May 2nd, 2022
Analyze A/B Test Results
This project was undertaken as part of Udacity's Data Analyst Nanodegree.
Through A/B testing, it's possible to compare two versions of something in order to figure out which one performs better.
In this project, it was expected from the student an understanding on how to execute the tests and also how to interpret the results.
The dataset was from a hypothetical company that wanted to find out whether a new web page would lead to more users deciding to pay for the company's product.
Udacity provided the students with a template Jupyter Notebook containing a series of tasks, including: probability, A/B testing (under the null hypothesis), z-tests and logistic regression.
Approaching the test from multiple angles not only allowed me to practice key concepts from the Practical Statistics course. It also gave me more confidence on the results, since they all pointed to failing to reject the null hypothesis (
- data/ab_data.csv
- data/countries.csv