For this problem, you will demonstrate that the deviance and Pearson 2 goodness of fit measures are not accurate when there are no replicates. For each setting, you will generate N = 1000 data sets, fit each using the correct binary logistic regression model, and save the deviance and Pearson X2 measures. You will then compare the distribution of each measure against the appropriate asymptotic 2 density. You will use the following model in all simulations: y|x Bernoulli(p) where x N(0, 1) and logit(p)= 0.35 x. Consider the following three settings: 1. n = 50 cases 2. n = 200 cases 3. n = 800 cases For each setting and goodness-of-fit measure, create a histogram and overlay the appropriate 2 density. Summarize what you find, making sure to address whether the distributions of deviance and Pearson X2 better fit the 2 density as the sample size increases