Graphic methods in causal biomedical research

Abstract

Background: the statistical graphs can be used for the exploratory data analysis, being indispensable in the study of multivariate relations.Objective: to update on the aspects related to the use of graphic methods in the study of causality in biomedical sciences.Methods: a bibliographic review was conducted using the specialized services available in the Internet: Pubmed/Medline, SciELO, SCOPUS, Springer, Web of Science, EBSCOhost and Google, from January through March 2019. The following descriptors in English, French, Portuguese and Spanish were used: statistical techniques, statistical graphs, causal diagrams, implicative graph, variable relations and implicative statistical analysis. The professional experience of the authors was included.Results: the mostly used graphic methods are presented, highlighting those that allow the presentation of the information such as the causal diagrams that visualize the relationship between multiple variables, as well as those that allow the exploration of data and multivariate relations to direct the subsequent analysis. The study shows the usefulness of the implicative graph to identify the factors that have an influence on an ending, making it possible to determine whether it is a risk or a protective factor, as well as the magnitude of the effect, by means of implicative intensity.Conclusions: the study showed the importance of the use of graphic methods in the study of causality. The superiority of the implicative graph is highlighted.

Keywords

• statistics as topic
• statistical analysis
• biomedical research
• analytical methods