AmbassadorGoose708
Please follow the instruction below and answer the question…
Please follow the instruction below and answer the question below.
DSCI 352 Assignment 4 – Regression
First, you will need to download the data file “employee-perf.csv” found here:
https://drive.google.com/open?id=14Ksw7uUdKDs_CLxqkCke5eLgadrGzp0l
Please place this file in a folder where you can access it with R and then open up R or R Studio.
Assume you own a software company. When you bring on new employees your hiring process is
1) to give them a technical aptitude test, 2) to bring them in for an interview (conducted by a team
of three hiring managers), and if preliminarily hired, 3) put them through a 6-week training
course. At the end of the course, you have the option to let the employee go with no further costs
or keep them on board. You keep a record of all your employees, including their interview scores
(an average score from 1-5 based on three hiring managers’ input from the interview), their
aptitude test score, the number of training classes they missed, and their annual performance
score. These records are found in the included file “employee-perf.csv”. You are hiring and want
to identify good candidates before having to keep them on the payroll for a full year (in order to
get a performance review).
To begin, first, set your working directory to where your “employee-perf.csv” is located and read
the data into a variable called employee.dat by executing the following command:
> employee.dat <- read.csv("employee-perf.csv")
1. Using the data in employee.dat, construct a linear regression model to predict the
Annual Performance Rating from all other relevant variables. DO NOT use any
variable selection methods such as stepwise selection, etc.
1) Place your regression model into a variable called employee.mod.
2) You may or may not need to leave out certain variables (i.e. columns) that are
irrelevant. (This does not mean you need to find the best combination of variables,
only that some columns may clearly be irrelevant).
3) In your model, make sure to list any variables (columns) in the same order they
appear in the spreadsheet. Also, don't forget to replace spaces ' ' with dots '.' when
specifying the variables in your model (i.e. Interview Score becomes
Interview.Score).
4) Please use either NO SPACES OR CONSISTENT SPACING. If you use
spaces, make sure to use no more than one space between characters.
(10 points)
Please copy and paste your R command to build the model and assign it to
employee.mod below (this should be a single line of R code using the lm
function).
Your input should be along the following lines:
employee.mod <- lm(...
2. In the model you created (employee.mod), which single metric tells you whether
your model has any predictive validity?
(10 points)
? The coefficient
? The p-value corresponding to the intercept
? The p-value corresponding to the F-statistic
? The adjusted R-squared
? The F-statistic
? The multiple R-squared
? The p-value corresponding to the coefficient
? The intercept
3. Specify the value in your model of the metric you identified in question 2. (10 points)
4. Does your model have any predictive validity? (10 points)
? Yes, with a high degree of certainty
? Yes, with a moderate degree of certainty
? No, with a high degree of certainty
? Inconclusive
? No, with a moderate degree of certainty
5. How much variance in Annual Performance Rating is explained by your model?
Please round your answer to two decimal places. (10 points)
.Answer………………………………………………………………..
6. Suppose you have the following candidates that have just made it through the training course
and you can either retain them or let them go. Their information is as follows:
Candidates Aptitude test score IntervInterview scores Misses training classes
Billgate 84 3.55 6
Mark Zuckerberg 37 4.72 1
Sergey Brin 86 4.61 3
This information is contained in the following spreadsheet, which you can download and make
predictions on using R’s predict function:
https://drive.google.com/open?id=1yN8v21QefygI1gPEk2PT4LrLeZwUjfAt
What is the predicted Annual Performance Rating for Bill Gates? Round your answer to two
decimal places. (10 points)
Answers:…………………………………………………………………………………………….
7. What is the predicted Annual Performance Rating for Mark Zuckerberg? Round
your answer to two decimal places. (10 points)
Answers:………………………………………………………………………………………………..
8. What is the predicted Annual Performance Rating for Sergey Brin? Round your
answer to two decimal places. (10 points)
Answer:……………………………………………………………………………….
9. If you have to choose two to hire based on your model, who would they be? (10 points)
? Bill Gates
? Sergey Brin
? Mark Zuckerberg
10. If you have to choose one to fire based on your model, who would it be? (10 points)
? Bill Gates
? Mark Zuckerberg
? Sergey Brin
11. By examining the model results, which cost-cutting policies (if any) would you
recommend, and why?(10 points)
? Eliminate the training classes, because they show no significant impact on employee
performance
? All three measures have a significant impact on employee performance; don’t eliminate
any.
? Eliminate the interview, because it shows no significant impact on employee performance
? Eliminate the aptitude test and interview, because in combination they show no significant
impact on the employee performance
? Eliminate the aptitude test and training classes, because in combination they show no
significant impact on the employee performance
? Eliminate the aptitude test, because it shows no significant impact on employee
performance
