So now that we've talked about general principles of tool development, we're going to focus on Computer Assisted Personal Interviewing or Electronic Data Collection. So CAPI stands for Computer Assisted Personal Interview. And here on the left, you can see an interviewer training, where paper questionnaires were employed. And I actually left the time stamp on this photo. See you can see that this is how things were done up until 2012. So until a few years ago, paper based questioners were really the standard practice for data collection. And this picture is really similar to what you would see in the field during data collection for a household survey. However, it's becoming quite rare to see this anymore, as most data collection activities have shifted to CAPI or Computer Assisted Data Collection, which you can see in the picture on the right. The idea for CAPI is that, it builds on its paper based predecessor, in that the starting place for a CAPI application is a paper based questionnaire. You still design a paper questionnaire and ensure that all the questions responses and skip patterns are correct. And then you have to program the electronic data application, which is the electronic version of the paper questionnaire. This front load some of the effort and workload. However, once that is done, there are a number of advantages that make a CAPI system beneficial over paper based data collection. So why is RADAR using CAPI data collection? Paper data collection systems are difficult to transport, query, and to share. And as compared to paper, electronic data collection has a number of advantages. First, it saves the interviewers and the survey participants time due to automated skip patterns that are pre programmed and the electronic questionnaires. It also improves data quality by implementing range and consistency checks and ensuring that interviewers are only able to enter correct valid responses. And we also are able to get the data sooner, which eliminates the need for separate data entry processing. It allows for quality course correction of the data if needed, and it takes less time between the survey implementation and having results available. And finally it streamline the analysis. So there are a number of different options for designing an electronic data collection system. CAPI platform options have become quite numerous in recent years, and so we wanted to provide you with resources to some of the more widely used options which include, CSPro, CommCare, SurveyCTO, Ona, RedCap, Magpi, Kobo Toolbox, Pendragon and Open Data kit or ODK. And each of these platforms have different functionalities, pros, cons, and so selecting the platform that is right for survey needs, requires investigating it to platform, and determining what's best available to suit your survey needs. For the RADAR questionnaire, we've selected the ODK platform for designing the electronic data application. While ODK is one of several platforms that allow you to do electronic data capture, it's the one we chose for radar for a number of reasons that meet the specific needs of the RADAR project. First of all, it's free and it's open source, it also has a pretty good user forum and support base and it's been around long enough to have people who can help you and who are willing to help for free. There are also some third parties that have developed businesses around the ODK platform, so you could also hire them to assist you more extensively. The forms are written and edited in Excel, which compared to a lot of other user interfaces, where people have to learn a new tool just to do their study. Excels already familiar to a lot of people, which makes it more easily used and adapted by a wide range of potential users. The development of ODK is ongoing. It began as somebody's sabbatical project who worked at Google in 2008 and it moved to a collaboration with the University of Washington in 2009. Now it's entirely supported by community of users similar to the way Wikipedia or the statistical analysis software are, is supported by a community of users. There is a steering committee that decides what new tools are needed and what needs to be revamped, but otherwise it's completely open source and in that way you can add to existing technology and use it as you need for your purposes. Okay, so we want to look at an overview of the open data kit or ODK system being implemented for the RADAR questionnaire. And this, I know is an overwhelming slide, but by the end of this session you're going to intuitively understand this. The main point here, is that the ODK system is a series of interconnecting tools that talk to each other. And the next few slides, we're going to go through each of these tools so you have a clear understanding of the ODK system. So a brief overview of the ODK tools, first, we have the questionnaires, which the ODK tool is called, the XLSForm. Next we have the server, which for the ODK system is called ODK Aggregate. Then we have tablets, which run of an app called ODK Collect. And finally, we have a data retrieval process that uses ODK Briefcase. So you can use just one of these tools or you can use several of these tools in concert and they're built to go together to create a complete Electronic Data Capture Application. We're going to talk about each of these tools a bit more in detail in the next slides. So the first tool we're going to talk about is the ODK