- Search for Research Summaries, Reviews, and Reports
- EPC Project
- Towards a Framework for Communicating Confidence in Methodological Recommendations for Systematic Reviews and Meta-Analyses
Research Report - Final – Sept. 9, 2013
Towards a Framework for Communicating Confidence in Methodological Recommendations for Systematic Reviews and Meta-Analyses
People using assistive technology may not be able to fully access information in these files. For additional assistance, please contact us.
We propose a framework for organizing and describing the rationale behind methodological recommendations, and for communicating one's confidence in them. We start by defining the background context in which the recommendations are made. We distinguish recommendations that are testable (in that their likelihood to hold can be informed by theoretical arguments or empirical data) from nontestable ones, which represent beliefs or assumptions that are not verifiable. Nontestable statements can be justified, but their validity cannot be demonstrated. Testable statements can be assessed in terms of the adequacy of their evidentiary basis.
Both testable and nontestable statements can be evaluated regarding their feasibility of implementation, the expected impact of following them versus not, and their congruence with the desired characteristics of the background context. Considering these four dimensions, one can indicate one's confidence (along a continuum) about how closely a methods recommendation should be followed: some recommendations may be perceived and communicated as mandatory items (minimum standards), while others as highly desirable but not mandatory items. Finally, giving specific methods guidance for addressing difficult or ill-defined problems can be premature pending more research or clearer problem specification. In such cases, describing the problem and laying out attributes of a satisfactory resolution can serve until actionable guidance can be offered.
We view the proposed framework strictly as a communication tool to describe rationale for the recommendations to the intended audience and not as a device to deduce the “correctness” of a recommendation. Nonetheless, application of the framework can facilitate the latter, because methodologists can most effectively and honestly critique recommendations whose rationale is transparent. We anticipate that this initial instantiation of the framework for making methods recommendations will evolve.