Logistic regression: In the continual effort of marketing analytics, Starbucks is monitoring patterns of customers purchasing specialty drinks by gender and time of purchase. The Specialty drinks...

Please see attachments. It's a logistic regression exercise of 5 small sub questions.


Logistic regression: In the continual effort of marketing analytics, Starbucks is monitoring patterns of customers purchasing specialty drinks by gender and time of purchase. The Specialty drinks dataset provides actual data from purchases made at the Harvard Starbucks store in spring 2013. Four randomly selected days sampled customer purchases morning and afternoon, recording the gender of the person making the purchase. The following logistic regression output was performed to determine the likelihood of a specialty drink purchase given the customer’s gender and whether the purchase was made in the morning hours or afternoon hours. Logistic regression results Dependent Variable: CoffeeCode (Success = 1) Independent Variable(s): GenderCode, TimeCoded Parameter estimates Variable Estimate Std. Err. Zstat P-value Odds Ratio 95% Low. 95% Up. Intercept -0.4055 0.3064 -1.323 0.1857 GenderCode 1.084 0.2474 4.382 <0.0001 2.957="" 1.820="" 4.801="" timecoded="" 0.584="" 0.2953="" 1.979="" 0.0479="" 1.794="" 1.006="" 3.200="" test="" that="" all="" slopes="" are="" zero="" statistic="" df="" value="" p-value="" g="" 1="" 23.817651=""><0.0001 log-likelihood = -206.33824 hosmer-lemeshow goodness-of-fit test statistic df value p-value hl-gof 2 1.8179331 0.4029 steps in performing a logistic regression: 1) find the estimated logistic regression equation: ln(estimated odds ratio) = bo+ b1(purchases last year)+ b2(extra cards) 2) calculate the estimated odds ratio estimated odds ratio = e(bo+ b1x1+ b2x2) 3) estimated probability of an event of interest: estimated probability of an event of interest = a. find the estimated logistic equation to predict whether a customer will purchase a specialty drink based on gender and time of day. b. interpret the estimated coefficient for either gender or time produced by the model. c. based on this model which combination of gender and time of day will produce the highest likelihood of a customer purchasing a specialty drink? d. predict the probability of a female customer purchasing a specialty drink in the afternoon. coffeespecialtytotal female6195156 morning424991 afternoon194665 male392564 morning23730 afternoon161834 total100120220 1 estimatedoddsratio estimatedoddsratio + log-likelihood="-206.33824" hosmer-lemeshow="" goodness-of-fit="" test="" statistic="" df="" value="" p-value="" hl-gof="" 2="" 1.8179331="" 0.4029="" steps="" in="" performing="" a="" logistic="" regression:="" 1)="" find="" the="" estimated="" logistic="" regression="" equation:="" ln(estimated="" odds="" ratio)="bo+" b1(purchases="" last="" year)+="" b2(extra="" cards)="" 2)="" calculate="" the="" estimated="" odds="" ratio="" estimated="" odds="" ratio="e(bo+" b1x1+="" b2x2)="" 3)="" estimated="" probability="" of="" an="" event="" of="" interest:="" estimated="" probability="" of="" an="" event="" of="" interest="a." find="" the="" estimated="" logistic="" equation="" to="" predict="" whether="" a="" customer="" will="" purchase="" a="" specialty="" drink="" based="" on="" gender="" and="" time="" of="" day.="" b.="" interpret="" the="" estimated="" coefficient="" for="" either="" gender="" or="" time="" produced="" by="" the="" model.="" c.="" based="" on="" this="" model="" which="" combination="" of="" gender="" and="" time="" of="" day="" will="" produce="" the="" highest="" likelihood="" of="" a="" customer="" purchasing="" a="" specialty="" drink?="" d.="" predict="" the="" probability="" of="" a="" female="" customer="" purchasing="" a="" specialty="" drink="" in="" the="" afternoon.="" coffeespecialtytotal="" female6195156="" morning424991="" afternoon194665="" male392564="" morning23730="" afternoon161834="" total100120220="" 1="" estimatedoddsratio="" estimatedoddsratio="">