questionnaire. So the ODK questionnaire or the XLSForm is an Excel sheet with certain worksheets and columns in a specific format. While all ODK questionnaires have the same basic format, they can range from very simple to quite complex in terms of additional structure and functions. There are typically three worksheets in an ODK questionnaire. First, a survey worksheet contains the questions and it includes three mandatory columns. We have the type of question, the variable name and the variable label. Each row of the survey worksheet corresponds to a question in the paper based questionnaire and then additional columns provide language and logic options for each question. There are also some rows for ancillary or operational variables that aren't in the paper based questionnaire, but are used to do calculations that may be used to determine a skip pattern, such as for example the number of eligible women. Next we have the choices worksheet, which contains the response options for the questionnaire. And it also contains three mandatory columns plus some additional optional columns, if you want to make your survey bilingual or trilingual. The list name contains answer choices, the name specifies a unique value within the group and the label shows the answer choice as it will appear on the form. And finally, we have the settings worksheet, the settings worksheet specifies a unique form name on the server. So XLSForms are Excel workbooks, but the way the ODK reads data is in a machine readable form called XForms. So we have to do a simple online conversion to get from the Excel workbook to the XForm. Here we've provided a link to an XLSForms to XForms, online converter, that can be used with the ODK Suite of tools. Now we'll talk about the next tool, servers. Servers have come a long way but they're probably still the biggest barrier or most challenging part of setting up your ODK data collection platform. Here we're going to use a bit interchangeably ODK Aggregate and the word server. ODK Aggregate is technically an app that sits on the server. So three things are required to set up a server. First is an IP address. Maybe you've seen these, they're often 10-digit codes that uniquely identify a computer. At the end of the day, servers are just a computer that you're touching with code instead of your keyboard. So basically, it's just a satellite computer, the IP address is the location of that computer. So in addition to the IP address, we need a domain name. The domain name is the address of a website that people type in the browser or the URL bar to visit a website. In simple terms, if your website was a house, then your domain name will be its address. The domain name is actually pointing to an IP address, but allows you to specify a domain name that is more directly related to your project and easier to remember than a series of numbers. And finally, ODK Aggregate is the app that is the container for the data that goes in and out of that central part of the ODK toolkit. There's a script in the ODK documentation that populates this for you. There are steps to set up these three components available as a resource. We have been using Digital Ocean for the IP address. You can buy domain names or you can get them for free at another website Free DNS. Once these three steps to setting up the server are completely, you will have a website that links two tablets where you can download the blank forms and upload completed forms. It also links to ODK Briefcase where you can download your data. So the server with ODK Aggregate is the central component that allows you to do everything on the cloud and also offline when you're not actually connecting to the server. Based on our experience fielding ODK surveys, we have a few recommendations for configuring your server. First, we've learned that because the ODK Aggregate is a script and it's generic, everybody's Aggregate server starts the same way. And you may want to change the default user name and password when you have your server for your study. We also recommend having two logins that have full access just in case somebody forgets their login information. We generally set up to super users to begin with and give one for the administrator and then one that goes to the field team or someone intimately involved in the data collection or analysis. We've also had some success setting up two servers, one is a testing server and then one is a production server. For the testing server, you can just use an IP address without a domain name as that server would only be used for the survey team to make sure the forms are working correctly while the production server would be where the final data would go. This also enables you to use both servers concurrently which can be helpful if for example, you are still debugging at the time of implementation, which inevitably happens with these types of studies. Just another quick note on servers, you don't have to use a domain name. We typically do for the production server for two reasons, it makes it easier fo people to remember and it doesn't script the data which is important for data security. So now moving on to tablets, the third component of our ODK system. The ODK Collect app is an Android app that replaces paper forms. Once you've installed the app, and unless you're getting your forms from the server or putting your data back on the server, it works without internet. So for the