How To... Calculate Effect Size (Two Samples t Test)

When you find a significant difference between two samples with a t Test, you normally report that you have found an "effect". But how meaningful is that effect? In this video you will learn step-by-step how to calculate Effect Size to determine the magnitude of the effect.

[@nesyjames6167]

Dr. Eugene thanks. This was helpful. Love d way u used variance so as to simply d calculations.

[@terrancegrant1664]

An excellent video, Eugene. Thanks.

[@carineokeh3543]

It was very helpful. Thanks a lot.

[@conchiorobiyirieba9225]

very helpful, many thanks

[@emermorrison951]

In the equation, you say pooled variance is, square root of 27.0 +21.5 divided by 2. Were 27 and 21.5 supposed to be squared first? I know that gives it a really high number, but that's what the formula says, right? What am I missing?

[@EugeneOLoughlin]

Hi Emer,
Please follow what I do: add 27.0 and 21.5 first. Then divide by 2. Then take the square root of that number - you should get a result of 4.92 as I do.
Hope this helps,
Dr E.

[@CamulosKnight]

There is a typo error in the table on the top left of the video. The value of 27 & 21.5 are actually for the sample variance (s-squared) and not the sample standard deviation (s) as shown. This is why the values can be directly substituted for the population variance (sigma-squared) in the equation on the top-right of the video without having to square them.

[@happyourssailing5690]

Eugene--can you look at your std deviation calculation. The pre test should be 5.20 & post test 4.63

[@happyourssailing5690]

Ignore...you were using "pool variance" not std deviation which is another variation measurement