Fitting a Least Squares Regression Line to Numerical Data (VCE SSCE General Mathematics): Flashcards

Modelling relationship between two numerical vars with a line

$a$ = y-intercept; $b$ = slope

Vertical distance between data point and regression line

Sum of squares of all residuals

Without squaring, residuals sum to zero

Numerical data, linear association, no clear outliers

$b = \frac{rs_y}{s_x}$

$a = \bar{y} - b\bar{x}$

Variable we want to predict (y)

$r = \frac{bs_x}{s_y}$