Prior research consistently shows a positive correlation between founders' education and experience and startup performance. However, this relationship may be driven by either treatment effects, whereby more educated and experienced founders attain superior performance, or selection effects, whereby these founders are more likely to secure resources, resulting in a positive correlation with startup performance. We investigate the role of these treatment versus selection effects in the context of startup accelerators, early filters for investment. Using an abductive approach, we find that founders' education and experience are not predictive of accelerators' initial acceptance, but increase a founder's likelihood of participation, conditional on acceptance. Our findings suggest that the observed positive correlation between founder human capital and startup performance is driven by selection, rather than treatment, effects. learn more on SSRN