![]() Therefore we estimate that $50\%$ of people in Region $1$ have good connectivity. I suppose what you are trying to say is that the ratio of premises with good connectivity to total premises is a useful predictor of the ratio of people with good connectivity to total people in the same region.įor example, in Region $1$ we find that $50\%$ of premises have good connectivity, Some premises in each region have good quality, and some have poor quality. I am assuming the individuals from a specific region have the quality of the internet connection of the region.Ī region does not have a quality of internet connection according to your figures. organisation B has 5000 individuals, 4500 come from region 4 with good quality of internet connection (66/90=73%) and 500 from a region 5 with lower quality of internet connection (33/123=27%), but I cannot just do 66+33/90+123 because the number of individuals coming from each region for that organisation is not 2500 & 2500, but 4500 & 500, as in, there are more people coming from region 4 than 5, so I need some sort of weighted average here. ![]() I would like an overall metric of quality of internet connection for each organisation. What I have is the quality of the internet connection per region and I know in which region each individual lives so I am assuming the individuals from a specific region have the quality of the internet connection of the region. I cannot link each individual to their premise. I would like to calculate the percentage of individuals/premises with decent internet for each organisation taking into account the number of individuals per region (that's my weight). I cannot link the individual to their premise, I only have access to a proxy - the number of premises in the region where the individual lives that have decent internet. *TOTAL_INDIVIDUALS that are part of the organisation In my case, I have something like: ORGANISATION TOTAL_INDIVIDUALS* REGION REGION_OF_TOTAL_INDIVIDUALS NUMBER_PREMISES_IN_EACH_REGION_WITH_DECENT_INTERNET TOTAL_PREMISES_IN_EACH REGION However, I am struggling to apply the same thinking to the weighted average of percentages. ![]() ![]() If we have the following (example taken from ): DO NOT LIKE CHOCOLATE LIKE CHOCOLATE TOTALSĪnd we want to calculate the percentage of people who like chocolate across the entire population, we do:ģ30,000/500,000=66% and NOT (90,000/100,000=90%) + (240,000/400,000=60%) / 2 which would give us 75%. I have a question that may be basic but I do not seem to find the solution for this and so would appreciate any advice please. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |