159 – Innovative MSc Programme for Maths Teachers in Kenya, Part 4

[00:00:07] David: Hi, and welcome to the IDEMS Podcast. I’m David Stern, a founding director of IDEMS, and it’s my privilege today to be here with both Lily and James. And we’re gonna have a three-way episode related to this MSc program in math innovation. Hi, Lily and James.

[00:00:26] James: Hi.

[00:00:27] Lily: Hi, David. It’s good to see you.

[00:00:29] David: So I recently recorded a few episodes with Mike Obiero about the MSc program in Maths Innovation, and he got stuck into explaining all the different ideas behind it. And when it came to the, the data side of it we discussed the fact that really James, we, he needed to refer to you.

And also Lily, you’ve got so deeply involved in actually designing the courses, I thought it would be re really useful to have the three of us discussing this together. But James, it’s been quite a long time that we’ve been discussing MScs in data research methods, and so on. And you’ve got a, you are getting a sort of a perspective on this, which I think is quite different from how a lot of other people are thinking about it.

Do you want to just share where you are at? .

[00:01:23] James: I think the most recent involvement was the the AIMS meeting. Where people from different universities across Africa came together to discuss an appropriate African MSc data science program. Yeah. Because most universities are at the point of developing and rolling out these degrees and I think the meeting there was mainly to yeah, try and contextualize the degree for, for an African con setting.

So I think going back to that meeting yeah, so I think the main thing that came out of our discussions in, in, in about the MSc in data science is the fact that I think, for our programs to be effective, we needed to start with data instead of methods like in most data science degree programs are doing.

And I think that is the sort of the main thing that was coming out of those of our discussions. And I can also go back into other MSc programs that are not data science, but I remember the research methods one in Maseno. I think that was one that is also closest to the data science.

And I think it, it took it also had the same approach where we were developing a degree program that was meant to be very useful in the practical world, in the real world. And I think, the approach there was also the same, that, we need a degree program that is strengthening people how to use data instead of focusing on how to use methods, which is the case with many statistics degree programs.

They focus a lot on the methods and they don’t focus a lot on the data.

[00:02:59] Lily: You, you’re saying this James, about many statistics programs. Is this kind of just in the African context or is this in your experience

[00:03:08] James: No, I think, yeah. In my, I think it’s a general problem for statistics degree programs they focus a lot on the methods all over the world. But I think in Africa, yeah. The problem is compounded by the fact that there isn’t access to, the data itself and also technology to be able to teach using data.

So like computers those ones are a need. They’re a challenge to access in Africa. So I think that probably compounds the problem that…

[00:03:40] David: Lemme take a step back on this for a second because I think there’s a lot of things you’ve already brought out that I think we’re gonna have to tease out here. One of them is you, you’ve discussed statistics, you’ve discussed research methods, you’ve discussed data science. There’s a whole range of things which not everyone will have been exposed to, the distinctions between them and so on.

[00:04:03] James: Yes.

[00:04:04] David: You’ve discussed the focus on methods versus data. Surely anything statistics related is data focused. Anything data science related is data focused. So again, articulating what you mean by that is gonna be really important. And then there’s this question, I’m really glad you brought up this AIMS meeting, which was a wonderful opportunity where educators from across the continent got together to try and ask the questions of what are the data science degrees that should be developed in the African context and how they should be developed.

It was a really powerful meeting. And it was a, I was lucky that I was actually teaching. Of course in the doctoral training school while that meeting was going on and I was able to come in and out and be part of it. But

[00:04:51] James: and your course was an interesting one. Problem solving in data science

[00:04:54] David: absolutely. Yeah.

[00:04:56] James: yeah.

[00:04:57] David: And it was a, and it’s of course, which since then, of course, you and Lily have gone on to teach in a similar context.

[00:05:04] James: Yes.

[00:05:05] Lily: which we’ve done a separate podcast on on that Rwandan doctoral training school course.

[00:05:11] David: That’s true. And Lily, one of the things that I want to draw out is this fact that this is not an Africa problem, and I think your experience and your personal experience on this relates to this, this issue about the data versus the methods, this is something you have personal experience on from, going through your education, becoming a statistician, and then recognizing that you actually didn’t have the data skills you thought you had.

[00:05:36] Lily: Yes, I remember when I, after I had done my bachelor’s degree and I was actually sat with you, David, in an office and someone came in for some kind of statistical consultancy help. I think that they were a master’s or a PhD student. And they came in and they asked the what should be a very simple question I felt.

