[Grieverheart]

This is correct. Sample size should be dependent on the strength of the effect. As a rule of thumb though, a sample of 50 people is very little. If physics used these kind of sample sizes and confidence levels, we would have discovered multitudes of new particles.