You are in charge of the marketing department. Your company offers insurance products via online and telemarketing channels. You are trying to improve the effectiveness of your telemarketing employees by examining the hiring practices. Specifically, you are trying to answer two questions:

1. Should we offer referral bonuses to our current employees?

2. A third party offers cognitive and personality testing (Specifically measures of Conscientiousness). Should we use those tests in our hiring process?

In order to examine these questions, you have collected referral, testing results, and other background information for your last 423 hires. The data you collected:

Sales:Employee Sales during their first month of employment (in $thousand; so 30.00 is $30,000).

Referral:Was the employee referred by another sales representative (0= non-referral; 1= referral)?

Months_Previous:Number of months the employee spent in their previous position.

Year_Education:Total years of formal education calculated from resume information.

Communication:Result of assessment measuring applicant communication ability. This measure has been in place for years and a higher score it better.

Years_Sales:Number of years of previous sales experience.

Cons.: Assessment of Conscientiousness based on third party testing (Conscientiousness is a dimension of personality). This is one of the assessments you are considering for the hiring process going forward. A higher score is better.

Cognitive:Assessment of Cognitive/Problem Solving ability based on third party testing. This is one of the assessment you are considering for the hiring process going forward. A higher score is better.

Please use this information to answer the following questions.

Now build a multiple linear regression using Sales as the Y and Cognitive, Cons., Years_Sales, Communication, Year_Education, Months_Previous, Referral as predictors. Instead of simplifying the model manually, run a forward stepwise regression with minimum BIC as the stopping rule and direction as forward. Include Sales as the outcome and all other variables as potential predictors. (Sales as the Y and Cognitive, Cons., Years_Sales, Communication, Year_Education, Months_Previous, Referral as predictors). Based on this model with no additional adjustments, please answer the following questions:

Questions:

1. Which predictors are dropped from the model?

2. All of the predictors in the model are statistically significant.

3. The resulting model explains more than a third of the variation in Sales.

4. The resulting model meets the assumptions of linear regression.

Answer set (Choose the best answer from here)

Cognitive, Cons., Years_Sales

True

False

Cons., Months_Previous

Referral, Year_Education

Cons., Years_Education

Run a second stepwise regression using Minimum AIC stopping rule, forward direction, Sales as the outcome and all other variables as predictors (Sales as the Y and Cognitive, Cons., Years_Sales, Communication, Year_Education, Months_Previous, Referral as predictors) Use the results with no additional adjustments to answer the following questions. (For the purposes of these questions, this model will be referred to as Model 2 and the previous model run using Minimum BIC and forward direction will be referred to as Model 1).

Questions

5. The r square of Model 2 is higher than Model 1. (True, False, Not Enough Information)?

6. All variables in Model 2 are statistically significant at a .05 level. (True, False, Not Enough Information)?

7. Model 2 has the same number of variables as Model 1. (True, False, Not Enough Information)?