Help with sample size

#1
Hello, I am a practicing Psychiatrist and I am conducting a study on Rural versus Urban Substance (drug) users in my hospital. It is a cross-sectional comparative study.

I browsed through some articles on sample size calculation for comparative studies and I calculated as follows:

1. One-year prevalence in a previous epidemiological study for rural and urban users was 43.2% and 56.8%.
2. A 3-month prevalence found by me in the hospital was 56% rural and 44% urban substance users.
3. Using 95CI and 80% power, I used the formula available online for comparing two proportions; n = (Zα/2+Zβ)^2 * (p1(1-p1)+p2(1-p2)) / (p1-p2)^2
hence n = (1.96+0.842)^2 * (.432(1-.432) + (.56(1-.56)) / (.432-.056)^2 = 236 for each group.

I want to know if this is the right way.
Link- https://select-statistics.co.uk/calculators/sample-size-calculator-two-proportions/
 

hlsmith

Less is more. Stay pure. Stay poor.
#5
One comment. have you already collected your data? Post hoc power calculations are considered a fax paus and frowned upon, since you likely already found a potential difference. Thus there is a form of selection bias going on. Since if the two proportions were null, you wouldn't move forward!