Math Home work

· This exam has four parts, which involve utilizing multiple analysis techniques to explore a human resources problem.

· The overall goal of this exam is to demonstrate your understanding of the key topics in this course and that you can apply them in a real-world situation.

· Worth: 25% of total mark for the course.

· Due: Wednesday, April 20th by 11:59pm

· Late submissions will not be accepted. Extensions will not be granted except in extreme circumstances. If you submit late, expect to receive 0% on the exam.

1. Most important rule: This is an individual exam. It must be completed by you and only you. The work you submit must be your own work. The normal expectations of students completing an exam apply. 

· You are not allowed to ask for help from another human, show your work to another human (other than when you submit to your instructor), or in any way gain any form of assistance from another human. 

· You cannot pay someone to do this exam for you. You cannot go to a website and share the document, get possible answers and use them in any way. You cannot go to a website, look for the exam, get possible answers and use them in any way.

· Communication about any portion of the exam in any way with any person other than Your instructor is strictly prohibited. 

· Asking for any kind of help for this exam is strictly prohibited. Think like you are writing an in-person exam. You couldn’t lean over to your buddy and ask them how to make a histogram. So you can’t do it on this exam. 

· Note: If two people submit work that is strangely similar, there is a very good chance that both students will receive 0 and will be sanctioned with academic misconduct.  Further to this, I will pay special attention to students who have worked together in the past to see if their answers are similar. 

2. It is “open” book. This means you can use any resource (other than a human) you want to complete the submission. This includes the textbook, course notes, course videos, and internet sites (excluding homework help sites like Chegg and CourseHero).

NOTE: If you directly use an outside source (e.g., paraphrase or quote), you still need to do a proper APA citation. What is described below is only for outside sources that you looked at for help and not for directly writing your final work. 

As I am absolutely convinced that most of you are using outside sources and failing to cite them, let us make it easy. If you look at a website outside of the class (i.e. not our textbook or from BB), insert the URL in the table below, state which part of the exam you used it for (1, 2, 3 ,4), and very briefly how you used it. In the first row, I’ve provided an example of what I’m expecting. Please delete it before submitting. 

If you claim you used no outside sources OR you directly used all outside sources, instead of submitting the above table, include the following sentence in your final work (see submission guidelines for where): 

“I, [insert full name], solemnly swear that I did not use any outside source (except those properly cited using APA referencing) to complete this exam. I understand that failure to indicate outside sources is an act of academic misconduct and could result in getting 0 on this exam.” 

The Craybill Instrumentation Company produces highly technical industrial instrumentation devices. The company has 45 sales regions, each headed by a sales manager.

The human resources (HR) director has the business objective of improving recruiting decisions concerning sales managers. The HR director determined that the primary method of evaluating the effectiveness of recruitment is the hire’s resulting “sales index” score, which is the ratio of the regions’ actual sales divided by the target sales. The target values are constructed each year by upper management, in consultation with the sales managers, and are based on past performance and market potential within each region.

At the time of their application, candidates are asked to take the Strong-Campbell Interest Inventory Test and the Wonderlic Personnel Test. The former test measures the applicant’s perceived interest in sales, while the latter measures their perceived ability to manage. For both tests, the higher the score the better. Due to the time and money involved with the testing, some discussion has taken place about dropping one or both of the tests.

The HR director decided to use regression modelling to predict the sales index (Sales) of the sales managers. To start, the HR director gathered information on each of the 45 current sales managers, including years of selling experience (Experience), and the scores from both the Strong-Campbell Interest Inventory Test (SCIIT) and the Wonderlic Personnel Test (Wonder). The attached Excel file contacted information on the 45 current sales managers.

Your goal is to perform analysis to determine: Can the sales index be predicted by the variables chosen by the HR director? If so, which variable or combination of variables is the most effective predictor.

Prior to doing the regression analysis, the HR director wants to get a sense of the quality of sales managers Craybill currently has. To do this, run an exploratory data analysis of the sales index and determine what story you want to tell about the sales managers.

· Goal: Answer the HR director’s question: “What is the quality of the current sales managers at Craybill?”

· How: Perform exploratory data analysis.

· Step 1: Run an exploratory data analysis (i.e., create visualizations and numerical summaries) for the sales index.

· Step 2: Choose one other variable (i.e., Wonder, SCIIT or Experience) to drill down into the sales index data. Run an exploratory data analysis (i.e., create visualizations and numerical summaries) for the sales index and your chosen variable.

· Step 3: Review your work and decide what story you want to tell about the quality of the current sales managers at Craybill.

· Step 4: Choose one visualization or set of numerical summaries (or both) from Step 1 and one visualization or set of numerical summaries (or both) from Step 2. Insert them in Table 1. Then write an answer/explanation to the HR director that explains the story and answers their question.

