Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How not to be a crank: rules for not being a science-dick

medium.com

Post reply on HN
Print view View original

61–70 of 103 posts

Re: How not to be a crank: rules for not being a science-dick

#61

Post by cwyers »

I don't know how much I agree with the premise of this article. People aren't resistant to retracting faulty research because the people who point out faulty research sound like cranks, they do it because their incentives are to maintain the work they've done. If you look at someone like Brian Wansink[1], what started off as unearthing methodological errors seems to have found actual fraud, although I'm sure Wansink wouldn't view it that way. And this massive peer review process not only didn't discover it, it's been an impediment to correcting it. At stake are millions of dollars in federal funding and more than a few careers. Do you think anyone involved is primarily concerned with tone? And I'm not sure how effective anyone can be at combating it by working within a crooked and broken system, where extracting rents from academic institutions for publishing their own work is the main incentive for publishers, and academics are incentivized to go along because their careers depend on numbers of publications and citations. You may have a high success rate at combating individual errors, but you aren't making any headway on the bigger issues like publication bias and the overall trend away from long, important work and towards shorted studies designed to provide an eye-catching conclusion so that you can get published and cited. It's like bailing out a boat with a drinking cup. You're still sinking.

1) https://slate.com/technology/2018/02/how-brian-wansink-forgo...

Re: How not to be a crank: rules for not being a science-dick

#62

Post by soVeryTired »

Nice article. For the last few years, I've had concerns that parts of economics have serious problems. But it's good to get a reminder of how to criticise well: points one and two (calm down, and don't be a dick) are particularly relevant to me.

For me, the stakes feel high enough that it's easy to get worked up and spit vitriol. But I've found that I'll lose the interest of anyone I'm talking to when I do.

Re: How not to be a crank: rules for not being a science-dick

#63

Post by psyc »

watwut wrote: post #57
psyc wrote: post #52

Earlier quoted context omitted.

I agree with you! But you've cherry-picked an issue that the majority of HN users agree with, so... In this case, I would say that the nuance lies not in whether it is good or bad, but in understanding the constellation of factors that would lead someone to hold the objectionable viewpoint. Reducing the issue down to the question of whether it's good or bad is itself the simplistic mentality that Fry was very explici…

I cherry picked issue whole thread started with. That is not cherry picking, that is staying on topic. There are plenty of reasons why one would be nazi. One is fear, other ressentment yet other that you was raised that way by parents. And yet others grown to be cruel, aggresive and sadistic and nazism makes them feel like it is good thing. However, being clear on whether it is good or bad or even dangerous and what…

Here, you are exploring nuance. I doubt it will cost anyone their life.

> However, being clear on whether it is good or bad or even dangerous and what is cost of they get their way is practically speaking super important.

Yes, it's important. It's important that children learn this, or sort it out by very early adulthood. That is what is meant by infantile. In my opinion, this is not a serious problem for the vast majority of civil folk. I do not understand the value in grown adults belaboring it. It's baby food. Eat grownup food.

Re: How not to be a crank: rules for not being a science-dick

#64

Post by captaincrowbar »

JasonFruit wrote: post #22

A certain amount of this boils down to, "Don't be mentally ill or have an otherwise disordered mind." I've known people who were the sort of crank he's talking about, and they were generally — though not always — unhealthily isolated people who had difficulties managing their personal lives. Most had a tenuous grasp on reality, especially the motivations of other people. There are cranks who could respond to these ru…

Well, I don't think the article was really aimed at actual cranks (who are, more or less by definition, immune to this kind of advice). It feels more aimed at scientists, or people of a scientific turn of mind, who may have found a genuine flaw in some widely accepted result but might not have the communication skills to get it across to the scientific community.

Re: How not to be a crank: rules for not being a science-dick

#65

Post by appleshore »

dghf wrote: post #24
appleshore wrote: post #21

Earlier quoted context omitted.

I disagree with the first statement. To me, it invokes concepts of relativism. Certainty in actions, when well prepared, has a magical and transformative effect on things. I am willing to entertain thought experiments about how I can be wrong but I know I am not about certain ideas and philosophies.

> I know I am not about certain ideas and philosophies Can you give some examples?

About secrets of the Universe and primordial first principles of life.

Re: How not to be a crank: rules for not being a science-dick

#66

Post by Thrymr »

ninkendo wrote: post #25
mitchty wrote: post #9

Earlier quoted context omitted.

My viewpoint on number one is simple: I'm always wrong Easier to move forward with that default assumption and try to prove I'm not as wrong as I thought and that "right" is the wrong attitude to hold.

How do you improve your knowledge if every path forward is still wrong? (In ML terms, how do you apply gradient descent if every path forward is up?)

You start with a prior, but you assume that it is not the optimal solution.

Re: How not to be a crank: rules for not being a science-dick

#67

Post by TheOtherHobbes »

zaphar wrote: post #46
Karunamon wrote: post #18

Earlier quoted context omitted.

I’m not sure that’s accurate. People hold all manner of irrational political beliefs and there is value in recognizing that irrationality.

Do you know the reasons for their beliefs? Can you articulate them? If not then you are conflating your own beliefs with the standard for rationality. You will also be really ineffective at convincing them they are wrong. Not because they aren't smart enough but because you are unable to communicate effectively to them.

What makes you think people understand the reasons for their own beliefs? How do you know the reasons they give you when asked are the correct ones? Why do you believe that challenging those statements respectfully will change their minds?

The actual requirement for persuasion is a believable narrative with a bit of moral weight, and a random selection of standard persuasive tricks that politicians and lawyers have been deploying with outrageous success for millennia.

Generally, being objectively correct is one of the least effective ways to change anyone's mind. You might have some hope in a scientific setting. But in popular debate, facts and rationality are almost completely powerless.

Re: How not to be a crank: rules for not being a science-dick

#68

Post by ScottBurson »

shaki-dora wrote: post #50
AlexCoventry wrote: post #34

Earlier quoted context omitted.

I don't understand why this is being downvoted. It seems to be the most insightful, empirically grounded, high-effort comment in the thread.

When an article on how not be seen as a crank makes you want to post a long comment about your research into Russian bots, you should probably think again.

This is exactly wrong. It's not having an opinion on a controversial topic that makes one a crank, it's whether that opinion is well supported and respectfully expressed.

In fact, mike_hearn's first link is to a blog post where he debunks claims of evidence that Russian bots influenced the Brexit vote. (He doesn't offer evidence that that didn't happen; what kind of evidence could exist? But he does argue that the alleged evidence offered by others that there was such influence is unconvincing.)

Re: How not to be a crank: rules for not being a science-dick

#69

Post by TheOtherHobbes »

TheAdamAndChe wrote: post #29
Karunamon wrote: post #18

Earlier quoted context omitted.

I’m not sure that’s accurate. People hold all manner of irrational political beliefs and there is value in recognizing that irrationality.

Yes, but there is also the real possibility that your beliefs are irrational as well. Everyone thinks they're right, and everyone has a train of thought that has led to that belief.

A key test for rationality is prediction.

If you believe that banning sex education in schools will decrease teen pregnancy, and it turns out that teen pregnancy rates actually increase when teens aren't educated, then your belief about sex education isn't rational.

Many - possibly most - public policy decisions are open to this kind of testing.

The fact that formal testing rarely happens, and that it's consistently ignored even when it does, suggests that politics itself is inherently pre-rational.

Arguably it's then irrational to expect it to be rational - but that's a deeper problem that doesn't just apply to politics.

Post reply on HN