“…In fact, this same analytic problem pops up in at least three different common categories of studies: (a) comparisons of two interventions where one is a newer alternative to standard practice that has some benefit (e.g., cost, flexibility, etc. ; e.g., Bettinger & Long, 2010; Johnson, Aragon, Shaik, & Palma-Rivas, 2000; Rickles et al, 2018), which is the focus here; (b) comparisons of two estimates of teacher quality (or a similar construct) where one estimate is created in such a way as to remove a possible bias (e.g., after randomly assigning students to classrooms; e.g., Bacher-Hicks, Chin, Kane, & Staiger, 2017; Chetty, Friedman, & Rockoff, 2014; Kane & Staiger, 2008; Steinberg & Garrett, 2016; Xu, Ozek, & Corritore, 2012); and (c) within-study comparisons that compare program effect estimates from experimental and quasi-experimental estimates of the same effect (Bloom, Michalopoulos, Hill, & Lei, 2002; Jacob, Somers, Zhu, & Bloom, 2016). As is clear from this wide range of work, the problems discussed in this response extend well beyond RHASW.…”