SDS 306 Lecture Notes - Lecture 17: Regression Analysis, Scatter Plot
16 Feb 2017
School
Department
Course
Professor
Document Summary
Quantitative variable condition: regression analysis cannot be used for qualitative variables. Straight enough condition: the scatterplot should indicate a relatively straight pattern. Outlier condition: outliers dramatically influence the fit of the least squares line. Condition: the data should not become more spread out as the values of x increase. The spread should be relatively consistent for all x. There should be no changes in the spread from one part of the plot to another. Plan: what is the relationship between sugar content and calories in cereal: variables: sugar content and calories are both quantitative, the scatterplot looks straight enough, no obvious outliers, the plot does not thicken. The spread is consistent from left to right. R = 0. 564 = 89. 5 + 2. 50x r2 = 0. 318 s = 19. 5 se = 16. 2. Slope: a serving of cereal has about 2. 50 calories per gram of sugar on average. Intercept: predict that a serving of sugar free cereal will have about 89. 5 calories.