Because I was like, I’ve got a bachelor’s degree in this, I should know the answer. And they said something about, what parametric model should I fit? And I was thinking, I don’t know what this means, I can’t tell this by looking at the data. I dunno what a parametric model is. It might have been a non-parametric model and…

And then just watching that kind of process and realizing like that even you didn’t know what it means, and actually you need to go through the data to be able to come to that other side. And then that experience actually looking at the data first. I think a lot of kind of the statistical learnings have come from that I’ve gone through have come from just experience rather than actually learning the, than actually at the university level.

[00:06:38] David: Yeah, it if you think of a particular model, or a particular test, often asking the question about that model or that test, you need to step back first and actually wait. We need to make sure we understand your data first, because that’s going to give us the answer. I, as a statistician shouldn’t answer that. It should come from your data and your problem. That’s where we will then draw out what is needed.

[00:07:04] Lily: But that was a very key thing. I think that was, at least for me, missing in that kind of education level was that you, like an expert, doesn’t look at the data and know immediately what it is. They don’t look at the data and immediately go, I know what the method here is. It’s that actually you still need to work on it and draw it out.

[00:07:21] David: And most importantly, if you start with the data, you get to multiple and different methods. So you don’t need specialists in an individual method.

[00:07:31] Lily: yes.

[00:07:33] James: David, I don’t know whether I remember my experience.

[00:07:35] David: Yes, go ahead.

[00:07:35] James: About the mean and the median

[00:07:37] David: Oh, I remember

[00:07:38] James: In the MSc class. When we were discussing about the mean and the median which sort of average should you use? And yeah, I think it was at that point, that I realized that, I hadn’t really understood when, these concepts.

I know how to calculate the median and I know how to calculate the mean, but, understanding the concept, when to use them when they’re appropriate, those are not very clear up to that point. And I think it’s because the teaching is so focused on the methods.

[00:08:04] Lily: To add to that from, and I, this my, but when I was teaching at AIMS with Francis, and at this point I’ve got a PhD in statistics, so I should definitely know things by now. Surely! And Francis puts up a question, this kind of question, I think it came from you, David, and I’m sure you’ve seen it, James, of of just the kind of plot of data showing the distribution of it.

And asked, estimate the standard deviation. And I was like, how? I was dumbfounded. I thought that Francis was joking and I thought like, how we’re not meant to know this. How can we tell this from a graph? Then he went through it. And that for me was a huge realization of not actually something like, like I thought that I understood these things, but clearly I didn’t.

[00:08:49] David: And most importantly, if you can’t estimate it, then you can’t interpret it because you don’t know what it means. You don’t know what it looks like.

[00:08:57] James: Yes.

[00:09:00] Lily: yeah. Yeah. So there’s definitely a lot of areas where I feel that definitely… and I’m sure, I hope that more realizations like that will continue to come out. That’s why I’ve been really enjoying writing these data science courses or these data skill courses for this MSc is because that kind of realization of oh, this is just that!

[00:09:23] David: Yeah and I think this is. This is something where what I’ve been really encouraged in some sense with is that there is a growing recognition, I believe, that it’s not all about the methods. There have been, and we’ve highlighted this in some of the responsible AI courses, Lily, that there have been huge scandals that have come because people have just applied methods without understanding the data and the biases and all the rest of it. And so I think this is all a big part of what I believe this approach that you’ve described, James, where you are wanting to put the data first and I think this is fantastic and I’m really encouraged that I think there’s a good chance of you having success in this.

I don’t know if you know the place where I first really encountered this, which again is one of my big realizations, was the New Zealand school curriculum where they managed to get data skills or statistics in from the first year of primary right the way through schooling. And they didn’t put in any hypothesis testing, any sort of modelling as most people would know it. But they did have bootstrapping in, and at first I thought, this is ridiculous, bootstrap, boot. You know what bootstrapping come from? Bootstrapping is because this is like trying to lift yourself over a fence by pulling on your bootstraps.

It’s impossible! It’s stupid. It’s a stupid thing to try and do. That’s where the term comes from and that’s what it is. Bootstrapping is the most silly statistical method. It is very useful because it is very widely applicable. It doesn’t have many assumptions behind it. The only assumption it has behind it is that the data you have is representative of the data you want to represent.

