STATS 250 Lecture Notes - Lecture 10: Scatter Plot
49
STATS 250 Full Course Notes
Verified Note
49 documents
Document Summary
Regression analysis: describing and assessing the significance of relationships b/w 2 quantitative variable. Bigger idea: what is the relationship b/w x and y: see it with a scatterplot, can create an equation and plug in a value to our equation to predict outcome. Generally we want to make a line of best fit to fit the points, and we do that with a least squares regression. If scatterplot suggest a linear dependence can find least squares regression line of y on x. Minimizes the sum of the squared vertical distances (the residuals") b/w data points and line. Describes linear relationship b/w y and x. Predicts y for a given value x. Y - y-hat = the residuals aka the error or how far the observed value is from the predicted value: observed - predicted value, the line minimizes the sum of the square of all residuals, note: Positive residuals: observations above the least squares line.
