I am using a mixed methods approach and need to look for correlations between the results of two quantitative data. A statistician once mentioned that I need 30 participants for correlation purposes and mentioned Andy Field. Does anyone know of an article or chapter I can read about participant numbers?
Thanks
The G*power program, which is readily available on the internet, will help you calculate sample size, according to the level of precision (confidence interval) you want for a correlation. In general, an N of 30 would be almost useless for detecting all but the strongest of correlations.
Hi Elanor,
Here's a link that might help:
https://books.google.com/books?id=WBSwhw-j1ZoC&printsec=frontcover&dq=sample+size+formula&hl=en&sa=X&ved=0ahUKEwjUxMiRisPaAhUH21MKHb1WCb4QuwUINDAB#v=onepage&q=sample%20size%20formula&f=false
Chapter 2 discusses sample size determination.
Have a great day!
--Adrian
It also depends on the population from which the sample is being drawn from. 30 would be too low for a population of say 200
- Badges
- Science topic