That’s the only assumption that you need for bootstrapping. You don’t need any other assumptions about your data. And that’s the key point. It always applies. It is never the right thing to use, but it is always something which is valid to use. And this insight to me that actually when we are talking about the training people with how to do data, how to work with data. If we don’t have something, if they don’t have a tool in their toolbox that they can always use, then what, how are they ever going to take the assumption seriously?

Because when the assumptions are not valid, they don’t know what to do and there’s nothing they can do. Whereas if you, if the only tool you have in your tool toolbox is bootstrapping, you can always use it. It is almost never the best thing to use. It’s pretty much, it’s not what we would recommend as being the right thing to use in almost any occasion, but you can always use it.

[00:12:42] Lily: I mean in terms that bootstrapping is something I didn’t come across until working with you, David. Like it’s not something that for me was taught in my degree and my understanding is that it’s just re sampling from the data that you have.

[00:12:57] David: Yeah. It is basically you resample from the data you have and and with repetitions, so you just get this sort of, the same data points, but different numbers of them and different whatever. And so it’s a really basic, silly method. But the fact that you can use this and you can always use this to therefore get an idea of, to do the equivalent of hypothesis testing to… it won’t give you the best results ’cause it doesn’t use the power of statistics, but it always applies.

And so that realization that actually everything I thought I knew about what was sensible to teach first and so on, just went out the window once I really understood the power of what the New Zealand educators had come up with. And this goes further of course, because this relates to the software.

We still don’t have software, which makes it easy to use bootstrapping in all cases. This is, this is something which I believe needs to be developed further, not because it’s needed in the analysis, but because it’s needed for education to, to enable people to become comfortable with data and to always have a fallback.

[00:14:15] James: I think in the African setting. Oh, probably in the general statistics education sort of context, I think the issue of software is a big one. Software for teaching. And I think that is partly, could be partly the challenge or is partly the challenge for, the fact that, the teaching is not data focused.

Because there isn’t easy software for people to use in African setting, then most softwares are commercial. The easy ones, like SPSS and all that. In the current setup we have R we have Python, but those ones are hard to teach to, teach in. For, and we have lots of lecturers who are still not well versed with them.

And therefore, using them to teach statistics in a data focused way then becomes a problem. So I think, yeah, the issue of software is a big one, and it plays a role in, in, in the fact that the teaching is more method focused than data focused.

[00:15:13] David: And I like the fact you brought up R and Python, which are fantastic. They’ve changed the world because they’re open. They’re incredibly powerful, but they put coding first and this is something which therefore, in a context where computer literacy is low, puts people at a great disadvantage.

And I’d argue that’s one of the big differences between a lot of the African contexts and a number of international contexts. In a number of international contexts the computer literacy is extremely high from a very young age, and therefore coding is something which is often brought in and it’s natural, and that’s not a big deal.

It’s not a big, as big a barrier in a lot of African contexts at the moment, digital literacy is still relatively low, and that means whether it’s, it might be higher for a younger generation, but it’s still low for the lecturers. And there’s, so digital literacy is a, is still an issue. And what I believe that is, as I often believe, is that challenge is actually an opportunity because I think coding is the wrong skill.

It is difficult in a context where digital literacy is very high to convince people you shouldn’t learn coding, and I don’t want to convince people you shouldn’t learn coding. I believe I learned to code when I was young and it’s helped me and many other people believe the same. So teaching coding is a good thing, but.

Relying on coding to teach data skills is wrong, and that’s that… Actually making that statement I think is very difficult, but it’s a statement I believe in that data skills are don’t, should not rely on coding skills. That’s just an anomaly for a short period of time. In the future, we will have better tools, which mean that the coding skills aren’t what’s needed for data skills.

[00:17:17] James: I think that challenge is probably yeah. Is, is an impediment to the change that we want to see in statistics education and data science training in Africa in the African context, I think so to speak. So if you think about the people who are teaching these courses at university yeah, because they find it hard to learn these skills.

I think they will, they struggle therefore, to make this change. I. To be able to teach this way. So I think there’s a lot of

[00:17:47] David: let me get clarity on what you’re saying when you.

[00:17:49] James: yes,

[00:17:50] David: What is these skills? Is it the coding skills or the data?

[00:17:54] James: Yes. I was following up from the coding skills.

[00:17:57] David: Okay.

[00:17:57] James: Yes,

[00:17:59] David: In my experience, within the African context, there are many lecturers who won’t touch it,

[00:18:06] James: yes,

[00:18:06] David: and there are some lecturers who make it part of their identity. Where I feel, what I feel is interesting with you is you are in between. You you have the coding skills, you’ve done things, you’ve learned those, but you are not advocating that everyone should be learning to code in that way. And that I think is really interesting. So let’s articulate that a bit more.

