I’m an OT doing a part time masters. We’ve had some stats teaching but the more complex stuff comes next year. I’m in the middle of writing my research proposal and getting stuck on sample size. I intend to do a survey using a validated attitude scale on staff working in healthcare. I will be using a form of stratified random sampling.
I’ve done a bit of research into G*Power and calculating sample size according to test, effect size etc using the very helpful Howitt and Cramer stats text book. I intend to use non-parametric tests – so G*Power has given me an idea of sample size for Mann Whitney U and I used one way anova to give me an idea about Kruksal Wallis.
But Spearman Rho and binomial logistic regression aren’t included.
So far G*Power results and estimate of response rate is suggesting a sample size of about 550. Is this going to be sufficient for Spearman Rho and binomial logistic regression?
Thank you for bearing with me this far – apologies for the length.