Thanks Prof. Singh

In my opinion, it must be a dyadic study.  You have to make dyads on one to one basis.

I would suggest you to determine exactly what you want to know regarding their "attitude" towards each other. Determine the characteristics you want to measure and develop the items. Then you give a set of self-administered questionnaire each to measure their perception towards their partner.

There are a lot of brilliant ideas up there. I will add please make a list of variables you want to include, and also compare them with the global variables then you pilot it through one to one interviews. Subsequently, formulate an interview schedule to be administered personally. This will give you a real data. A questionnaire may misguide you! I hope it helps!

Thanks Mr. Awang & Mr. Arif

I have decided one variable i.e. Emotional Expressivity & I wanted to find out what husbands think about emotional expressivity of wives and vice versa. I was just thinking about the tests that should be used for analysis. I was thinking of 't test' but then whether 'Independent t-test' of 'Paired t-test' will be suitable.

 Rahul Sir 

Please elaborate the method of creating dyads.

Which attitudes are you referring to?

I am referring to the views of husband towards their wives emotional expressiveness and wife's views towards their husband's expressiveness, which implies what husbands think about their wives, how expressive wives are according to them and vice-versa.

Try the theory of planned behavior with a qualitative method or mixed methods.  The TPB is a good predictor of attitudes.

I strongly believe in order to understand attitude toward each other, the best way is to tap their day to day behaviors in specified contexts.