I hate when a study comes out and people automatically make blanket statements.
It depends on the type of evaluation and many other factors.
With something like a survey, it’s a quantitative analysis. So, questions should be relatively objective. And inferences should only be made in abnormalities. For example, if 60% of students say the professor is too difficult, I’d ignore it. If 90% of students say the professor is too difficult, I’d probably side with the students.