Ordinal data in regression? My data is almost only ordinal data. My dependent variable is a 4 point scale:  
1 = completely agree
2 = agree to some extent
3 = disagree to some extent
4 = completely disagree   
I want to use regular OLS using SPSS. Should I use ordinal regression? What is the difference between ordinal regression and linear regression? When I try to use linear and my dependent variable, SPSS refuses to run the analysis. What should I do?
 A: Since your response is ordinal then you should use ordinal regression. At a very high level, the main difference ordinal regression and linear regression is that with linear regression the dependent variable is continuous and ordinal the dependent variable is ordinal.  
Now you can usually use linear regression with an ordinal dependent variable but you will see that the diagnostic plots do not look good. When you say SPSS won't run the linear regression what do you mean? Are you getting an error? 
A: I agree with Lauren Goodwin's answer. 
Additionally, in SPSS, all the variables are classified on the basis of their measurement as - nominal, ordinal, and scale. Your dependent variable is perhaps described as ordinal in your variable list. To run a linear regression model, SPSS requires that the dependent variable be scale. This is perhaps the reason why SPSS is not running your analysis (or is showing an error).
A: Ordinal regression is the way to go (since it's implemented in SPSS Statistics).
The model is somewhat multinomial regression, with the resulting variable ordered in some way. Now, depending on the function to fit (logit or probit), the order should be lowest or highest category first. For more info, ofc, check the documentation.
