STAT 2000 Lecture 9: February 1

• School: University of Georgia
• Department: Statistics
• Course: STAT 2000
• Professor: Dunlap

Document Summary

3. 4 cautions in analyzing associations (cid:1877) = (cid:884)(cid:891)(cid:891). (cid:886)(cid:890)+(cid:886). (cid:885)(cid:887)(cid:4666)(cid:883)(cid:887)(cid:890)(cid:4667)=(cid:885)(cid:890)(cid:889). (cid:890)(cid:884) (cid:1859) (cid:1858: 1. (cid:1877) = (cid:884)(cid:891)(cid:891). (cid:886)(cid:890)+(cid:886). (cid:885)(cid:887)(cid:1876) x=calories (cid:1877) =sodium, 2=(cid:882). (cid:890)(cid:889)(cid:882)(cid:891) 87. 09% of the variation in response (sodium) has been explained, 2. Extrapolation: using a regression line to predict y-values outside of observed range of data: be cautious of outliers; look out for regression outliers, regression line is often puled toward any regression outliers. Overestimated values will have a negative residual; underestimated values will have a positive residual. Slope= talks about it as a one unit increase. If r" is close to zero; there is a weak linear relationship: example: if the college rate increases by 1% then the poverty rate decreases by. Use parameter estimates to find the least squares regression line. Positive residuals will be above the regression line; negative residuals will be below the regression line.