Hi
I'm modelling the theory of planned behaviour with four variables (subjective norm, perceived behavioural control, affective attitude and cognitive attitude) predicting behavioural intention.
As I measure these variables following a marketing message, I am wondering where to place past behaviour in the model.
I've had advice that I should control for past behaviour - and presumably this means add it as another predictor of behavioural intention?
But my own feeling is that past behaviour is a background factor influencing all of the other variables and shouldn't be included as a control variable.
Does anyone have similar experience of modelling an experiment where they have considered the role of past behaviour?
Many thanks for your thoughts.
Howard