You’ve got the skills to do the coding, but you’ve recognized that, or you’ve identified that there are other skills which people can learn irrespective of whether they teach the coding skills which you value more highly, which relate to the data, these data skills. And so when you are putting forward, the data should come first, where’s that come from? What is it that you’ve experienced that’s helped you to to make that jump, that leap? I know for example, the Stats Made Simple experience was a very deep one for you.

[00:19:14] James: Yes. Yes, exactly. That’s where I was understanding the basic concepts, like the mean, the media, the standard deviation that Lily was talking about. So building that understanding of those concepts is actually something that. Yeah, that I have been advocating in my teaching but I’ve been using data to to teach those, to teach that understanding.

And I think at that level I’ve not needed, those hard coding skills to be able to make people understand those concepts. So building that statistical understanding.

[00:19:50] David: Let me see if I can articulate this correctly.

[00:19:52] James: Yes.

[00:19:54] David: So I think what you are saying as I understand it, is that you recognize that the coding skills you got were independent of some of the data skills you that you’ve, that you value,

[00:20:11] James: Yes.

[00:20:13] David: by disentangling these two there, then you believe that more people can gain the data skills.

[00:20:21] James: Yes, exactly. Yeah, disentangling the two sort of helps. But if you put it forward that you need the coding to be able to understand, then it becomes a difficult route and approach to follow in our context.

[00:20:35] David: I, I think it’s true everywhere, and my theme is that actually it’s easier to see when there are relatively low levels of data liter, sorry, of computer literacy of digital literacy, that if you put the barrier of the coding skills, then you are actually limiting access to the data skills.

[00:20:58] James: Yes.

[00:20:59] David: And I believe that separation needs to happen everywhere. That actually this, the importance of the data skills is much bigger is much greater than the coding skills. Coding skills are great, but it’s always going to be a small proportion of the population that has them, whereas the data skills that we’re talking about, this is something which, once you’ve understood it, these are accessible.

It’s just they’re often put behind these barriers, either now the coding barriers or historically the mathematical barriers. And actually recognizing that the data skills themselves, if treated in their own right, then they are more accessible possibly than any of the others.

[00:21:50] James: Yes, exactly.

[00:21:53] David: And that’s, and this is where, I still remember in some of those discussions where you are arguing quite often for, let’s put data first, and this is what you’ve done. In this MSc program, the MSc and math innovation, this is what we’re supposed to be discussing. So let’s get back to that for a second.

The, there you’ve got this data stream where you’ve got two components to it. You’ve got the data skills, which is really James, where you are putting the data first. Then you’ve got the responsible AI, which Lily is where sort of these misconceptions where a lot of the emphasis is on how do we make sure that if people, when people use AI, because they should be using AI because we need AI in all sorts of ways, it’s helping wonderful advances happen, but we need to use it responsibly.

And those are the, that’s all part of this data side, these two sides to it, the responsible AI and the sort of data skills. Lily, do you have thoughts on that or do you want to add some comments on that?

[00:23:03] Lily: I hope that the, like throughout the course, the kind of different ways that AI is being used, ’cause I think it’s very easy to say we need AI and and it’s being used in, but maybe people don’t have, don’t have kind of concrete examples of that in their context. I think AI can seem like this kind of mystical being, or you might immediately nowadays think of… as I accidentally can have the tendency to do, act immediately think of kind of generative AI, Chat GPT and whatnot.

Whereas actually AI is being used in lots of ways and I hope that in the course those kind of different ways that AI can help, and has the potential to help and does impact in a really positive way, comes out and brings some clarity.

[00:23:51] David: Absolutely. And just to articulate this, that the large language models or generative AI more generally, these are what has led to this recent sort of charge for AI, but AI more generally and the machine learning methods behind it have a much wider set of applications and recognizing that these are just part of the tools that you should have in your toolbox. This is James, where we come back to, your data should come first. If you have these tools in your toolbox and other tools, then you have a better chance of actually drawing out the right information.

Actually being able to go into a particular data set or a particular source of data and draw out the information which is useful, and understand what it can tell you and what it can’t tell you. Okay. I want to go slightly further. James, in this MSc program is designed particularly, Mike has conceived it as he says, first and foremost for teachers, for educators. Why is it that for teachers and educators, these data skills and thinking about responsible AI is something which you are also keen that this is the right audience for these ideas.

