Have you seen people taking statistics classes? It’s almost universally hated because, for some reason, people struggle with it. I really enjoyed it because it’s probably the most practical form of math you can take. I was only one class away from getting a minor in statistics, but wasn’t going to take an additional semester just for that minor.
I really enjoyed my stats classes. It was a lot of really great information and general knowledge.
I do think having that background makes me simultaneously more and less trusting of polls. The recent performance of polls in the midterms and across various races this year are quite surprising to me. They haven’t done a good job of capturing the abortion backlash trend against Republicans.
There would have to be a reason however for their deficiency, and that’s where the stats background is helpful. I’m not going to baselessly dismiss polls because I dislike the results, I need reasons. And the best explanation I can think of is the simple random sample. The people they’re polling don’t constitute a true SRS, because I wager there’s some conflating factor skewing results. Landlines are going to favor older populations. Texting people to take polls is going to have the bias of who actually goes on to take them.
The biggest enemy to statistics is that there’s no law forcing people to answer questions if they’re chosen for a poll.
Have you seen people taking statistics classes? It’s almost universally hated because, for some reason, people struggle with it. I really enjoyed it because it’s probably the most practical form of math you can take. I was only one class away from getting a minor in statistics, but wasn’t going to take an additional semester just for that minor.
I really enjoyed my stats classes. It was a lot of really great information and general knowledge.
I do think having that background makes me simultaneously more and less trusting of polls. The recent performance of polls in the midterms and across various races this year are quite surprising to me. They haven’t done a good job of capturing the abortion backlash trend against Republicans.
There would have to be a reason however for their deficiency, and that’s where the stats background is helpful. I’m not going to baselessly dismiss polls because I dislike the results, I need reasons. And the best explanation I can think of is the simple random sample. The people they’re polling don’t constitute a true SRS, because I wager there’s some conflating factor skewing results. Landlines are going to favor older populations. Texting people to take polls is going to have the bias of who actually goes on to take them.
The biggest enemy to statistics is that there’s no law forcing people to answer questions if they’re chosen for a poll.
Sure, but look at the polling method. They used landlines and cellular.