Okune 58:57
Okay. Okay, I think just the last statements are now after this conversation, if you think there's any benefit to digitally archiving such data, like research data in Nairobi? Any benefit to doing the work or do you think...?
It is beneficial. We get to know what is happening in different fields. So yes, it is beneficial. Qualitative data really has not been documented so much like quantitative data. If you need quantitative data, it's so easy to get access to it even some work that has been done by other organizations, as long as you're partnering with them you'll get it. But getting qualitative data has not been easy. I can't remember any organization that has offered to provide raw qualiative data. It's not easy. So I think for me, this is a useful thing. Maybe you're right, sharing in Nairobi, because most of the organizations are based in Nairobi anyway. Yeah, but I believe we'll be sharing information, not just covering subjects in Nairobi, not just subjects in Kenya, but maybe East Africa. Or all the other countries that they do their work.
Angela Okune 1:00:40
Yeah, I think the geography is interesting, because, as you said, most research companies that are based here are not only doing studies of Nairobi...
Angela Okune 1:00:51
But I think because everyone is at least has some sort of office here to allow for that physical meeting and to build that trust within let's say this data community, is important. What do you think are the risks to digitally...to archiving this kind of data?
Like any other digital thing, we've heard of, how do you call it? It's not cyber crime or what but, such kind of thing can happen. Stealing of information can happen if there are no proper protocols on how to use it. And again, if you put the restriction on money, I'm very sure you would minimize that. But again, there needs to be clear communication flow on where the payment is supposed to be done, and how the payment is supposed to be done. Because some people are very clever, might come up with their own ways and con people using the same platform. Yeah, so security of the information has to be very high. Ensured, you need to ensure that the security is very high. Like, externally, it's not very easy to tamper with any of the systems that you've put in place.
Angela Okune 1:02:23
And in terms of anonymity and all of this, do you think like, for example those in Lubaga.
Yeah, Lubaga, Tanzania.
Angela Okune 1:02:33
Like, is that important to have? Should the name of the town be removed? Do you need more information? Do you need...do you not need...do you prefer maybe that's even not there? How do you feel about proper names?
The locations are very important because this is not information for everybody in Tanzania. So we go back to the background, at least a summary, a paragraph on what was happening. And maybe you can mention not just Lubanga, but other villages where it was done. And maybe now for the transcripts, you need to label them, this is Lubanga. This is here. This is here. Yeah, that'll be important. Yeah, but a bit of a background.
Angela Okune 1:03:22
But the format like this, participants anonymized and then the location specified. It's okay?
That's fine. Yeah.
Angela Okune 1:03:29
Okay. What do you think other people in your organization would say about sharing data?
Angela Okune 1:03:37
Do you think they would share the same sentiments like you?
[laughs] That's hard to say...that's a very difficult question because you might not know what the perception they might have about this. That's purely mine. How I feel based on the difficulties we have in getting secondary data. Somebody else might have a different opinion about the same. That's why it's qualitative. All answers are correct.
Angela Okune 1:04:12
Yes of course.
So for me to talk on behalf of anybody else might be difficult. Yeah.
Angela Okune 1:04:21
But yourself you feel like you...
It's a good idea. Yeah, it's a good idea. It's a good start point for qualitative data. And if it can be more structured, I'm very sure when you talk to other experts in this field, they might even come up with more suggestions on how you organize the data. Ummm, yeah, so if it's well packaged, well organized. This is very useful for me. Yeah, for me, it's very useful.
Angela Okune 1:04:57
Great! Well, thanks. Thank you for your time. I don't know if you have any questions? We can turn the recorder off and you can ask them.
AO: This discussion took place at the working place (research office) of the person interviewed located in Nairobi, Kenya on Tuesday, May 14, 2019 from 10:23 AM - 11:28 AM. The discussion was guided by an amended version of this set of questions, which I had prepared in advance. This was the first audio recorded interview I had done as part of this project and we kept to the guide pretty closely. No refreshments were provided. We were seated in a closed meeting room. As per the interlocutor's wishes expressed during the consent form process, I anonymized all proper names mentioned in the interview including project names, organization names, personal names. At the time of conducting this interview, I had known the interlocutor for five months and we had interacted a couple of times regarding qualitative research. We had already had an in-depth more personal session (unrecorded) a few months prior to this meeting where she explained her background, training, and interest in qualitative work and I explained my interest and project to her.
This transcript is part of a broader essay ("Researching in/from Nairobi") on expectations, values and experiences of those producing qualitiative research data in and about Nairobi as part of Angela Okune's dissertation project.
Angela Okune, "TRANSCRIPT: 190514_001 RESEARCHING IN/FROM NAIROBI", contributed by Angela Okune, Research Data Share, Platform for Experimental Collaborative Ethnography, last modified 20 March 2020, accessed 25 May 2022. https://www.researchdatashare.org/content/transcript-190514001-researching-infrom-nairobi