How to test mean differences for two questions

[agustosba]

Hi all!

I have two questions in my survey measured by 7-point response scale (1=extremely low, 7=extremely high)
1) measuring the impact of product A
2) measuring the impact of product B

I wanna test if product's A impact is statistically different from product's B.

The right analysis is paired-samples t-test?

Thank you for your time.
--
Aysen

[Art Kendall]

If I understand your post:
Your question can be framed as a repeated measures analysis.
This is also called a within-subjects or within-respondent analysis.
The same respondent was asked the same question under two conditions.

As framed there is no second group of respondents. So the analysis is for the SAME people.

2 groups measured on one variable ==> ordinary t-test
1 group  measured twice on one variable.  (twice can refer to two time, or two conditions such as a variation in the stimulus question, etc).