Airlines or Snail-lines?

Investigation OverviewThis investigation was conducted to understand the impact of flight delays. When conducting my analysis, I compared the amount of on time flights against delayed flights, the reasons behind the delays, the effects of departure times as well as the performance of different carriers. To perform this investigation, I utilised univariate, bivariate and multivariate exploration which allowed me to understand trends within my analysis.

Dataset OverviewThe dataset used for my analysis was retrieved from the United States Bureau of Transportation Statistics (BTS) and includes flight data of carriers generating excess of 0.5% of total domestic travel in America. My dataset was refined to the 2019 calendar year. During this period there was 7.4 million(m) scheduled flights and 1.598m hours of delays.

On-time flight compared to delayed flightsWhen analysing flight status, I organised the data into three categories: on-time, delayed (arrival time at the gate > 14 minutes), cancelled or diverted. I discovered that 79.2% of flights were on time, 18.7% were delayed and 2.1% were cancelled or diverted. During 2019 there were 5.9m flights on-time, 1.4m flights delayed and 154k flights cancelled or diverted.

Distribution and proportion of delay types

With 1.598m hours of delays during 2019, it was important to understand the types of delays and how they were proportioned. Late aircraft delays, carrier delays and National Aviation System delays (NAS) made up 94% of the total delays. When comparing this to the distribution delays, late aircraft delays had the largest interqualtile range (IQR) of 34 minutes (IQR3 - IQR1) and 50% of their delays fall between 15 and 49 minutes. The distribution of carrier and NAS delays were both skewed to the right with IQR ranges of 26 minutes (7-33mins) and 22 mins (8-30mins).

When looking at the visualisation of the distribution of delays, you will notice the y axis ends at 105 mins. I chose to remove outliers from the data set to make the visual easier to interpret.

Effects of Departure Time

Departure time has a significant impact on overall delay time. During my investigation, I found that many flights were delayed even though they departed from the airport on-time or earlier. I decided to compare delayed flights that had left the airport on time against flights that were delayed before departing the terminal. Of the 1.4m delayed flights, 88% of flights are delayed due to them departing late with an average delay time of 75 minutes. Whereas, 12% of flights were delayed when they left on-time or ahead of schedule with an average delay time of 28 minutes. If a flight leaves the airport delayed, its delay time is 167% greater than flights that have left on time or earlier.

Airline Carrier Performance

In order to see how individual carriers performed, I needed to find the percentage of delayed flights, as well as, the average delay time. Based on the total number of flights, 18.7% were delayed with an average delay time of 69 minutes. Using this information, I found that Frontier Airlines Inc, JetBlue Airways and Atlantic Southeast Airlines had the highest percentage of delays (all three carriers were ~25% delayed). When looking at average delay time in minutes, Atlantic Southeast Airlines flights were delayed for 87 mins, Skywest Airlines inc were delayed for 85 mins and JetBlue Airways flights were delayed for 81 mins. When combining these two variables, Atlantic Southeast Airlines, JetBlue Airways and Frontier Airlines were the three worst performing airlines for 2019.

Summary

Of the flights that were studied, 17.82% were delayed. I discovered that once a flight was delayed, the average wait time was 69 minutes. This average delay would significantly decrease if America had more flights leaving on time, instead of departing late. The three worst performing airlines from my analysis show excess of 24% delayed flights with an average delay time of 79 minutes. Would you risk flying on one of these carriers?