|
@KateLaskowski | |||||
|
I'm starting the year off with something I didn't expect to ever do: I'm retracting a paper. I recently discovered major problems in the raw data associated with it and so the results shouldn't be trusted. journals.uchicago.edu/toc/an/current… pic.twitter.com/1Rq87kY0QY
|
||||||
|
||||||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
I collaborated on this paper with Jonathan Pruitt & PO Montiglio. We were testing predictions from the social niche specialization hypothesis in social spiders. 2/
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
The paper argued that we found support that repeated social interactions increased behavioral variation and had positive effects on individual and group success. This was so cool and I was so excited about these results! 3/
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
However, I recently fielded a question from a colleague about the available data. There were an unexpected number of duplicate values that I had never previously noticed. Unfortunately these problems spanned colonies, treatments and time points 4/
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
Making it biologically impossible to explain and Jonathan, who collected the data, couldn't figure out how these patterns emerged either, meaning we couldn't just somehow correct it. 5/
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
The prudent thing was then to try to remove the corrupted data. But in this much reduced dataset, we could no longer find support for our original results. 6/
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
The main point is therefore we can no longer stand behind these results. I'm deeply sorry (& embarassed) I didn't catch these problems when I recieved the data for analysis. 7/
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
Finally some of my sincerest thanks to @DanielBolnick at Am Nat for being so absolutely supportive during this whole process 8/8
|
||
|
|
||
|
Graham Coop
@Graham_Coop
|
17. sij |
|
Sorry to hear this Kate, that sucks. But a big thank you for correcting the record!
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
Thanks Graham. Yeah it is not awesome, especially because this was one of my favorite papers in one of my favorite journals 😭
|
||
|
|
||
|
Christos Ioannou
@C_C_Ioannou
|
17. sij |
|
Nice one, @KateLaskowski. Are the other papers based on data "collected in the laboratory of the third author" being checked? My PhD student has a couple of papers submitted that heavily cite their work.
|
||
|
|
||
|
Kate Laskowski
@KateLaskowski
|
17. sij |
|
Yes, I am going through all my papers on this topic and should have more information on that front soon.
|
||
|
|
||