Despite efforts to teach logic and critical thinking in the classroom, students will often give the answer that they think is expected. Perhaps a discredited theory from the 18th century can help students see how different conclusions can be drawn from the same experiment, suggests Mike Tingle