I have some data to analyze and the outcome we are looking for is whether or not a woman had a positive pregnancy test. So the possible answers are only "Yes" or "No", which could be codified into "1" or "0". I thought I would use a t-test, but how would you address the standard deviation? Would that be a standard deviation of zero?
Binomial test for the "proportion of women with positive pregnancy tests", Fishers Exact test for comparing proportions between groups, Logistic regression when covariates should be considered.
Jason -
It sounds like you may be looking for a binomial distribution, for which you can search on the internet, and also find the mean and standard deviation, and confidence intervals explained. It will also talk about contingency tests, but without a full knowledge of the power in each application, those results can be misleading. I encourage you to consider the confidence intervals instead.
Jim
PS - Ah. Don't know why Jochen's answer did not show up on my screen when I first answered this, but I see he has a good answer for you.
Article Practical Interpretation of Hypothesis Tests - letter to the...
Hi Jason
You really should try to find a local statistician to provide you with advice. In the meanwhile, have a look at this beginners guide to medical statistics:
http://www.bmj.com/about-bmj/resources-readers/publications/statistics-square-one
Adrian