This repository is dedicated to providing from-first-principles implementations of one-way and two-way analysis of variance (ANOVA), as well as Tukey’s Honest Significant Difference (HSD) test. This resource is particularly suited for graduate and advanced undergraduate students engaged in programs with a strong emphasis on computational statistics, as well as industry practitioners with a focus on the computational aspects of statistical analysis.
The LinkedIn articles cited below, along with the repository, cover the following topics:
-
One-Way ANOVA: Exploration of its assumptions, the rationale underpinning its methodology, implementation from first principles, and an illustrative examples.
-
Two-Way ANOVA with Interaction: A comprehensive implementation of a two-way ANOVA model including an interaction term, accompanied by a practical example.
-
Type I Error in Multiple Mean Comparisons: Investigation of Type I error rates in the context of multiple comparisons, supported by computational examples.
-
Tukey’s HSD Test: A conceptual overview and from-scratch implementation of Tukey’s HSD test, along with a practical demonstration.
-
Supplementary Visualizations: A collection of bonus visualization scripts to support exploratory data analysis (EDA) preceding the ANOVA procedures.
Tyuryaev, Vadim. 2024a. “A Deep Dive into ANOVA (Part 1).” 2024. https://www.linkedin.com/pulse/deep-dive-anova-part-1-vadim-tyuryaev-luyvc/.
———. 2024b. “A Deep Dive into ANOVA (Part 2).” 2024. https://www.linkedin.com/pulse/deep-dive-anovapart-2-vadim-tyuryaev-diopc/.
———. 2024c. “A Deep Dive into ANOVA (Part 3).” 2024. https://www.linkedin.com/pulse/deep-dive-anovapart-3-vadim-tyuryaev-c6rtc/.