Las variables en bioestadística. Lo indispensable.
Understanding Variables in Statistics
Classification of Variables
- The concept of variables is crucial in statistics, often causing confusion due to their various characteristics and definitions. However, the classification can be straightforward.
- Variables are categorized based on how they are measured, primarily into two families: qualitative (cualitativas) and quantitative (cuantitativas). Qualitative variables measure qualities, while quantitative variables measure quantities.
Types of Qualitative Variables
- Qualitative variables can be further divided into nominal and ordinal types. Nominal variables only show names; for instance, gender can be classified as male or female.
- Nominal variables can also be polyatomic, allowing multiple classifications such as medical diagnoses (e.g., diabetes, hypertension).
- Ordinal qualitative variables indicate an order or ranking but cannot undergo mathematical operations. They are best described using percentages or proportions.
Types of Quantitative Variables
- Quantitative variables measure amounts and are split into continuous and discrete categories. Continuous variables allow decimal values (e.g., weight), while discrete ones do not permit fractions (e.g., number of children).
- Another classification for quantitative variables includes interval and ratio types. Interval variables have an arbitrary zero point (e.g., Celsius temperature), whereas ratio variables possess an absolute zero point (e.g., Kelvin temperature).
Dependent and Independent Variables
- Beyond measurement methods, another way to classify variables is by their effect: dependent, independent, and confounding variables.
- A dependent variable's behavior relies on another variable known as the independent variable. Confounding factors may influence the relationship between these two.
- For example, weight may appear dependent on height; however, this relationship could be affected by other factors like ethnicity that must be considered to avoid erroneous conclusions about their correlation.