I do not know how to separate rational from irrational beliefs. Can I use the theoretical framework of rational-behavioral therapy? Coding the data based on the dimensions of irrational beliefs can be a way of this?
I don't think you can or should try to differentiate between what you think might be irrational. It would be more satisfactory to ask about the basis of their beliefs and code these accordingly. Only then can you avoid your subjective interpretation which would probably make the study unpublishable