AP Statistics: Exploring Two-Variable Data

This chapter transitions from analyzing single variables to exploring the relationship between two quantitative variables. The primary goal is to describe the form, direction, and strength of an association using graphical and numerical methods. You will learn to create and interpret scatterplots to visualize patterns, and then calculate the correlation coefficient, r, to measure the strength and direction of a linear relationship.

Building on this, the chapter introduces the least-squares regression line (LSRL) as a model for linear data. You will be expected to find the equation of this line and use it to make predictions. A significant part of the work involves understanding the components of the model, such as the slope and y-intercept, and interpreting them in the context of the data. You will also analyze the quality of the model by examining residuals and the coefficient of determination, r-squared.

The main difficulty often lies in the conceptual understanding, particularly the critical distinction that correlation does not imply causation. Students must also be precise when interpreting statistical output from computer software. This chapter builds directly on the skills of describing data distributions from the previous unit on one-variable data, extending concepts like center and spread to the context of residuals and prediction error.

Formulas for AP Statistics: Exploring Two-Variable Data

Correlation Coefficient: $r = \frac{1}{n-1} \sum \left( \frac{x_i - \bar{x}}{s_x} \right) \left( \frac{y_i - \bar{y}}{s_y} \right)$ Measures the strength and direction of a linear relationship between -1 and 1.
Slope of the Least-Squares Regression Line: $b = r \frac{s_y}{s_x}$ b is the slope, r is the correlation, s_y and s_x are the standard deviations.
Y-intercept of the Least-Squares Regression Line: $a = \bar{y} - b\bar{x}$ The LSRL must pass through the point of averages (x̄, ȳ).
Equation of the Least-Squares Regression Line: $\hat{y} = a + bx$ ŷ is the predicted value of y for a given value of x.
Residual: $\text{residual} = y - \hat{y}$ The difference between the actual observed value (y) and the predicted value (ŷ).
Coefficient of Determination: $r^2$ The proportion of variation in the response variable (y) explained by the explanatory variable (x).
Standard Deviation of the Residuals: $s = \sqrt{\frac{\sum (y_i - \hat{y}_i)^2}{n-2}}$ Measures the typical size of a prediction error (residual).