Taleb's grandma seems pretty down to earth.
The Scent of Bad Psychology
11–20 of 26 posts
Re: The Scent of Bad Psychology
#12Earlier quoted context omitted.
https://osf.io/8pc9x/ As per the large amount of links in the article, the author seems to have passed the test themselves, then talked about what heuristics they used to get their result (that matched the studies linked). Did you read the article? One can argue with their heuristics, of course, but dismissing them because they're open about their background is a classic middlebrow dismissal. This isn't presented as…
> Did you read the article? > classic middlebrow That is classic unwelcome behavior on HN. Maybe you can keep your judgements of other commenters and hot take reactions to yourself and stick to the subject. I didn't criticize them for being open about their background, but because they lack expertise to offer expert opinions. Again, if they wrote an article on rules of thumb for evaluating code or on reading radiolog…
Is there anything in the actual content of the article you disagree with?
Re: The Scent of Bad Psychology
#13It's been said before but needs to be said again: this is happening everywhere in the biomedical and related fields. The neurosciences, oncology... the list goes on and on. Anyone take a look at AI research lately? How much tweaking is going on there? How much is your big data finding due to the idiosyncracies of your particular dataset? Psychology, as has historically always been the case--meta-analysis itself bloom…
I’ve recently been at a master students’ thesis defense where a method was discussed that is supposed to speed up algorithms by deciding whether to run them on GPU or CPU. The method is to look at the program code with deep learning algorithms.
Fair enough. There’s apparently a set of benchmarks that other papers used and were reused here to compare against. Yet the student was unable to answer my question as to how bias in this dataset was treated or even what percentage of the algorithms was faster on GPU to begin with, they only showed how much better their deep learning fit the data. As in, if 99% is faster on GPU I could as well just provide that as a static answer and be better than any machine learning.
I frankly find it tragic that this kind of stuff is not ironed out in academia. This was was in a very highly ranked university.
Re: The Scent of Bad Psychology
#14The author offers rules of thumb for evaluating research, but is a business school grad (MBA? BBA?) and "a data science amateur, from the Latin amare – to love." Am I misunderstanding something? On what do they base these instincts and intuitions for evaluating research papers? If they wrote a blog post on rules of thumb for evaluating code, would you take it seriously? The "scent" here may be something on my shoe ..…
https://osf.io/8pc9x/ As per the large amount of links in the article, the author seems to have passed the test themselves, then talked about what heuristics they used to get their result (that matched the studies linked). Did you read the article? One can argue with their heuristics, of course, but dismissing them because they're open about their background is a classic middlebrow dismissal. This isn't presented as…
As per the article it seems that even rigorous science isn't even _that_ rigorous these days.
Re: The Scent of Bad Psychology
#15The author offers rules of thumb for evaluating research, but is a business school grad (MBA? BBA?) and "a data science amateur, from the Latin amare – to love." Am I misunderstanding something? On what do they base these instincts and intuitions for evaluating research papers? If they wrote a blog post on rules of thumb for evaluating code, would you take it seriously? The "scent" here may be something on my shoe ..…
>On what do they base these instincts and intuitions for evaluating research papers? If they wrote a blog post on rules of thumb for evaluating code, would you take it seriously?
Depends, since I tend to evaluate arguments on their merits, not on the author's credentials.
If one doesn't have the background to evaluate those arguments, I don't think they're supposed to be reading them anyway (whether they are from an "amateur" or a Nobel winning pro).
Re: The Scent of Bad Psychology
#16Earlier quoted context omitted.
https://osf.io/8pc9x/ As per the large amount of links in the article, the author seems to have passed the test themselves, then talked about what heuristics they used to get their result (that matched the studies linked). Did you read the article? One can argue with their heuristics, of course, but dismissing them because they're open about their background is a classic middlebrow dismissal. This isn't presented as…
> Did you read the article? > classic middlebrow That is classic unwelcome behavior on HN. Maybe you can keep your judgements of other commenters and hot take reactions to yourself and stick to the subject. I didn't criticize them for being open about their background, but because they lack expertise to offer expert opinions. Again, if they wrote an article on rules of thumb for evaluating code or on reading radiolog…
>That is classic unwelcome behavior on HN.
I'd say middlebrow dismissal is even more classic: https://news.ycombinator.com/item?id=4693920
>I didn't criticize them for being open about their background, but because they lack expertise to offer expert opinions.
You didn't criticize the author's opinions or any specific signs of insufficient expertise however. Just their credentials.
Credentials != expertise (one can be an expert by self-studying, in fact in CS many are).
Re: The Scent of Bad Psychology
#17Taleb's grandma seems pretty down to earth.
Taleb himself is pretty awesome.
