How to analyse data with two replicates per group
Dear Simon, IMHO you should use individual as a random effect. If a systematic difference between both eyes are relevant / plausible, then add a variable coding for left or right to the fixed effect. Otherwise, don't include it in the model. Create a subset of the data which has no missing values in all variables used in your full model. Fit all your models on this subset. Then you can compare AIC. Best regards, ir. Thierry Onkelinx Statisticus / Statistician Vlaamse Overheid / Government of Flanders INSTITUUT VOOR NATUUR- EN BOSONDERZOEK / RESEARCH INSTITUTE FOR NATURE AND FOREST Team Biometrie & Kwaliteitszorg / Team Biometrics & Quality Assurance thierry.onkelinx at inbo.be Havenlaan 88 bus 73, 1000 Brussel www.inbo.be /////////////////////////////////////////////////////////////////////////////////////////// To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of. ~ Sir Ronald Aylmer Fisher The plural of anecdote is not data. ~ Roger Brinner The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data. ~ John Tukey /////////////////////////////////////////////////////////////////////////////////////////// <https://www.inbo.be> Op do 24 jan. 2019 om 17:30 schreef Simon POTIER <sim.potier at gmail.com>:
Dear all, Thanks in advance for your help. I will try to be precise, do not hesitate to tell me if you need more details. I have a dataset measuring some features of raptors eyes (let says here eye size for exemple). I have measured both eyes (left and right eyes) of 47 individuals. If I want to compare eye size with age and sex, what type of analysis should I use: 1) Should I analyse both eye separately (which can be interesting in term of vision)? 2) Can/should I analyse both eye together including Individual as a random effect? In other terms, I was wondering whether including individuals as random effect (with only two levels for each individual (left and right eye)) is possible? One more question, I want to compare the AIC of my models, but I get this error message: Warning message: In AIC.default(mm.edge, mm.edge2, mm.edge3) : models are not all fitted to the same number of observations I understand that the problem come from the fact that I have some NA in some fixed effect. How can I deal with this please because it is not possible to compare AIC? Best regards, Simon Potier -- ******************************************************* Simon Potier Department of Biology Lund University S?lvegatan 35 S-22362 Lund Sweden ******************************************************* phone +336 11 31 67 16 sim.potier at gmail.com https://www.simonpotier.fr ******************************************************* [[alternative HTML version deleted]]
_______________________________________________ R-sig-mixed-models at r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-sig-mixed-models