I have data from several Likert Scales which look at how people feel about object a - e. I want to compare the means of these results and see whether people feel differently about object c and object e (for example. I want to test whether this difference is significant. Does anyone know what kind of test I can carry out and how? Thank you for any help!
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
1
This sounds like a within-subjects test, to me. If you want to compare differences within a single sample, I recommend a paired-samples (some times called dependent samples) t-test. It tests whether the difference between two sets of scores from the same cases are different from 0 on average. If you want to compare more than 2 distributions of scores simultaneously, a repeated measures ANOVA would be appropriate, with your within-subjects factor being the item response code (e.g., a, b, c, etc. in your example above).
These tests are fairly robust to violations of their assumptions, but if your data are truly ordinal (i.e., scale of 1-5) alternative, non-parametric tests may be more appropriate.
-
$\begingroup$ Thank you for the answer. I did run some paired sample tests. I also reorganised my data and ran a one way ANOVa and then independent T-tests. For non-parametric tests what would you recommend - Mann Whitney U tests? $\endgroup$– DjangertCommented Dec 16, 2018 at 14:50