Scatter plots in descriptive statistics serve as a fundamental visualization mechanism for assessing bivariate relationships between continuous numerical variables within multivariate data analysis. The concept relies on the formal definition of plotting paired observations (x, y) to reveal patterns such as positive or negative association strength and directionality without imposing parametric constraints like linearity or homoscedasticity. As a descriptive subfield of exploratory data science, this method establishes the necessary theoretical groundwork for identifying potential correlations that inform subsequent inferential modeling techniques.
Prerequisite chain context: requires Correlation between Metric Variables in Statistics.