########################################################### ########################################################### ############ PHONEMIC ANALYSIS OF KOREAN NAMES ############ ########################################################### ###################### LISA SULLIVAN ###################### ###################### July 13, 2020 ###################### ########################################################### ########################################################### ###################################################### ############ READ FILES FOR ANALYSIS ################# ###################################################### read.delim("phonological_full_coded_2020_05_07.txt",header=T,as.is = T, comment.char = "#" )->dat_phonological_full read.delim("phonological_s1_coded_2020_05_07.txt",header=T,as.is = T, comment.char = "#" )->dat_phonological_s1 read.delim("phonological_s2_coded_2020_05_07.txt",header=T,as.is = T, comment.char = "#" )->dat_phonological_s2 ############################################ ############ CART ANALYSIS ################# ############################################ ######## Factor Everything ########## dat_phonological_full$open_syll_cnt=factor(dat_phonological_full$open_syll_cnt) dat_phonological_full$initial_snd_type=factor(dat_phonological_full$initial_snd_type) dat_phonological_full$final_snd_type=factor(dat_phonological_full$final_snd_type) dat_phonological_full$syll_1_type=factor(dat_phonological_full$syll_1_type) dat_phonological_full$dark_v_cnt=factor(dat_phonological_full$dark_v_cnt) dat_phonological_full$light_v_cnt=factor(dat_phonological_full$light_v_cnt) dat_phonological_full$front_v_cnt=factor(dat_phonological_full$front_v_cnt) dat_phonological_full$back_v_cnt=factor(dat_phonological_full$back_v_cnt) dat_phonological_full$high_v_cnt=factor(dat_phonological_full$high_v_cnt) dat_phonological_full$low_v_cnt=factor(dat_phonological_full$low_v_cnt) dat_phonological_full$round_v_cnt=factor(dat_phonological_full$round_v_cnt) dat_phonological_full$diphthongs=factor(dat_phonological_full$diphthongs) dat_phonological_full$w_cnt=factor(dat_phonological_full$w_cnt) dat_phonological_full$j_cnt=factor(dat_phonological_full$j_cnt) dat_phonological_full$stops=factor(dat_phonological_full$stops) dat_phonological_full$sonorants=factor(dat_phonological_full$sonorants) dat_phonological_full$aspirated=factor(dat_phonological_full$aspirated) dat_phonological_full$tense=factor(dat_phonological_full$tense) dat_phonological_full$plain=factor(dat_phonological_full$plain) dat_phonological_full$Sex=factor(dat_phonological_full$Sex) dat_phonological_s1$initial_snd_type=factor(dat_phonological_s1$initial_snd_type) dat_phonological_s1$syll_1_type=factor(dat_phonological_s1$syll_1_type) dat_phonological_s1$dark_v_cnt=factor(dat_phonological_s1$dark_v_cnt) dat_phonological_s1$light_v_cnt=factor(dat_phonological_s1$light_v_cnt) dat_phonological_s1$front_v_cnt=factor(dat_phonological_s1$front_v_cnt) dat_phonological_s1$back_v_cnt=factor(dat_phonological_s1$back_v_cnt) dat_phonological_s1$high_v_cnt=factor(dat_phonological_s1$high_v_cnt) dat_phonological_s1$low_v_cnt=factor(dat_phonological_s1$low_v_cnt) dat_phonological_s1$round_v_cnt=factor(dat_phonological_s1$round_v_cnt) dat_phonological_s1$diphthongs=factor(dat_phonological_s1$diphthongs) dat_phonological_s1$w_cnt=factor(dat_phonological_s1$w_cnt) dat_phonological_s1$j_cnt=factor(dat_phonological_s1$j_cnt) dat_phonological_s1$stops=factor(dat_phonological_s1$stops) dat_phonological_s1$sonorants=factor(dat_phonological_s1$sonorants) dat_phonological_s1$aspirated=factor(dat_phonological_s1$aspirated) dat_phonological_s1$tense=factor(dat_phonological_s1$tense) dat_phonological_s1$plain=factor(dat_phonological_s1$plain) dat_phonological_s1$Sex=factor(dat_phonological_s1$Sex) dat_phonological_s1$height=factor(dat_phonological_s1$height) dat_phonological_s1$backness=factor(dat_phonological_s1$backness) dat_phonological_s1$dark_bright=factor(dat_phonological_s1$dark_bright) dat_phonological_s1$vowel_length=factor(dat_phonological_s1$vowel_length) dat_phonological_s2$initial_snd_type=factor(dat_phonological_s2$initial_snd_type) dat_phonological_s2$syll_2_type=factor(dat_phonological_s2$syll_2_type) dat_phonological_s2$dark_v_cnt=factor(dat_phonological_s2$dark_v_cnt) dat_phonological_s2$light_v_cnt=factor(dat_phonological_s2$light_v_cnt) dat_phonological_s2$front_v_cnt=factor(dat_phonological_s2$front_v_cnt) dat_phonological_s2$back_v_cnt=factor(dat_phonological_s2$back_v_cnt) dat_phonological_s2$high_v_cnt=factor(dat_phonological_s2$high_v_cnt) dat_phonological_s2$low_v_cnt=factor(dat_phonological_s2$low_v_cnt) dat_phonological_s2$round_v_cnt=factor(dat_phonological_s2$round_v_cnt) dat_phonological_s2$diphthongs=factor(dat_phonological_s2$diphthongs) dat_phonological_s2$w_cnt=factor(dat_phonological_s2$w_cnt) dat_phonological_s2$j_cnt=factor(dat_phonological_s2$j_cnt) dat_phonological_s2$stops=factor(dat_phonological_s2$stops) dat_phonological_s2$sonorants=factor(dat_phonological_s2$sonorants) dat_phonological_s2$aspirated=factor(dat_phonological_s2$aspirated) dat_phonological_s2$tense=factor(dat_phonological_s2$tense) dat_phonological_s2$plain=factor(dat_phonological_s2$plain) dat_phonological_s2$Sex=factor(dat_phonological_s2$Sex) dat_phonological_s2$height=factor(dat_phonological_s2$height) dat_phonological_s2$backness=factor(dat_phonological_s2$backness) dat_phonological_s2$dark_bright=factor(dat_phonological_s2$dark_bright) dat_phonological_s2$vowel_length=factor(dat_phonological_s2$vowel_length) ######## Full Names ######## ### Construct & plot the full tree demtree_2s = rpart(Sex~open_syll_cnt+dark_v_cnt+light_v_cnt+front_v_cnt+back_v_cnt+high_v_cnt+low_v_cnt+round_v_cnt+diphthongs+w_cnt+j_cnt+stops+sonorants+aspirated+tense+plain, data = dat_phonological_full) plot(demtree_2s, compress= T, branch = 1, margin = 0.1) text(demtree_2s, use.n=T,pretty=0)
