Field experiment on eBay

Testing whether response times reveal eBay sellers' evaluations of offers

Untitled

Collaborators: Ian Krajbich

Programming: R, eBay API, python

Skills: Data Cleaning, Field Experiment, Linear and Logistic Mixed Effects Regressions

🔭 About


One of my previous projects using observational eBay data showed that seller’s and buyer’s response times on eBay varies with how good the offer is, namely responders are fast ( 🐇 ) to reject ( 👎 ) bad offers and slow ( 🐢 ) to reject ( 👎 ) good offers and they are fast ( 🐇 ) to accept ( 👍 ) good offers and slow ( 🐢 ) to accept ( 👍 ) bad offers. Although we controlled for possible confounding factors such as seller and item characteristics, we wanted to know if this results reflected a causal relationship between offer size and response times (RTs). So we decided to run a field experiment on eBay to test this idea by acting as buyers and making random offers to sellers on eBay.