I am doing a study using Factor Analysis (Principal Component analysis) and linear regression. My goal is to find out how likely is it to become a user of a specific product. In my questionnaire are 15 Statements about the product - the answers are re-worked to the scale: 1-Wrong answer, 2-Don´t know, 3-Right Answer. These are grouped into 4 groups due to Factor Analysis. Now I would use the factor loading as Independent variable for the linear regression.
The problem is: Since I don´t know the distances between 1-Wrong and 2.DK and 3-Right, I should do the regression 2x (First time: Variables with the scale 1-Right,0-Wrong+DK, Second time with the Scale 1-Wrong,0-Correct+DK) and compare the options how robust they are. How can I combine this idea with my factor analysis? Can I get the factors, recount them somehow and do the linear regression? Thank you for your help.