STA 6166 – Exam 3 – Fall 2010 PRINT Name _________________________True or False:Q.1. In linear regression, it is possible for an independent variable to be significant at the 0.05 significance levelwhen it is the only independent variable, and not be significant when it is included in a regression with other independent variables.Q.2. A study reports a 95% Confidence Interval for the Relative Risk of a disease for Group A relative to Group B to be (0.50 , 0.70). We conclude that Group A has a higher probability of getting the disease than Group B.Q.3. A simple linear regression is fit, and we get a fitted equation of �50 10Y X= +. Our estimate of the increasein the mean of Y for unit increase in X is 60.Q.4. In simple regression, if X is temperature (in Fahrenheit) and Y is distance (in Yards) and a colleague wishesto transform X to Celsius and Y to Meters, the regression coefficient for X will remain the same for the two regressions, but the correlation coefficient will change.Q.5. In logistic regression, we model the probability of a particular outcome for a binary response variable as a function of independent variable(s).Q.6. In multiple regression, it is possible for the error sum of squares to increase when we add an independent variable to an existing model.Q.7. A multiple regression model is fit, relating Gainesville House Prices (Y, in $1000s) to 4 predictors: BEDrooms, BATHrooms, an indicator (dummy) variable for NEW, and SIZE (ft2). A subset of the results are given in the following tables.p.7.a. Complete the ANOVA table.p.7.b. Complete the CoefficientsTable.p.7.c. Compute R2 ____________________________________________________p.7.d. Compute the predicted price (in $1000s) for a 3 Bedroom, 3 Bathroom, not new home that is 3000 ft2.p.7.e. We fit a reduced model with only NEW and SIZE, and obtain R2=0.7226, and SSR=733543.3. Test H0: Bed = Bath = 0 at the = 0.05 significance level:p.7.e.i. Test Statistic:p.7.e.ii. Reject H0 if the test statistic falls in the following range ________________________________p.7.e.iii True/False: After controlling for NEW and SIZE, neither BEDS nor BATHS is associated with house prices.Q.8. A simple linear regression is to be fit, relating fuel efficiency (Y in gallons/100 miles) to cars weight (X, in pounds), based on a sample of n=45 cars. You are given the following information:( )�( )2 21 1 12( ) 13069326 ( ) 13385 ( ) 16.52739 3.4 2.835n n ni i i ii i iX X X X Y Y Y YX Y Y Y= = =- = - - = - == = - =� � ��Compute the following quantities:p.8.a. �1b ______________________________________________________p.8.b. �0b ______________________________________________________p.8.c. Residual Std. Deviation se _______________________________________p.8.d. Estimate of mean efficiency for all cars of x*=2000 pounds ___________________________________p.8.e. 95% Confidence Interval for all cars of x*=2000 poundsLower Bound = __________________________ Upper Bound = _____________________________p.8.f. Regression Sum of Squares SSR = ___________________________________________p.8.g. Proportion of Variation in Efficiency “Explained” by Weight ___________________________p.9. A study considered the effectiveness of wearing helmets among skiers and snowboarders in Canada. The researchers considered 2 groups in a Case/Control Study: Group 1 (Cases) – Skiers/Snowboarders suffering head or neck injuries (219 out of 824 wore helmets) Group 2 (Controls) – Skiers/Snowboarders suffering other injuries (929 out of 3295 wore helmets)p.9.a. Fill in the following contingency tablep.9.b. Odds of wearing a helmet: p.9.b.i. Among Head/Neck Injury ___________________________________________________ p.9.b.ii. Among Other ____________________________________________________________p.9.c. Odds Ratio (Head/Neck relative to Other) _____________________________________________p.9.d. 95% Confidence Interval for (population) Odds Ratio:Lower Bound _________________________ Upper Bound __________________________________p.9.e. Conclusion based on your confidence interval Higher odds of wearing helmet among those with head/neck injury Higher odds of wearing helmet among those with other injury Cannot conclude odds of wearing helmet differ by injury typeHave a nice semester
View Full Document
Unlocking...