Week 7 Case Study (Case Study 6) The QuickFix manager was very pleased with your final model to predict vehicles served He decides to hire you for a second project He says that they would like to develop a model to refine who they interview for repair specialists They are opening several new locations and the interview process is timing consuming He'd like you to develop a model to restrict the pool of applicants to those individuals who they would be most likely to hire In other words, he liked to limit his interviews to quality candidates He provides you with a data file of previous interviewees The data file includes information on whether the interviewee was hired as well as three variables that might be useful predictors for developing the model The variables are as follows Hired (Interview) This is the binary decision of whether a previous interviewee was hired (1 Candidate was Hired, 0 Candidate was Not Hired) It is the response variable that you need to predict It is assumed that if a model predicts a value of 1, the candidate should be interviewed and if a model predicts a value of 0, the candidate should not be interviewed Your job is to develop two models using the available data, determine which model is a better fit, and make a recommendation about whether the model fit is good enough to meet the manager's needs Assessment Score This is a hands on diagnosis and repair performance task that all applicants must complete It is designed to assess their applied repair knowledge and skill Possible scores range between 0 and 50 It is one of three predictor variables for the model Years of Exp This is the candidate's years of experience as a repair specialist It is one of three predictor variables for the model ASE Certified The National Institute for Automotive Service Excellence (ASE) certifies automotive technicians who have successfully passed a series of exams and have at least two years of relevant hands on work experience (1 Candidate is ASE Certified, 0 Candidate is Not ASE Certified) ASE certified technicians are easily recognizable by a shoulder patch with the Institute's logo ASE Certified is a dummy variable and is one of three predictor variables for the model You will use the Interview worksheet in the QuickFix Interview Case Study Data xlsx workbook for this case study You decide to perform the following analyses to develop a predictive model Linear Probability Model Run a Linear Probability Model to predict Hired (Interview) using the Assessment Score, Years of Exp , and ASE Certified variables using Excel Label your results in an Excel workbook using the prompt number Write the regression equation for the Assessment Score, Years of Exp , and ASE Certified model using the variable names, intercept coefficient, and slope coefficients from the regression output Write your answer in the box below Is there individual significance for each variable of the Linear Probability Model assuming alpha 0 05 Write your answer in the box below Use the Linear Probability Model to calculate the predicted Hired (Interview) probability for all previous interviewees in the data set Remember that you can do this manually in Excel or you can get predicted values using the Residuals option for regression in Excel Label your results in an Excel workbook using the prompt number Convert the predicted Hired (Interview) probability to a predicted binary decision (1 Candidate should be hired interviewed, 0 Candidate should not be hired interviewed) for all previous interviewees in the data set using a success probability cutoff of 0 50 Label your results in an Excel workbook using the prompt number Create a confusion matrix for the Linear Probability Model using the actual Hired (Interview) decision and the predicted binary decision from the Linear Probability Model Label your results in an Excel workbook using the prompt number What percentage of the actual hired not hired decisions were correctly predicted by the Linear Probability Model Write your answer in the box below Logistic Regression Model Run a Logistic Regression Model to predict Hired (Interview) using the Assessment Score, Years of Exp , and ASE Certified variables using R Label your results in an Excel workbook using the prompt number Write the Logistic Regression equation for the Assessment Score, Years of Exp , and ASE Certified model using the variable names, intercept coefficient, and slope coefficients from the regression output Refer to Section 17 3 of the textbook for examples of logistic regression equations Write your answer in the box below Is there individual significance for each variable of the Logistic Regression Model assuming alpha 0 05 Write your answer in the box below Use the Logistic Regression Model to calculate the predicted Hired (Interview) probability for all previous interviewees in the data set Remember that you can do this manually in Excel or you can get the predicted values from R Refer to Section 17 3 of the textbook for the formula for calculating predicted probabilities for a logistic model Label your results in an Excel workbook using the prompt number Convert the predicted Hired (Interview) probability to a predicted binary decision (1 Candidate should be hired interviewed, 0 Candidate should not be hired interviewed) for all previous interviewees in the data set using a success probability cutoff of 0 50 Label your results in an Excel workbook using the prompt number Create a confusion matrix for the Logistic Regression Model using the actual Hired (Interview) decision and the predicted binary decision from the Logistic Regression Model Label your results in an Excel workbook using the prompt number What percentage of the actual hired not hired decisions were correctly predicted by the Logistic Regression Model Write your answer in the box below Comparing the Linear Probability Model to the Logistic Regression Model Using the results of confusion matrices from the Linear Probability Model and Logistic Regression Model, which model is the better fit Provide the measure that you used to make your decision Write your answer in the box below In the box below, write a concise summary report for the general manager Your report should include a description of the methodology, the results, and conclusions recommendations Two to three sentences each for methodology, results, and conclusions recommendations should be sufficient Your recommendation to the general manager should include whether the best model from Case Study 6 is likely to be useful for making predictions for his intended business purpose To make the recommendation, consider the likelihood of correctly predicting the candidates that should be hired with the better fitting model Write your answer in the box below

