These explorations, which occur before running confirmatory statistical analyses, can shed greater light on key issues relevant to missingness. Various researchers have suggested analysts can better understand missingness in their data through exploratory analyses, including visual and numerical summaries ( Cheng, Cook, & Hofmann, 2015 Buja, Cook, & Swayne, 1996) akin to classical exploratory data analyses ( Tukey, 1962). Yet, outside of some statistical tests (e.g., Little, 1988), there is little guidance for forming and examining theories about missingness ( Tierney & Cook, 2018). Much of the literature on missing data has focused on the implications of this assumption ( Pigott, 2019 Little & Rubin, 2002). data with missing values) is that analysts understand what data are missing and why (i.e. A vast literature on missing data methodology highlights the ways missingness can bias statistical inferences, examines conditions under which these biases can be corrected and proposes various statistical procedures to do so ( Rubin, 1976 Pigott, 2019, 2001a Schafer & Graham, 2002 Graham, 2009 Little & Rubin, 2002 van Buuren, 2018).Ī key assumption of many analysis methods for incomplete data (i.e. However, meta-analyses and meta-regressions frequently contend with missing data ( Pigott, 2019). type of therapy provided), how or on whom it was implemented, or the context in which it was studied (see Cooper et al., 2019). For example, a meta-analyst may examine how the effectiveness of interventions is related to the type of treatment (e.g. Alternatively, meta-regression is a statistical model analogous to standard linear regression, wherein effect estimates are regressed on covariates pertaining to those effects, including study- and effect-level information ( Hedges, 1982a, 1982b Cooper, Hedges, & Valentine, 2019). Traditional meta-analyses summarize the results of ensembles of studies of interventions, typically reporting the average impact or variation across impacts. Systematic reviews of substance abuse research hold great promise for better understanding the effects of interventions ( Tanner-Smith, Wilson, & Lipsey, 2013 Tanner-Smith et al., 2016 White et al., 2010 Newbury-Birch et al., 2018 Ramsey et al., 2019 Yuvaraj et al., 2019).
0 Comments
Leave a Reply.AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |