Checking the relationship between the crash rates of first and second period to calculate if the risk is persistent and is predictable based on its crash history.
Introduction The case study was done to assess the safety of airlines based on its track record, using the airline safety data hosted by FiveThirtyEight, which provides the safety records of major commercial airlines, over a span of 30 years. The context of this analysis is the outcome of certain academic studies that high-profile crashes can shift passenger demand away from the airlines involved in the disasters. The dataset divides the 30-year period into two halves and gives information about the number of crashes, fatal accidents and fatalities occurred over these years along with available seat kilometers (ASKs) of 56 different airlines. The prime objective of the case study was to check whether there was a relation between the crash rates of first time period and that of the second which would imply that the risk is persistent and is predictable based on its crash history.
Furthermore, other factors within the limit of the dataset were analyzed and type of ownership of the airline was identified as a potential predictor. ‘Are nationalized airlines safer than privatized ones?’ emerged as a sub question for the analysis.
The report documents the work undertaken to find out if the safety of airlines could be assessed based on its track record and type of ownership. The information pertaining to the crash history of airlines, the steps taken to convert this information into a risk score which indicates the risk of flying in that particular airline and the inferences upon completion of exploratory data analysis are discussed in length.