### Run 10-part cross-validation to prune the full tree plotcp(demtree_2s)
### Prune the tree and replot the data demtree_2s = prune(demtree_2s, cp = 0.015) plot(demtree_2s, compress = T, branch = 1, margin = 0.1) text(demtree_2s, use.n=T,pretty=0)
### Get an estimate of the predictive power of the tree based on the current data predict(demtree_2s, type = "class") -> predValues_2s #Predicted Values preds_2s = data.frame(obs=dat_phonological_full$Sex, pred=predValues_2s) #Observed vs Predicted xtabs(~obs+pred, preds_2s) #Cross-tabs of observed vs predicted
## pred ## obs F M ## F 608 366 ## M 307 663
(780+485)/nrow(dat_phonological_full)# Proportion correct
## [1] 0.6507202
######## Syllable 1 ######## demtree_s1 = rpart(Sex~syll_1_type+initial_snd_type+dark_bright+backness+height+round_v_cnt+vowel_length+stops+sonorants+aspirated+tense+plain, data = dat_phonological_s1) plot(demtree_s1, compress= T, branch = 1, margin = 0.1) text(demtree_s1, use.n=T,pretty=0)
### Run 10-part cross-validation to prune the full tree plotcp(demtree_s1)
### Prune the tree and replot the data demtree_s1 = prune(demtree_s1, cp = 0.016) plot(demtree_s1, compress = T, branch = 1, margin = 0.1) text(demtree_s1, use.n=T,pretty=0)
### Get an estimate of the predictive power of the tree based on the current data predict(demtree_s1, type = "class") -> predValues_s1 #Predicted Values preds_s1 = data.frame(obs=dat_phonological_s1$Sex, pred=predValues_s1) #Observed vs Predicted xtabs(~obs+pred, preds_s1) #Cross-tabs of observed vs predicted
## pred ## obs F M ## F 622 352 ## M 365 605
(672+542)/nrow(dat_phonological_s1)# Proportion correct
## [1] 0.6244856
######## Syllable 2 ######## demtree_s2 = rpart(Sex~syll_2_type+initial_snd_type+dark_bright+backness+height+round_v_cnt+vowel_length+stops+sonorants+aspirated+tense+plain, data = dat_phonological_s2) plot(demtree_s2, compress= T, branch = 1, margin = 0.1) text(demtree_s2, use.n=T,pretty=0)
### Run 10-part cross-validation to prune the full tree plotcp(demtree_s2)
### Prune the tree and replot the data demtree_s2 = prune(demtree_s2, cp = 0.02) plot(demtree_s2, compress = T, branch = 1, margin = 0.1) text(demtree_s2, use.n=T,pretty=0)
### Get an estimate of the predictive power of the tree based on the current data predict(demtree_s2, type = "class") -> predValues_s2 #Predicted Values preds_s2 = data.frame(obs=dat_phonological_s2$Sex, pred=predValues_s2) #Observed vs Predicted xtabs(~obs+pred, preds_s2) #Cross-tabs of observed vs predicted
## pred ## obs F M ## F 604 370 ## M 244 726
(604+725)/nrow(dat_phonological_s2)# Proportion correct
## [1] 0.683642
####################################################### ################# UNIVARITE ANALYSIS ################## ####################################################### ############ Center Variables ############# contrasts(dat_phonological_full$open_syll_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$initial_snd_type)=contrasts(dat_phonological_full$initial_snd_type)-1/2 contrasts(dat_phonological_full$final_snd_type)=contrasts(dat_phonological_full$final_snd_type)-1/2 contrasts(dat_phonological_full$syll_1_type)=contrasts(dat_phonological_full$syll_1_type)-1/2 contrasts(dat_phonological_full$dark_v_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$light_v_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$front_v_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$back_v_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$high_v_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$low_v_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$round_v_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$diphthongs)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$w_cnt)=contrasts(dat_phonological_full$w_cnt)-1/2 contrasts(dat_phonological_full$j_cnt)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$stops)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_full$sonorants)=cbind(zero=c(3/4,-1/4,-1/4,-1/4), one=c(0,2/3,-1/3,-1/3), two_three=c(0,0,1/2,-1/2)) contrasts(dat_phonological_full$aspirated)=contrasts(dat_phonological_full$aspirated)-1/2 contrasts(dat_phonological_full$tense)=contrasts(dat_phonological_full$tense)-1/2 contrasts(dat_phonological_full$plain)=cbind(zero=c(3/4,-1/4,-1/4,-1/4), one=c(0,2/3,-1/3,-1/3), two_three=c(0,0,1/2,-1/2)) contrasts(dat_phonological_s1$initial_snd_type)=contrasts(dat_phonological_s1$initial_snd_type)-1/2 contrasts(dat_phonological_s1$syll_1_type)=contrasts(dat_phonological_s1$syll_1_type)-1/2 contrasts(dat_phonological_s1$dark_v_cnt)=contrasts(dat_phonological_s1$dark_v_cnt)-1/2 contrasts(dat_phonological_s1$light_v_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$front_v_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$back_v_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$high_v_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$low_v_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$round_v_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$diphthongs)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$w_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$j_cnt)=contrasts(dat_phonological_s1$light_v_cnt)-1/2 contrasts(dat_phonological_s1$stops)=contrasts(dat_phonological_s1$stops)-1/2 