Comments by "nexus1g" (@nexus1g) on "Veritasium"
channel.
-
1
-
1
-
1
-
1
-
1
-
@cheesecakedelicious You did commit an argument ad hominem fallacy by stating, "I'm sure he might be smart and charismatic and personable, but... turns out those qualities don't actually jive that strongly with good research, which is kind of orthogonal if not negatively correlated to his personality," when these things have nothing to do with one another. It's also a strawman argument. You made the assumption because I said he was an awesome person was a supporting argument as to why he's a good researcher. He's a good researcher because he publishes research which adds a lot of quality to the body of science.
Skepticism is useless and is tantamount to just being a contrarian as it begins with a doubt to the truth when the truth is an unknown. Critical analysis is useful, and it doesn't begin with a presumption of untruth. When you begin with a presumption of untruth, then it's very easy to bias an analysis through that lens--especially when you're motivated to do so. When you begin with genuine curiosity, you presume neither fact nor fiction, which assists in establishing better and varying assumptions through which to view research.
You're arguing yourself into epistemological nihilism. You can read the paper. You can determine who has quoted the paper and why. Which you apparently have done the former, given your next paragraph. Why not present the ratio directly? Because data isn't presented to the public in that way. It's presented in a way that it must be thought about to certain levels to understand. For instance, a 2018 headline from CNN reads, "Walmart's CEO earns 1,188 times as much as the company's median worker". According to media industry research, some 60% of people will read no further than that. I'm using this headline because it stuck out as particularly egregious use of numbers. I checked the SEC filings for Walmart and looked a bit deeper into it to see if the CEO's salary impacted worker's wages. As it turns out, his entire $24M compensation package divided among the 2 million workers would amount to about 46 cents per two-week check. So what was the point of the headline--or even the existence of the article itself? The design of the experiment is very good in how it reflects the way numbers are abused in popular communication to convey a certain emotion to the reader versus seeking to be accurate, requiring astute readers to think deeper than being presented information as accurately as possible.
That is the world of information we all navigate on a daily basis, and so experiments to better understand how people perceive the data they receive are well-designed when they reflect that environment. And those binary outcomes are what people walk away with. Many will read the aforementioned CNN headline and walk away with the belief that Walmart's CEO is taking so much compensation that he's bankrupting his workers. Others will delve deeper into the assumption that the CEO's compensation hurts workers which is allowed to hang without resolution. The research was to determine whether political leaning impacted how problems were approached, and whether education (science comprehension theory) mattered or politics (identity-protective cognition thesis) mattered here. They controlled for the education with the politically neutral question regarding skin cream. Then they tested against that control with politically charged questions (e.g. guns). This is a very well-designed experiment.
1
-
1
-
@cheesecakedelicious 2. More people need to read their Descartes then. With sound reasoning, there's no need to be a nihilist. Just remember what George Box said: "All models are inaccurate, but some are useful." As long as we're able to use the body of science to a useful degree to at least some ends predict what will happen in our world, it's of consequence and useful.
3) "Did the authors actually think science-comprehension theory would win? I argue, no, they didn't, not seriously. They deliberately baked their expectation of a result into the design of the experiment. "
By what reasoning do you believe this? They controlled with an apolitical question and the results showed numeracy mattered, creating a divide between the well-educated and the less-educated. Then they asked a political question and didn't see that same divide. How is that baking any assumption into the design of the experiment?
I feel like you're assuming because they got a certain result, they then must have had bias in the design and therefore got the result they sought. Are you sure that you're not biased towards yourself, with a concern that this study might reflect you, causing you to attempt to undermine the study for identity protection?
People, by and large, do believe outrageous, easily-disprovable things, no matter their education level and given all the time in the world for them to consider, ponder, and verify. For instance, do you know how many people on the left (the statistically well-educated partisan alignment) still believe that US defense spending accounts for over half of all government spending because of a headline that caught fire many years ago?
1
-
1