Help me with my final year psychology report and I'll love you forever!

[helenwilliams]

Hi, I hope someone can help me!

I'm doing a psychology experiment and due to the nature of the experiment it has been very hard to get many participants tested. I need to know if I've got enough to draw a solid conclusion from.

I have 4 participants doing 4 conditions, and there are 5 repetitions of each condition for each person. I need to find out if there is a significant difference between the averages of each condition. I'm using a within subjects ANOVA, and I guess I'll need to use t-tests to determine the differences between each individual condition?

Anyway, my main question is - will 4 participants be enough?

Thanks for your help!

 

[staryeyed]

The number of participants you will need in order to find a significant effect if it exists will depend on the effect size you expect the phenomenon you are testing to have. The stronger the effect the less people you need to detect it. There are ways of calculating the suggested n based on your estimate of the size of your effect but in order to find out what you expect your effect size to be you are going to have to dig around in your literature to see what other people have found. An easier option is to look at the sample size of studies that are similar to yours. You will probably want a sample size that is fairly similar to the size of sample being used by other researchers who are studying this topic.

Of course you could always just run an analysis on what you have so far and see if it is significant. If so then you don't need to bother running more subjects. If not is it trending in the correct direction. If so then running a few more subjects may increase those trends. On the other hand if you do not see any sort of pattern in your data now, and breaking your back to get subjects is only going to get you one or two more data points then it probably isn't worth running more subjects.
Long story short sample size is a bit of a subjective judgment.