contrasts(dat_phonological_s1$sonorants)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_s1$aspirated)=contrasts(dat_phonological_s1$aspirated)-1/2 ###contrasts(dat_phonological_s1$tense)=contrasts(dat_phonological_s1$tense)-1/2 contrasts(dat_phonological_s1$plain)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_s1$height)=cbind(high=c(2/3,-1/3,-1/3), low_mid=c(0,1/2,-1/2)) contrasts(dat_phonological_s1$backness)=cbind(front=c(-1/3,-1/3,2/3), back_central=c(1/2,-1/2,0)) contrasts(dat_phonological_s1$dark_bright)=cbind(neutral=c(-1/3,-1/3,2/3), bright_dark=c(1/2,-1/2,0)) contrasts(dat_phonological_s1$vowel_length)=cbind(mono_diph=c(-1/3,2/3,-1/3), j_w=c(1/2,0,-1/2)) ##### ******** NO TENSE IN SYLL 1 contrasts(dat_phonological_s2$initial_snd_type)=contrasts(dat_phonological_s2$initial_snd_type)-1/2 contrasts(dat_phonological_s2$syll_2_type)=contrasts(dat_phonological_s2$syll_2_type)-1/2 contrasts(dat_phonological_s2$dark_v_cnt)=contrasts(dat_phonological_s2$dark_v_cnt)-1/2 contrasts(dat_phonological_s2$light_v_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$front_v_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$back_v_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$high_v_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$low_v_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$round_v_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$diphthongs)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$w_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$j_cnt)=contrasts(dat_phonological_s2$light_v_cnt)-1/2 contrasts(dat_phonological_s2$stops)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_s2$sonorants)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_s2$aspirated)=contrasts(dat_phonological_s2$aspirated)-1/2 contrasts(dat_phonological_s2$tense)=contrasts(dat_phonological_s2$tense)-1/2 contrasts(dat_phonological_s2$plain)=cbind(zero=c(2/3,-1/3,-1/3), one_two=c(0,1/2,-1/2)) contrasts(dat_phonological_s2$height)=cbind(high=c(2/3,-1/3,-1/3), low_mid=c(0,1/2,-1/2)) contrasts(dat_phonological_s2$backness)=cbind(front=c(-1/3,-1/3,2/3), back_central=c(1/2,-1/2,0)) contrasts(dat_phonological_s2$dark_bright)=cbind(neutral=c(-1/3,-1/3,2/3), bright_dark=c(1/2,-1/2,0)) contrasts(dat_phonological_s2$vowel_length)=cbind(mono_diph=c(-1/3,2/3,-1/3), j_w=c(1/2,0,-1/2)) ######### Full Names ######## # Number of syllables lm_2s_syll_type = glmer(Sex~open_syll_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_syll_type)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ open_syll_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1755.8 1783.7 -872.9 1745.8 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.8265 -0.3946 -0.0299 0.3757 3.2022 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.504 2.550 ## syllable_1 (Intercept) 2.637 1.624 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.09386 0.31328 0.300 0.764 ## open_syll_cntzero 2.38199 0.48293 4.932 8.12e-07 *** ## open_syll_cntone_two 1.66310 0.36958 4.500 6.80e-06 *** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) opn_s_ ## opn_syll_cn -0.064 ## opn_syll_c_ -0.150 0.810
dat_phonological_full %>% ggplot(aes(open_syll_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Open Syllables") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Open Syllable Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Dark Vowels lm_2s_dark_v = glmer(Sex~dark_v_cnt + (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_dark_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ dark_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1775.3 1803.1 -882.6 1765.3 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6200 -0.3950 -0.0352 0.3785 2.6749 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.927 2.632 ## syllable_1 (Intercept) 3.220 1.794 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.3962 0.3321 1.193 0.2329 ## dark_v_cntzero -1.0847 0.5026 -2.158 0.0309 * ## dark_v_cntone_two -0.6858 0.3731 -1.838 0.0660 . ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) drk_v_ ## drk_v_cntzr -0.219 ## drk_v_cntn_ -0.236 0.861
dat_phonological_full %>% ggplot(aes(dark_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Dark Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Light Vowels lm_2s_light_v= glmer(Sex~light_v_cnt + (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_light_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ light_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1779.7 1807.6 -884.9 1769.7 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6629 -0.3997 -0.0347 0.3769 2.6094 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.407 2.722 ## syllable_1 (Intercept) 3.107 1.763 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.268648 0.338389 0.794 0.427 ## light_v_cntzero -0.008313 0.495251 -0.017 0.987 ## light_v_cntone_two -0.099583 0.445895 -0.223 0.823 ## ## Correlation of Fixed Effects: ## (Intr) lght__ ## lght_v_cntz -0.228 ## lght_v_cnt_ -0.266 0.811
dat_phonological_full %>% ggplot(aes(light_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Light Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Front Vowels lm_2s_front_v = glmer(Sex~front_v_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_front_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ front_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1772.0 1799.8 -881.0 1762.0 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.7754 -0.4031 -0.0301 0.3761 2.4809 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.672 2.583 ## syllable_1 (Intercept) 3.261 1.806 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.2925 0.3739 -0.782 0.43409 ## front_v_cntzero 1.4522 0.5706 2.545 0.01093 * ## front_v_cntone_two 1.3494 0.4735 2.850 0.00438 ** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) frnt__ ## frnt_v_cntz -0.500 ## frnt_v_cnt_ -0.498 0.858
