I need more info.
Are the two samples independent or not...
A paired difference experiment, where we look at the husband and wife together, might be the best test.
I would like to know how I can compare two means of interval variables. I would like to know for example if males and females were different in their marital satisfaction (interval). Do I need to use T-test, Anova, or anything else?
This is called a paired difference experiment.
You subtract one set from the other and do a one sample t test with n-1 degrees of freedom.
I just googled 'paired difference experiment' and this looks fine...
Teach/Me Data Analysis
BUT all it is, is a one sample t test now.