Answer to Question #40953 in Statistics and Probability for William Marigold

Question

How do I prove or disprove the null hypothesis for a set of data, if the data values are simply 1 or 2. The population is 30 and the standard deviation is 0.49

Answer 1

Dear visitor
Unfortunately, the statement of your question is a bit confusing.What do you need to be done? Please, give us the details so we could help you.
The answer depends on the choice of test. If you test interaction between variables then apply association test or kappa statistic. If not interaction, it is necessary to know how many variables is under consideration: if more than two variables, then apply logistic regression; if one variable, then it is necessary to know what values you expect. If expected values are less than 5, then you apply Fisher`s test. If expected values are greater than 5, then you have to know whether observations of sample are dependent or not. You use chi-square test in case of independent observations. You use McNemar test in case of dependent observations.