3 ms·
The paper states: > By focusing on grant proposals that fell just below and just above the funding threshold, we compare “near-miss” with “near-win” individual
by glial 7y ago
The paper states:
> By focusing on grant proposals that fell just below and just above the funding threshold, we compare “near-miss” with “near-win” individuals to examine longer-term career outcomes.
This is an example of a 'regression discontinuity analysis'. The underlying assumption is that applications just-above and just-below the NIH funding threshold have roughly the same merit, and falling on either side of the threshold can be explained by noise in the review process. So they're explicitly comparing scientists whose applications are equal with respect to merit - meaning the people dropped were not 'bad candidates' and the people funded were not 'better candidates'.
Since candidates are (assumed to be) equal in terms of merit, comparing the eventual outcomes of the two groups is a way to infer the causal effect of early funding or rejection on career outcome.
- konschubert 7y agoYes, I think I understood that. My point is that the simple event of a near-miss creates "evolutionary pressure" in the aftermath of the event, removing weaker candidates from the pool (their professors fire them, ...) and thereby biases that pool for success.
- mattkrause 7y agoThe applications just below and above the payline are presumably of similar merit, but the people are not necessarily matched and there’s only a tenuous link between person-quality and grant-quality (to the extent that either are real constructs that can be measured even sorta accurately).