sophiasha / investigate_dataset Goto Github PK

The conclusions that were made by looking at the data focus on correlations. Exploring other information related to each variable, more concrete conclusions could be made. Using general knowledge about the Titanic and the results from the data, each variable that was compared could be logically considered to relate to each other.

Since the data was a sample of everyone on the Titanic, the conclusions may not be accurate to the entire data set. For future research in order to determine this, a T-Test would need to be calculated since the population standard deviation is not known. If each value is calculated as significant, then the conclusions would be accurate

Cabin had too many missing values and any results from the data might not be accurate for the whole set, so it was omitted. The only real conclusions to be drawn would be if it was assumed the beginning letter was for each floor, then determine if higher floors had better survival. Logically, more expensive floors would be higher up and for high class people, so Cabin number was not needed even if there was no null values. Instead, Pclass and Fare logically covered any results that cabin number would lead to Age was used even though there was a decent amount of missing values. This could cause some inaccuracies in the analysis since random values were used, but ultimately the trends in the graphs were good enough to draw conclusions from.

Based on the plots by age and by passenger class, it was clear that no matter what sex or class the children still had a high chance of survival. The Age plot split by sex also clearly showed that the majority of survivors were female. First class passengers also clearly had priority over other people. So it is safe to say that the women, children, and first class were most likely to survive