4 ms·
"most of it comes from personal experience." Sample size = 1? Almost sounds like younger you was better off Personal experience is a pretty terrible way to fo
by Lewton 10y ago
"most of it comes from personal experience."
Sample size = 1? Almost sounds like younger you was better off
Personal experience is a pretty terrible way to form political opinions, it leads to stuff like overestimating the rate of crime, the cost of terror etc.
- Klockan 10y ago> Personal experience is a pretty terrible way to form political opinions, it leads to stuff like overestimating the rate of crime, the cost of terror etc. Isn't it the other way around? Few have significant personal experiences with crime or terrorism so it is only when they read about it that they start thinking that it is a big issues.
- Lewton 10y agoI'd classify reading something in a newspaper under "personal experience", the news are mostly written in a way to invoke emotional responses. In contrast to, say, looking at the actual data and crunch the numbers yourself to try and get an educated view This is obviously not very practical but at least then you could hold your opinions with good conscience
- Fnoord 10y ago> Personal experience is a pretty terrible way to form political opinions ..and the very belief that personal experience is worth more than theory and data. In normal human beings personal experience is related to the amygdala which is related to emotions. We study a great deal in our youth. Sure, we don't all go to uni or college. Do we stop learning after that? Not completely, but we peak at age 35. After that we still get more experienced. Smarter? Generally, not so much. The senior adults and elder are voting based on their experience, and this clouds their judgement. While it is partly by choice (look at how intelligent Noam Chomsky for example still is), it is also inevitable. What we see now is what I suspect the onset of the babyboom generation. Once they've passed away we will see more balance between the two camps.
- hliyan 10y agoPerhaps a single verified data point is better than a series of hypotheses? Because most, but not all books about politics, are based on hypotheses and not accurate data. Also, if everyone voted personal opinion, shouldn't the voting process average everything out, rather than the kind of wild media-driven opinion surges/swings we observe today? (Not being rhetorical, I'm actually asking if that's a possibility)