7 ms·
Economists, policymakers, and scientists studying how to improve science output have themselves been hamstrung for decades because economists have to measure so
by dluan 7y ago
Economists, policymakers, and scientists studying how to improve science output have themselves been hamstrung for decades because economists have to measure something, and for so long it's been nothing but bibliometrics.
Until we can come up with a way to measure the other outputs of science productivity, we're stuck with this citation-based machine that has all of this institutional and cultural inertia behind it. Which is why I've come to believe this kind of change or new introduction of value isn't going to come from within academia. E.g. novelty is in some fields directly at odds with what makes you a "productive" professional scientist.
Last year I was in the NBER's Science of Science Funding working paper session, and most of the datasets discussed are still heavily focused on patents, citations, and bibliometrics (https://projects.nber.org/drupal/SOSF/data https://projects.nber.org/drupal/SOSF/data).
- gumby 7y agoIt’s the same as using lines of code as a productivity metric.
- jackcosgrove 7y agoI think the answer is that trust should be part of the equation. "If you can't measure it, you can't manage it" is a description of how to keep track of a low-trust system. Scientists are generally conscientious people, and bean counting is demoralizing. If there is some organizational solution, I think it should be to keep organization size small enough that it can be governed by personal relationships and trust. Otherwise social trust takes decades to build up, and I don't think there is a quick fix.