Question 1. (18 points) In this question, we will use the U.S. Census Data on Voting and look at correlations of voting outcomes and demographic characteristics.

(a) Go to the following website and download Table 1. Make sure you work with the sheet labeled "Table 1." https://www.census.gov/data/tables/time-series/demo/voting-and-registration/p20-583.html

Create scatter plot that shows the relationship between the percent of US citizens (both sexes combined) who reported voting in the 2018 election and age [in Excel, look under Charts.]. Have voting (your dependent variable) on the y-axis, and age, yearly from 18 to 79, on the x-axis. Add a linear fit and the fitting equation to the graph . Include your graph on your answer

(b) Interpret the relationship between voting and age based on your answer to (a). Explain why this relationship is not causal (give a detailed explanation, with examples). [Your answer should be 1]

(c) Use the same Table 1, but now create dataset with the percent of US citizens who reported voting by age (yearly from 18 to 79) and sex (male or female). Hence, you should have 3 variables: 2 Sex (equals 1 if male, 0 if female), Age (18 to 79), and Reported Voting (percent) for each agesex group. Run a linear regression of Reported Voting in 2018 on Age and Sex [use Data > Data Analysis; see the link above for step-by-step instructions]. Include your regression table on your answer (don't worry about anything labeled SUMMARY OUTPUT or ANOVA, you only need the table showing coefficient estimates, standard errors, etc.).

(d) Interpret your regression output: 1. What do the coefficients on Age and Sex mean? 2. What do the standard errors tell you? 3. The p-values shown are for the null hypothesis that the given coefficient is 0. What do these p-values tell you?