class <- read.csv("S:\\dept\\Brady\\ALMMUSSP\\Chapters\\Data, Syntax, and Output\\Chapter 4\\classroom.csv", h = T) library(lme4) model4.1.fit.lmer <- lmer(mathgain ~ 1 + (1|schoolid) + (1|classid), class, REML = T) summary(model4.1.fit.lmer) ranef(model4.1.fit.lmer) ranef(model4.1.fit.lmer)$schoolid ranef(model4.1.fit.lmer)$classid model4.1A.fit.lmer <- lmer(mathgain ~ 1 + (1|schoolid), class, REML = T) anova(model4.1.fit.lmer, model4.1A.fit.lmer) model4.2.fit.lmer <- lmer(mathgain ~ mathkind + sex + minority + ses + (1|schoolid) + (1|classid), class, na.action = "na.omit", REML = T) summary(model4.2.fit.lmer) model4.1.lmer.ml.fit <- lmer(mathgain ~ 1 + (1|schoolid) + (1|classid), class, REML = F) model4.2.lmer.ml.fit <- lmer(mathgain ~ mathkind + sex + minority + ses + (1|schoolid) + (1|classid), class, REML = F) anova(model4.1.lmer.ml.fit, model4.2.lmer.ml.fit) model4.3.fit.lmer <- lmer(mathgain ~ mathkind + sex + minority + ses + yearstea + mathprep + mathknow + (1|schoolid) + (1|classid), class, na.action = "na.omit", REML = T) summary(model4.3.fit.lmer) model4.4.fit.lmer <- lmer(mathgain ~ mathkind + sex + minority + ses + housepov + (1|schoolid) + (1|classid), class, na.action = "na.omit", REML = T) summary(model4.4.fit.lmer) # OBSOLETE CODE # Use MCMC samples from the posterior distributions of the parameter estimates for the fitted model # to evaluate plausible values for the parameters: is 0 included between the 2.5% quantile and the # 97.5% quantile for a given parameter? library(coda) set.seed(101); samp1 <- mcmcsamp(model4.4.fit, n = 1000, deviance = TRUE) colnames(samp1) # has "deviance" if (require("coda", quietly = TRUE, character.only = TRUE)) { densityplot(samp1) qqmath(samp1) xyplot(samp1, scales = list(x = list(axs = 'i'))) print(summary(samp1)) print(autocorr.diag(samp1)) }