- Q2 For The Titanic Data Set The Ordinal Logistic Regression Was Used To Predict The Likelihood Of Ticket Class Pclass 1 (19.34 KiB) Viewed 34 times
Logistic Regression Table Odds 95% CI Predictor Coef SE Coef Ratio Lower Upper Const(1) -6.25401 0.384759 Const(2) -3.84132 0.296997 Sex male -0.226486 0.193839 0.80 0.55 1.17 Age 0.0641692 0.0067241 1.07 1.05 1.08 Fare 0.103117 0.0073133 1.11 1.09 1.12 0.78 Ties Total Measures of Association: (Between the Response Variable and Predicted Probabilities) Pairs Number Percent Summary Measures Concordant 141882 88.9 Somers' D Discordant 17429 10.9 Goodman-Kruskal Gamma 0.78 312 0.2 Kendall's Tau-a 0.49 159623 100.0 a) Explain what does the concordant percentage (88.9%) mean and what does it say about the quality of the model? b) Are all three predictors significant in the model? If not, which one(s) is insignificant? c) Explain the meaning of partial slope of the gender variable. Which gender was more likely to gain higher-class ticket? d) Calculate the predicted probabilities: P(pclass=1), P(pclass=2) and P(pclass=3) for a female passenger whose age was 40 years and paid 30£ for the fare? e) According to the probabilities in part d), in which class the female passenger is more likely to be placed? f) Use your calculated probabilities in part d) to predict the class for a male passenger whose age was 40 years and paid 30£ for the fare?