- cross-posted to:
- [email protected]
- cross-posted to:
- [email protected]
You’re just now noticing?
It’s not only weaponized - it’s also that polling methodologies are still largely focused on random cold-calls from unknown numbers, which basically nobody under 40 picks up for because we don’t have land lines, and have moreover been inundated by spam calls and texts CONSTANTLY for quite a while now. Any polls run under that methodology will absolutely have a very strong demographic skew towards older people as a result. Thus, the data becomes increasingly inaccurate as you restrict to respondents younger than about the Gen X midpoint.
Yeah pretty much. Though the skew towards older people isn’t exactly the problem - they track the ages of respondents and weight the results accordingly. The issue is instead that with such a small and unusual proportion of young people answering their phones, the data on that cohort is unreliable.
Older people are more likely to vote though. Getting more polling data from 20 somethings would likely skew your results.
What you want is “likely voter polling” as those who vote are the only ones who affect the outcome you’re trying to predict.
Run any poll you see past this: https://projects.fivethirtyeight.com/pollster-ratings/
Polls are like “science journals.” You only want to pay attention to the reputable ones.