Introduction:

multivariate statistical techniques are those that examine multiple characteristics considered in the same individual, and that because they are interrelated do not make sense to determine their effect in an isolated way. They are more frequently applied in observational studies in relation to the etiology and prognosis of a disease, since they permit to explain, predict and control confusing variables to obtain greater consistency in statistical inference.

Objective:

this paper is aimed at describing the multivariate techniques that can be applied for the study of causality in biomedical sciences.

Methods:

a review of the literature on the subject was carried out, searching in bibliographic databases such as Pubmed/Medline, SciELO, SCOPUS, Web of Science, EBSCOhost, Google, among others. The descriptors used in the search strategy were those provided in the key words of the article, which were used in English, French, Portuguese and Spanish, combined with the logical operators. Out of more than 200 digital documents found, the most current ones were chosen, and a report was prepared where the techniques were grouped according to their objective and types of variables applied in the analysis.

Results:

descriptive, explanatory and mixed techniques were presented, considering the different types of regression and structural equations as the most applied techniques in the study of causality.

Conclusions:

The use of implicit statistical analysis in the determination of risk factors and prognosis was proposed.

        · resumo em Espanhol     · texto em Espanhol     · Espanhol ( pdf )