dat_phonological_full %>% ggplot(aes(front_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Front Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Back Vowels lm_2s_back_v = glmer(Sex~back_v_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_back_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ back_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1773.3 1801.1 -881.6 1763.3 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.5336 -0.3993 -0.0340 0.3702 2.6728 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.739 2.596 ## syllable_1 (Intercept) 3.195 1.787 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.5693 0.3695 1.541 0.1234 ## back_v_cntzero -1.2824 0.5583 -2.297 0.0216 * ## back_v_cntone_two -0.6399 0.5195 -1.232 0.2180 ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) bck_v_ ## bck_v_cntzr -0.472 ## bck_v_cntn_ -0.487 0.836
dat_phonological_full %>% ggplot(aes(back_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Back Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of High Vowels lm_2s_high_v = glmer(Sex~high_v_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_high_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ high_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1779.8 1807.6 -884.9 1769.8 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6283 -0.4011 -0.0353 0.3775 2.5993 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.375 2.716 ## syllable_1 (Intercept) 3.112 1.764 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.24337 0.33959 0.717 0.474 ## high_v_cntzero 0.07557 0.50699 0.149 0.882 ## high_v_cntone_two 0.01690 0.36671 0.046 0.963 ## ## Correlation of Fixed Effects: ## (Intr) hgh_v_ ## hgh_v_cntzr -0.256 ## hgh_v_cntn_ -0.281 0.856
dat_phonological_full %>% ggplot(aes(high_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("High Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Low Vowels lm_2s_low_v = glmer(Sex~low_v_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_low_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ low_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1779.6 1807.5 -884.8 1769.6 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6421 -0.3981 -0.0351 0.3752 2.6226 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.377 2.716 ## syllable_1 (Intercept) 3.123 1.767 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.1756 0.3927 0.447 0.655 ## low_v_cntzero 0.2307 0.5793 0.398 0.691 ## low_v_cntone_two 0.1153 0.6026 0.191 0.848 ## ## Correlation of Fixed Effects: ## (Intr) lw_v_c ## low_v_cntzr -0.540 ## lw_v_cntn_t -0.516 0.802
dat_phonological_full %>% ggplot(aes(low_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Low Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Round Vowels lm_2s_round_v = glmer(Sex~round_v_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_round_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ round_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1773.0 1800.9 -881.5 1763.0 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.5141 -0.3995 -0.0341 0.3708 2.6729 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.707 2.590 ## syllable_1 (Intercept) 3.196 1.788 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.5744 0.3666 1.567 0.1171 ## round_v_cntzero -1.3218 0.5545 -2.384 0.0171 * ## round_v_cntone_two -0.6904 0.5148 -1.341 0.1799 ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) rnd_v_ ## rnd_v_cntzr -0.460 ## rnd_v_cntn_ -0.478 0.835
dat_phonological_full %>% ggplot(aes(round_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Round Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Diphthongs lm_2s_diphthong = glmer(Sex~diphthongs+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_diphthong)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ diphthongs + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1779.4 1807.3 -884.7 1769.4 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.7076 -0.4017 -0.0345 0.3761 2.5881 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.429 2.726 ## syllable_1 (Intercept) 3.063 1.750 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.18730 0.38175 0.491 0.624 ## diphthongszero 0.26148 0.56209 0.465 0.642 ## diphthongsone_two 0.05778 0.48659 0.119 0.905 ## ## Correlation of Fixed Effects: ## (Intr) dphthn ## diphthngszr -0.501 ## dphthngsn_t -0.499 0.835
dat_phonological_full %>% ggplot(aes(diphthongs,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Diphthongs") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Diphthong Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of /w/ Diphthongs lm_2s_w = glmer(Sex~w_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_w)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ w_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1771.9 1794.2 -881.9 1763.9 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6014 -0.3978 -0.0353 0.3686 2.5868 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.341 2.709 ## syllable_1 (Intercept) 3.010 1.735 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 1.1530 0.5117 2.253 0.0242 * ## w_cnt1 2.1387 0.8972 2.384 0.0171 * ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) ## w_cnt1 0.751
