b) Write out the R code for how you would get the estimates of this model. c) The following is the output for the model: # # ## Coefficients: # # Estimate Std. Error z value Pr(>|z|) ## (Intercept) 1. 5220 2.3142 0.658 0. 5108 ## PriceCH 2.2321 1. 4682 1.520 0. 1285 ## PriceMM -2. 7813 1. 1042 -2.519 0.0118 . ## Store7Yes -1. 2020 0. 2804 -4.286 1.82e-05 .# ## - -- ## Signif. codes: 0 '.*#' 0.001 '.*' 0.01 '.' 0.05 ' .' 0.1 ' ' 1 # # ## (Dispersion parameter for binomial family taken to be 1) # # # # Null deviance: 429. 17 on 320 degrees of freedom ## Residual deviance: 398.62 on 317 degrees of freedom ## AIC: 406.62 # # ## Number of Fisher Scoring iterations: 4 Write out the full equation of the fitted model, with the given coefficient estimates, Hint: for this model do we model the response directly, i.e. " - Bo + Biz + ...? d) Which predictors appear to have a significant association with the response variable? e) For the predictors that are significant, how would you generally describe the relationship with the response? That is 'as (predictor name increases, the .., increases / decreases. f) How would you interpret the following output where we are predicting if they will purchase the Minute Maid orange juice given the price for Minute Maid is 2.10 and the sale is at store 7. oj. glm2 = glm(Purchase - PriceMM + Store7, data - samp.oj . family = "binomial") predict (oj.glm2, newdata = data. frame (PriceMM - 2.10, Store7 - "Yes"), type = "response") # # ## 0.2020863