Dunning-Kruger and other bogus memes | Hacker News Reader

Dunning-Kruger and other bogus memes

danluu.com

52 pointsIllotus22 comments

Opinion: a log like decrease in happiness vs income is basically consistent with "more money doesn't make you happier". Sketching a graph that is completely flat is misleading however. The only question is, at what point on the log scale is the increase small enough to disregard (compared to other life factors).

Another one from the article: I'm arguing without empirical evidence of the superiority of good type systems (Note: I try to stay away from claims about what those are). I also think it's not possible to make controlled studies on this area - which is why it's such a contested issue. Studies use either students and/or toy problems, or irrelevant datasets such as bugs reported against github repos etc. The only way top make a valid study is to have N teams implement and maintain a product with the same spec over K years and then compare the results and try to factor out developer skill. I say it's not a feasible problem to solve. We'll have to stick to arguing type systems without evidence to back up the argumebts. But that is the best kind of arguing, after all.

> Opinion: a log like decrease in happiness vs income is basically consistent with "more money doesn't make you happier".

I think you mean log-like increase.

Assuming so, then more money does make you happier (although at a decreasing rate per dollar per year.)

The GP is saying it may be negligible at a certain point, compared to any other conceivable factor.

Sorry yes of course. Log increase. Was thinking the first derivative (rate the happness increases with wealth is negative).

So I write this as an academic who actually has read and cited Drs. Dunning and Kruger's (in)famous work. This is a little off the cuff so it isn't going to have the citations it should.

The generalized thing that the authors are talking about is information literacy. It is the process that includes not just understanding the information but identifying the need for information, locating it, understanding the information, evaluating the information, and then applying the information to some affect. Interestingly, it has found a home primarily within the libraries because...well they are really solid place to go for how information is organized.

“Making search easier for students can therefore be a double-edged sword: while it enables students to get to information faster and easier, it can also reinforce unreflective research habits that contribute little to the overall synthesis of a research paper or academic argument.” [0]

More broadly from research I have been involved in on student information literacy, self report data is spectacularly garbage because of underlying misconceptions students' hold about what they did. One semi-famous observational study showed that students strongly conflate finding a piece of information with understanding it. That was a 'whoa' moment for me. Cognitive conflation of access to a piece of information and deep coherent understanding of it. Houston...well you know the rest.

I personally attribute that in large part to fundamental problems in how we as a society think about education. This is philisophical about how we fail to differentiate transmission of information from the development of insights and understanding. We teach information as if it is both of those things. Science, as much as people scream otherwise isn't facts. Facts are the result of science. There is a lot of other work showing it is really hard to get faculty to change teaching practices. The reality is it doesn't matter. Society, not just teachers, need to think about information different and think about knowledge different ly for us to break out of this loop.

[A]Dimmock, N. (2013). Hallmarks of a good paper. In N.F. Foster (Ed.), Studying students: A second look (7-17). Chicago: ACRL.

Looks like the difference between perceived and actual scores fit the popular understanding.

The people who perform the worst have the most inflated perception of their performance, while the people who perform the best underestimate their performance.

I think people take Dunning-Kruger and apply it to their everyday anecdotes. They are thinking in terms of relative differences rather than absolute, and thus the pop-sci understanding may not be that far off from the original.

Seems paradoxical that the Dunning-Kruger effect is being applied to the Dunning-Kruger effect.

Yeah, it's almost impressive how all 3 of their conclusions are so off the mark it descends into parody.

Not paradoxical. Just ironic.

This was my first thought.

I'm sure the noble prize awarded to the Dunning and Kruger's work in psychology was just a result of a bunch of imbeciles who didn't realize it was a "bogus meme".

This article was absolute drivel.

I just did some quick googling. It seems they received the Ig-Nobel award, not an actual Nobel prize. This award is a bit strange, I don't fully understand... It's some kind of parody:

https://en.m.wikipedia.org/wiki/Ig_Nobel_Prize

It's simply there to highlight the wide and wooly range of things being practiced under the umbrella of science. I find it pretty heartening - the world needs research on slug poop and the psychology of sexbots, too.

In regards to software engineering, The Leprechauns of Software Engineering has lots of good examples of this type of thinking.

https://leanpub.com/leprechauns

The most interesting thing about the graphs isn't the slope, but that "perceived score" is far more tightly coupled to the individual's perception of their overall ability than their actual score. But that could be an artefact of how and when questions about ability were asked (conjecture: I haven't read the original paper)

"Perception of their overall ability" seems to have a different scale that the 2 others, so the important point is not the actual values but the correlation measure.

The money/happyness one is particularly interesting to me. If money does not equal happyness why am I so happy when I get more? I make more than 75k so obviously 75k isn't enough.

It highly depends on where you live. If you live in the Bay area, 75k will not make you as happy as 60k in the mid west because rent is so high.

That's always seemed a particularly silly one to me that doesn't in the aggregate comport to any experience of the real world. There's presumably a declining utility curve at some point but that's not the same as saying there isn't a correlation. Money does help make a lot of issues/problems go away.

The fact that you want more is irrelevant. You may be unhappy and believe that more money will make you happier, but it could be just a false belief or a bias.

But the question presumes that somehow you are unhappy because you don't have enough money. What if you are already happy but more money makes you more happy?

Money is not a cure for unhappiness seems like a more accurate phrase.

The point was that more money gives diminishing returns to happiness.

Would you really be twice as happy to win a $2 billion lottery jackpot as you would a $1 billion jackpot? Certainly, you'd be bouncing off the walls from the thrill either way, but I doubt you'd be twice as happy.

When it comes to money, there are diminishing returns on the happiness you get from what the money lets you do. At the low income levels, getting a little more money lets you not live paycheck-to-paycheck, losing sleep every night because you're not sure if you're going to be able to pay rent. At the extreme levels, it merely determines if you buy a 200 ft yacht or a 30 foot yacht.

I think part of the problem is that it's really hard to create a metric that defines happiness.

Oh man, it’s the HN link I was born to read.

The way people throw around Dunning-Kruger in particular is so aggravating. You can pretty much always tell they never got anywhere even close to reading the research.

Reply on news.ycombinator.com