Need help: Can i genaralize to the population, or do i have to weight my data?

#1
I have a sample of 194 households from 3 villages. The number of hh from each village is 61, 67 and 66. I used a 90% confidence level and 10% margin of error to determine sample sizes. (total household population for all three villages is 4194 households). The total sample size for all three villages is significant at 99% confidence level and 9% margin of error, however, i have kept my generalizations at the 90% confidence level. To determine confidence intervals i have used a z-score of 1.65. Now there are a couple of questions i have when generalizing to the population:

1. When i make analyses, do i have to weight my data to the hh population of the specific village ? (which significantly reduces the values of the data). Or is my sample so evenly spread across the three villages that i don't need to do this?

Weights of villages:
Village 1 = (34% of total household population)
village 2 = (38%)
village 3 = (28%)

2. Or can i simply analyze the data from all three villages as a total, and apply the confidence intervals to make my generalizations?

I would really appreciate your help!!!
 
Last edited: