Use theUS_CRime.txtdata to extract 4 columns of data: Murder, Robbery, Larceny, and Burglary for South Carolina. Your task is to use multiple regression and predict the Murder count based on the other three variables. Create the appropriate regression table and answer the questions.Numerical answers are rounded so choose the answer that matches the best:

1. What is the standard error and what would be a one sentence interpretation of this number?

• a. 35; It is the number that correlates the number of murder with the numbers of robberies, larcenies and burglaries.
• b. 35; Our prediction for the number of murders when there are no robberies, larcenies and burglaries
• c. 17; It is the number used to create the prediction interval of the number of murders.
• d. 35; Our prediction for the number of murders will be accurate plus/minus 35 murders.
• e. None of these

2. What is the coefficient next to Burglary and what would be a one sentence interpretation of this number?

• a. 0.004468; For each murder, the number of burglaries increases by this amount.
• b. 0.004468; In a case of zero burglaries, we expect this amount of murders.
• c. 2.28E-07; It is the number that correlates the number of murder with the numbers of robberies, larcenies and burglaries.
• d. 0.004468; For each burglary the number of murders increases by this amount.
• e. None of these

3. What is the statistical interpretation of the number 288.37?

• a. It helps us to create the prediction interval.
• b. It is the how much the number of murders will increase if there are no robberies, larcenies, and burglaries.
• c. When there are no robberies, larcenies and burglaries, this is the expected number of murder.
• d. It tells us the total number of robberies, larcenies, and burglaries when there is no murders.
• e. None of these

4. Give an interval prediction of the murder count in South Carolina if we recorded 65000 counts of robbery, 715000 counts of larceny, and 215500 counts of burglary.

• a. [2105.30, 2442.43]
• b. [5605.58, 5677.30]
• c. [1288.07, 1359.79]
• d. [1612.30, 1697.95]
• e. None of these

OpenSATdata.

Multiple Regression

1. Click onData analysis, thenRegression.
2. In the dialog box for the inputX-variables highlightall three columns: High School GPA, SAT, and Letters (and do not forget to includeLabels)
3. In the dialog box for Y-variable highlight College GPA and click on the tabLabelsand then OK.
4. This table should appear:

Multiple Regression

Also seeVideo.

1. Since the letters had a P-value of 0.65 which is bad (that is larger than 0.05) we will need to re-do the analysis.
2. Repeat the regression but now using ONLY two columns for the input X-variables highlighted columns. High School GPA and SAT scores--NOT the Letters (and do not forget to includeLabels).
3. In the dialog box for Y-variable highlight College GPA and click on the tabLabelsand then OK.
4. This table should appear:

Use theSATdata and via Multiple regression select the two variables that predict College GPA the best. Make another table with these two variables and answer the questions based on this table.Numerical answers are rounded so choose the answer that matches the best:

5. What is the predicted College GPA for the student with High School GPA = 2.79, SAT = 900, and Letters = 8?

• a. 1.96
• b. 2.32
• c. 2.14
• d. 2.23

6. What is the standard error of this prediction? Give the interval prediction.

• a. 0.587; [1.64, 2.82]
• b. 0.587; [1.56, 2.73]
• c. 0.587; [1.74, 2.91]
• d. 0.587; [1.37, 2.55]

7. What is the number, -0.08813, and what is its interpretation?

• a. X-intercept: It shows how much college GPA decreases when high school GPA and SAT are equal zero.
• b. Y-intercept: The number in this case offers no statistical interpretation.
• c. X-intercept: For every 10 points dropped on the SAT a student's college GPA decreases by 0.8813.
• d. None of the above.

8. If student A's High School GPA is 0.65 point higher than student B's, and on the other hand, student B's SAT score is higher than student A by 200 points, which of the two students will have higher predicted College GPA and by how much?

• a. Student B by 0.05 points
• b. Student B by 0.01 points
• c. Student A by 0.02 points
• d. Student A by 0.03 points