I am doing a project on HIV patients to explore their quality of life (WHOQOL-HIV-BREF) regarding 7 variables of which 3 variables are scale generated consisted of 18 items. As a prevalence study of an unknown proportion of the population, I wanted to use P=50%, z=95% at 5% margin of error. But someone suggests me to use a 10% margin of error.

I need a clear idea of this, when and why I have to change the margin of error.

More Md. Arif Billah's questions See All
Similar questions and discussions