Hello, I wonder if someone can help with the following. I have a study where individuals completed an exposures questionnaire at 2 time points. They responded y/n to whether they had experienced a list of exposures. I want to look at the reliability of their responses after ca. 18 months. The closest test I can think of is McNemar change test, but the HO is that the rate of change will not differ in either direction, which is not what I am after. My HO is that the change, in either direction, will not differ from zero. Does anyone know which test I should use? thanks Catherine ---------------------- Catherine E. Unwin [log in to unmask] %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%