Margins of error in Howard County polling, part 1
HoCo Rising recently highlighted a new poll on the Howard County executive race. This occasioned much comment among HCR’s readers on the political implications of the polling numbers. Rather than add to that discussion (fun though it may be) I want to address a different issue, namely how confident can we be that these or any other poll numbers are actually correct, or at least close to the truth? For example, the stated “margin of error” on the county executive polling is 5%; what does that actually mean? Also, even in the full results no margin of error was stated for the results broken down by county council district; can we mere mortals figure that out ourselves? I thought it would be fun to explore these questions, and in the process rediscover a bit of the statistics knowledge I had lost since college. If anyone else is interested in this topic so much the better. ...