In the Kenyan context, you know what, why are we taking teachers and training them in data skills and responsible AI?

[00:25:29] James: In my view, I, several reasons. One is yeah, we, statistics, concepts begin to, in Kenya, I think in the, and in teachers here we are meaning high school teachers mainly. I think they, the high school curriculums, there is some statistics com, components there in, in their curriculum.

And I think having teachers who are understanding the data and the statistics… if the, if we have teachers who have, understanding of the data and these statistical understanding skills, I think they can begin to impact those skills to school kids right from high school before even they joined university.

So I think we have a big market there. Teachers all over Kenya are looking to further their education, especially maths teachers here. I think we are talking about maths teachers for this program and I think, capturing that market, they would begin to impact the skills to, high school kids.

And I also see maths teachers from high school once they have their masters, they often transition into university teaching. Some of them get into the university and yeah. I think, yeah, I think that’s a way of also building a community of educators who are putting data first in their teaching.

So I think, yeah, that’s why we have a big market for teachers for this degree program. Yeah.

[00:26:48] David: And even when you were a student, many of your colleagues, your masters, when you were teachers.

[00:26:54] James: They were teachers. Yeah. Yeah, exactly. Yeah.

[00:26:57] David: and I think that the, one of the things which you’ve drawn out, you’ve drawn out the example of mean, median. This is taught at high school.

[00:27:07] James: Exactly. Yeah.

[00:27:09] David: What if that understanding that you only got when you were a postgraduate student of those basic concepts, what if that could become, ingrained in the school level education?

What if those concepts which are taught then are taught with that depth of understanding? Why might you want to use one or the other? What does this mean that critical thinking about data? This is broadly data literacy.

[00:27:35] James: Exactly. Yeah.

[00:27:39] David: And the best way to get that into schools is through teachers.

[00:27:43] James: Yeah.

[00:27:44] David: And so anyway, this is a, there is a wonderful opportunity there. I have one other element on that just from my personal experience in Kenya, was that when one of your colleagues, Zach did try to change the statistics part of the high school education, the biggest problem he came up against was the teachers he was working with were saying, oh, that’s great. What about the rest of the maths curriculum? And so because the statistics is within the mathematics curriculum, you need it embedded in a broader initiative to be able to have scalable impact.

That was a really deep insight for me within the Kenyan context.

[00:28:28] James: Yeah, exactly. I was wondering whether this program would also be appropriate for… I was, before I became a lecturer, I was a graduate assistant. And yeah, I had to do an MSc as part of my, yeah, of my job. So I, yeah, I was just wondering because I’m thinking that, yeah, we will have teachers at high school who are teaching statistics the right way, or data, building this data literacy.

What about those in the university?

[00:28:57] David: Absolutely. I think you mentioned yourself that the transition, once you have a master’s, if you go on and do a PhD, then you have that. So that is a route, but, and you know as well as I do that graduate assistants are hard to come by right now.

[00:29:12] James: Yeah, exactly. Yeah. Yeah.

[00:29:14] David: But if we take the equivalent of that, this is partners like, them who are offering these long-term internships, these apprenticeships, which could include this sort of master’s program as part of it, as a way of training the next generation of lecturers and getting people into those.

This is absolutely possible and I think it’s really what I would hope would happen is that this won’t just be a master’s for teachers, but particularly for the data skill side, this could be data managers all over the place. This could be a whole range of people who would gain, because, imagine you are a data manager as part of an agricultural research project somewhere.

You do this MSc, not only are you learning the data skills you need to do your data job better, but you are also learning the education skills, which means that you can do the capacity building of others better, because that’s often a big part of your role is that work with others on their data and play a capacity building role.

And so there’s all sorts of things. So I, there’s a number of audiences that I would be really excited. This program could serve them well. And that’s the intention I believe Mike has. And, your vision that he’s adopting of the data focused education for the data stream, which doesn’t exist within the Kenyan curriculums and more broadly, it’s been very hard to do.

I think you are trying to launch at Maseno University, a data science program in this vein, and that may appear in the next couple of years, but it doesn’t exist yet. And so it is not in competition. And of course if that data science degree did exist, because these are open educational resources, what you are building now can fit into that.

Now that degree might be more focused. It might be face to face, and so therefore it has other things, which are going for it. But I think this is exactly where that synergy could then come in.

[00:31:14] Lily: This is very exciting. It’s very exciting to see it, in the bigger, in this kind of bigger picture. And to understand a lot more about the context from, from you and James.

