QTM 100 Lecture Notes - Lecture 23: Type I And Type Ii Errors, John Tukey, Weighted Arithmetic Mean
Document Summary
Estimated linear regression equation: to predict a value of y for a specific value of x. Predict an average response (a confidence interval for the mean: more variability in individual responses than in average responses, so these intervals take different forms. Variability about the line: estimate the constant variability about prediction line: residual standard deviation (r output calls this the residual standard error) Predicting mean y: confidence interval for the mean: Predicting individual y (more variability than mean y: prediction interval can be approximated by. The multiple linear regression model x x y. = intercept - this is the predicted value of y when both x1 = 0 and x2 = 0. = slope associated with x1 when controlling or adjusting for x2 interpretation: controlling for other x variable, an. 1 additional 1 (x-axis unit)/(y-axis unit) is associated with blank slope increase o. = slope associated with x2 when controlling or adjusting for x1.