So, what you're saying is this: If students are bribed, they will - gasp! - give instructors higher scores. Shame upon students to be so much like other homo sapiens sapiens! How dare they? Actually, let me rephrase that: As far as I know, students were told that the chocolate had nothing to do with the instructor. The study has then proved yet another shocking fact about students: They - oh no! - like chocolate. They like chocolate so much, in fact, that it will put them in a good mood and cause them to see the their bitter instructors in a rosier glow.
Again: despicable!
Seriously, though - shame on you for latching on so willingly to such a silly result. Next thing you'll try to discredit student evaluations by telling us about a study that proved how perfectly fine professors get lower evaluations just because they beat their students.
Again: Duh.
I repeat: All students are humans. Humans can be influenced by unethical behaviour surrounding research design. Giving students chocolate before the evaluations is unethical. Instructors should not do it.
Any more questions?