Question

Week 7 Case Study (Case Study  6) The QuickFix manager was very pleased with your final model to predict vehicles served  He decides to hire you for a second project  He says that they would like to develop a model to refine who they interview for repair specialists  They are opening several new locations and the interview process is timing consuming  He'd like you to develop a model to restrict the pool of applicants to those individuals who they would be most likely to hire  In other words, he liked to limit his interviews to quality candidates  He provides you with a data file of previous interviewees  The data file includes information on whether the interviewee was hired as well as three variables that might be useful predictors for developing the model  The variables are as follows  Hired (Interview)   This is the binary decision of whether a previous interviewee was hired (1 Candidate was Hired, 0  Candidate was Not Hired)  It is the response variable that you need to predict  It is assumed that if a model predicts a value of 1, the candidate should be interviewed and if a model predicts a value of 0, the candidate should not be interviewed  Your job is to develop two models using the available data, determine which model is a better fit, and make a recommendation about whether the model fit is good enough to meet the manager's needs  Assessment Score   This is a hands on diagnosis and repair performance task that all applicants must complete  It is designed to assess their applied repair knowledge and skill  Possible scores range between 0 and 50  It is one of three predictor variables for the model  Years of Exp   This is the candidate's years of experience as a repair specialist  It is one of three predictor variables for the model  ASE Certified   The National Institute for Automotive Service Excellence (ASE) certifies automotive technicians who have successfully passed a series of exams and have at least two years of relevant hands on work experience (1 Candidate is ASE Certified, 0  Candidate is Not ASE Certified)  ASE certified technicians are easily recognizable by a shoulder patch with the Institute's logo  ASE Certified is a dummy variable and is one of three predictor variables for the model  You will use the Interview worksheet in the QuickFix Interview Case Study Data xlsx workbook for this case study  You decide to perform the following analyses to develop a predictive model  Linear Probability Model Run a Linear Probability Model to predict Hired (Interview) using the Assessment Score, Years of Exp , and ASE Certified variables using Excel   Label your results in an Excel workbook using the prompt number  Write the regression equation for the Assessment Score, Years of Exp , and ASE Certified model using the variable names, intercept coefficient, and slope coefficients from the regression output  Write your answer in the box below  Is there individual significance for each variable of the Linear Probability Model assuming alpha   0 05  Write your answer in the box below  Use the Linear Probability Model to calculate the predicted Hired (Interview) probability for all previous interviewees in the data set  Remember that you can do this manually in Excel or you can get predicted values using the Residuals option for regression in Excel  Label your results in an Excel workbook using the prompt number  Convert the predicted Hired (Interview) probability to a predicted binary decision (1 Candidate should be hired interviewed, 0  Candidate should not be hired interviewed) for all previous interviewees in the data set using a success probability cutoff of 0 50  Label your results in an Excel workbook using the prompt number  Create a confusion matrix for the Linear Probability Model using the actual Hired (Interview) decision and the predicted binary decision from the Linear Probability Model  Label your results in an Excel workbook using the prompt number  What percentage of the actual hired not hired decisions were correctly predicted by the Linear Probability Model  Write your answer in the box below  Logistic Regression Model Run a Logistic Regression Model to predict Hired (Interview) using the Assessment Score, Years of Exp , and ASE Certified variables using R   Label your results in an Excel workbook using the prompt number  Write the Logistic Regression equation for the Assessment Score, Years of Exp , and ASE Certified model using the variable names, intercept coefficient, and slope coefficients from the regression output  Refer to Section 17 3 of the textbook for examples of logistic regression equations  Write your answer in the box below  Is there individual significance for each variable of the Logistic Regression Model assuming alpha   0 05  Write your answer in the box below  Use the Logistic Regression Model to calculate the predicted Hired (Interview) probability for all previous interviewees in the data set  Remember that you can do this manually in Excel or you can get the predicted values from R  Refer to Section 17 3 of the textbook for the formula for calculating predicted probabilities for a logistic model  Label your results in an Excel workbook using the prompt number  Convert the predicted Hired (Interview) probability to a predicted binary decision (1 Candidate should be hired interviewed, 0  Candidate should not be hired interviewed) for all previous interviewees in the data set using a success probability cutoff of 0 50  Label your results in an Excel workbook using the prompt number  Create a confusion matrix for the Logistic Regression Model using the actual Hired (Interview) decision and the predicted binary decision from the Logistic Regression Model  Label your results in an Excel workbook using the prompt number  What percentage of the actual hired not hired decisions were correctly predicted by the Logistic Regression Model  Write your answer in the box below  Comparing the Linear Probability Model to the Logistic Regression Model Using the results of confusion matrices from the Linear Probability Model and Logistic Regression Model, which model is the better fit  Provide the measure that you used to make your decision  Write your answer in the box below  In the box below, write a concise summary report for the general manager  Your report should include a description of the methodology, the results, and conclusions recommendations  Two to three sentences each for methodology, results, and conclusions recommendations should be sufficient  Your recommendation to the general manager should include whether the best model from Case Study  6 is likely to be useful for making predictions for his intended business purpose  To make the recommendation, consider the likelihood of correctly predicting the candidates that should be hired with the better fitting model  Write your answer in the box below

SolutionInn · Accepted Answer

The Answer is in the image, click to view ...

Question: Week 7 Case Study (Case Study #6) The QuickFix manager was very pleased with your final model to predict vehicles served. He decides to hire

Step by Step Solution