@FinnLindgren
As we spoke about this afternoon, I created a model to predict tick abundance in mainland Scotland - It has a zero inflated Poisson distribution (zero inflated Poisson1) with two random effects, the effect of the site of tick collection and the number of samples per site (apart from the fixed effects).
The dataset contains information on tick counts obtained from tick surveys in which sites for tick collection were selected over Scotland and in each site several samples were performed. The table contains information on 686 sites with multiple samples in each site, a total of 10611 samples (each sample is a row). The measure of interest is the count of ticks per sample.
The model is:
Model <- bru(count ~ Intercept + frost + rain + deer + forest + mysmooth + site + sample
One of the questions is: inlabru model of tick relative abundance is extremely influenced by the random effects, the site and the number of samples per site, what I mean is that the predictive values have higher influence from the random effects than from the fixed effects (climatic, hosts and habitat variables). If I predict to the geographical space without considering the two random effects, I obtain values of tick per sample of 0 – 10, but if I add the random effects to the equation, the model predicts values 8 times higher (0 – 80). The question is: what is happening internally in the model? Why this huge effect of the random effects, especially the site effect?
Thank you.
Rita
@FinnLindgren As we spoke about this afternoon, I created a model to predict tick abundance in mainland Scotland - It has a zero inflated Poisson distribution (zero inflated Poisson1) with two random effects, the effect of the site of tick collection and the number of samples per site (apart from the fixed effects). The dataset contains information on tick counts obtained from tick surveys in which sites for tick collection were selected over Scotland and in each site several samples were performed. The table contains information on 686 sites with multiple samples in each site, a total of 10611 samples (each sample is a row). The measure of interest is the count of ticks per sample. The model is: Model <- bru(count ~ Intercept + frost + rain + deer + forest + mysmooth + site + sample
One of the questions is: inlabru model of tick relative abundance is extremely influenced by the random effects, the site and the number of samples per site, what I mean is that the predictive values have higher influence from the random effects than from the fixed effects (climatic, hosts and habitat variables). If I predict to the geographical space without considering the two random effects, I obtain values of tick per sample of 0 – 10, but if I add the random effects to the equation, the model predicts values 8 times higher (0 – 80). The question is: what is happening internally in the model? Why this huge effect of the random effects, especially the site effect? Thank you. Rita