One-Way ANOVA [Analysis of Variance] simply explained

This video is about analysis of variance, or ANOVA for short. We discuss what an ANOVA is and why you need it. We look at the hypotheses and the prerequisites for an analysis of variance, and I show you how to calculate an ANOVA and what the equations behind an ANOVA are. Finally, we discuss how to interpret the results and take a look at the post-hoc test. First of all, there are different types of analysis of variance. The simplest and most common form is the one-way analysis of variance. And this video is all about the one-way analysis of variance.

► EBOOK

► ANOVA statistics calculator

► Tutorial

► F-distribution table

► Video Test for Normality

► Video Levene's test

► Video Kruskal-Wallis-Test

► Video Two Way ANOVA

00:00 What is an analysis of variance?
01:53 What are the hypotheses in an analysis of variance?
02:29 What are the assumptions of an analysis of variance?
05:55 How is an ANOVA calculated?
12:16 How is an analysis of variance calculated with DATAtab?
13:00 What is a post-hoc test?

[abdelgaderalfallah]

Simple, easily to digest and break down and for utmost benefits.... thank you ma'am I wish you were my professor back in the medical university.
Keep it up, Will you?
👏👏👏👏👏👏👏👏👏

[Questtoknowwithdraftab]

I appreciate your effort and time you spend in making this video. Nicely explained

[phdpablo]

Excellent. Congrats for pedagogical explained.

I like the videos a lot, I have learned a lot. I would love for you to make a video on what sample size to take for a T test or ANOVA. Or rather how many repetitions should I use?

[jayvirvaghela]

superb explanation 🎉🎉 thank you so much🙏🙏

[llKaiserxll]

Fantastic visuals, great way to explain the topic.

[tundeoyebanji]

I love the way you simplify statistics. Would you mind doing something on propensity score matching? I'm not certain that the test is on datatab yet.

[EJ_comedy]

Can you please do a video on principal component analysis?

[abhishekchandrashukla]

Hey, you said the normality assumption becomes less important as the sample size increases. From what I have seen in the internet, if the sample size > 30 and the data is not normal then we can still use a T-test or an ANOVA test as they are more robust, unless the distribution is very skewed.

We can use T-tests or ANOVA test because:
1. We can use Box-cox transformation if the data is not extremely skewed.

Is my understanding correct, or there is something else I am missing?

If you could clarify an another question, I would really appreciate it.
Let's say we want to check if the heights differ between two groups significantly. We draw 10 persons from each group and measure their height.
Theoritcally we know that heoghts are normally distributed, but when we look at the sampled data (both groups) they do not follow a normal distribution. So in this case should be use a T-test or Mann Whitney U-test.

Finally, why is that Hypothesis tests that make assumptions (parametric tests) are more powerful than non-parametric tests that do not make assumptions, when the data satisfies the assumptions of parametric tests?