dat_phonological_full %>% ggplot(aes(w_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Diphthongs") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("/w/ Diphthong Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of /j/ Diphthongs lm_2s_j = glmer(Sex~j_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_j)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ j_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1776.7 1804.6 -883.4 1766.7 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.7554 -0.3994 -0.0335 0.3747 2.5721 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.472 2.733 ## syllable_1 (Intercept) 2.959 1.720 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.07062 0.41251 -0.171 0.864 ## j_cntzero 0.93861 0.60554 1.550 0.121 ## j_cntone_two 0.44428 0.56571 0.785 0.432 ## ## Correlation of Fixed Effects: ## (Intr) j_cntz ## j_cntzero -0.597 ## j_cntone_tw -0.587 0.842
dat_phonological_full %>% ggplot(aes(j_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Diphthongs") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("/j/ Diphthong Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Stops lm_2s_stops = glmer(Sex~stops+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_stops)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ stops + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1759.7 1787.6 -874.8 1749.7 1939 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.4598 -0.3974 -0.0331 0.3618 2.6461 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.915 2.630 ## syllable_1 (Intercept) 2.861 1.692 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 1.0578 0.4030 2.625 0.00867 ** ## stopszero -2.4090 0.5969 -4.036 5.44e-05 *** ## stopsone_two -1.3533 0.6049 -2.237 0.02528 * ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) stpszr ## stopszero -0.560 ## stopsone_tw -0.527 0.805
dat_phonological_full %>% ggplot(aes(stops,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Stops") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Stop Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Sonorants lm_2s_sonorants = glmer(Sex~sonorants+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_sonorants)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ sonorants + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1779.3 1812.7 -883.6 1767.3 1938 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6234 -0.3965 -0.0355 0.3753 2.6478 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.495 2.738 ## syllable_1 (Intercept) 3.019 1.738 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.2211 0.3473 0.637 0.524 ## sonorantszero -0.7928 0.6009 -1.319 0.187 ## sonorantsone -0.2905 0.4772 -0.609 0.543 ## sonorantstwo_three -0.2218 0.5235 -0.424 0.672 ## ## Correlation of Fixed Effects: ## (Intr) snrntsz snrntsn ## sonorantszr -0.136 ## sonorantson -0.263 0.855 ## snrntstw_th -0.315 0.622 0.790
dat_phonological_full %>% ggplot(aes(sonorants,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Sonorants") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Sonorant Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Lenis Obstruents lm_2s_plain = glmer(Sex~plain+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_plain)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ plain + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1767.0 1800.4 -877.5 1755.0 1938 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.5057 -0.3949 -0.0362 0.3746 2.7030 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.344 2.519 ## syllable_1 (Intercept) 3.065 1.751 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.9280 0.4788 1.938 0.052613 . ## plainzero -2.6796 0.7483 -3.581 0.000342 *** ## plainone -2.3975 0.8262 -2.902 0.003709 ** ## plaintwo_three -2.7056 1.4868 -1.820 0.068808 . ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) planzr plainn ## plainzero -0.518 ## plainone -0.691 0.875 ## plaintw_thr -0.727 0.678 0.905
dat_phonological_full %>% ggplot(aes(plain,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Lenis Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Fortis Obstruents lm_2s_tense = glmer(Sex~tense+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_tense)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ tense + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1774.6 1796.9 -883.3 1766.6 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6473 -0.3997 -0.0343 0.3751 2.6038 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.288 2.700 ## syllable_1 (Intercept) 3.175 1.782 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -8.699 21.116 -0.412 0.68 ## tense1 -17.969 42.229 -0.426 0.67 ## ## Correlation of Fixed Effects: ## (Intr) ## tense1 1.000
dat_phonological_full %>% ggplot(aes(tense,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Fortis Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Aspirated Obstruents lm_2s_aspirated = glmer(Sex~aspirated+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_aspirated)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ aspirated + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1775.1 1797.3 -883.5 1767.1 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6305 -0.3995 -0.0341 0.3728 2.5979 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.348 2.711 ## syllable_1 (Intercept) 3.040 1.744 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.7217 0.4257 1.695 0.0900 . ## aspirated1 1.1961 0.6995 1.710 0.0873 . ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) ## aspirated1 0.648
dat_phonological_full %>% ggplot(aes(aspirated,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Aspirated Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
