analysis, bivariate
(NL: analyse van twee variabelen)
Bivariate analysis is a collection of techniques used to investigate the relationship or association between two variables. Depending on the research question, one of the variables is considered the independent variable, and the other the dependent variable.
Suitable techniques for visualizing or analyzing two variables depend on the measurement level of the variables involved. In this course, we limit ourselves to the techniques summarized in the tables below.
Data visualisation techniques
| Independent variable | Dependent variable | Technique |
|---|---|---|
| Qualitative | Qualitative | Clustered bar chart |
| " | " | Stacked bar chart |
| " | " | Mosaic plot |
| Qualitative | Quantitative | Box plot |
| " | " | Density plot |
| Quantitative | Quantitative | Scatter plot |
Techniques for data analysis
| Independent variable | Dependent variable | Test | Measure |
|---|---|---|---|
| Qualitative | Qualitative | Independence test | Cramér's V |
| Qualitative | Quantitative | T-test | Cohen's d |
| Quantitative | Quantitative | Correlation coefficient |
- In the case of two quantitative variables, in this course we only investigate whether a linear relationship exists between the two variables.
- Note that there is indeed a statistical testing procedure to investigate linear correlation, but it is not covered in this course.