🇳🇱
Wow Trumps approval rating is suprisingly high, even higher than Obama's was around this time in his first term....
Sadly looks like Trump is going to get reelected. Hoping Pete wins though so I can vote against trump. Only thing that would cause me to not vote democrat is if one of the evil three win (Biden, Warren, Bernie)
@freemo this isn't the first time you've claimed Trump has a higher approval rating than Obama did. I wonder if you are looking at a certain pollster who makes this claim, because the aggregate/average doesn't really support it:
@khird I mostly stick to the well respected polls with good controls. Obviously just massively importing every poll known to man is not a good way to get accurate data.
Ironically though even on the link you gave it shows one point where trump bear out obama's approval rating briefly just before the impeachment (which was also the last time I noticed that Trump's approval rating was higher than obama's). So not really a good counter example on two accounts.
Also notice how the list of polls on that site are largely inclusive of non-professional polls taken by agencies like news media which we all know have extreme bias and certainly dont conduct very scientifically accurate polls. However if you look at the polls from agencies devoted to scientifically accurate or professional polling without a political bias you see a very different trend.
Data science, particularly in polls, means you have to know your source. You cant just pool together every known bias poll and expect the results to not be bias.
@freemo right, 538 weight pollsters by quality & adjust for house effects. They do a much more methodologically-sound job than just blindly "importing every poll known to man."
Trump briefly cleared his predecessor's rating once, but he's 4-7 points back now. So, given that you made the claim today, I wondered if you're maybe looking at a single pollster - Rasmussen, for example - with a big Republican house effect.
Edit: I see you mean Gallup. Still gonna go with the aggregate myself, to increase effective sample size, but I appreciate you providing the extra detail.
@khird They certainly made an attempt to weight the polls. But a **very** poor attempt at that. They rated polls from agencies with **very** poor track records of accuracy as being A+ polls for example.. Sorry but everyting smells of bad practice here.
You cant just verage together polls that are 90% from bias non-professional sources like a news agency with polls from professional sources with dubious weighting, no peer-review, and no objective process and expect that to be meaningful.