######### Syllable #1 ########## # Initial Sound lm_s1_initial_snd = glmer(Sex~initial_snd_type + (1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_initial_snd)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ initial_snd_type + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2369.0 2385.7 -1181.5 2363.0 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.0116 -0.7856 -0.2091 0.7411 3.1583 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.712 1.647 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.1608 0.2323 -0.692 0.489 ## initial_snd_typeV -0.1952 0.4637 -0.421 0.674 ## ## Correlation of Fixed Effects: ## (Intr) ## intl_snd_tV 0.619
dat_phonological_s1 %>% ggplot(aes(initial_snd_type,fill=Sex)) + geom_bar(position="fill") + xlab("Sound Type") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Initial Sound Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Syllable Type lm_s1_initial_syll = glmer(Sex~syll_1_type+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_initial_syll)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ syll_1_type + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2354.1 2370.8 -1174.0 2348.1 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.0745 -0.7792 -0.2082 0.7314 3.2046 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.178 1.476 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.01856 0.16809 -0.110 0.912 ## syll_1_typeOpen -1.35091 0.33807 -3.996 6.44e-05 *** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) ## syll_1_typO -0.091
dat_phonological_s1 %>% ggplot(aes(syll_1_type,fill=Sex)) + geom_bar(position="fill") + xlab("Syllable Type") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Syllable Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Height lm_s1_height = glmer(Sex~height+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_height)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ height + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2369.6 2391.9 -1180.8 2361.6 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.0269 -0.7818 -0.2105 0.7403 3.1439 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.699 1.643 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.1343 0.1850 -0.726 0.468 ## heighthigh 0.3245 0.3894 0.833 0.405 ## heightlow_mid -0.4812 0.4554 -1.057 0.291 ## ## Correlation of Fixed Effects: ## (Intr) hghthg ## heighthigh -0.023 ## heightlw_md 0.178 -0.127
dat_phonological_s1 %>% ggplot(aes(height,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Height") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Height") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Backness lm_s1_backness = glmer(Sex~backness+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_backness)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ backness + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2371.1 2393.4 -1181.6 2363.1 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.0050 -0.7862 -0.2081 0.7409 3.1842 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.719 1.649 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.11043 0.18737 -0.589 0.556 ## backnessfront -0.11543 0.41100 -0.281 0.779 ## backnessback_central -0.03756 0.44058 -0.085 0.932 ## ## Correlation of Fixed Effects: ## (Intr) bcknss ## backnssfrnt 0.102 ## bcknssbck_c 0.185 -0.128
dat_phonological_s1 %>% ggplot(aes(backness,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Backness") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Backness") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
#Vowel Type (Dark - Light - Neutral) lm_s1_darkness = glmer(Sex~dark_bright+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_darkness)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ dark_bright + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2369.1 2391.4 -1180.6 2361.1 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.8996 -0.7827 -0.2007 0.7378 3.1056 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.691 1.64 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.08165 0.19137 -0.427 0.670 ## dark_brightneutral 0.07024 0.45194 0.155 0.876 ## dark_brightbright_dark -0.58705 0.40843 -1.437 0.151 ## ## Correlation of Fixed Effects: ## (Intr) drk_br ## drk_brghtnt 0.310 ## drk_brghtb_ -0.025 0.020
dat_phonological_s1 -> dat_phonological_s1b dat_phonological_s1b$dark_bright = as.character(dat_phonological_s1b$dark_bright) dat_phonological_s1b %>% mutate(dark_bright=if_else(dark_bright == "Bright", "Light", dark_bright)) -> dat_phonological_s1b dat_phonological_s1b$dark_bright = factor(dat_phonological_s1b$dark_bright, levels = c("Light", "Dark", "Neutral")) dat_phonological_s1b %>% ggplot(aes(dark_bright,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Type") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Rounding lm_s1_round_v = glmer(Sex~round_v_cnt+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_round_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ round_v_cnt + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2369.1 2385.8 -1181.6 2363.1 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.0325 -0.7854 -0.2099 0.7420 3.1749 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.727 1.651 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.01361 0.34294 -0.04 0.968 ## round_v_cnt1 0.12142 0.40532 0.30 0.765 ## ## Correlation of Fixed Effects: ## (Intr) ## rond_v_cnt1 0.846
