The entire assignment must be done in SPSS. If you know how to use SPSS running the model should not take very long! The purpose of this assignment is to perform logistic regression,interpret the results, and analyze whether or not the informationgenerated can be used to address a specific business problem.

For this assignment, you will use the "Adult Incomes" data set from the Topic Materials.

The marketing department is interested in creating advertisingdirected primarily at high-income individuals, and they it has come toyou seeking very specific customer data. The director of marketingexplains that individuals with large amounts of disposable income tendto purchase luxury items. Therefore, understanding what predictors arecorrelated with high income can be very useful for a marketingdepartment because it can help it tailor messages to the high-earningcohort. For example, individuals that earn capital gains tend to behigh-income earners, and advertisements for luxury items can be targetedtoward them on realty or investment websites.

As a member of the analytics team, you have been asked to determine alist of predictors and their relative impact on the likelihood of anindividual being a high-income earner. Individuals earning more than$50,000 annually are considered high income earners. In your summary,include discussion of how the marketing department can use your resultsto devise a smart advertising strategy.

Question 1: Partition the data to create atraining data set (70%) and test data set (30%). With a cut-off of 0.5,run logistic regression with "Income" as the target and the followingpredictors: "Capital_Gain," "Hours_Per_Week," "Sex," "Age," and "Race."Show the model summary and variables in the equation. Which probabilityis being modeled? Include the "Model Summary" and "Variables in theEquation" outputs when submitting the answer.

Question 2: Is "Race" a statisticallysignificant predictor when modeling whether incomes are greater than$50,000 annually? Explain your answer. Use a 5% significance level.

Question 3: Rerun the model without "Race"while still using a cut-off of 0.5. Show the model summary and variablesin the equation. Write the equation showing the probability as afunction of the predictors. Interpret the meaning of the coefficientsfor "Age" and "Sex." Include the "Model Summary" and "Variables in theEquation" outputs when submitting the answer.

Question 4: Given that approximately 26% ofthe individuals in the data have incomes greater than $50,000 annually,rerun the model in Question 3 with a cut-off of 0.26. Show theclassification tables and percent correct for each predicted outcome(>50K and <=50K) for the training data and test data. Why is thepercent that is correct usually lower when the test data are used?Include the "Training Classification Table" and "Test ClassificationTable" outputs when submitting the answer.

Question 5: Consider the followingindividual: Age=30, Sex=Female, Hours_Per_Week=40, Capital_Gain=$0.Based on the logistic model from Question 4, what is the probability ofthis individual earning more than $50,000 annually? What would be thepredicted class for this individual? Explain your answer.

Question 6: Based upon your analysis, whatare the predictors that can determine whether or not an individual wouldbe considered a high-income earner? Discuss how the marketingdepartment can use this information in formulating its advertisingstrategy? Present your findings in the form of a 250-word executivesummary that includes relevant data, charts, and tables to validate theconclusions presented.

General Requirements:

Submit the answers to Questions 1-5 and the executive summary as Word documents. The professor would like us to show how P=1/(1+e^-())- works into the work by plugging in our numbers from Variables in the Equation chart and Classification table created in SPSS modeler. (He explained this concept in class and I am just not grasping how these numbers work into the equation and how this equation works into the model! Hopefully after studying your completed work on this assignment I will be able to make better since of it!) He wants all charts from SPSS to be loaded into homework. Thanks for your help!