True or False? 1 The correlation coefficient r is a more...

This preview shows page 1 - 2 out of 6 pages.

STA 6126 – Fall 2007 – Exam 4 PRINT NAME ____________________________________ True or False? 1. The correlation coefficient (r) is a more appropriate measure to report than the slope of the regression line (b) when describing the association between two variables when there is not a clear independent and dependent variable. 2. The slope of the regression line and the correlation coefficient both must lie between -1 and +1. 3. Authors of a report state that the coefficient of correlation for their analysis is r=0.5. This means that using X to predict Y reduces prediction error by 50% as opposed to not using X for the predictions. 4. Simpson’s Paradox refers to the situation where overall the association between X and Y is one direction, but when we control for a factor Z, the X-Y association is in the opposite direction for each level of Z. 5. A regression equation is fit, relating weekly food expenditures (Y) to number of household members. The prediction equation is Y-hat = 25+60X. The estimated increase in mean weekly food expenditures increases by $60 for each extra household member. 6. For a regression relating salary (Y) to work experience (X), the Total sum of squares (around the sample mean Y-bar) is 2000 and the Error sum of squares (around the fitted line Y-hat) is 500. The coefficient of determination r 2 is 0.25.

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture

  • Left Quote Icon

    Student Picture