I want to compare food consumption between three treatment groups. Each treatment group had 20 individuals, however, due to logistical constraints they were group housed in cages with 2-3 individuals per cage. I have information on the average mass of each individual, but the food consumption information is per cage. What is the best way to compare food consumption?

If I assign the average food consumption per cage to each individual, this is making the assumption that all individuals ate equal portions. If I treat cage as the experimental unit, I lose information on the number of individuals per cage, and individual mass, etc.

What is the best way to analyze this data without loosing information or making assumptions? Would it be acceptable to assign a weighted food consumption proportion to each individual? e.g. if individual A is 20 grams and individual B is 30 grams, and total cage food consumption is 100g, then assign A a food consumption value of 40 grams and B a food consumption value of 60 g?

  • 1
    $\begingroup$ Could you be a bit more specific about the context of your question? It doesn't seem like "best way to analyze this data" is a well defined problem without more knowledge about what you're trying to study. What are the treatments? What is the treatment effect you're interested in studying? Additionally, it would be helpful to be more precise with your description of how you chose treatment. Did you do simple randomization or some blocking design? Do individuals in the same cage have the same treatment status? It's hard to give a good answer without some clarity on these sorts of issues. $\endgroup$ – stats_model Apr 8 at 3:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.