Starting a Company Can Turn You into a Psychopath | Hacker News Reader

Starting a Company Can Turn You into a Psychopath

blog.statwing.com

32 pointsglaugh21 comments

The title "Starting a Company Can Turn You into a Psychopath" is complete nonsense when paired with the linked article. Correlation is not causation.

With that being said, I've seen founders become lots of things both good and bad. The experience of doing something as challenging as creating a new business is certainly the type of endeavor which will likely forever alter your view of the reality in which we reside.

those two curves look the same to me. "Average" is a terrible metric, especially with the spread. My eye estimates the "p-value" for difference between those curves is about .8, depending on what distribution you think it is.

Well, p-value is <0.00001, but that's because it's a pretty big sample. You're right that the effect size is "small" (Cohen's d), which as you suggest you can basically see from the curves.

"Advanced" tab here: https://www.statwing.com/open/datasets/88ed02bd6dda2f813ba98...

Edit: I think I'm not allowed to respond to the reply to this comment for flamewar prevention reasons, so putting it here. You can also look at the nonparametric version of the test if you don't buy that parametric is good here (though I'd disagree because of the central limit theorem). p-value there is 0.00004

I don't believe that p value. the non-founder curve looks fine, but the founder curve has bins with some pretty serious deviations from expected, so you are overestimating your confidence in your curve fit.

The curves seem to look a bit deceiving, probably something in how our eyes work - switch to the 'count' tab instead of the 'percent' tab. It's much easier to see then just how much 'flatter' the founder curve is compared to the non-founder curve, and this flatness pushes up the deviation, but it pushes up the average far more and puts the confidence interval well out of range of the non-founder curve.

Judging based off just the visual curve is not a great idea as our brains are wired up for matching patterns and not for working out these kind of differences properly. It's why math is such a good idea when making decisions.

you're kidding right? Because I can make the non-founder curve just as "flat" by yanking up the scale on the percentage graph to 10,000%. The reason why it looks "flat" is because when you go to "count" the website scales both curves the same y-axis. I think you mean to say something like "kurtosis" - which is, if anything, obscured to the eye by the process of flattening by scaling.

And math is not a good idea when you make assumptions like normality of curves which are absolutely not normal. In this case, using the t-test to calculate a p value.

I think reification of poorly done math is a bigger problem than math by eye.

Agreed. Sort of a satire of how articles will take some research and draw wild conclusions about it for the headline.

We also considered "Psychopaths start companies" and "Some third variable relates to both psychopathy and starting companies", all of which are equally valid hypotheses :)

You missed the [the title of this post is tongue in cheek] part.

Humor is impermissible in HN titles!!1!

Of course, you could also assume that having an altered view of reality leads one to believe it's reasonable to strike out on their own (and in some cases start their own business). Again, doesn't imply causation, but it's possible way to flip the relationship on its head.

Posting on Hacker News Can Turn You Humorless

I think you are mistaking psychopathy with how obsession and ambition can lead people morally astray.

One can't be fixed, the other sometimes can be.

Given that sociopathy/psychopathy is an actual pathology and has some organic basis, I'm not really sure the title is accurate. I don't think anyone has indicated that sociopathy is developed during your adulthood.

Getting real for a second, all jokes aside, sociopaths are a real and severe threat to society. You should shun and avoid sociopaths at all costs. You might be tempted into thinking "but I need my CEO to be...". The answer is NO. Such a CEO will do nothing, add no value, and ultimately steal all the money of the company, and not feel the least bit guilty or remorseful. Because sociopaths don't have those feelings.

So unite brethren against the non-human sociopaths among us!

PHP programmers are psychopaths. Who is surprised? :-)

It doesn't seem clear that these results are even statistically significant. As there are no calculated errors associated with the final values (which there should be for any study involving numbers) and the first sample is about 1/10th the size of the second. It doesn't seem like you can even draw the conclusion that there are more psychopaths proportionally within the founder group, let alone establish a causal relationship. TL;DR Study does not really say anything statistically meaningful.

This is exactly what I logged in to write. For a website called statwing.com there is a lack of statistics in this article.

At the top is the disclaimer: "this is not a scientific study". Then they present their data and bring up some new questions based on their findings and link to their stats tool.

The problem is: their are no conclusions to be drawn from this, no new questions that arise, no discussion we can have- its the back of a cereal box.

If they do a little more work and show that they have found something meaningful- then we'll talk.

Our bad. Should have put some p-values in the text itself. All those differences are statistically significant, typically at <0.001 levels.

All the statistical output, including confidence intervals, is available via clicking through the links to the dataset itself, available here: https://www.statwing.com/open/datasets/88ed02bd6dda2f813ba98...

Edit: We added a couple notes about this in the blog post, thanks for the feedback.

Most of this sounds absolutely useless. I mean, we could spend all day coming up with statistical hypotheses like:

"People who drink Folgers' coffee are more likely to be psychopaths"

"People who drive red cars are more likely to be psychopaths"

"People who eat out every 3rd Sunday of the month on leap years are more likely to be psychopaths"

...and we wouldn't be any smarter for it. Aren't there any more useful ways to show off your analytics platform?

Looks like the same thing can be said if you are a student...

So the question is: Did drug dealing turn Walter White into a psychopath, or was his psychopathology lurking and accelerated when he got lung cancer?

Reply on news.ycombinator.com