placeholder

Hasnain says:

Interesting read on scientific paper evaluations across various communities.

And a harrowing indictment of one security conference. I can’t speak to the details here but the argument rings true.

“The papers I reviewed for CCS were all about finding bugs, and each and every paper bragged that they found more bugs than the state of the art. But none of my submissions came with even a whiff of discussion on whether the results would be generalizable, or why the algorithms would be correct. Not a single one would discuss statistical significance or effect size. Data sets, replication packages, threats to validity or even discussions of limitations were all absent.

“At any SE or PL conference, these CCS papers would have sunk in an instant””

Posted on 2021-07-30T07:29:27+0000