Aggregators take note: We found that many polls in these contests are sponsored by or affiliated with GOP campaigns, increasing potential bias of unadjusted polling averages
Thank you for the explanation. It would helpful to include this kind of information in all polling reports, since, too often, it's hard to tell when one is actually seeing a statistically significant difference.
Fascinating analysis, but what's is the sample size and margin of error? Even a ten point spread can be an illusion if the margin of error is +/- 5 points. I'm sure the polls showing a 3 point advantage for the Dem in VA are within the margin. I think one of the reason people are suspicious of polls is that they mistake specific numbers for ranges. Thanks.
The average of polls has a lower margin of error than the individual polls. The MOE for the polling average, in terms of standard sampling error, does not extend to zero; the Democrats are definitely in the lead _among these polls_.
If you’re asking about the predictive MOE (from stuff like uniform bias and a bad likely voter filter), yes, there’s a chance the polls are wrong and Dems lose.
Thank you for the explanation. It would helpful to include this kind of information in all polling reports, since, too often, it's hard to tell when one is actually seeing a statistically significant difference.
Now that we're starting to get into election season, would you spend a few paragraphs to explain "weighting" in polls?
Fascinating analysis, but what's is the sample size and margin of error? Even a ten point spread can be an illusion if the margin of error is +/- 5 points. I'm sure the polls showing a 3 point advantage for the Dem in VA are within the margin. I think one of the reason people are suspicious of polls is that they mistake specific numbers for ranges. Thanks.
The average of polls has a lower margin of error than the individual polls. The MOE for the polling average, in terms of standard sampling error, does not extend to zero; the Democrats are definitely in the lead _among these polls_.
If you’re asking about the predictive MOE (from stuff like uniform bias and a bad likely voter filter), yes, there’s a chance the polls are wrong and Dems lose.
So G you have a prediction on how wide the margins of victory will be if the favored Dem candidates win their governorships?
Conditional on them winning, I think the win margins would be in the high single digits or low double digits