 How many participants do you need in your study? This will be a short demonstration of an a priori power calculation in g-power, something missing from a lot of studies but something that's actually a lot easier than you might imagine. So the first step is to download g-power. So we're going to go to g-power.hu.de. We'll then scroll down to download and you can click to download g-power in either Windows or Mac. Once that's downloaded and installed, we can then open up g-power and take it from there. The first thing is to decide on the statistical test that we want to use. I find it's easiest to use the test bar at the top. So under there for correlations and regression, we've got various correlations and through to multiple regression, etc. If we want to compare means, we've got different varieties of t-tests, ANOVA, MANOVA, ANCOVA, etc. So to keep it simple, I'm going to start with an independent samples t-test. So we're comparing means of two independent groups. We've just got a small number of options that we can change. To start with, we're going to do a two-tailed test. We then need to state the effect size that we want to be able to reliably detect. If we hover over it, we get the default effect size conventions of small, medium and large. We can either use these, we can take an effect size from the literature, or ideally we can state what the smallest effect size is that would be considered clinically or practically meaningful and then set our study up to be able to detect that. But for now, I'm going to take a medium effect size of 0.5. Our alpha level is 0.05 as usual. We then want to know how many participants would be required to be able to detect an effect size at least as large as the one we've stated, a certain percentage of the time. That percentage of the time is our power, and it's commonly set to 80%. I'm going to set this to 0.8. Then all we're saying here is that we've got even ratio of our two groups, so the same number of participants in each group. If I click calculate, we can see that detecting that effect size 80% of the time would require 64 participants in each group, so a total of 128 participants in our study. That's it in a few simple steps. That is the a priori power calculation. We can explore it a little bit more so we could see, for example, if we were only interested in detecting large effects, I can change that to 0.8. Our required sample size has dropped to 26 in each group and 52 in total. If we had a strong reason for performing a one-tailed test that was justified and ideally pre-registered, then we can see that again it's reduced to 21 participants per group and 42 in total. We can also use this to explore different study designs prior to the study. For example, if it was possible to use the same group in both conditions and perform a paired samples t-test, we can see how many participants would be required for that. Under tests means again, instead of two independent groups, this time we would have two dependent groups with matched pairs go for the same values that we had in the last calculation and we're now down to a total sample size of 12 participants. That's it for a brief demonstration of an a priori power calculation in G-Power. Hopefully you found that useful. If you've got any questions or anything else you'd like to see, please leave a comment. If you want to be notified of any of the tutorials, then just click subscribe and the little bell next to it. Thanks a lot.