Simulation for De la Fuente (2021)
Year: 2026
Type: preprint
Abstract: This simulation projects looks at the robustness of age effect in the De la Fuente (2021, Journal of Animal Ecology) paper " Foraging networks and social tolerance in a cooperatively breeding primate (Callithrix jacchus)" Hypothesis 1b. In the project, I first refitted the original MCMCglmm with original Individual dataset provided by the author. I then refitted an extended model by added Individual ID to the random effects. Next, I simulated the node degree by keeping other main effects while setting age effect = 0. I simulated 100 datasets and refit each of them to both the original model and the extended model (200 run in total), and recorded the age effect in each model. I defined "significant" result as p < 0.05 and 95% CI excluding 0. Finally, I calculated the false positive rate of the age effect. My reproduction should be slight different from the original result. This is because without the original codes, particularly the prior, my result of refitting the original models won't reach the exact same result, but as least the significance and posterior mean of each variable should be close. For the simulation I expect the false positive rate of the age effects to be low for both original and extended models, specifically <5%. The authors claimed that they ran a variety of models to reach the most appropriate one. Thus even though they didn't add Individual ID in the random effect to account or individual variation, the model should be robust.
Source: OSF Preprints (OSF Preprints)
Author: Fei Xu
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7225895200
