Projects for metropolis: 
11 of
1
shown.



Hierarchical linear model of MPA effects
by brice.semmens, last updated 2/5/08, sharing set to public
These are a series of Matlab routines to carry out a hierarchical linear model of fish population trajectories that includes a linear adjustment for a "reserve effect" at sites designated as MPAs. %The model is set up as follows: %FIRST LEVEL %MODEL:***** y~ intercept + abundance_index * (S_g + R_t +R_nt) + error *** %where: %S_g is the slope for a given species in a given geographic region (drawn %from a normally distributed species specific hyperparameter) %R_t is a species specific additive adjustment to the slope that is drawn %from a normally distributed hyper parameter representing the reserve effect on harvested species. Activated by a dummy variable. %R_nt is the species specific additive adjustment to the slope that is drawn from a normally distributed hyper parameter %representing the reserve effect on nonharvested species. Activated by a dummy variable. %both intercept and error terms are nuisance parameters and do not need hyperparameters. I suppose you could put in an %error hyper parameter with an informative prior just to keep the model in check. %SECOND LEVEL (hyper parameters all normally distributed) %S_g_hat  These are slopes for a given species from which S_g's are drawn %R_t_hat  This is the reserve effect for targeted species from which R_t's across targeted species are drawn %R_nt_hat  This is the reserve effect for nontarged species from which %R_nt's across nontargeted species are drawn %
Projects for metropolis: 
11 of
1
shown.

