Sunday data/statistics link roundup (11/2/14)

Better late than never! If you have something cool to share, please continue to email it to me with subject line “Sunday links”.

1. A DrivenData is a Kaggle-like site but for social good. I like the principle of using data for societal benefit, since there are so many ways it seems to be used for nefarious purposes (via Rafa).
2. This article claiming academic science isn’t sexist has been widely panned Emily Willingham pretty much destroys it here (via Sherri R.). The thing that is interesting about this article is the way that it tries to use data to give the appearance of empiricism, while using language to try to skew the results. Is it just me or is this totally bizarre in light of the NYT also publishing this piece about academic sexual harassment at Yale?
3. Noah Smith, an economist, tries to summarize the problem with “most research being wrong”. It is an interesting take, I wonder if he read Roger’s piece saying almost exactly the same thing  like the week before? He also mentions it is hard to quantify the rate of false discoveries in science, maybe he should read our paper?
4. Nature now requests that code sharing occur “where possible” (via Steven S.)

5. Great [Better late than never! If you have something cool to share, please continue to email it to me with subject line “Sunday links”.

6. A DrivenData is a Kaggle-like site but for social good. I like the principle of using data for societal benefit, since there are so many ways it seems to be used for nefarious purposes (via Rafa).
7. This article claiming academic science isn’t sexist has been widely panned Emily Willingham pretty much destroys it here (via Sherri R.). The thing that is interesting about this article is the way that it tries to use data to give the appearance of empiricism, while using language to try to skew the results. Is it just me or is this totally bizarre in light of the NYT also publishing this piece about academic sexual harassment at Yale?
8. Noah Smith, an economist, tries to summarize the problem with “most research being wrong”. It is an interesting take, I wonder if he read Roger’s piece saying almost exactly the same thing  like the week before? He also mentions it is hard to quantify the rate of false discoveries in science, maybe he should read our paper?
9. Nature now requests that code sharing occur “where possible” (via Steven S.)
10. Great](http://imgur.com/gallery/ZpgQz) cartoons, I particularly like the one about replication (via Steven S.).