“…Fourth, in many instances publications, researchers, or institutions under investigation are not a random sample of the whole population of publications but highly selective. This particularly holds for studies on the causal effects of certain events, such as academic honors (e.g., Azoulay, Stuart, & Wang, ; Chan, Frey, Gallus, & Torgler, ) or scientific scandals (e.g., Azoulay, Furman, Krieger, & Murray, ; Lu, Jin, Uzzi, & Jones, ), on citation impact. Hence, a simple comparison of average citations received by Nobel Papers with citations of average papers in a field would be “naïve” (Morgan & Winship, , p. 44) as it assumes the absence of nonrandom selection into treatment (for a general discussion, see Imai, King, & Stuart, ).…”