When comparing answers on two questions by the same people, is a paired samples t-test appropriate (details)?

I asked the SAME people to rate how much they talked about subject A and how much they talked about subject B on a scale from 1 to 5.
To know if there is a statistically significant difference in their scores, should I use a paired samples t-test? Is that the appropriate test to run?

Thank you!

2012-12-12T01:37:16Z

So I guess that means I cannot use it?

In any case, here are more details on the survey I did -
-I asked people to rate from 1 to 5 how much they talk about their interests and how much they talk about the woman's interests (when they are "hitting on" a woman).
-The same group of men said that on average, they rate themselves at a 1.69 when talking about their interests and at a 3.38 when talking about the woman's interests.

So if the paired t-test is still not appropriate, IS there a test to see the statistical difference between those two numbers? If so, I'm sure there is, what would it be?

(For example, if they rated 2.12 for one and 2.19 on the other, I'm sure it would not be statistically significant).

satish_newase1232012-12-12T01:29:52Z

Favorite Answer

Paired t test measures the significance difference between two different treatment on same person: Suppose apply treatment A on patient and after some time apply treatment B on same patient which gives two responses x and y respectively. If we want to test whether both test have same effects or significant difference between A and B then we can use a Paired t test.

Suppose we assume that the two subject as treatment with same student then we can use paired t test.