I’ve been using the margin of error (MOEs) formula to calculate MOEs for some of my work, and the calculation process is pretty straightforward. Here is the formula I’ve been using for reference during this discussion: http://calculator.mathcaptain.com/margin-of-error-calculator.html.
I have three groups of interest for this study – although I’ll only be calculating MOEs for two of them for this analysis 1) Non-problem gamblers (est. population represented 4,665,005), and 2) At-risk/problem gamblers (est. population represented 497,841).
I want to sub-divide my sample and calculate a margin of error for each strata, which makes it a little more complicated. For example, I want an MOE for each age group of a specified sex and gambling status (see attached database). Would the “n” in the formula for this calculation represent the total “n” of the individual strata (non-problem gamblers and at-risk/problem gamblers combined) –or– do I use the sample “n” for each type of gambling status (i.e., separate out n’s for non-problem gamblers and at-risk/problem gamblers)? For example, 24 male respondents (19 non-problem gamblers and 4 at-risk-problem gamblers) aged 18-24 said they fell into either category, so should I use the 24 for the “n” or the 19 for the “n” when I create a MOE for that group? You may also want to double check my other formulas; they look right, but I can always use a second set of eyes.
I appreciate any help in the matter!
Calculating Standard deviation and Grouping data between +/- 3*sigma into one category must help.
You have a nominal matrix, so, if you apply an ANOVA, you'll produce variances for the total population and all the subgroups as well.
- Badges
- Science topic
- Similar topics
- Mathematics
- Applied Mathematics