Welcome to Moore Statistics Consulting LLC

This is a quick commentary on sample size calculations. The estimation of your sample size reflects how many people you need to show statistical significance. Your formula will vary depending on the outcome variable. If the outcome variable is dichotomous you would use frequencies to generate your sample size. If you outcome variable is continuous you would need another equation. There are sample calculations specifically for certain hypothesis tests. Generally, you should have access to power sample size software to make your predictions. There are statistical programs like GPOWER,  NQUERY , and PASS. But there is an art to it. You do factor in the scenario where you lose some of your sample size due to outside forces. Maybe the participants decide to quit your study and disappear without a trace. But the main stumbling block seems to be the effect size. When you calculate your sample size, there is an aspect of it that can vary. The effect size is set by the researcher or physician conducting the study. It represents the minimum effect the investigator is looking for in their experiment. . That effect can be the difference between two treatment groups or something else.

Calculating sample size in a statistics 101 course is pretty straight forward in that you know the effect size. You are given particular scenarios and you are expected to pluck out the numbers for your sample size calculation. It’s like walking though a haunted  house ride at an amusement park- you know what to expect no surprises really. Publications and grant work are another story.


If you enjoyed this blog there are more to follow.

Liked it? Take a second to support Moore Statistics Consulting LLC on Patreon!
Become a patron at Patreon!

Leave a Reply

Your email address will not be published. Required fields are marked *