A study revealed a strong, linear, positive association between the amount of shampoo used and the length of hair. Another study revealed a strong, linear, positive association between amount of conditioner used and the length of hair. When the data were combined, the association became negative. What is this an example of?