Reference no: EM132400911
Labor Market Outcomes of Graduates of for-profit and non-profit Colleges
Please download the dataset "associate_degrees.dta" and open in Stata. The dataset contains information on the employment status of 400 recent recipients of Associate's degrees (2-year post-secondary degrees in the United States). 200 graduates are a random sample from a for-profit university and 200 are a random sample from a non-profit. The for-profit university not only gave you information on the employment status of its students, but also gave you salary data for those students who reported being employed.
PART I: CONFIDENCE INTERVALS
1. What fraction of for-profit university graduates are unemployed in your sample? What fraction of non-profit graduates are unemployed?
2. Please calculate the 95% confidence intervals for the population unemployment proportions of for-profit university graduates as well as of non-profit graduates. Show all steps in your work: (i) what is the center of the CI? (ii) what is the SE of the distribution of the sample proportion? (iii) is this the actual SE or is it an estimated SE? (iv) which table do you use? t or z? (v) how many SE's does your CI need to stretch?
3. If you believe that 10% of unemployed graduates of for-profit schools default on their loans, what is the 95% CI for the proportion of the population of for-profit graduates in default?
4. Please calculate the 90% and 95% confidence intervals for salaries of the population of employed for-profit university graduates (note that all relevant sample size conditions are met). Again, show all steps in your work: (i) what is the center of the CI? (ii) what is the SE of the distribution of the sample mean? (iii) is this the actual SE or is it an estimated SE? (iv) which table do you use? t or z?(v) how many SE's does your CI need to stretch?
PART II: MANAGEMENT
5. How would you use these results to inform policy or organizational strategy? What would be the next steps?
6. Suppose you work for a manager who says one day, "I got a budget increase allowing me to collect twice as much data; that's great because our CI estimates will be twice as precise." How would you respond?
7. Suppose you work for a manager who says one day, "Let's just calculate our CIs with 90% coverage probability instead of 95%; this will make the CIs narrower, and so more precise." How would you respond?