1.  Read Studenmund Ch. 13 and the EViews "Help" information regarding binary dependent variables.  Refer to the help facility as often as necessary.  Peruse the "quality of life" survey questionnaire that was distributed in class.  Recall that the variable numbers correspond to questions in the survey.

2. Choose a binary variable that you wish to better understand, and create a suitable dummy variable to represent it.  Specify an appropriate logit model that "explains" the variation in the dependent variable.  Predict signs for each of the coefficients in your model (+, -, or ? where appropriate).  Briefly justify your predictions.

3.  Retrieve the file qualitylife99b.wf1 from the ECN 32 share on the public-eba share of server Ricardo.  Estimate the model that you specified in (2).  Conduct hypothesis tests for each hypothesis that you specified in (2).  State the formal null and alternative hypotheses, the decision rule, the test-statistic, and the results of your test.

4.  Compute estimated probabilities (that the dummy dependent variable is 1) for each observation in your dataset.  Pick one observation from your dataset and interpret the probability in words, i.e., explain what it means.

5.  Compute the mean values for the X's used in your model.  Compute the probability that the dependent variable would equal 1 for a person who had X's that equalled the means of the X's from your sample.  Explain in words what this means.

6.  Compute the marginal affect of one of the X's on the probability of the dependent variable being 1, holding constant the other X's in the model.  Interpret this in words, given variables involved in your study.

7.  Compute Rp-squared, and explain in words what this means.

8.  No class on Friday, due to Project Jericho.

9.  E-mail your answers and supporting information (as necessary) by Tuesday, May 9.