https://ahmed-azeez.github.io/tags/statistics/Recent content in Statistics on Ahmed Azeez | PortfolioHugoen-usSun, 19 Jul 2026 00:00:00 +0000https://ahmed-azeez.github.io/2026/07/19/multicollinearity-explained/Sun, 19 Jul 2026 00:00:00 +0000https://ahmed-azeez.github.io/2026/07/19/multicollinearity-explained/Multicollinearity occurs when two or more independent variables in a regression model are highly correlated with one another. At first glance this may not seem like a problem — after all, if two variables are related, shouldn't they both help explain the outcome? The issue is that the model struggles to tell which variable is actually responsible for the observed association. The Problem: Two Variables, One Signal Imagine trying to determine whether income or wealth has a stronger relationship with a health outcome.https://ahmed-azeez.github.io/2026/07/12/efa-reporting-standards/Sun, 12 Jul 2026 00:00:00 +0000https://ahmed-azeez.github.io/2026/07/12/efa-reporting-standards/Exploratory Factor Analysis (EFA) is a multivariate statistical method used to determine the underlying dimensions, factors, or latent variables within a set of observed variables. To ensure your findings are replicable and interpretable, specific technical details must be transparently reported. The Foundation: Justification & Data Before diving into the numbers, researchers must justify the use of EFA over other methods like Confirmatory Factor Analysis (CFA). This is typically necessary when the factor structure is previously unknown or when developing a new scale.https://ahmed-azeez.github.io/2025/05/16/diagnostic-tests/Fri, 16 May 2025 00:00:00 +0000https://ahmed-azeez.github.io/2025/05/16/diagnostic-tests/Before you run a single regression or ANOVA, there is a step that separates rigorous analysis from shaky conclusions: diagnostic testing. Think of it as a pre-flight checklist for your data. Skip it, and you risk landing in entirely the wrong place. Why Diagnostics Matter Most statistical methods rest on assumptions — about how data are distributed, how variables relate to one another, and how errors behave. When those assumptions break down silently, the model keeps running and happily produces numbers that mean very little.