STAT 205 Lecture Notes - Lecture 33: Probability Plot, Statistical Hypothesis Testing, Total Variation
Document Summary
Simple linear regression: using sample data to study the linear relationship between two variables (x & y) Assumptions behind linear regression a. k. a. the simple linear regression (slr) line: there is a linear relationship between x & y. = random error component, (cid:4666)0, (cid:4667)where is the variation of y about the regression line (only works if variation of y at all points in the sample is the same) Because is expected to be 0 the formula for the linear regression line/line of means is: 0 and 1 can only be known if you have knowledge of the population, therefore we use estimates in our calculations. When used with x they can provide information about the mean value of y which is simply a point on the line of means. Because (cid:4666)0, (cid:4667) the greater variability ( 2) of the greater error there will be in predicting 0 and 1 and as a result, y.