Ensuring Prevention Science Research is Synthesis-Ready for Immediate and Lasting Scientific Impact

Abstract

Synthesis of evidence from the totality of relevant research is essential to inform and improve prevention efforts and policy. Given the large and usually heterogeneous evidence available,reaching a thorough understanding of what works, for whom, and in what contexts, can only be achieved through a systematic and comprehensive synthesis of evidence. Many barriers impede comprehensive evidence synthesis, which leads to uncertainty about the generalizability of intervention effectiveness, including: inaccurate terminology titles/abstracts/keywords (hampering literature search efforts); ambiguous reporting of study methods (resulting ininaccurate assessments of study rigor); and poorly reported participant characteristics, outcomes, and key variables (obstructing the calculation of an overall effect or the examination of effect modifiers). To address these issues and improve the reach of primary studies through theirinclusion in evidence syntheses, we provide a set of practical guidelines to help prevention scientists prepare synthesis-ready research. We use a recent mindfulness trial as an empirical example to ground the discussion and demonstrate ways to ensure: (1) primary studies are discoverable; (2) the types of data needed for synthesis are present; and (3) these data are readily synthesizable. We highlight several tools and practices that can aid authors in these efforts, such as creating a repository for each project to host all study-related data files. We also provide step-by-step guidance and software suggestions for standardizing data design and public archiving to facilitate synthesis-ready research.

Methods for testing publication bias in ecological and evolutionary meta‐analyses

“…Overall, we suggest that all future meta‐analyses in ecology and evolution should test for publication bias, and try to identify related biases. For meta‐analysts to achieve this goal, all empiricists need to report their statistical results, including their sample sizes and estimates of uncertainty ( SE and SD ), transparently and compressively (Gerstner et al., 2017; Hennessy et al., 2021).…”

Estimating the reproducibility of social learning research published between 1955 and 2018

“…Yet ostensibly, materials should be ready to share at the review stage of the publication process, as reviewers regularly request access to the data. Widespread data science tools and data archiving technology dramatically reduce the costs of data curation today [54]. Indeed, if data are well archived, well documented and made available at the time of publication, it should demand negligible time and effort to deal with, or obviate the need for, a data request.…”

A framework and case study to systematically identify long‐term insect abundance and diversity datasets

“…This example is a useful illustration of how search strategies will fail to capture relevant studies when not well designed to thoroughly explore the literature. Authors of primary research articles often do not consider the role their work might play in future syntheses (Hennessy et al, 2021), and as such may not describe their research using the keywords that synthesists and meta‐analysts subsequently use to search the literature. To address this problem, we used a semi‐automated approach that allowed us to expand our selection of search terms based on methods designed to identify missing synonyms in fields that lack standardized keywords (Grames, Stillman, et al, 2019), rather than simply relying on an initial preselected set of terms.…”