Re: The Scent of Bad Psychology
#18Earlier quoted context omitted.
> Did you read the article? > classic middlebrow That is classic unwelcome behavior on HN. Maybe you can keep your judgements of other commenters and hot take reactions to yourself and stick to the subject. I didn't criticize them for being open about their background, but because they lack expertise to offer expert opinions. Again, if they wrote an article on rules of thumb for evaluating code or on reading radiolog…
> That is classic unwelcome behavior on HN. I'd say middlebrow dismissal is even more classic: https://news.ycombinator.com/item?id=4693920 > I didn't criticize them for being open about their background, but because they lack expertise to offer expert opinions. You didn't criticize the author's opinions or any specific signs of insufficient expertise however. Just their credentials. Credentials != expertise (one can…
What would be productive is discussing the expertise of the author, or if you don't think that's of interest, not discussing it at all. I notice you posted the original article; I'm not criticizing you for posting it but perhaps this is getting personal?
Insulting commenters with derogatory names is unwelcome and lowers the quality of discussion; I don't care if you have a link to it. Tangentially, it is a good example of how even the smallest behavior by a leader, such as pg, set a norm for many followers: people are digging up his posts 6 years later like they are scripture - wisdom, law and conduct to study. pg is just a person, not a prophet; YC and HN are wonderful, but his every act is not wisdom or law. (No insult to pg; the same could be said of Linus Torvalds, Steve Jobs, and everyone else who ever worked in SV or lived on Earth, including me.)
> Credentials != expertise (one can be an expert by self-studying, in fact in CS many are)
Credentials are a strong, widely used sign of expertise. The CS experts who lack educational credentials are practitioners, professionals in the field; they have other credentials. The author is neither educated in the field nor is a professional or practitioner in it. Everyone else weighs those things heavily. Do you hire or take the advice of people with neither education nor experience in the domain?
The article in particular requires expertise: It's a series of rules of thumb, things that are gained only through extensive experience doing the task described. And it requires doing it with success - not all experience is good experience - and apply good analysis to that experience. Rules of thumb for physics research from someone who washed out are not the same as those from Stephen Hawking; how about rules of thumb from someone who never has studied nor practiced physics, but read some links? Again, if they wrote 'rules of thumb for evaluating code' on the same basis, it would be ridiculous.
In general, everyone knows that the Internet is full of people spewing nonsense they know nothing about, but even smart people often seem to fail to apply that to particular instances, especially when other psychological drives apply - such as when what is said moves or inflames them. The downvotes and criticisms (this is an observation, not a complaint) seem to be because my comment disrupts that drive (in this case, the inflammatory narrative about social sciences). That's ok and what I expected. My prediction is that if I wrote the same about a different topic, then the response would be different: If what I wrote agreed with the crowd (e.g., if the author said psychological research was high quality and I questioned their expertise), then I'd expect many to agree with me, but for the same bad reason. Inflammatory stuff is, disappointingly, still hard to talk about rationally.
Re: The Scent of Bad Psychology
#19Is priming really bullshit? I watch batters go through all sorts of little rituals to get their heads in the game.
If you wash 5 times a day before prayers, yeah, washing your hands is going to put you in a mindset.
Or maybe I just don't understand what priming is.
Re: The Scent of Bad Psychology
#20The author offers rules of thumb for evaluating research, but is a business school grad (MBA? BBA?) and "a data science amateur, from the Latin amare – to love." Am I misunderstanding something? On what do they base these instincts and intuitions for evaluating research papers? If they wrote a blog post on rules of thumb for evaluating code, would you take it seriously? The "scent" here may be something on my shoe ..…
> On what do they base these instincts and intuitions for evaluating research papers? If they wrote a blog post on rules of thumb for evaluating code, would you take it seriously? Depends, since I tend to evaluate arguments on their merits, not on the author's credentials. If one doesn't have the background to evaluate those arguments, I don't think they're supposed to be reading them anyway (whether they are from an…
> I tend to evaluate arguments on their merits, not on the author's credentials
Of course, let's not make this binary, one or the other. Everyone uses both to varying degrees, and other signals too (if the post used bad English, for example, many would doubt the author's credibility).
In this case, I lean more heavily on expertise because the content is a product of expertise - rules of thumb. If they wrote 'I can create cold fusion in a teacup', I'd lean heavily on the merits - can they do it or not?
> If one doesn't have the background to evaluate those arguments, I don't think they're supposed to be reading them anyway
That greatly narrows everyone's reading material and awareness of the world. Somehow I have to decide what to do about many things in life, almost all of which I lack expertise in.
In the end, there is no way to avoid having to decide whether to trust the author or not - or more generally, to trust others or not, from consultants to bosses to employees to that wonderful person you might want to marry.