3 ms·

Trend has been dramatic though: https://news.gallup.com/poll/1597/confidence-institutions.aspx https://news.gallup.com/poll/1597/confidence-institutions.as...

Trend has been dramatic though: https://news.gallup.com/poll/1597/confidence-institutions.aspx https://news.gallup.com/poll/1597/confidence-institutions.as... Great deal/Quite a lot Confidence in Newspapers went from about 35-40% in the 80s to about 22% today (COVID likely a major influence on 2022's 16% number). Very Little number also notable.

0xbadcafebee 4y ago
Before Trump was elected, the polls predicted Hillary would win. Before Biden was elected, the polls predicted Trump would win. Polls can be manipulated, misconstrued, and simply poorly executed. Leading questions, doublebarreled questions, double negatives, question order, margin of error, misread data, "random" sampling, focus groups saying what you want to hear, flukes like a surge due to novelty, and more, all skew polls in ways that don't reflect reality. Polls are more shamanism than science. So why did the numbers change? The media has been telling us we shouldn't trust them for years, and people are reporting back what they've been told. It's also possible that when people say they "don't trust the news", what they really mean is they don't trust what's being reported - as in, if they're reading about politics, they don't trust the politics. You might suspect the newspaper is lying, or you might suspect Trump is lying - but either way, it's "the news" you don't trust, so you start to doubt the institution that's bringing you the lies.
- timbit42 4y ago
  The problem with polls is they only show, assuming they are accurate, what percentage of people will vote a certain way. They don't indicate who will win. Due to the electoral college, many votes are effectively thrown out. This is the reason for the discrepancy. It would increase democracy in the US if the electoral college was eliminated. FiveThirtyEight.com uses polling data, but also looks at the electoral college, and accurately predicted that Trump could beat Hillary and that Biden could beat Trump, although the odds that Trump would beat Hillary were low, they did predict it was possible.

0xbadcafebee 4y ago

Before Trump was elected, the polls predicted Hillary would win. Before Biden was elected, the polls predicted Trump would win. Polls can be manipulated, misconstrued, and simply poorly executed. Leading questions, doublebarreled questions, double negatives, question order, margin of error, misread data, "random" sampling, focus groups saying what you want to hear, flukes like a surge due to novelty, and more, all skew polls in ways that don't reflect reality. Polls are more shamanism than science. So why did the numbers change? The media has been telling us we shouldn't trust them for years, and people are reporting back what they've been told. It's also possible that when people say they "don't trust the news", what they really mean is they don't trust what's being reported - as in, if they're reading about politics, they don't trust the politics. You might suspect the newspaper is lying, or you might suspect Trump is lying - but either way, it's "the news" you don't trust, so you start to doubt the institution that's bringing you the lies.

timbit42 4y ago
The problem with polls is they only show, assuming they are accurate, what percentage of people will vote a certain way. They don't indicate who will win. Due to the electoral college, many votes are effectively thrown out. This is the reason for the discrepancy. It would increase democracy in the US if the electoral college was eliminated. FiveThirtyEight.com uses polling data, but also looks at the electoral college, and accurately predicted that Trump could beat Hillary and that Biden could beat Trump, although the odds that Trump would beat Hillary were low, they did predict it was possible.