Introduction:

statistical analysis is divided into two components: descriptive analysis and inferential analysis, statistical tests, are divided into parametric and nonparametric. Choosing an appropriate statistical test, however, is challenging.

Objective:

to propose a tool, which serves to select a statistical test to those not experts in biostatistics, for biomedical research.

Methods:

a narrative literature review study was conducted. It was based on a review of the literature on the main statistical tests currently used, available in the literature of bibliographic databases such as Pubmed/Medline, SciELO, SCOPUS, Springer, Web of Science, EBSCOhost among others. This tool was validated by expert criteria.

Results:

a scheme was designed, which can be used in two ways: the first is to follow the scheme until discovering which is the appropriate statistical technique for your data and the second: following the scheme backwards, determine if the test was a logical choice for the analyzed data. To use it, you must identify dependent and independent variables, then you will arrive at a synthesis measure or a useful point estimate for your data, which will be followed by a general classification of statistical tests.

Conclusions:

the proposed guideline is an attempt to provide a useful guide to select a statistical technique for data analysis in biomedical research.

        · abstract in Spanish | Portuguese     · text in Spanish     · Spanish ( pdf )