4 ms·

> Overall this seems to be a rather obvious result. "Popular people are more likely to have their pull requests accepted" That is not the result at all; they a

> Overall this seems to be a rather obvious result. "Popular people are more likely to have their pull requests accepted" That is not the result at all; they are leaving social factors (follower count etc) aside to specifically examining the personality traits, and have rather counterintuitive conclusions; - Firstly, personality seems to matter, significantly - Neuroticism correlates positively with pull request acceptance [1] - High extraversion doesn't increase the likelihood of a pull request being accepted, but low extraversion does [1] - The personality trait differential between the requester and closer is important; the larger the gap the better [2] [1] "In sum, we observe requesters who are high on Openness, Conscientiousness, and low on Extraversion have a higher likelihood of getting the pull request accepted. Similarly, a closer who are high on Openness, Conscientiousness, Extraversion, and Neuroticism accepts more pull requests." [2] "We also noted the absolute difference in personality traits between the requester and the closer results in positive effects, suggesting that diversity in personality is beneficial in open source projects."

bawolff 6y ago
So people who are sterotypically nerds do better at sterotypically nerdy things? Did the study account for the possibility that maybe those personality traits corresponded to being members of a shared cultural context, which helped their chances of code being merged because people like people like themselves? Edit: but the bit about personality difference contradicts what i said, so maybe not.
- acituan 6y ago
  > So people who are sterotypically nerds do better at sterotypically nerdy things? I don’t get what makes you think that that is the conclusion. Study doesn’t look at the cluster of open source contributors against “rest of the world”, there is no stereotype being established here. It is examining concrete project dynamics (e.g. pull request acceptance) against personality traits within open source developers. Big five is such a widely applicable factor analysis that I don’t think there is one particular configuration that would define a “nerd” except maybe for introversion-extraversion axis. Since the study shows meaningful differences on other axes too, I don’t think it is about nerdiness at all.

bawolff 6y ago

So people who are sterotypically nerds do better at sterotypically nerdy things? Did the study account for the possibility that maybe those personality traits corresponded to being members of a shared cultural context, which helped their chances of code being merged because people like people like themselves? Edit: but the bit about personality difference contradicts what i said, so maybe not.

acituan 6y ago
> So people who are sterotypically nerds do better at sterotypically nerdy things? I don’t get what makes you think that that is the conclusion. Study doesn’t look at the cluster of open source contributors against “rest of the world”, there is no stereotype being established here. It is examining concrete project dynamics (e.g. pull request acceptance) against personality traits within open source developers. Big five is such a widely applicable factor analysis that I don’t think there is one particular configuration that would define a “nerd” except maybe for introversion-extraversion axis. Since the study shows meaningful differences on other axes too, I don’t think it is about nerdiness at all.