[00:31:25] David: And maybe let’s just finish ’cause I’m conscious we’re almost out of time. Let’s just finish with two last questions and I’ll start with you, Lily. You have been actually supporting writing quite a lot of the material around this, and as you said you’ve gained quite a lot from that first draft that you are creating, which is being iterated on.

And what is it that you would say you found is different from, really different from the approach that you had when you were studying statistics?

[00:32:02] Lily: A lot of it, I’d say, would be that kind of, interactivity perhaps, or that kind of trying to make it interactive through the discussion boards. I think a lot of times with statistics or mathematics in, in those degrees, at least in kind of the degrees I did, you would, you might work together on a tutorial.

You might work together on a practical, but you wouldn’t have discussions around things. I know friends of mine that were in kind of humanity degrees would have these discussions. And so what I find quite exciting is this kind of notion that we can bring that into this area to try and really draw out some more kind of powerful points.

[00:32:43] David: That’s a really interesting one. Because it is this fact that within the skills, once you take the data focus, there’s actually a lot of complexity behind it. And so it’s a lot less about right answer, wrong answer, and a lot more about actually, thought processes and. And that’s where having the discussions which help deepen your thought processes is what it’s about. So it’s a lot less black and white, right or wrong, and a lot more this is complicated. This needs a bit deeper thought than I, than you would’ve thought. And maybe it’s more than you can actually do to just get right or wrong.

There’s no black and white, there’s shades of grey.

[00:33:24] Lily: Yeah, and I think that, that can be as a mathematician, you expect things, going into a maths kind of environment or into a maths teaching level, you expect things to be black and white. You expect that you’re gonna be taught okay, right and wrong. Whereas it’s quite interesting that it’s not like that.

[00:33:43] David: Abso and this is a really I didn’t expect that, but this is great because the data skills, a lot of what it is it’s about recognizing that it’s not about getting it right or wrong, it is about actually doing what you can within the constraints you have. And there’s almost always more that could be done or other things that could be looked at.

But once you’ve got data, what can you do with that data? Thank you. That’s, that surprised me. Good,

[00:34:10] Lily: It surprised me too, but anyway.

[00:34:13] David: James. So the one view that I want to just finish us off on is, within the Kenyan context that you understand very well. If this degree program is successful and these data skills get more widely out there, what do you hope that this might mean? What might happen if these data skills were more widely available in the population because the teachers have them and they pass them on to the students, and you now have a wide range of people that have these basic data skills.

How might this change the society? The…

[00:34:52] James: Yeah, so I think if, yeah, I think that probably what I’m thinking is with the population, with the skills, then what I’m seeing is probably informed decision making. I think so. People will be able to interpret things correctly or information correctly and, apply it in their context.

So I think that’s probably the bigger picture that I’m seeing. That is coming out of, if this degree program rolls out very successfully. I think even the, this will be what catapults the university education also to change. And yeah, I think generally we’ll have a society that is data literate and, being able to, use information to make decisions.

[00:35:39] David: So let me just see if I can understand what you’re saying. I think this is at the heart of actually making informed decisions is being able to interpret the data behind them. And so you are saying that if we have a population where those skills are widely available, it should help in that and you’re in some sense saying that at the moment, and I’ve seen this in many ways, the there’s decisions are often made very hierarchically and they can often be not informed by data because the data doesn’t exist. It’s not there, the structures.

So you actually have quite a lot of instability that then comes and is generated because the data on the situation on the ground is not informing the decision making in the ways it could and that this is being, and I think I agree, having that informed decision making, being something which is really widespread and becomes a cultural part of the society because it’s baked into the education systems. This may enable a layer of growth, which is, incomparable to what we could imagine without it is what I’m hearing.

[00:36:49] James: Yes, exactly. Yeah.

[00:36:51] David: No. Interesting. And I think it’s really interesting that you’ve chosen that informed decision making, because it is something where this isn’t a challenge in Kenya. This is a challenge everywhere in the world, and this is understanding how do you build structures so that societies make well-informed decisions based on the realities.

And the data that comes through and actually having good data, that’s a global challenge. And so it’s nice to think that actually a country like Kenya could become at the forefront of this, who knows? A nice thing to have.

[00:37:25] James: Who knows? Yeah. Yeah.

[00:37:28] David: Great. Thank you both.

[00:37:28] James: Great.

[00:37:28] David: It’s been an interesting discussion, so thank you and look forward to talking again soon.