 Thanks, John. Thanks, everyone. Welcome to day two of CSB conference. It's exciting to be back here and really inspired by all that has been happening since yesterday. I am going to be talking about low-income data diaries, and my name is David Celacio Poku and the key thing I'll be exploring is how low-tech data experiences can inspire accessible data skills and tool design. So a little bit about myself. I call myself a data plumber. And what that really entails is I think about my role as helping organizations, individuals, communities to figure out how they can use data for the artwork to make the right decisions. If we can go to the next slide. In case you don't know what I look like, it's a very blurry picture of half of my face in darkness. But my work since 2014 has span working with civic tech entrepreneurs, to civil society organizations, journalists, governments, all across different countries and over five continents. I do have a background in biology and computer science, but one of the things I learned really quickly was, despite all the power that data and technology can build, that people actually needed that to deliver the work that matters to society were not able to access that. So shifted my efforts to see what it would take to unlock these skills for these reformers in the work that they do. As you can see some past affiliations with a tech incubator, the School of Data, Open Knowledge Foundation, and Open Contracting Partnership have contributed to a lot of these experiences I will be talking about. We'll go to the next slide. So if we take the world today, being the real world we have, people or stakeholders are trying to make decisions. So that's the the arrow that's towards the bottom left, bottom right. And there's this complex set of steps that they have to go through in order to get to that. But we do agree that data is valuable. Data is a key resource. Some people use the term data is the new oil. I do not prefer that. I tend to say data is actually oxygen, because it can be a limited resource, but it's essential to the way we go about doing things. So if then if data is oxygen in the digital age in order to thrive in the digital and the information age, then what role can we play for reformers that are actually doing things that can have a significant impact on society? So my naive approach or my thought is we need to then move from a data-chopped society, pun intended, to one in which data enables action. And that's what I set up to do starting out from 2014 in my young and adventurous path. Data literacy. So what is data literacy? There are several definitions, but thinking about literacy, it's the ability to read, to write, to be able to access or be comfortable with the resources you need to go up around working with data or using data. And that's what in the next slide that is captured a little bit. So this is a very good report that was done by the Center for Humanitarian Data, and I encourage most of you to check it out. And they did a survey of different humanitarians or people who work in the field, and this is one of the definitions that they came up with and some of the insight that they learned. But I think there's a good summary of what that is. Next slide. So data literacy, we need two things. We need data skills and we need data tools. Here we are going to mix all these things up and then we'll get literate reformers or people working in the data space, data stakeholders, and we will solve the challenge of the digital age. So that was my thinking. Next slide. The way I came to encounter how data literacy was being done was that we were trying to move stakeholders from where they were in terms of their skills and the tools to a point where they could access the skills that existed or the tools that existed for them to become data literate. So when it comes to data skills, it's about app skilling. So what does that entail? Examples may be where a data stakeholder is given skills that allow her to access existing data literacy paradigms. So for example, teaching a journalist how to scrape a PDF with tabula. When it comes to tools, I call that app tooling, where you try to get a stakeholder or encourage them or support them to access existing data tools. So an example could be requiring a civic technology to sign up for a bank card or a bank account in order to access cloud services, the advanced features or even the core features of cloud services. So in just a summary, if we look at data stakeholders on one end and then you look at the data skills and the data tools that they need, what the current approach or the traditional approach which I call data literacy 1.0 is trying to do is trying to get them to move from where they are to where the data skills and the data tools are. So what happens when this way of going about data literacy meets low-tech reality? So here are some dyes that I've come up with that capture some of the experiences that I did not expect in doing this. So Dear Diary, this week we ran a data journalism training for 30 radio generalists. All was well until we got to the end of the data visualization module and one participant asked me, how would you communicate a data visualization through radio? Next. Dear Data Diary, my team has been working with the U.S. City Mayor's Office to implement the open contracting data standard to improve public contracting. Significant technical effort is required to kick this off, which the Mayor's Office does not have the personnel and the budget for. How do we demonstrate the full potential of the OCDS with these constraints? Dear Data Diary, we are about to kick off a three-month virtual data journalism training for six freelance journalists. My goal for today's call was to agree on the dates, the curriculum I'd like to design. I quickly learned that you can see all these bunch of things in terms of not having the right computers to be able to actually access the tools and the skills that we want to not having the right setup and even being worried about what time they'll be spending and the amount of money they'll be using if they were involved in this training for too long. Another one, at today's introduction to Data Science Meetup, only five out of the 30 participants were able to use the Amazon Web Services or Google Cloud Platform option for their model deployments. The rest did not have a bank card to create a cloud account. How do we increase access to such critical data resources for such participants? And then finally, Dear Data Diary, today was the first day of the training. Our plan was to learn how to use filters, formulas, and eventually pivot tables for data analysis. Yeah, what we ended up doing for six out of the eight hours, learning how to install software, find files in Windows, and navigate spreadsheets interfaces. So as you can see, I discovered that the traditional data literacy approaches did not work in these low income contexts or low tech contexts. And there are a number of reasons for this. I can't capture everything, but there are historic and current income constraints that limits the ability of stakeholders in low income contexts to access the current resources in the way we approach things. And then also, there is this promise of long term benefits you can gain from data literacy. And it's worth investing the resources and the time to get there. And that may not always work for people in these contexts. So what are some of the gaps that low income context presents and what I learned from being in these contexts? And there are two that I call low tech, but I wanted to break down what low tech is standing for. So there's the low technical data skills or knowledge, which has different aspects, little or no experience, or exposure to non-smart