Science education research has shown that students use causal reasoning, particularly the model'agent instrument object', to explain or predict the outcome of many natural situations. Students' reasoning seems to be based on a small set of few intuitive rules. One of these rules quantitatively correlates the outcome of an experiment with the characteristics of objects involved, suggesting that an increase in one quantity should bring about an increase in a related quantity. In this article we present and discuss five simple experiments to engage students and illustrate that an increase in one property doesn't always lead to an increase in another.