Actually, even more important than sample size is having a truly representative random sample. One can do a good valid study with a sample as small as about 50 subjects providing it is representative of the population and the appropriate statistical method is used. Any good stats text also has different tables for doing studies with small samples. Unfortunately most people without advanced statistical training don't know this and are overly impressed by large sample sizes alone.
Last edited by bobphilo; 07-30-2018 at 01:03 PM.
|