If Twitter had real and unique users, you could directly estimate candidate strength by the number of followers.
You could also look at likes, engagement, and so on.
This may be an interesting complement to traditional scientific polls, which endeavor to collect a random sample via phone calls.
But people don’t pick up the phone for strangers anymore, so perhaps it moves to social media.
This may be an interesting complement to traditional scientific polls, which endeavor to collect a random sample via phone calls.
But people don’t pick up the phone for strangers anymore, so perhaps it moves to social media.
Now, of course not everyone who follows an account supports that account. Just like a retweet is not an endorsement.
But if Twitter could roll out “real and verified” users at scale, it could add a separate field to indicate if you were a supporter and not just a follower.
But if Twitter could roll out “real and verified” users at scale, it could add a separate field to indicate if you were a supporter and not just a follower.
