- Download this Excel document. We will spend some time trying to understand what influences graduation rates.
- Graph a histogram for $/Student and SAT score.
- Copy & paste each graph into a word document.
- Describe whether or not each variable appears to be normally distributed.
- Estimate the mean, median, and mode based on the graph for both variables.
- Calculate the mean, median, and mode for %Grad.
- Create a scatterplot between %Grad and SAT score with %Grad on the x-axis. Do you believe there is a relationship? Explain why or why not and the direction of the relationship (e.g., positive, negative, or no relationship)
- The average SAT score in the United States is 1030. Conduct a hypothesis test which compares the distribution of this dataset to the national average at the 5% level. Is there a statistically significant difference?
- Conduct a hypothesis test comparing the average SAT for a Lib Arts institution and a Univ. Are there statistically significant differences at the 5% level?
- Conduct a multiple regression (copy and paste all three regression tables into the Word document):
- The equation for the regression is Grad% = %PhD + Top 10% + $/Student + Acceptance + SAT + School_Type. You may need to recode the School_Type into a binary variable.
- What variables explain differences, with statistical significance at 5%, in the graduation level at colleges. Explain, in plain language, the relationship between each significant variable and the graduation rate.
- Grinnell College, located in Iowa, has a 1244 SAT, 67% acceptance rate, 22,301 $/student, 65% were in the top 10%, and 79% of faculty had PhD’s. Calculate the predicted value of their graduation rate. What’s the difference between the predicted graduation rate and the actual graduation rate (e.g., calculate the number)?
- How much of the variation in %Grad is explained by this model?
- Email me the answer sheet (Word) and the results (either Excel and/or SPSS).
