Statistical Sciences 2244A/B Lecture Notes - Lecture 17: Confidence Interval, Statistical Parameter, Binomial Distribution
Document Summary
Magnitude and direction of how strong the relationship is by how close the points lie to a. Imaginary line : tight values to line = high correlation, but does not relate to slope. Clicker: a plant population is known to have a weak positive relationship between age (years) and height (inches). Minimizes the sum of (squared) vertical distances between the data and the line. Software best decides the line: by minimizing the area from the line . Model 1 regression: where we try to match a line to lines by looking at least squares data. Where ! is the predicted value a and b are sample estimates. We want the least possible squares when we add up the squares. Meaning there is no slope and y does not change as x changes. If we reject the null hypothesis then we can say there is a linear relationship.