offers an empirical examination of this question for a subpopulation of biomedical academic scientists who received research … change using individual-level panel data. We find that the biomedical scientists who eventually choose to found or join a for …

, suggesting that professional imprinting takes place in the early stages of a scientist’s academic career. Based on a sample of …), we find that scientists with industry-oriented co-authors are more likely to be involved with industry (personal peer … effect). Moreover, we find that the scientist’s involvement increases with the orientation of the scientist’s department …