food operating system. What do I mean by that? Some people actually have never used Windows, Linux, or iOS computers. So getting used to that interface is a challenge before you start training. Low English proficiency, or even other popular Western languages are a challenge. And no programming background. That's just a few examples of the low technical data skills or knowledge. And then there's another dimension, which is low access to technological tools or processes to be able to build these data skills. So examples could be no former bank account to access some of these resources, because they require sometimes you to sign up for an account and have a credit card, limited internet connectivity, slow computing device, or not at all, manual data processes and systems, as we saw in the city news office. So these are not edge cases. And the reason why I say that is that there can be a perception that, oh, this is for places in either in the African context or in the Latin American context or in specific areas where there's a low income society. But this can actually exist in high income areas too. So the point I wanted to make low tech context do not only exist in low income countries and it can happen as a result of limited access to high computing power due to emergencies. There is the COVID-19 pandemic has left people working from home. They may not have as much access to computing resources that they had before, limited budget due to prioritization for other goals, and then reduce access to high tech tools because of increased demand. So these low tech contexts can become significant or applicable in even high income countries and areas. So the thing I learned quickly was our current approach to data literacy education is largely inaccessible to a large population of data stakeholders. And they tend to function in a limited context and with limited resources. So what do we do about this? Can we have a new data literacy approach that will pay attention to low tech context? And I believe that there is some benefit that low tech context presents to the way we go about data literacy. It can lead to increased access to more data stakeholders. There can be increased in contextual innovation and affordability, increased resilience and simplicity. The fact that you have to simplify the way you deliver skills and tools means that they become much more resilient and also simple. So this was the thing that we realized throughout the years and have been adapting my approach to testing it out to see how we can get much closer to data stakeholders and where they are with their skills. So here are some quick lessons and thoughts. So again, this is the same thing that I showed before where we have data stakeholders and data skills and data tools on extreme ends. Can we develop a data literacy approach or framework in which we now move these tools and skills to the data stakeholders? Or we can develop an approach that gets them started, but we also meet them halfway. So that can also be a hybrid approach. So what have I been doing as I'm learning these things? We've had to adapt our data literacy approaches to maximize long-term success. It's not about me. So learn that quickly. And some ways include basic skills. So providing additional basic modules around introduction to computers, the web, spreadsheets as part of training modules that people can have access to. Prepare for slow on no internet at all. So one approach that we've been testing out, we'll have a computer that we can turn into a local area network so that it can broadcast significant files and not rely on the external internet and also create a Docker image or file that has all the required software and then show people how to just install that once and then they don't have to be installing Java and other things in order to use specific software. Assess competencies. So have pre-training surveys and quizzes to be able to know if people actually are at a level of competency. You need them to be in order to deliver the skills that they need to move forward. And then plan for longevity. So embed data supports, particularly what we see in this media houses and CEOs despite them even being in trainings. When they go to their media houses, they are not able to leverage on the skills right away. So embed data supports that can actually get them going over time. And then also check in on participants or people you've trained. It's easy to have a one or two day training and then forget and write reports about how people have become data journalists. But it's much more important to check in and see how much progress they've made over time. So I take inspiration from other industries and other spaces. So this is a company in Ghana called Dex Technologies and they designed and created a science set that was aimed at solving the problem of under-resourced science classrooms. It's about not more than $20 for such a science set that has almost all the equipment you need to run a science lab for each student. And the amazing thing about it, because it's so resilient and robust and affordable, they actually also have a partnership or a contract with 500 schools in the UK to supply this to them. And then finally, another company, Zipline. It's a medical delivery drone company. And they tested out their model in Rwanda in Ghana on how to deliver critical, life-saving medical essentials to hospitals that were not accessible. And they've built this out in such a way that now even the US in this time of COVID-19 wants to use their services because there are some people in rural areas who cannot come to hospitals because they are high areas of potential infection. And so showing this that low income, low tech experiences can teach us about how we can build much more accessible, much more robust and resilient approaches to data literacy and hopefully make the space much more accessible and get to that point of empowering data stakeholders as we want. I just want to give a quick shout out to some organizations that have been part of this journey, School of Data, Open Knowledge Foundation, Open Contracting Partnership, and obviously the Cap and Trees, all of them doing great work around building that new way of data literacy. And we're all still learning. I'm still learning. There are also some tools that are out there that I didn't put on here. So everything from Open Data Kit to Kobo Toolbox to even Kaggle, that makes free learning courses and without requiring credit cards or you signing up with a bank account, all these are making it much easier for us to get to data stakeholders. And with that, I will pause. So the last slide, just leaving the question, so how would you communicate to data visualization through radio? Yeah, that's great. These are great questions and really appreciate you bringing the discussion of how we communicate data, data literacy, bringing it in such an illustrative way that brings this discussion of completely destroying the rubric of the skills training and the people have to have that power dynamic of accessibility, really breaking that down. So thank you very much. We did have a couple questions, but we're out of time. So I am going to transition over to the next speakers. And for those that weren't here last week, or sorry, yesterday, last week, for those that weren't here last yesterday, what we'll be doing is we'll take these questions that we've received while we were listening to David. We'll put them over in Slack and there is a channel in Slack that's the CSV5 Q&A. And David will be available to respond to them in Slack. So if you'd like to talk to him through Slack or if you'd like to pose more questions, head on over to Slack. So thank you very much, David. And I'm going to move over now to Emmy and Daniel.