Quote:
Originally Posted by Aaron W.
This is true. But that doesn't mean much in terms of criticizing a particular theory.
Of course it does. If something is a priori likely to be wrong, we should be treating it extremely skeptically. I love how you have a problem with me (without question on the right side of this debate), but not with the people who uncritically present this as meaningful/probably true.
Quote:
As "gospel"? There are only a few posts in this thread. Where has anyone treated it as such? No wonder you sound like you've got no clue what anyone is saying. It's because you've got no clue! (Surprise!)
It's been uncritically repeated all over the mass media.
Quote:
You've just given yourself license to slap a gigantic "NUH-UH" on every piece of literature (published or otherwise). That's a highly controversial move.
That's precisely what science is. Peer review, replication, etc, are core elements of research. None are present here. I'm very entitled to slap "this is probably total bull****" on any single study, especially conclusions that use imprecise language data snooped with poor significance.
Quote:
Basically, you don't actually understand social science research. You've got in your head that "authoritarian" is some concept that is so narrowly defined as to require a significant amount of categorization to identify it. But that's not the reality of the structure of many of these surveys.
Is it going to be "perfect" in categorizing people? Nope. But four binary questions that address the same topic is enough for a binary categorization. There will probably be noise, as there always is. But if this is the heart of your criticism, then all I do is shrug at your ignorance.
The heart of my criticizing is data snooping, poor significance, unreplicated, and yes, the fuzziness of this particular concept, compounded by the fact that he only asked four correlated questions relating to child rearing. The study is meaningless.
Quote:
Or I could be both bloviating and scientifically minded, understanding that my concept of scientifically minded requires critical thinking. I pointed to that other thread for a reason. That reason is that it exemplified the level of think that you are capable of. And it's not very high. You've re-affirmed my belief by your participation in this thread.
You're good at being a bore, not so good at criticizing.
Quote:
Given an initial position of ambivalence, the chance it's correct is 50%. What information do you have to lower that from 50% when the only available information appears to push it above 50%?
Most things do not have causal connections, so 50% is just lol. Of everything he data snooped, only authoritarism was signficant. According to Aarow W.'s genius theory, it was 50% likely that sex was correlated with Trump support, 50% likely that age was correlated with Trump support, 50% likely that race was correlated with Trump support, 50% likely that education was correlated with Trump support, etc.
Your setting of baseline probability at 50% is really silly.
As for this study, it's WAY lower than 50%. It's probably below 10%. "Correct 50% of the time" is the baseline for large scale, blinded, peer reviewed, non data snooped, published, replicated multiple times, hard data (i.e. medical measurements) conclusions. To get to 75% you need to get to meta review level. This particular claim is:
- Written by a PhD student
- Soft data
- Unreplicated
- Poorly captured (four questions on child rearing)
- Data snooped, heavily
- Appears unpublished and not peer reviewed
- Not blinded
The odds that it's true are well below 10%.
Last edited by ToothSayer; 02-20-2016 at 04:04 AM.