Designing good clicker questions is tricky. I want them to be challenging enough that the students have to think quite a bit, but because correct answers count for marks, I want most students to get them right most of the time.
Usually I let the students discuss each question with each other before they answer. One way to improve both the benefits of the consultations, and the proportion of the answers that are correct, might be to first present the question not for marks, asking students to answer without consulting their neighbours. Then show them the range of answers (?) without indicating which is correct. Then ask them to consult their neighbours before answering again, this time for marks.
I have a couple of little books about using clickers in the classroom (gifts from the textbook rep). One is specifically about science teaching - I'll see what suggestions it has.