Build and evaluate a simple regression model to predict the sales index using the training set . To start, pick one of the three variables (Experience, SCIIT, or Wonder) that you believe will be useful at predicting the sales index.

· Goal: Build and evaluate a model to predict the sales index.

· How: Perform simple linear regression analysis:

· Step 1: Choose one variable as a predictor for the sales index.

· Step 2: Run the residual analysis to determine if the assumptions of regression are valid for this model.

· Note: Even if they are not all valid, continue with the analysis.

· Assume that ID#s are handed out in numerical order from earliest hire to most recent hire.

· Step 3: Build the regression model.

· Step 4: Evaluate the regression model using the testing data set.

· To evaluate the model, you’ll use something new called the root mean square error or RMSE. It is found by taking the average of the square of the residuals, and then square rooting the result.

· In general, the closer RMSE is to 0, the better the model is at making predictions. The larger the RMSE is, the worse the model is at making predictions. The RMSE is an absolute measure which means the size of the data needs to be considered when interpreting it.

· Step 5: Make predictions.

To do the above, fill in the following table. Leave the titles and labels alone. But anything in [] should be removed or replaced.

In Part 3a, you have done a thorough analysis for one possible predictor of sales index. Here, we want to address the HR director’s question of “Can the sales index be predicted by the variables chosen by the HR director? If so, which variable is the most effective predictor?”

· Goal: Determine if any of the chosen variables can predict the sales index. If so, which variable is the most effective predictor?

· How:

· Step 1: Run the regression analysis for each variable: Experience, SCIIT, or Wonder.

· Step 2: Compare the three models.

· Step 3: Write a paragraph that explains the results of the analysis to the HR director.

Build two multiple regression models using the training data set and determine which one is better. Each model needs to have at least two independent variables. Once you build your models, fill in the following table.

· Goal: Build two multiple linear regression models and compare them to determine which is better.

· How:

· Step 1: Choose the variables for Model 1. Run the regression analysis for this model using Excel. Plug in the necessary values into the table below.

· Note: For Multicollinearity, choose how you want to measure this.

· Step 2: Repeat Step 1 for Model 2 with different variables.

· Step 3: Compare both models using the results in Table 5. Determine which model is better. Then write a paragraph that presents your argument and provides evidence to support your decision.

Your submission needs to follow these guidelines. 

Submit one Word doc or PDF file and one Excel file. 

The Word doc needs to follow this format:

· Section 1: Filled in Outside Sources table OR the statement that you did not use any outside sources. If you have used any sources directly in your work, insert the reference list here (in addition to the table). 

· Section 2: Table 1

· Section 3: Table 3

· Section 4: Table 4

· Section 5: Table 5

· Section 6: Table 2

See the Word document “Final exam submission (Word).docx” for the required layout.

The Excel doc should have a minimum of five sheets that demonstrate your computations to complete the exam. Do not provide explanations or any additional information. You are providing the Excel spreadsheet simply so we can see your work. Your spreadsheet needs to include the following sheets:

· “Part 1”: Your exploratory data analysis

· “Part 2”: Your training set and testing set

· “Part 3 – Experience”: Your regression analysis for Experience.

· “Part 3 – SCIIT”: Your regression analysis for SCIIT.

· “Part 3 – Wonder”: Your regression analysis for Wonder.

· Note: One of the tabs for Part 3 will include the in-depth analysis for Part 3a. While the other two will include a shorter analysis for Part 3b.

· “Part 4 - Model 1”: The regression analysis for model 1

· “Part 4 - Model 2”: The regression analysis for model 2

You can have more sheets than this but ensure that you have labelled the sheets appropriately to help your instructor find the information.

See the Excel document “Final exam submission (Excel).xlsx” for the required layout.

Here is how you’ll be marked on the final exam

· Superior performance– A+: The answer is correct, complete, and demonstrates a very strong understanding of the relevant course content.

· Excellent – A: The answer is correct, complete, and demonstrates a strong understanding of the relevant course content.

· Good - B: The answer is mostly correct and complete, with no errors or only small ones. Understanding of relevant course content is generally demonstrated. Above average performance.

· Satisfactory - C: The answer is mostly correct and complete, with either multiple small errors or a significant error. Basic understanding of course content is demonstrated.

· Marginal Performance - D: There is more than one significant error. The response suggests a lack of understanding of course content.

· Fail – F: There are multiple errors and overall, the answer does not demonstrate understanding of the course content.

· Not done: The component is missing.

Plagiarism is any act where you present work as your own when it is not. When plagiarism is found, a letter is sent to the Office of Student Conduct. 

Cheating is when you do something which gives you an unfair advantage over other students. 

When you submit anything at MRU with your name on it, you are stating that you are comfortable with all the work presented and you agree that it is your work. Make sure you review your work to ensure it is your own prior to submitting. 

Here are two common scenarios that I have seen in the past.