dat_phonological_s1 %>% ggplot(aes(round_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Round Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Length & Diphthong Type lm_s1_diphthong = glmer(Sex~vowel_length+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_diphthong)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ vowel_length + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2367.7 2390.0 -1179.9 2359.7 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.9526 -0.7853 -0.2091 0.7430 3.1677 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.629 1.621 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.4633 0.3635 1.275 0.2025 ## vowel_lengthmono_diph -0.9318 0.5754 -1.619 0.1053 ## vowel_lengthj_w -1.9042 1.0715 -1.777 0.0755 . ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) vwl_lngthm_ ## vwl_lngthm_ -0.845 ## vwl_lngthj_ -0.727 0.691
dat_phonological_s1 %>% mutate(vowel_length = if_else(vowel_length == "monophthong", "monophthong", "diphthong")) %>% ggplot(aes(vowel_length,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Length") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Length") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
dat_phonological_s1 %>% filter(vowel_length != "monophthong") %>% ggplot(aes(vowel_length,fill=Sex)) + geom_bar(position="fill") + xlab("Diphthong on/off-glide") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Diphthong Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Stops lm_s1_stops = glmer(Sex~stops+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_stops)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ stops + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2365.0 2381.7 -1179.5 2359.0 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.1659 -0.7838 -0.2086 0.7452 3.1748 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.517 1.587 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.1346 0.2082 0.647 0.5179 ## stops1 0.8728 0.4156 2.100 0.0357 * ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) ## stops1 0.525
dat_phonological_s1 %>% ggplot(aes(stops,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Stops") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Stop Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Sonorants lm_s1_sonorants = glmer(Sex~sonorants+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_sonorants)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ sonorants + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2365.5 2387.8 -1178.7 2357.5 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.0965 -0.7812 -0.2000 0.7361 3.2369 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.575 1.605 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.2011 0.3799 -0.529 0.597 ## sonorantszero -0.5145 0.6112 -0.842 0.400 ## sonorantsone_two 0.7126 1.1076 0.643 0.520 ## ## Correlation of Fixed Effects: ## (Intr) snrnts ## sonorantszr -0.777 ## sonrntsn_tw -0.874 0.809
dat_phonological_s1 %>% ggplot(aes(sonorants,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Sonorants") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Sonorant Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Lenis Obstruents lm_s1_plain = glmer(Sex~plain+(1|syllable_1) , data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6)))
## Warning in checkConv(attr(opt, "derivs"), opt$par, ctrl = control$checkConv, : Model ## failed to converge with max|grad| = 0.0472614 (tol = 0.001, component 1)
summary(lm_s1_plain)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ plain + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2367.4 2389.6 -1179.7 2359.4 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.9860 -0.7846 -0.2154 0.7406 3.1201 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.572 1.604 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.6805 0.0015 453.7 <2e-16 *** ## plainzero -1.5501 0.0015 -1033.3 <2e-16 *** ## plainone_two -2.3794 0.0015 -1586.1 <2e-16 *** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) planzr ## plainzero 0.000 ## plainone_tw 0.000 0.000 ## convergence code: 0 ## Model failed to converge with max|grad| = 0.0472614 (tol = 0.001, component 1)
dat_phonological_s1 %>% ggplot(aes(plain,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Lenis Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# No Fortis Obstruents in S1 # Number of Aspirated Obstruents lm_s1_aspirated = glmer(Sex~aspirated + (1|syllable_1), data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s1_aspirated)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ aspirated + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2368.6 2385.3 -1181.3 2362.6 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.9822 -0.7854 -0.2076 0.7421 3.1701 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 2.695 1.642 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.1512 0.3713 0.407 0.684 ## aspirated1 0.5756 0.7424 0.775 0.438 ## ## Correlation of Fixed Effects: ## (Intr) ## aspirated1 0.872
dat_phonological_s1 %>% ggplot(aes(aspirated,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Aspirated Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
######### Syllable #2 ########## # Initial Sound lm_s2_initial_snd = glmer(Sex~initial_snd_type + (1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_initial_snd)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ initial_snd_type + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2004.9 2021.6 -999.5 1998.9 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.7155 -0.6458 -0.1493 0.5680 4.1888 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.315 2.705 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.05646 0.30355 -0.186 0.852 ## initial_snd_typeV -0.16780 0.60702 -0.276 0.782 ## ## Correlation of Fixed Effects: ## (Intr) ## intl_snd_tV 0.520
