 In this presentation, we will discuss organization units. But we have already introduced this concept when performing data analysis. We will now discuss some of the key design principles when creating these within DHIS2. So by now, we should be aware that organization units represent the administrative or organizational units within a particular organization. This could be facilities, clinics and hospitals within a health system, provinces and districts that define a specific administrative unit within a hierarchy, and this could also be specific divisions or departments within an organization. We know that the organization units are located within a hierarchy and reflect the administrative structure and levels within a particular organization. Data can also be aggregated and rolled up through the hierarchy. We have seen this in the different reporting tools that are available within DHIS2. An example hierarchy is training land. It has four levels, national, region, district and facility. We've been using this hierarchy in various examples and should be familiar with it by now. So let's take this hierarchy and apply it to some key design principles. One key consideration when we're designing the organization unit hierarchy is that we do not want to mix multiple concepts. As an example, we may have different organization unit levels that help us to define our hierarchy. This could be for example, facilities and villages. We should not mix projects in this hierarchy as well. Projects might move over time or might expand to other locations. Because they are transient in nature, we want to make sure they're captured in different ways. Applying them directly in the organization unit hierarchy is not a correct application. Therefore, we want to make sure that we are not mixing multiple concepts in these hierarchies. Organization units can also be classified into groups. These groups can further be grouped together into group sets. We will demonstrate this concept in a moment. Organization unit groups and group sets are particularly useful for data analysis. We can add in these individual group sets as data dimensions within any of the analysis tools. There are some rules here, however. An organization unit can only be a member of one group per group set. This is because if these organization units are members of multiple groups within a group set, you will have duplicated data values when you perform your data analysis. These groups and group sets allow for building alternative hierarchies for the purposes of data analysis. Let's go through a couple examples. The first example we can discuss are types of facilities. For example, in training land, I have a number of different facility types. I have health centers, dispensaries, private hospitals, specialist hospitals. We can take all of the facilities that meet this criteria and divide them into individual groups. These groups can be further grouped together as the type of facility. So within the type of facility group set, I have my individual groups of dispensary, health center, private hospital, specialist hospital, and referral hospital. Each of these groups are further subdivided into the individual facilities that make them up. Here is an example where this is useful for analysis. We can see that we've added in these individual facility types to our data analysis. For now, we've just added in three types, dispensary, primary health care centers, and health centers. This allows us to separate our analysis by these facility types. Here is another example of organization unit groups and group sets. We still have the same facility hierarchy, but now we're applying them to the groups of urban, rural, and peri-urban. This group set makes up the location of these individual facilities. Let's take a look at this example in DHIs too. Here we can see similar separation of these data items. We have now peri-urban, rural, and urban separations of these data elements. We can see these values of change to reflect the organization units that make up these individual facility locations. These are quite useful in terms of separating our data during data analysis. They allow for these alternative hierarchies to be built and applied to our data. Just remember that key rule that we discussed. Organization units cannot be members of multiple groups within a group set. For example, if Crow Health Center was part of both the urban and the rural group, if I display values by urban and rural and compare them together, I will have the value for Crow Health Center being counted twice. We want to make sure to avoid this when we're creating these groups within DHIs too. Now that we've discussed some of these principles, let's get into DHIs too and show you how to create these organization units.