4 ms·

Well, reproduced research is inherently more valuable than novel but unsupported research. Otherwise why fund research at all? Just hire alchemists or poets, at

• esrauch 8y ago
  The current state of research seems to reflect that there is a balance: almost none of the most cited papers go really far out of theirs way to be trivially reproducible, even in computing where doing so is unusually much easier. If a paper is very novel and impactful but difficult to reproduce, it doesn't seem to matter much for citation counts as long as people just believe it to be true. And yet there's no revolution among funding sources to send money away from those and instead towards researchers who spend more time on it.