5: The government conducts the census every 10 years, but not everyone responds. Follow-up workers interview the people who don't mail back their form. To intelligently allocate workers, we need to figure out which quantities are associated with the response rate. Use EX2. CENSUS. Most columns give the percentages of each neighborhood with a certain property, e.g., Suburban gives the percentage of that neighborhood that is considered "suburban", Hispanics gives the percentage of individuals that identify as Hispanic, Age45to65 gives the percentage of the individuals within that age range. Other columns like VacantUnits and RentingHH give the fraction of housing units or households with that quality. a) We want to predict ResponseRate. What five variables are most strongly correlated (may be positive or negative) with ResponseRate. Hint: use all. correlations (). b) The variable most correlated with the response rate is HomewownerHH, the fraction of households that own their home. Calculate r2 and interpret its value. c) Is the association between HomewownerHH and ResponseRate statistically significant? Explain. Do you think it holds practical significance? Explain. d) Explore the association between logMedianHouseValue (the logarithm of the median home value) and MedianHHIncomeCity (the median income of households in the city). Use associate () (let x be MedianHHIncomeCity ) and describe the direction and form of the stream of points. Does the stream have "unusual features"? Which measure (Pearson vs. Spearman) should be used to measure the strength of the association and why