In: Statistics and Probability
Prevention and detection of drinking and driving are major concerns of traffic police. In the U.S. in 2012, nearly one-third of all traffic deaths involved drivers with excessive blood alcohol concentrations (BAC). Investigative authorities employ breath-alcohol analyzers, i.e. breathalyzers, to measure breath alcohol concentrations (BrAC) because the tests can be easily applied and do not require an attending physician to extract blood. Thresholds for alcohol-impaired driving are based on blood alcohol concentration, so it is important to understand the relationship between breath and blood alcohol concentrations. Researchers in Germany measured both the blood and breath alcohol concentrations for 17 drivers that were stopped for traffic violations. The data are shown in the table and are also available in a separate file on Canvas. (Note: German authorities use different units of measurements than U.S. authorities, so the numbers presented in the table are not equivalent to the U.S. threshold of 0.08 BAC for drunk driving.) Use the data to develop a regression equation that will predict the value of BAC (blood alcohol) by the using value of BrAC (breath alcohol).
Note: We predict Y by using X. Make sure you have X and Y correct in your calculations. You can lose a lot of points if X and Y are backward.
(a) What percent of the variability in blood alcohol is explained by the regression?
(b) Fit a linear regression equation that can be used to predict BAC from BrAC. (You will have to provide the estimated intercept and the estimated slope.)
(c) Use the estimated regression equation to predict BAC when BrAC = 0.28.