Function `glmer()` uses by default the Laplace approximation, which is not optimal for dichotomous data. A better alternative is the adaptive Gaussian quadrature. You can use this method by setting argument `nAGQ` of `glmer()` to a higher number (e.g., 11 or 15) or alternatively using the [**GLMMadaptive**][1] package. In your example, it gives: <!-- language-all: lang-r --> library("GLMMadaptive") helpmeobiwan <- list(NestPlot = c(1, 0, 0, 0, 0 ,0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,0, 0, 0, 0, 0, 0, 0, 0, 0),NumDeadJun = c( 0.1409216, -0.1932639,-0.5274494,-0.5274494, 0.1409216, -0.5274494, -0.5274494 , 0.4751071, -0.5274494 , 2.1460347 ,-0.5274494, -0.1932639, 0.8092926, -0.5274494, -0.5274494 ,-0.5274494 ,-0.1932639, 0.1409216, -0.5274494, -0.5274494 ,-0.5274494, -0.5274494 ,-0.5274494, 0.1409216,-0.5274494, -0.5274494 ,-0.5274494, 0.1409216, -0.5274494, 0.1409216, -0.5274494, -0.5274494, -0.5274494, -0.1932639, -0.1932639, -0.5274494, 0.4751071 , 0.1409216 ,-0.5274494, -0.5274494, -0.5274494, -0.5274494, -0.5274494, -0.5274494, -0.5274494, -0.5274494, -0.1932639, -0.5274494, -0.5274494 ,-0.5274494 ,-0.5274494, 0.1409216, -0.5274494, -0.5274494, -0.1932639, -0.5274494, -0.5274494, -0.5274494, 0.1409216, -0.5274494, -0.5274494 ,3.1485912 , 2.4802202, 1.4776637, -0.5274494 , 2.8144057, -0.5274494, -0.5274494, 1.1434781, 3.8169623, 3.8169623 ,-0.1932639, -0.5274494 ,1.4776637 , 1.8118492, -0.5274494),RandomPair = c( "Madera2" , "Starfire1", "Madera2" , "Madera3" , "Starfire1" ,"Starfire1", "Starfire2", "Madera1" , "Madera3" ,"Starfire2" ,"Starfire2", "Madera1", "Madera2", "Starfire1", "Starfire1" ,"Starfire1", "Madera1", "Madera2" , "Starfire1", "Starfire1", "Starfire1", "Madera1" , "Starfire1", "Starfire1", "Madera1", "Madera1" , "Starfire1", "Madera2" , "Madera1", "Madera2" , "Madera1" , "Madera1" , "Starfire1" ,"Starfire1", "Starfire1" ,"Starfire1" ,"Madera2" , "Madera2", "Starfire2" ,"Starfire2", "Starfire2" ,"Madera3" , "Madera3" , "Madera3" , "Madera3" , "Madera3" , "Starfire2", "Starfire2", "Starfire2", "Starfire2" ,"Starfire2", "Madera3", "Madera3" , "Starfire2", "Madera3" , "Madera1" , "Starfire2" ,"Starfire1", "Madera2" , "Madera3" , "Madera3" , "Madera2" , "Madera3" ,"Starfire2", "Madera3", "Starfire1", "Madera3" , "Starfire2", "Starfire1", "Madera3", "Starfire1", "Starfire2" ,"Madera1" , "Starfire2", "Starfire2", "Madera1" )) helpmeobiwan <- as.data.frame(helpmeobiwan) fm <- mixed_model(NestPlot ~ NumDeadJun, random = ~ 1 | RandomPair, family = binomial(), data = helpmeobiwan) summary(fm) #> #> Call: #> mixed_model(fixed = NestPlot ~ NumDeadJun, random = ~1 | RandomPair, #> data = helpmeobiwan, family = binomial()) #> #> Data Descriptives: #> Number of Observations: 76 #> Number of Groups: 5 #> #> Model: #> family: binomial #> link: logit #> #> Fit statistics: #> log.Lik AIC BIC #> -46.2248 98.44959 97.27791 #> #> Random effects covariance matrix: #> StdDev #> (Intercept) 0.0477673 #> #> Fixed effects: #> Estimate Std.Err z-value p-value #> (Intercept) -0.1568 0.2829 -0.5544 0.579304 #> NumDeadJun -1.2274 0.4917 -2.4961 0.012558 #> #> Integration: #> method: adaptive Gauss-Hermite quadrature rule #> quadrature points: 11 #> #> Optimization: #> method: hybrid EM and quasi-Newton #> converged: TRUE [1]: https://drizopoulos.github.io/GLMMadaptive/