Prof. Jane Miller presented at the National Institutes of Health Library’s recent Scholarly Publishing Symposium “Elevate your Research Visibility” where she discussed her work to improve the clarity of communication about quantitative research titled “Clearly Communicating the ‘Importance’ of a Research Finding: Beyond Statistical Significance.”
Too often, researchers present quantitative results as if statistical significance were the only determinant of whether those results are “important.” This presentation starts with a brief review of hypothesis testing, then considers what questions inferential statistics can and cannot answer, including statistical significance, causality, causal order, direction of association, practical importance, whether the independent variable is modifiable, and generalizability of the results. She then reviewed the essential elements for communicating about quantitative research and suggest related resources.
The NIH staff members who attended the session asked useful questions about how the guidelines she offered could be used by journal editors, article or grant proposal reviewers, and others who encounter authors that focus mostly on statistical significance, to the exclusion of other key facets of “importance.”