Titanic Case Study

1707 Words4 Pages

Passengers of the Titanic
The relationship between the median ticket fare cost and whether the passenger survived or not.
Question:
I wonder if there is a difference in the median ticket fare cost (in the time’s British pound) between passengers that died and passengers that survived on the Titanic from a sample of the passenger data from the Titanic when it sank in 1912.

The purpose of this question is to investigate whether passengers who paid more for their ticket had a better chance of surviving. I did some research into this to see if there is a difference between the median ticket cost of people who died compared to who survived on the titanic.
I choose these variables because I believe that depending on how much a passenger spent could …show more content…

This means they may offset the average value for this data had I chosen the mean over the median. Therefore, ultimately the median in this scenario will be a more accurate representation of the average. iNZight’s ‘Visual Inferencing Tool’ will be what I use to display using the data. It will present this as a box and whisker graph. I will then analysis the data distribution discussing skews, inter-quartile range, range, shape etc. I will make a first judgement based on what I see being presented. From there I will create a difference between medians bootstrapping confidence interval, this is so I can produce reliable intervals that will potentially provide evidence for my question. iNZight will also be used for all statistics and …show more content…

This bootstrap confidence interval was made using iNZight’s Bootstrapping Confidence Intervals VIT Module. This was set to do 1000 re samples which included 1000 bootstrap distributions in order to create those confidence intervals above that show with 95% certainty that the difference between the medians for ticket fares for the passengers who survived compared to the passengers that died will fall within this range. I did 1000 resamples because this creates a more accurate confidence interval than a few amount of resamples would’ve done.
The bootstrapping confidence interval created a range of £11.5 < MSurvivors – Mdied < £18.5. Due to this confidence interval not contain £0 it can be concluded with 95% accuracy that the difference between the medians of the real population will fall within this confidence interval. This is similar to what I expected from my analysis. This brief use of the bootstrapping confidence interval to create a statistical analysis

Bibliography (References):

Open Document