Hacker News Reader

Comment by coding123 | Hacker News Reader

I ABSOLUTELY agree with this, and this is completely lost on people. For every study that says coffee is bad there's 10 more that say it's good, simply because people want it to be good, NOT because it's actually good. The problem is, we humans play right into that as if the number's game actually means something - 1 out of 10 studies says something is bad - the other 9 say it's good. So rather than actually work off the truth (the objective fact of whether coffee is good or bad) we go off the numbers instead. Now take coffee and replace that with any number of other things.

View on HN

Isn't this more a distinction between science and science reporting. Coffee is complex and interacts in ways that can be described as good and bad. The press reporting blows by the wind of the press release that day. Hence the Daily Mail and others reporting lots of common things both causing and protecting us against cancer.

>Isn't this more a distinction between science and science reporting

In a way yes, but if we include the summaries of findings that scientists deliver to the public/governments/etc as "science reporting", then science reporting is 100% of science as we know it (unless we happen to practice it).

Consider dietary advice that has since found to be BS, that was passed in the 70s and 80s as "scientific fact", not just by the press, but by scientific advisory boards (to governments etc) comprised by actual scientists.

Right but wasn't that advice basically the state of the field at the time? That we know different now doesn't mean the scientists then should have acted on information unavailable to them.

Right? Might as well be critical of public schools not teaching object oriented programming in the 1950's. And that's why my parents can't program, because schools were inferior...

Well, how about refraining from teaching BS state of the art?

Newton's theory wasn't proven wrong -- just inefficient at larger units / speeds, and was superseded by a more accurate theory that still gives the same results (more or less) for regular everyday observations.

Contrast with the kind of dietary advise of the 70s/80s which has been totally reversed. It wasn't "correct, but not entirely accurate", it was irrelevant, or worse, opposite, of what actually holds (according to what science says today).

Newton's theory was proven wrong: astronomers have observed movement patterns of celestial bodies which Newton's theory did not predict. But that doesn't change the fact that the first-order description of the planets' movements that Newton's theory produced is correct. And it doesn't change the fact that Newton's theory describes very accurately all types of motion across which we come in our daily life.

Facts always are true or false forever. If something can be considered true one day and false the next one, it never was a fact. It's just a theory.

Maybe this is why facts don't change our minds. Given exact evidence that it was wrong, which we all accept, and then turn around and say, oh yeah, we got it right this time. We're all trying to feed into our biases, not change our minds. I'm not trying to point out what you're beliefs in low-fat vs low-sugar diets have changed, I'm trying to point out the fact that you see evidence that science was wrong, and you still believe in it.

I'm not a luddite, I think everyone should believe in what is most likely true, I mean what else can we fucking do right? What I'm saying is don't turn "science facts" into a religion and believe it blindly. I think actual scientists do a good job at this, because they're actually testing it.

We fail when we convert science reports into articles and then into policy so quickly simply because it feeds into the biases of the people in charge at the time.

'your beliefs'

>Right but wasn't that advice basically the state of the field at the time?

Then perhaps it wasn't a field that should feel confident enough to give public policy/health suggestions?

>That we know different now doesn't mean the scientists then should have acted on information unavailable to them.

Most of the issue was because BS studies, referenced but not replicated papers, assumptions and the like. Not because they lacked some more precise instruments or such.

It's not like Einstein's theories, which complete Newtons, giving more accurate predictions at edge cases, while Newton's are still accurate enough for everyday calculations (e.g. ballistics).

It's like claiming X (e.g. on fat) and finding out that it's actually -X, or that X is irrelevant.

Sometimes the state of the field at the time was inherently biased. See "How the Sugar Industry Shifted Blame to Fat" [0] for example. Science is not always a progressive quest for the truth.

[0] https://www.nytimes.com/2016/09/13/well/eat/how-the-sugar-in...

Reality not just science is complicated.

One 'fact' I remember was men are more likely to physically harm their spouse, women are more likely to kill their spouse. [used as an example, may not be correct] If it's true that means domestic violence is biased in both directions. In other works Men are not more, less, or just as violent as women. Instead violence is a insufficiently complex model for the behaviors being discussed.

The hardest part of science is our desire to simplify past the point of accuracy. Consider, cities albino alter their local climate in more ways than just temperature. However, temperature is easy to model and complex interactions with various layers of the atmosphere and wind patterns is so guess what's ignored about this interaction. You don't teach middle school students about chimera when they study DNA, but it's a very real thing that actually happens.

This seems to extend in every direction. Do we want trade policy for today, next year, our children's future?

That, and bad science. Especially in medicine where studies with N=7 white college students are somehow considered acceptable.

Good and bad are opinions. 9/10 studies saying coffee causes earlier mortality (not saying this is true, just giving an example) is a fact. For me the enjoyment of coffee might outweigh loosing a year of my life, and wind up saving the government money on my pensions and healthcare in the long run. So is this good or bad?

This fails because things can't simply be broken down into a sets of "good" and "bad," and people really only do this because they want to justify things they already believe. Those who want to actually learn and understand things tend not to be satisfied with "coffee is good" or "coffee is bad." Neither of those represent the results of any kind of expertise.

If you're reading a study who's conclusion is 'x is bad' then I question whether you're actually reading any scientific literature to begin with. No scientist in their right mind would try and support an argument so broad.

It sounds like you're just reading some news article's terrible summary of an already terribly summarised abstract.

I agree with what you're saying here. I'm just trying to point out how lay people take in so called "science". In fact most people are just reading headlines and confirming their own bias, and then they tell all their friends that their new coffee drinking habit is confirmed as good by science. All I'm saying is that at the end of the day, the science that was paid for is distilled into sound bites and propaganda for someone's agenda.

(I actually love coffee and just finished my latte).

>For every study that says coffee is bad there's 10 more that say it's good, simply because people want it to be good

Or, maybe, it actually is good. Why do you assume some agenda on behalf of the 9, not the one?

There's an old joke along the lines of, if an old, white haired scientist tells you something is possible, he's probably right; if he tells you something isn't, he's probably wrong. My extension to that is, if 100 scientists tell you something, it may be right or wrong, but that's the way you should place your bets.

Reply on news.ycombinator.com