3
$\begingroup$

I have 4 data groups: - 3 groups from a 5-point Likert scale (1 ~ 5) - 1 group is binary (Incorrect/Correct -- converted to 0, 1)

I have performed Pearson's r and Spearman's rho with the Likert responses easily enough and they appear to be related to a scatterplot. However, the binary data is giving some strange numbers that don't appear to be showing a true relationship (I'm new to using SPSS so this may be the problem).

An alternative I want advice on in the comparison of means. I am aware of paired t-tests but I don't know if they are suitable or if something else is better suited. I'm especially interested in the relationship between, e.g. a) Perception of x= M:3.12, b) Realisation of x= M:3.95

$\endgroup$
2
  • 5
    $\begingroup$ I don't entirely follow your question. You may find it helpful to read this blog post: how to ask a statistics question. Regarding your question about whether binary responses should be coded as {0,1}, or {1,5}, it will not make any difference--you can try it for yourself & see. $\endgroup$ Dec 28, 2012 at 4:32
  • 1
    $\begingroup$ Thanks. I have updated my original question and shifted the focus to a comparison of means, as in a paired t-test or an alternative. $\endgroup$
    – Andrzej
    Dec 28, 2012 at 8:43

1 Answer 1

2
$\begingroup$

If you want to compare two groups on three different Likert scales, each scored 1 to 5, you probably want three ordinal logistic regressions, each with a Likert scale as the dependent variable and the group as the independent variable.

I do not know how to do ordinal logistic in SPSS, but it surely can be done.

$\endgroup$
0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.