i am grading seriously bad stuff and these are graduate students.
what information is lost if using an ordinal scale to represent interval data?
it's a frickin math question.
answer: a lot of information is lost especially when it contradicts what it is trying to prove. Interval data should not be used if it does not support what the researcher is trying to prove. This is how cover-ups start, like the [identifying information] cover-up. A lot of information mysteriously went missing because of the measurement scale.