Hi,

Apologies if this is very obvious, I've managed to get confused about it. If i want to check if two categorical exposure variables are independent of each other prior to putting in a log regression model, can I just do a Chi2? And if it's significant, they are not independent.

Thanks,
Carla