I'm experiencing difficulties with running a mixed model in SPSS (version 26).
My experiment is a 2x2 factorial design in which we provide light and/or larvae to chickens. I have individual (continuous) measurements from behavior tests as outcome parameters. There are no repeated measures. But since the chickens are housed in groups (pens), and the light and larvae are provided on pen level, these individual measurements within a pen are not independent. Therefore I need to run a linear mixed model with light, larvae and their interaction as fixed effects and pen as random effect.
However, SPSS gives me the following warning:
"The final Hessian matrix is not positive definite although all convergence criteria are satisfied. the MIXED procedure continues despite this warning. Validity of subsequent results cannot be ascertained."
It probably says so because the pens are linked to the light/larvae treatment. But it should be possible to run this analysis in some way, right? Can anyone help me find out what I'm doing wrong?
Thank you so much!
