>4% adjusted A lot of times false precision bugs me but here I think you are comparing fine differences and need a bit more. Using your figures, (36/33)/(329/

>4% adjusted A lot of times false precision bugs me but here I think you are comparing fine differences and need a bit more. Using your figures, (36/33)/(329/309), I get 2.5%, not 4%.