bivariable

Bivariable analysis refers to the examination of the relationship between two variables. It is used to assess whether and how the variables move together and to quantify the strength of their association. Unlike multivariable analysis, it does not adjust for other variables.

When both variables are continuous, common tools are the Pearson product-moment correlation coefficient and simple linear regression. The Pearson r measures linear association, while scatter plots visualize the pattern; if the data are not normally distributed or the relationship is monotonic but not linear, Spearman rank correlation or Kendall tau may be used. For two categorical variables, the relationship is often summarized in a contingency table and tested with a chi-square test of independence or, when sample sizes are small, Fisher's exact test. For one continuous and one categorical variable, differences in the continuous outcome across categories can be tested with t-tests or ANOVA; the strength of association can be summarized with point-biserial correlation or eta-squared.

Interpreting bivariable results requires caution: correlation does not imply causation, and observed associations may be influenced by confounders or biases. Outliers can distort the magnitude of the association, and small samples reduce statistical power. Visualization, such as scatter plots or contingency tables, aids interpretation and helps check underlying assumptions.

Bivariable analyses are common in many fields, including epidemiology, psychology, economics, and social sciences, serving as a preliminary step before more complex multivariable modeling.