STAT 302

Include this page as a cover for your assignment. Last version Jan 17, 10:15am.

Last Name:

First Name:

Student ID:

• Due Friday January 25th, 2013 at 4:30pm in the boxes outside the Statistics Workshop.

• Clearly identify your assignment including ﬁrst name, last name and student ID, otherwise you may be deducted points.

• Justify all your answers and show all work. When applicable write down any formula(s) you are using and show all work. No answers which include only a ﬁnal number will be given credit.

I acknowledge that the answers on this assignment are my own and that I haven’t incurred in any type of academic dishonesty.

Sign here:

1

Question 1.

(Based on problem 8 chapter 5 from the textbook.) The data set ch05q08.jmp (available on webct) contains data on starting salaries (SAL) of a group of 30 college graduates who have recently entered the job market, along with their cumulative grade point averages (CGPA).

A)

Using JMP, ﬁt a regression model that explains SAL with CGPA. Use the Fit

Model function with the Minimal Report emphasis as explained on page 34 of Regression using JMP. Do not use the Fit Y by X function.

Label the axes of the scatterplot and provide a suitable title that also includes your ﬁrst and last names in parenthesis, for example ”Nice graph (Jorge Rodr´ ıguez)”. Display the

”Regression Plot”, ”Summary of Fit” and ”Parameter Estimates” options (close all other parts of the output), PRINT this output and submit it with your assignment. Your name has to be directly typed into the output. Missing or handwritten names will produce a score of 0 for this question.

B)

CGPA.

Write down the least squares equation that relates SAL (dependent variable) and

C)

Provide the sample means and sample variances of SAL and CGPA. No need to do it by hand. Use JMP to obtain these values.

D)

How many more dollars can a graduate expect to receive on average as a starting salary for each 0.1 increment in his/her CGPA? Also provide a 95% conﬁdence interval for this average increment in starting salary.

E)

Is the additional starting salary for each unit increment in CGPA signiﬁcant? Conduct a test of hypothesis and show all four steps. You can use JMP to help you with some of the calculations. We are considering a whole unit increment, not 0.1 as for the previous part.

F)

What is the expected starting salary for a graduate with a CGPA of 3.1?

G)

Request 95% conﬁdence and prediction intervals for all the values of CGPA in the sample. Create a new column that has as title your ﬁrst and last names (this column won’t have any values, only dots will be displayed on its cells).

PRINT the section of the spreadsheet with the values of SAL, CGPA, the intervals and the last column with your name for all 30 observations. You should have 7 columns. It is possible to display all this information in only one page. If some columns are very wide, reduce their width.

H)

Calculate by hand an interval that with probability of 95% includes the starting salary of a newly and randomly chosen graduate with a CGPA of 3.5. Even though you need to calculate this by hand, some of the required elements can be obtained with JMP.

2

Question 2.

(Based on problem 20 chapter 5 from the textbook.) Use data set ch05q20.jmp (available on webct) to solve the following questions. Researchers have studied the ecology of ponds in rural Bangladesh. Of particular interest in such studies are the zooplankton, phyloplancton and copepod counts (per liter of water) in these ponds.

Copepods are a particular type of zooplancton that are thought to be a natural reservoir of cholera bacteria, but copepod counts can be diﬃcult to obtain in laboratory analyses; total zooplankton counts are easier to obtain.

A)

Using JMP, ﬁt a regression model that explains copepod counts with zooplankton counts. Use the Fit Model function with the Minimal Report