Citation Context

... fit with thesLaplace approximation. The Laplace approximation, similar to maximum likelihoodsestimation, is an iterative method of parameter estimation useful in analyses of Poissondistributed data (=-=Bolker et al., 2008-=-).sIn social sciences it is not uncommon for count data to exhibit variance greatersthan the mean (in other words, greater variation than would be expected, assumingsPoisson-distributed data). This is...

Citation Context

...rms in particular. Out of the random factors,sYear seemed to have only a minor effect on the function, so I left it out as well. ID and position datasremained.sI used generalized linear mixed models (=-=Bolker et al. 2009-=-) and a model inference within ansinformation-theoretical framework (Burnham & Anderson 2002), to asses simultaneously the role ofsweather variables and the nesting success of both study and control a...

Citation Context

... switching behavior using a Generalized Linear Mixed Model (GLMM) with switching as a binary dependent variable (0=no switch, 1=switch) and player identity as a random effect affecting the intercept (=-=Bolker et al., 2009-=-). We assumed a beta binomial distribution (to correct for over-dispersion) with logit link function (Gelman & Hill, 2007). Independent variables were the frequency change in demonstrator choices for ...

Citation Context

...the forest edge. GLME’s have fewer assumptions than traditionalsregression (Pinheiro and Bates 2000) and provide a more flexible approach to analyzingsnon-normal data when random effects are present (=-=Bolker et al. 2009-=-). In GLME, fixedseffects are factors that describe experimental treatments or are sources of systematicsvariance; random effects are associated with particular experimental units that aresselected at...