The ANES asks the respondent if she is married, and if she is not and there is another adult living in the household, they ask the respondent if she has a partner. Use the recoded marital status partner variable as your dependent variable in a logistic regression model. When you run the model, select cases so that the recoded marital status variable < 2 (this will limit the model to those who are married and those who are partnered). Use two independent variables: age and whether the respondent is gay, lesbian, or bisexual.

Create a table of your results. Then run predictions for

a. a 50-year-old heterosexual

b. a 50-year-old nonheterosexual Turn these predictions into odds, and then calculate the odds ratio of nonheterosexual versus heterosexual. Does it match up with the corresponding result in the output? After completing this exercise, be sure to go back into the Select Cases box and get all of your cases back.