Mike LaCour faked data in Science and related discussions

[Felix L. Garcia]

Dear all,

Last week, I got into a lengthy discussion about the faked data in LaCour & Green 2014 with a buddy at UCLA who is very familiar with the lab that the data set came from, as well as the early signs that things were going wrong, and the extent of the fallout for other investigators and students associated with that lab. This conversation prompted me to look more closely not only at the original work, but also at the professional and public discourses that continue to unfold around this issue. Here is a link to a Dropbox folder containing a (non-exhaustive) group of sources including the original publication, related This American Life podcast, David Broockman's exposé, LaCour's response, retraction final decision from Science editor-in-chief Marcia McNutt, and a few other pieces from the NY Times and NY Magazine. I especially liked the even handed reality check around the ultra-competitive world of academia, by UMich sociology doctoral candidate Drew Foster. I hope that sharing these might help stimulate some discussion around our role as scientists and responsibility to the public, and how something like comes to happen in the first place.

Best,

Felix

-- 

Félix L. García

[Raquel Bibi]

Dear Felix,

We need not look as far away as UCLA for falsification in datasets.  I can present evidence that our very own faculty engage in behavior that equal LaCour.  Are you interested in commenting on these data as well?

Best,

Raquel
Sent from my iPhone

--
You received this message because you are subscribed to the Google Groups "City College Methods Shop" group.
To unsubscribe from this group and stop receiving emails from it, send an email to City-College-Metho...@googlegroups.com.
To post to this group, send email to City-College...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Felix L. Garcia]

Hi Raquel,

At the moment I'm not able to comment on data sets that I'm not already familiar with, which right now are limited to my dissertation data. That said, I think the point you're making underscores the fact that this really is a problem that happens more often than we'd like to admit, and one which the field of academia unfortunately promotes and/or condones in subtle ways. I think one of the best ways that we can all help the field as a whole is to push for more journals to publish all results from well-conducted studies, not just findings that are statistically significant and have enormous effect sizes. Another thing we can do is become more involved in mentoring, which I think is the responsibility of both faculty and students: Mentors can contribute by checking in with ourselves about whether our expectations for our students are realistic given the amount of time we are able to dedicate to them. As students, it is our duty to make sure to go to mentors when we need more guidance, and to judge for ourselves whether a particular mentor is able to provide the guidance we are needing. If not, the sooner we accept that, the sooner we can move on. Of course, this is easier said than done.

Best,
Felix