This might have been the wrong way to run this experiment. The fact that the participants publicly shared their opinion, and then they would also publicly had

This might have been the wrong way to run this experiment. The fact that the participants publicly shared their opinion, and then they would also publicly had to signal that they were wrong could simply be such a big psychological factor, that the topic at hand did not matter. It would’ve been better to vote anonymously, or even better ask one class at the beginning and a different class at the end of the lecture, then share the statistics with both classes at the next lecture.