In the previous chapter we fit the additive model to this same data. Whereas the additive model was a three parameter model ( , , and ) the interaction model requires us to estimate one additional parameter ( ). Let's take a brief look back at what the additive model looks like, and see what we have "bought" with this extra parameter. In the table below we've put two jitter plots. On the right is the one we have just been discussing, with the interaction model predictions overlaid. On the left is the same jitter plot, with the same data, but this time with the predictions of the additive model. Additive Model Interaction Model Look carefully at the two graphs. What differences do you notice between the two models? in a short answer

tip_ percent control smiley face condition control smiley face gender =e= female == male