Q1) A recent study observed a random sample of 50 people who had lost their jobs, and recorded the number of weeks it took each person to find...

Q1) A recent study observed a random sample of 50 people who had lost their jobs, and recorded the number of weeks it took each person to find another job. The following variables were also recorded for each person in the sample.

1. Age (years) when job was terminated

2. Education (years of school + college/university) when job was terminated

3. Marital status (= 1 if married, 0 if not) when job was terminated

4. Head of household status (= 1 if head of household, 0 if not) when job was terminated

5. Tenure (years): how long spent working at firm, when job was terminated

6. Manager (= 1 if terminated job was in management, 0 if not)

7. Sales (=1 if terminated job was in sales, 0 if not)

You will find this data set in the Excel spreadsheet "Layoff Study Data".

Analyze the data and develop the best multiple regression model you can, to predict how many weeks a laid off employee will need to find a job. Use the following method (from chapter 9):

i. Begin with the full regression model containing all seven potential independent variables.

ii. Drop insignificant variables from the model, until only significant variables remain.

 Eliminate independent variables one at a time, by selecting the variable with the largest p-value in excess of 0.05. Re-estimate the regression model after dropping a variable, and repeat. Keep going until the p-values of all remaining independent variables are 0.05 or less. Show your model selection process, along with the final regression model chosen.

Which variables remain in your final model, i.e., which factors significantly affect the length (weeks) of unemployment before finding a new job? Why do you think each of these remaining variables might help explain how quickly a laid off employee finds another job?

For your final model, interpret clearly the meaning of: (a) the estimated value of R-Squared and (b) the estimated value of the regression coefficient of each independent variable remaining in the model.

How many weeks of unemployment does your model predict for a laid off person who is 37 years old, with 15 years of schooling, not married, is head of a household, and was employed for 8 years at his last job in a non-management position in sales.