We have a survey with several questions with 5 likert scale points and we would like to compare the answers to those of another likert like question with 10 points.
The approach we thought of is a linear regression considering the questions as ordinal categorical independent variables and the 10 points question as quantitative dependent variable.
All the questions will be adjusted for some factors (age, etc...)
The idea would be to first select significant questions (p > 0.2) using multiple regressions for each question plus confounders. Then put signicant question in a multiple regression.
So I'd like to ask you:
- Is this approach a correct way to select the question which best correlate?
- It's a too big assumption to use a linear regression with a 10 point scale?
- Are we using the correct way to perform feature selection to eliminate questions which don't correlate?