“…The results are mixed when “child‐driven” versus “relationship‐driven” models are contrasted and findings may depend on the specific outcome variable being assessed (e.g., social/emotional vs. cognitive/academic) as well as on the sources of information for TCRs and/or child behaviors (e.g., teacher report, parent report, direct observation, and performance based). However, reports from numerous studies have suggested that teachers' TCR perceptions were at least partially influenced by child behaviors assessed before teachers rated their TCRs (i.e., child‐driven effects; e.g., Birch & Ladd, 1997; Guan et al., 2020; Wu et al., 2018; Zatto & Hoglund, 2019). However, in other studies, investigators have reported that TCRs predict children's subsequent social and academic outcomes even when controlling prior child behaviors, temperamental characteristics, and/or cognitive abilities (i.e., relationship‐driven effects; e.g., Cadima et al., 2016; Cadima et al., 2019; Varghese et al., 2019).…”