This was an interesting one: The Washington Post got a lot of clicks with a new “poll” that showed Clinton and Trump are neck-and-neck in Texas. The only problem: It wasn’t really a poll. It was a online survey via the project’s co-sponsor, Survey Monkey. Because it’s a survey and not a poll of randomly-selected respondents, it wasn’t possible to calculate a Margin of Error. To put this into perspective, I compared the results with the most recent statewide polls and then added a national average to boot. I hope this made the story more clear.
Sept. 7, 2016
Comments