Exploratory Graphing of Titanic Data

Today I engaged in some exploratory graphing of a dataset of survivors of the Titanic. Here are my findings.

This graph shows survivability by class and sex aboard the Titanic.

Two things I immediately note is that the difference between survivability between First- and Second-Class passengers that are also Female is very little: First Class Female passengers survived at 96% while Second Class Female passengers survived at 88%. My question from this data is whether those passengers (that are Second Class and Female) also came with some form of family or not. If they did, why did they allow their male counterparts to perish at significantly higher rates? If not, that would explain the significant difference between the two sexes graphed.

Next, I notice that Male passengers in the Second- and Third-Class survived at equally low rates. I’d like to investigate whether they suffered similar issues. Was the design of the ship equally as an issue for the Second Class’s escape from the sinking as it was for the Third Class?

This next graph shows survivability by age category and class. For this data graph, the following classifications were used as they closely related the norms of the period in which the incident took place.

Child: Birth – 12
Adult: 12 – 44
Midage: 45 – 59
Elderly: 60+
NA: Unknown age

This graph has interesting cross-sections. First, I notice that children in the Second Class survived at an astounding rate of 100%. Every single child of the Second Class survived. The percentage of children in the First Class drops to 80%. This is the only category where passengers of the Second Class survived at higher rates than the First Class: why? I can’t conjecture a single explanation for this phenomenon.

Another disturbance arises in the elderly class: more elderly Third-Class passengers survived than elderly Second-Class passengers. This imbalance is cleared for adult passengers though, so why did the elder working-class steerage passengers end up surviving over their slightly wealthier counterparts? How could they do so when their adult handlers (assuming they needed handlers) perished at traditional rates by class?

The information here is interesting and requires further research: which is why exploratory graphing exists.