The “Unusual Episode” Data Revisited

Survival on the Titanic: Illustrating Wald and Lagrange Multiplier Tests for Proportions and Logits

“…We are not the first to realize that this event attracts students' attention. Dawson (1995) and Simonoff (1997) used cross-tabulations and logistic regression to relate the Titanic's passengers survival rate to age, gender, and economic status. To maintain student interest, they presented their results without disclosing the nature of the disaster and asked students to try to identify the historical event that put these people at risk.…”

Graphical Methods, Analytic

“…A mosaic plot in its original form is constructed by dividing up the horizontal and vertical axes in turn by each new variable. Figure 9 shows a diagram for the Titanic data set (Dawson, 1995). Initially the horizontal axis was divided up according to the numbers in each class, then the individual columns were divided vertically by age, then the resulting cells were each divided horizontally by gender.…”

Logistic Regression, Categorical Predictors, and Goodness-of-Fit: It Depends on Who You Ask

“…Consider the data summarized in Table 1, which were the original motivation for this article. The data give the survival percentages, separated by economic status (first class passengers, second class passengers, third class passengers, or crew), age (adult or child), and gender (male or female), for the 2,201 people on board the ocean liner Titanic when it struck on iceberg and sank on April 15, 1912(Dawson 1995. Logistic regression can be used to model these data through the use of indicator variables or effect codings.…”