4 ms·
Good thing we don't have empirical research in software development. This means we avoid a replication crisis.
by goto11 5y ago
Good thing we don't have empirical research in software development. This means we avoid a replication crisis.
- bhk 5y agoHah! This paper describes a deep-dive into papers on database research that made quantitative claims, and about half of the investigated papers had problems with replications of some of their claims. The Repeatability Experiment of SIGMOD 2008 https://pages.saclay.inria.fr/ioana.manolescu/PAPERS/SIGRecRepeatability.pdf https://pages.saclay.inria.fr/ioana.manolescu/PAPERS/SIGRecR...
- Leherenn 5y agoBenchmarking is often close enough.