This repository outlines my approach to selecting the appropriate statistical test for determining the p-value, which helps assess whether to reject the null hypothesis. In simpler, business-oriented terms, the aim is to evaluate whether a specific change—such as a website layout modification, button placement, or feature update—directly contributes to improvements in key performance indicators (KPIs).
For instance, in digital business contexts, we're often interested in metrics like Average Revenue Per User (ARPU), Average Session Duration, and Average Order Value (AOV). Through carefully designed A/B tests, we measure whether these adjustments lead to statistically significant improvements, or if the changes are simply due to random chance.
The goal is not just to perform the test but to gain actionable insights, ensuring that decisions are data-driven and aligned with business objectives. By following a structured hypothesis-testing framework, we can confidently validate the impact of our design changes and optimize user experience to enhance overall business performance.