there must be a dependent variables in regression analysis. if you regard your data as categorical data, you can try Chi-square analysis (Contingency table analysis).

I think you must first better clearly define your question so that you can choose an appropriate analysis. You can use regression analyses to determine the association between an independent and dependent variable, but your result does not have causation. So, the reporting and discussion of your findings must clearly express that there are correlations/relationships, but you do not know causality. Since you are using a ranking system, depending on the distribution of your data, you may try an ANOVA with a categorical independent variable (normally distributed), or a spearman rank test (not-normally distributed).

You may also try to use pairwise correlation as you do not have a dependent variable. It should be able to give you the association between any two variables.

The purpose of regression is to measure the degree of association between a DV and one or more independent variables. A close association does not necessarily indicate cause-effect. It's simply a statistical association that is established. If you want to investigate cause-effect relationships you are better off with an experimental design. Rachel gives good advice in her answer to the question. 

Regression analysis (logistic, multiple..etc) is essentially intended to find and measure association between a given dependent variable (an outcome) like for example the risk of death among a group of cases of disease. The researcher tries to find which of a set of possible risk factors (explanatory variables) could predict the death or otherwise. This means that in regression analysis we need an outcome and a set of determinants to see which one is significant and which one is not.

You have several variables that you are looking at.  Make one of them the dependent variable.  Toot around with it. 

Hello, 

As others have noted, for regression analysis a dependent variable is required. The following link, http://statisticalhorizons.com/prediction-vs-causation-in-regression-analysis, is a good overview of Prediction vs. Causation in Regression Analysis, which you might find helpful.

http://statisticalhorizons.com/prediction-vs-causation-in-regression-analysis