actual interviews, you don't need any connectivity and you can store data on the app and then when you're back in an internet area you can send them to the server. ODK Collect is flexible and it supports many data types including alpha, numeric and numeric data, as well as gPS, location, audio, image and video files. They also update this periodically as the user community uses it more and finds things that they like to change your ad. The latest release of the app is available on the Google Play Store and if you're interested in an earlier release, you can find it on GitHub. You connect to ODK aggregate to get the blank forms and to upload completed forms. You will need to enter your ODK aggregate URL, username and password the first time you connect to the server on your tablet or device. Once you're connected, you can get forms from the server onto your tablet and send finalized forms from your tablet back to ODK Aggregate. While on or offline you can fill in blank forms. Finally we're going to talk about data retrieval. So far we have talked about the XLSForm which gets loaded onto the server and then from the server gets loaded onto the tablet. Data gets collected on the tablet and then is sent back to the server. And now we want to download all of the data we've collected from our survey. To do this data retrieval, we need to use a tool called ODK Briefcase which is a desktop application used for exporting forms from ODK servers. ODK Briefcase can be downloaded from GitHub at the link provided here and then installed on your desktop computer. Once installed on your desktop,tThere is a two step process for data retrieval, which requires you to first pull the form you want and then second export the form. So a bit more on data retrieval. You pull the form you want because you will get access to the whole server and you might not want all of the forms that are there. For example, you may only want forms for your survey of interest or for a particular version of your questionnaire. And this step is where you make the selection of which forms you would like to export. And remember a form is equivalent to a single complete questionnaire. Next you export the selected forms and you will see you have an aggregated data set that you can use to look at your data and start data cleaning and analyses processes. So now here we are back in our ODK System Overview slide and this should be less overwhelming. Now that we've talked through all the various parts and you can see how they fit together to create a complete electronic data capture system. So configuring the RADAR Coverage XLS Form. For some organizations, the only configuration needed to use the RADAR Coverage XLS Form will be connecting the tablets to a server. But additional configuration from ODK Collect's main menu may be helpful to do a number of activities. Including changing the device language, text, font or navigation options during an interview. Changing map settings, setting defaults when editing, finalizing and importing forms. Controlling whether ODK can collect anonymous usage data and managing administrative settings. For guidance on changing survey content for the RADAR Coverage XLS Form, please refer to the RADAR Coverage ODK documentation in appendix D which includes specific information for adapting the coverage survey XLS Form. I wanted to give a little information about resources available for working with ODK generally. And then the RADAR ODK questionnaire, more specifically. ODK has developed a website that has announcements and links to its documentation materials, which is the first link provided here. In addition, ODK has also developed documentation with detailed instructions on how to do various tasks. We've provided here, a link to the general documentation site plus a few links that are likely to be helpful to most RAdar tool users. One is on guidance for making forms in Excel and another for the online converter to go from Excel to ODK XForms. The guidance for making forms in Excel contains information that may be useful if for example, you want to add a different type of question that you haven't used before but you're not sure what it's called for the type column in the survey worksheet. You could look it up here. The second link is where you would convert your questionnaire when you're done and ready to move forward. For the RADAR questionnaire, we've developed a generic core questionnaire in ODK that has all of the available modules including the household, woman, man and child. If you are choosing to use the RADAR questionnaire as the basis for your household survey, you can adapt this ODK tool alongside your paper questionnaire which is a huge time savings to not have to start from scratch with a complex ODK survey design. We have also created RADAR specific documentation which supplements ODK documentation with specific processes and recommendations for how we advise you to adapt the RADAR ODK questionnaire for implementation. We hope that this documentation will provide somewhat of an orientation to the RADAR ODK tool because the ODK documentation materials are great, but they do assume an introductory level of ODK knowledge. So these lessons have focused on survey tool developments and next will be moving on to preparing for field work, which will be focused on training of data collectors, preparing for field data collection and moving forward with implementing your survey.