The purpose of this case assignment is to apply logistic regression analysis to a human resources analytics problem, determine solutions for addressing organizational challenges, and communicate recommendations to organizational stakeholders. Problem Top Industries includes several production facilities in selected mid-sized cities in the upper Midwest. The company’s core product is a grain sweep used to clean out the bottom of multistory grain storage bins. Since customer orders have been steady and relatively predicable over the years, Top Industries has a voluntary overtime policy. However, in anticipation of increased business opportunities for the company, the director of human resources has been interested in knowing if Top Industries production employees would be willing to work paid overtime. On an annual basis, Top Industries administers an anonymous corporation-wide survey to its employees in an effort to understand employee sentiment regarding wages, work environment, management, etc. Interestingly, questions related to “employee willingness to work paid overtime” has not historically been part of the survey. Employees were asked if they had actually worked overtime at Top Industries during the past year. Some of the questions on the survey relate to employee job satisfaction, employee satisfaction with immediate manager, and employee identification with the organization. The director of human resources recently came across a study that actually reported that these sentiments are generally related to employee willingness to work overtime. Top Industries has three production sites of interest for this case study: 1. Site A is a production facility that has significant production activity. Despite the corporate voluntary overtime policy, about half the workers at this site voluntarily work paid overtime. Site A is located in a different city from Site B and Site C. 2. Site B is a production facility that is located about 5 miles from the new production facility, Site C. Site B has a steady production operation, and, consequently, no overtime has historically been needed from its production workers. However, the director of human resources will be asking for volunteers to transfer to Site C. 3. Site C is a new production facility, located about 5 miles from Site B. It is anticipated that paid overtime work will be needed at Site C. The director of human resources would like to know if employees from Site B (where historically little overtime has been needed) would be willing to transfer to Site C (where it is anticipated that overtime work will be needed). At first glance, the director of human resources thought that a simple survey of Site B workers, asking them if they would be willing to transfer to Site C and work paid overtime, might suffice. However, the director of human resources was concerned with response bias. Furthermore, the director would rather have workers who work paid overtime really want to work paid overtime. Thus, the thought was to use the Site A data to build/validate a model to predict employee willingness to work paid overtime. Since employee willingness to work overtime is a “yes” or “no” variable, it has been deemed that a logistic regression model will suffice. Since the annual corporate survey data related to employee job satisfaction, employee satisfaction with immediate manager, and employee identification with the organization are available, this information can be used to build a model to understand employee willingness to work paid overtime. The fact that the Site A data also indicates if the employee worked paid overtime makes it possible to build a model. Therefore, the Site A data are to be used to build (i.e., train) and validate a logistic regression model to gain insight into the employee willingness to work paid overtime. The data available for Site B contain the annual corporate survey data related to employee job satisfaction, employee satisfaction with immediate manager, and employee identification with the organization. Note that since employees at this site have not historically needed to work overtime, no overtime data is available. Once the logistic regression model for employee willingness to work paid overtime has been developed based on the Site A data, the Site B data will be applied against the model. It will then be possible to predict employee willingness of Site B production workers to work paid overtime. Logistic Regression Analysis The “Human Resources Case Study Data” file contains the Site A and Site B data. Review the “VariableDefinitions” sheet in the file, which describes the coding for the datasets. This case assignment has three primary tasks: 1. Create the logistic regression training/validation model based on the Site A data. 2. Predict employee willingness of Site B production workers to work paid overtime based on the training/validation model. 3. Discuss/summarize the results in a formal PowerPoint presentation file and provide a working IBM SPSS Modeler *.str file. Use IBM SPSS Modeler to create (i.e., train) and validate a logistic regression model based on the Site A data. Using the Site A data from the “Human Resources Case Study Data” file, perform a logistic regression analysis to build a logistic regression model for overtime. Note that “JobSat,” “SatisMgr,” and “OrgIdent” are the independent (i.e., predictor) variables. The Modeler nodes needed, in order of sequence, are: Excel Source Node, Partition Node, Type Node, and Logistic Node. For selected nodes, please note the following mandatory settings: 1. Partition Node: Specify “Train and test.” Specify 80% for the Training partition size; specify 20% for the Testing partition size. Check the “Repeatable partition assignment” box and click on the “Generate” button. This will ensure that the training and validation data partition will remain constant each time the model is run. 2. Logistic Node: In the “Fields” tab, ensure that the “Partition” field is specified in the Partition area. In the “Model” tab, check the “Use partitioned data” box and specify “Backwards Stepwise” as the method. In the “Expert” tab, ensure that the “Classification cutoff” is set to 0.5. Select additional items in this tab as needed for the assignment. 3. After running the model and creating a Model Nugget Node, connect an Analysis Node to the Model Nugget Node. In the “Analysis” tab, ensure that the “Separate by partition” box is checked. 4. Run the Analysis output node. 5. Take note of the model results. After creating the logistic regression model above, use IBM SPSS Modeler to predict employee willingness of Site B production workers to work paid overtime, based on the Site B data. The Modeler nodes needed, in order of sequence, are: Excel Source Node, Model Nugget Node (copied from the model developed using the Site A data), and Table Node. For selected nodes, please note the following: 1. Excel Source Node: Reference the Site B data in the “Human Resources Case Study Data” file. 2. Model Nugget Node: This node is copied from the model developed using the Site A data. 3. Table Node: When run, this node will contain the results of the prediction. 4. Take note of the classification (i.e., predicted) results. Write a two page paper that summarizes the setup and results of your logistic regression analysis. The analysis setup portion for the Site A Training/Validation model summary should include the following and provide screenshots to illustrate the content: 1. Indicate which predictors best explain overtime and why. 2. Final logistic regression equation (stated in mathematical form). 3. An assessment of the overall model efficacy and its ability to classify overtime. 4. The results of the Validation of the Training model. 5. Additional information as applicable to enhance the summary of the findings. For the analysis setup portion of the paper for the Site B Prediction summary, address the following: 1. Identify which specific Site B employees were classified as being willing to work overtime. 2. Calculate the overall percentage of Site B employees that would be willing to work overtime. Include the results of each analysis in your paper. The use of graphs, charts, and supporting data is required. Interpret the results of each analysis and draw general conclusions from the results. Make recommendations for the organization and address the organizational challenges that may be encountered based upon your recommendations. Organize the Paper in the following way (incorporate the content-specific information described above accordingly): 1. Introduction and case background. 2. Objectives for each analysis. 3. Approach or method of analysis and justification for selecting the approach or method. 4. Results of each analysis. 5. Supporting graphs, charts, data, and spreadsheets for each analysis 6. Interpretation of the results for each analysis. 7. General conclusion of each analysis. 8. Recommendation to the organization, including addressing organizational challenges that may be encountered based upon the recommendation. 9. Description of additional human resources analytics approaches that might useful for Top Industries. Submit the word doc and completed IBM SPSS Modeler *.str file to me.
The post Human Resources Analytics Case Study
Assignment status: Solved by our experts