The goal of any data analysis project should be one (or both) of two things:
- to derive practical, actionable insights from the data
- to validate or reject a hypothesis based on that data
In the previous couple of posts, we saw that flight delays show an exponential decrease as time progresses, i.e. while the average delay time across all delayed flights might be close to an hour, more than 50% of flights are delayed by less than 30 minutes, and as time progresses, the number of delayed flights continues to drop.
Generally speaking, there is a 90% chance that your flight arrives within two hours of its posted arrival time. This led us to conclusion that you can be reasonably assured of making a connection if you left an hour or two between your two flights.
Let’s ask another question of the data. Do you want to minimize the chances of your aircraft being delayed? If so, what time of day is best for flying?
The best time of day to minimize your chances of a delay is the morning time. We split the day into five groups. The data show