A group of notebooks analyzing the results of A/B Testing on open datasets
A/B testing is important for discovering what works when trying something new without having to rely on what boils down to a guess. To perform A/B testing, two or more versions of a product, be it a landing page or the GUI of a full application, must be presented to a large set of users so their interactions may be observed and compared across versions
Unfortunately, this makes developing novel data for A/B testing difficult, as it takes both time and a large user base to reach a reasonable number of test subjects from which one may derive a viable result.
This repository will circumvent this issue by using only open and publicly available datasets. Analyses will be sequestered to sub-directories based on the set and my be broken out into their own repository if it makes sense in the future.
This repository is publicly available and has not governing license at its top level. All analysis documents are also publicly available and without license, baring that which is inherited through use of data. This does not mean that the data used for the analysis, which was not generated directly as a part of this project, is unlicensed or follows a singular licensing agreement. Refer to the data source for license information if attempting to use the data or derivative analysis for any purpose.