 quality success and part of the series on the data quality online academy. So some discussion about minmax ranges. What we know is that good quality routine health data is stable and consistent and does not yoyo. It doesn't go large variations up or down and any variations can and should be explained. Full number of first antenatal visits should be constant month in, month out with maybe some seasonal variation. Head count should be consistent. The setting of minimum and maximum ranges provides a guide as to what the data should not exceed and if used correctly assists in preventing or limiting extreme outliers. So the idea is that the minmax ranges gives you a safety cushion within the data that can work that would be expected, would be usable and outside that you have to start and think and see what's going on. This means that ranges should be wide enough to accept most values and narrow enough to accept values that are not likely. A range of 1 to 1000 is not a minmax range. A range from 10 to 12 is not a minmax range. Minmax ranges are not punishment. In other words, people shouldn't keep on entering the data, keep on tripping over the messages that this is out of range. In most cases with extreme outliers, we've noticed that there are either no minmax values preset or if they are set, they are just ignored, press enter, enter and continue. So some considerations about minmax ranges, they are used during data entry. We also know that there are quite a few data elements which can show a fair degree of fluctuation that we would consider normal. Every planning supplies are a good example. Vaccine supplies, drug supplies, this month you got nothing, next month you get 2000. That can be considered normal and they are not problematic. Where you've got a disease or medical condition that runs at very low numbers and slowly there's an uptake. Minmax ranges may help pick that one up. Wherever you have expected seasonal increases, be careful when using minmax. An example is malaria or diarrhea in children in the rainy season. So there are times when you have an expected seasonal increase. And in order to access minmax ranges, you need access or rights to the data entry screen and according to the user role policy in most countries that's only granted to the people who enter data, the data capturing staff. So if you double click on any value, you will see the last 12 months of data. And so the ranges promotes good data quality and alerts to data that does not fit the norm, doesn't fit the normal pattern. And 12 months. If you double click on a value, you get what is called the data information window. Top left hand corner is the comment. And we see that this comment field is almost never used. Perhaps mainly because it's not seen. And above the comment is a star, which if you click on it will save it and allow you to run a follow-up analysis. On the right, we see where you can set the minmax ranges. You can manually set them and save them. And we see here the graph in data format for the last 12 months. If we look at this data information window, no ranges have been set. And there's a substantial amount of variation. And if you eyeball the values, do you think that we could manually think about the ranges for this specific data element for this specific facility? And possibly open to debate, our minimum could be four because that would track everything except a zero. And the maximum would be 18, which would track everything above the 20 line. Just remember that this range here is only for this data element for this facility. It's not carried over anywhere else. The other thing that we find on the data information window is an audit trail. And every time somebody changes the value, it keeps track of it so you can see when values were changed and updated and who did that. This is how you set about making minmax ranges in bulk. And you find this under data administration, minmax generation. You select a data set, select an OU level, restrict your selection. Do not try and run it for the whole country at once. Run it for one district or a small area. You run it for more than that. It will crash because remember it has to go through every facility and every data element. And you click generate and that will automatically make it calculate. So the exercise for this session is either access to your own country, DHIS-2 or the demo database. If you're looking for your own country, DHIS-2 instance, confirm if minmax ranges have been set for your country. And then if you have a staging or testing instance, have a look at it. Generate minmax values for one data set and a limited org unit level. If possible, open the data entry screen and view the data information window. We saw that picture earlier on. Run minmax violations for selected data set and OU level. Remember the minmax violations is in the data quality app. Write a short report on the violations. In other words, where they have been triggered. What is the action plan? Now that you've got these violations, what are you going to do about it? Are we just going to ignore it? Or do you think there is something that you could do? Maybe manually change some of the values, discuss with the facility staff. What is the action plan? And if you're using the demo database, follow the same instructions, except for you, it's not a staging or testing instance. Thank you very much.