Okay, so i have to decide whether i should use chi-square or Cochran's Q test for one of my study.
If i assume meat as experimental unit (EU) and impose 6 different wood smoke (like apple smoke, cedar smoke, hickory smoke etc.) on EU and consumers (people) act as observational unit (OU). I have 10 consumers and each one will taste each treatment (6 treatments) and respond liking (9 point scale), and other variables such as if that particular treatment was smoky, fruity, burnt, tender etc. Should i use chi-square because treatments were independent.
Another case, if i assume consumer as EU and apply all 6 treatments (say 6 types of apples) on it, so now i think i am violating independent assumption, so i should use Cochran's Q test. Again each consumer will taste all varieties and respond liking, and other response variables such as sweet, sour, bitter etc.
