 This video is about how to use weights in Nesta, first log into the UK Data Service web page, then get to Nesta through Get Data, explore online, and following the links through to Nesta. I'm going to use the 2017 British Social Editude Survey in this example and I will find the data sets by dropping down from Research Data Sets on the left. Weighting the data will adjust for survey design and non-response and aims to make the data representative of the general population. So as an example, I'm going to select Variable Description, Household Grid and Demographics and have a look at the variable how many members are there in the household. Selecting on that variable moves the descriptives into the centre of the screen, we can see at the top right of the tab Description. Now, if we want to weight it, we just use the weight icon, which is on the top right. There are two possible weights automatically selected after selecting the weight icon. By reading the documentation located on the catalogue page for the British Social Editude Survey, you can decide which weight you need to use. In this example, we will use the final BSA weight by selecting that and moving it to the variable selected box and then selecting OK. A table comes up and you can see that we are now in the Tabulation tab at the top of the page. We're going to use the same variable before, how many members in the household. Selecting that and add to row. Accept the UK Data Service Access Control and then the table comes up with the weighted frequency. We can tell that the variable is now weighted because there is a note at the bottom of the table which tells us the weight is on. Also, the frequencies are no longer whole numbers, which is an indication that they have been weighted. Looking at the percentages here, for example the number of people living in the household with only one person. When it's weighted, it's 17.5%. If we go back to the description, you can see the same variable that we have here, an N of 1181, a whole number and a percentage of 29.6%. So that's vastly different from the weighted results. When you weight data, it's not generally going to make such a big difference. However, this is an example of weighting to correct the survey design. During the data collection, the interviewer only asked for information from one member of each household. And that's why there is such a wide difference between the weighted data and unweighted data for this particular variable. Thank you for listening.