New from me: When the first COVID-19 vaccine was authorized last December, a lot of Americans were on the fence about getting it. Since then, there's been a dramatic shift toward vaccine acceptance -- but the remaining holdouts could be more stubborn.
fivethirtyeight.com/features/more-…
fivethirtyeight.com/features/more-…
Plus, some lessons from vaccine surveys that apply to polling in general:
-Snapshots, not predictions
-People are bad at predicting future behavior
-Question framing matters
-Look at the undecideds!
fivethirtyeight.com/features/more-…
-Snapshots, not predictions
-People are bad at predicting future behavior
-Question framing matters
-Look at the undecideds!
fivethirtyeight.com/features/more-…
This is yet another polling story where the presence/absence of some sort of "not sure" option is crucial. Polls that didn't have an explicit "not sure" option showed higher initial support for vaccination than ones that did, and less substantial movement over the next few months
Less a case of a "right" or "wrong" format than one where different versions of the data tell a fuller story when looked at in aggregate.
Polls providing options beyond "yes" and "no" captured the public hesitation and potential for movement, while those that pushed respondents provided an early indication of the direction those people might ultimately break toward.
• • •
Missing some Tweet in this thread? You can try to
force a refresh
