In this exercise the students will compare a model that only contains position 1 to the regression model from DAX 4. They will then use this new model to test if the company discriminates by height. The students may refer to previous instructions for more details on how to run the analysis in SAS. A video is attached that demonstrates how to compare similar results. The conclusions will likely be the same as in the video, but the numbers will likely differ slightly. YOU MUST HAVE THE CORRECT CORRESPONDING NUMBERS TO YOUR DATA TO RECEIVE CREDIT.
This lab is meant to emphasize the evaluation of the model accuracy. There is a trend towards the increased use of machine learning which emphasizes model evaluation vs testing statistical samples. Regression is a commonly used model in machine learning.
Instructions
Use your labor data set to create a data set that contains only position 1.
Use this data set to run all analysis to test assumptions (summary descriptive statistics with max min, mean, etc, Regression Analysis Results Table that shows F, Pr, and R-square you produced in your analysis, correlation analysis ,and scatter plots of continuous variables (height, years of employment, and performance)
Note: An Example of Regression Analysis Results Table (Example Only – Not DAX 5 results) that must be submitted to prove a Regression Analysis was conducted. Example Regression Results Tables.pdfPreview the document
Run a linear regression of $/hr in SAS using Location as a classification variable and with performance and years of employment as continuous variables.
Run a linear regression of $/hr in SAS using Location as a classification variable and with height, performance, and years of employment as continuous variables.
Use the outputs to complete the memo. NOTE THE ASSUMPTION TESTS CAN BE LIMITED TO THE MODEL WITH HEIGHT.
Attachments
DAX 5 Video (Links to an external site.)
DAX 5 Memo 2020.docxPreview the document
Deliverable
Business Memo
Rubric
DAX5 (Fall 2020)
DAX5 (Fall 2020)
Criteria Ratings Pts
This criterion is linked to a Learning OutcomeBusiness Memo submitted on time.
10.0 pts
Full Marks
0.0 pts
No Marks
10.0 pts
This criterion is linked to a Learning OutcomeCorrect salutation, date, subject line, etc.
5.0 pts
Full Marks
0.0 pts
No Marks
5.0 pts
This criterion is linked to a Learning OutcomeParagraph 2 provides an explanation of any steps and discusses if the linear model is appropriate to run on this data. Tables and graphs should be referenced.
20.0 pts
Full Marks
15.0 pts
Mostly Correct
10.0 pts
Partially Correct
5.0 pts
Mostly Incorrect
0.0 pts
No Marks
20.0 pts
This criterion is linked to a Learning OutcomeAppropriate tables and graphs. This should include graphs/plots to verify the assumptions, a summary table, the parameter estimates table, and a correlation table.
20.0 pts
Full Marks
15.0 pts
Mostly Correct
10.0 pts
Partially correct
5.0 pts
Mostly incorrect
0.0 pts
No Marks
20.0 pts
This criterion is linked to a Learning OutcomeParagraph 3 summarizes the results discussing whether the model is significant, states which parameter estimates are significant and interprets them, discusses the root MSE, R-Squared, and if the results are meaningful. Tables and graphs should be referenced. Comparisons should be made between the DAX 4 and DAX 5 model and between a Model with and without height.
25.0 pts
Full Marks
19.0 pts
Mostly Correct
13.0 pts
Partially Correct
7.0 pts
Mostly Incorrect
0.0 pts
No Marks
25.0 pts
This criterion is linked to a Learning OutcomeParagraph 4 is a summary Paragraph with a conclusion, a recommendation, and the impact of the recommendation.
The conclusion is the plain language answer to the research question. It should also provide a recommendation to include position 2 in the model or not and for what action to take against a potential lawsuit for height discrimination.
20.0 pts
Full Marks
15.0 pts
Mostly Correct
10.0 pts
Partially Correct
5.0 pts
Mostly incorrect
0.0 pts
No Marks
20.0 pts
Total Points: 100.0
