Ordinal dependent variable - regression help!

demwill

New Member
First of all, I am a complete newbie to stats from a non mathematical background, but I need to analyse survey data for my dissertation!

So I have a dependent variable that is ordinal - it's a 5-point likert item (strongly disagree - strongly agree). My independent variables are composite variables that have been calculated from 3 or 4 likert-items and as these values can be, for example, 3.2, 1.9, 4.5 etc, I have treated the differences between 1-5 as meaningful and thus interval.

To look at strength of relationships between variables I have conducted a bivariate analysis using Spearman's rho.

I want to conduct a regression analysis, but now I am lost! I'm under the impression that I should conduct an ordered logistic regression, but being a complete stats newbie under serious time constraints, I have no idea how to do this (or if I should do this).

I'm also under the impression that I have to do a regression in certain steps, as I am testing a conceptual framework where a couple of the concepts not only are expected to influence the overall dependent variable, but also some of the other concepts. (I hope this makes sense).

I would really, really appreciate some help!