A/B Testing at WorldQuant University
In this project, WQU needs our help to improve its programs! The team have noticed that many applicants to the DS Lab never complete the admissions quiz. In this project, We’ll design and conduct an experiment to see if we can increase quiz completion.
First, we’ll explore the DS Lab applicant pool, and formulate a research question and hypothesis. Then we’ll run our experiment and analyze the results. Finally, we’ll build an interactive web application so that others can run the same experiment.
This project is an example of A/B testing, also known as a randomized controlled experiment. In the private sector, A/B tests are used to improve things like email marketing and product pricing. In politics, they can be used to test campaign messages. Scientists of all kinds use them in their research.
We were able to:
- Build a choropleth map to show the distribution of ADSL students around the world.
- Create a custom Python class to implement ETL processes.
- Design an experiment and analyze the results using a chi-square test.
- Build an interactive web application that follows a three-tiered design pattern.
This project consists of 5 Notebooks :
- Meet the DS Lab Applicants
- Extract, Transform, and Loa
- Chi-square test
- Experiment Web Application
- Admissions in the MScFE 🎓🗞
You can find the whole project blog here