1

I am planning an experiment and need some feedback on the best way to design/ analyze the data. I want to test the effect of 2 drugs on males and females. Within females, I want to test the effect of estrous cycle (2 levels; in or out). Should I just cross ("sex" with three levels against drug with two levels). Should I nest estrous into sex even though males would not fave this factor?

kjetil b halvorsen
  • 63,378
  • 26
  • 142
  • 467
Robert C Spencer
  • 11
  • 2
  • Welcome to [stats.se]! Please take a moment to view our [tour]. By asking about in or out of the estrous cycle, are you defining out as before menarche, after menopause, or after full hysterectomy with ovaries removed? You may also want to consider the estrous cycle one of many noise factors to be aware of, and conduct a robust designed experiment with an inner and outer array where you could account for various hormones, weight, diet, &c. – Tavrock Mar 11 '17 at 05:52
  • @Robert C Spencer If you are interested in checking the effect of estrous cycle, why shoul you not exclude male - "Even though males would not have this factor.? " I do not have knowlege about medical science. More details with some description about the problem shall be apprecited. –  Mar 11 '17 at 08:11
  • 1
    males do not go through estrous. Thus a nested design would have 2 levels of estrous nested within females, but no corresponding levels for males. – Robert C Spencer Mar 12 '17 at 20:55
  • For the analysis part see https://stats.stackexchange.com/questions/372257/how-do-you-deal-with-nested-variables-in-a-regression-model/372258#372258 – kjetil b halvorsen Jun 09 '21 at 15:39

0 Answers0