4. Global Warming Transformed For this exercise, you will need to access the "Child Mortality" and "CO, Emissions" data set. The 2006 version of these data were collected for 191 countries, and archived by Dr. Hans Rosling. They are in the back of this book, and also available at www.statconcepts. com/datasets, where they can be downloaded as a text or Excel file. The scatterplot for the original pair of variables "Child Mortality" and "CO2 Emissions" is shown below. 200 180 16 140 120 O 100 Child Mortality (0-5 year old dying per 1,000 born) o o 8 20 O OO O O O 10 20 30 40 60 70 CO, Emissions (tonnes per person) a. Construct a scatterplot of log,, (Child Mortality) versus log,, (CO, Emissions). How does this plot compare to the one for the original variables shown here? You need to be detailed with your answer. b. What kind of association do you see in your newly constructed scatterplot? i. Negative, extremely weak o ii. Positive, moderately strong iii. Positive, extremely weak iv. Negative, moderately strong 8 292 Beyond the Numbers: A Student-Centered Approach for Learning Statistical ReasoningChapter 12 | More than One Variable: Association and Correlation c. Use software to compute the correlation coefficient for these two variables. What is it? i. 0.44 ii. -0.44 iii. 0.80 iv. -0.80 d. Is it wise to use the correlation coefficient "r" to summarize the association between this new pair of variables? Why or why not? i. Yes, because the scatterplot is essentially a straight line ii. Yes, because the data are ratio level iii. No, because the data are ratio level iv. No, because the scatterplot is curvilinear and not a straight line