The course introduced data-driven puzzles through a wide array of fields, including political, economic, historical, social and biological. For their final projects, students chose a data set from a previous project or found one online to analyze.
Craig and his partner discovered that Airbnb makes their data available online. That became the impetus for their project, to study variables that affect price and location of Airbnb rentals in the Boston area, near Craig's hometown.
To begin, they needed to find data that was 'clean;' data that is consistent and free of errors. Next, they determined if there were ethical concerns in analyzing the data. Data ethics concern privacy and security, among other issues. Since the data was freely and publicly available online, they deemed it to be ethically sound to use.
After thinking about the data and the Boston area, Craig and his partner hit upon the idea of plotting national landmarks and correlating them with the location and price of the rentals.
They created three statistical models and analyses.
“We came to several conclusions from our data,” said Craig. “One of our major questions was what affects price. We found that as distance from the closest landmark increases by one mile, price decreases by about $30.”
They also concluded that if you are trying to price your own Airbnb rental, you should consider the number of people you can accommodate, landmark distance, and the mean distance from other Airbnbs.
“Finally, we found that there was a high correlation between landmark distance and mean distance from Airbnbs,” said Craig. “It was fascinating to see just how close Airbnbs tend to be to landmarks.”
Craig wants to work in finance after graduation. And he offers some free advice. “Our research shows that new hosts should try to identify up and coming landmarks so you can charge a higher price later on. For renters, if you're willing to travel outwards from a central location, the price of your Airbnb will drop significantly.”