Unformatted text preview: y-intercept parameter <- a slope parameter <- b so Yi and Xi are values that are measured rather than estimated. We use these to estimate a, b, and u. why do we care about the predicted if we have the observed? The Ui tells us a lot of stuff. It tells us how well the line fits the data. You want the sum of square residuals to be close to 0. What does this tell us about politics?...
View Full Document
- Summer '08
- Statistics, Trigraph, Yi, line fi ts