That's what I don't understand. The polls say that Republicans are set to win, but all I hear are Democrats are motivated due to Roe vs Wade and WV v EPA. I did read an article by Nate Silver that said they don't look at current polling to predict elections they use historical election data. But doesn't current polling data effect the election?
That's what I don't understand. The polls say that Republicans are set to win, but all I hear are Democrats are motivated due to Roe vs Wade and WV v EPA. I did read an article by Nate Silver that said they don't look at current polling to predict elections they use historical election data. But doesn't current polling data effect the election?
Current polling is not particularly informative WRT election results. The election is in November, which means that we are 4 months out. Polling will become more indicative as we get closer to the actual election.
A lot of things can happen between now and November. Small things can have large effects. Any polling more than 30 days outside of the election is kind of pointless when it comes to predicting results.
I am less skeptical about polling than a lot of people because I have actually worked with it in my Grad student says (creating, administering, and analyzing). It is, in general, fairly accurate (even when people do not think it was accurate), at least at the national level.
I haven't read anything current by Silver, but my memory says that it is one of those things that changes over time.
They do a lot of analysis on correlation, and may have found that June polling has much less to do with November results than other underlying factors. That'll change when we're talking about October polling, but even then, their model takes a lot of factors into account and assigns them weighting that changes over time.
That's what I don't understand. The polls say that Republicans are set to win, but all I hear are Democrats are motivated due to Roe vs Wade and WV v EPA. I did read an article by Nate Silver that said they don't look at current polling to predict elections they use historical election data. But doesn't current polling data effect the election?
Current polling is not particularly informative WRT election results. The election is in November, which means that we are 4 months out. Polling will become more indicative as we get closer to the actual election.
A lot of things can happen between now and November. Small things can have large effects. Any polling more than 30 days outside of the election is kind of pointless when it comes to predicting results.
I am less skeptical about polling than a lot of people because I have actually worked with it in my Grad student says (creating, administering, and analyzing). It is, in general, fairly accurate (even when people do not think it was accurate), at least at the national level.
I haven't read anything current by Silver, but my memory says that it is one of those things that changes over time.
They do a lot of analysis on correlation, and may have found that June polling has much less to do with November results than other underlying factors. That'll change when we're talking about October polling, but even then, their model takes a lot of factors into account and assigns them weighting that changes over time.