dat_phonological_s2 %>% ggplot(aes(initial_snd_type,fill=Sex)) + geom_bar(position="fill") + xlab("Sound Type") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Initial Sound Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Syllable Type lm_s2_initial_syll = glmer(Sex~syll_2_type+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_initial_syll)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ syll_2_type + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1997.5 2014.3 -995.8 1991.5 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.7894 -0.6390 -0.1379 0.5717 4.2804 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.933 2.633 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.2388 0.2670 -0.895 0.37105 ## syll_2_typeOpen -1.4943 0.5373 -2.781 0.00542 ** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) ## syll_2_typO 0.310
dat_phonological_s2 %>% ggplot(aes(syll_2_type,fill=Sex)) + geom_bar(position="fill") + xlab("Syllable Type") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Syllable Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Height lm_s2_height = glmer(Sex~height+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_height)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ height + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2006.5 2028.8 -999.2 1998.5 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.6462 -0.6451 -0.1467 0.5678 4.2073 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.306 2.703 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.05742 0.26945 -0.213 0.831 ## heighthigh -0.03269 0.54196 -0.060 0.952 ## heightlow_mid -0.49930 0.69269 -0.721 0.471 ## ## Correlation of Fixed Effects: ## (Intr) hghthg ## heighthigh -0.150 ## heightlw_md 0.252 -0.188
dat_phonological_s2 %>% ggplot(aes(height,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Height") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Height") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Backness lm_s2_backness = glmer(Sex~backness+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_backness)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ backness + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1995.6 2017.9 -993.8 1987.6 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.6188 -0.6541 -0.1396 0.5638 4.1200 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.366 2.523 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.1240 0.2554 -0.486 0.62719 ## backnessfront -1.8856 0.5867 -3.214 0.00131 ** ## backnessback_central 0.9698 0.5722 1.695 0.09010 . ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) bcknss ## backnssfrnt 0.216 ## bcknssbck_c 0.157 -0.106
dat_phonological_s2 %>% ggplot(aes(backness,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Backness") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Backness") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
#Vowel Type (Dark - Light - Neutral) lm_s2_darkness = glmer(Sex~dark_bright+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_darkness)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ dark_bright + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1995.7 2017.9 -993.8 1987.7 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.5969 -0.6527 -0.1384 0.5645 4.1370 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.403 2.53 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.2320 0.2544 -0.912 0.36165 ## dark_brightneutral -1.7915 0.5919 -3.027 0.00247 ** ## dark_brightbright_dark -0.9176 0.5597 -1.639 0.10115 ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) drk_br ## drk_brghtnt 0.257 ## drk_brghtb_ 0.044 -0.020
dat_phonological_s2 -> dat_phonological_s2b dat_phonological_s2b$dark_bright = as.character(dat_phonological_s2b$dark_bright) dat_phonological_s2b %>% mutate(dark_bright=if_else(dark_bright == "Bright", "Light", dark_bright)) -> dat_phonological_s2b dat_phonological_s2b$dark_bright = factor(dat_phonological_s2b$dark_bright, levels = c("Light", "Dark", "Neutral")) dat_phonological_s2b %>% ggplot(aes(dark_bright,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Type") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Rounding lm_s2_round_v = glmer(Sex~round_v_cnt+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_round_v)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ round_v_cnt + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1997.8 2014.5 -995.9 1991.8 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.5513 -0.6537 -0.1468 0.5636 4.2003 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.639 2.577 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 1.0214 0.4475 2.282 0.02247 * ## round_v_cnt1 1.5001 0.5395 2.781 0.00543 ** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) ## rond_v_cnt1 0.831
dat_phonological_s2 %>% ggplot(aes(round_v_cnt,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Vowels") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Round Vowel Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Vowel Length & Diphthong Type lm_s2_diphthong = glmer(Sex~vowel_length+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_diphthong)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ vowel_length + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2002.6 2024.9 -997.3 1994.6 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.6679 -0.6446 -0.1473 0.5682 4.1859 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.218 2.687 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.7612 0.4761 1.599 0.1098 ## vowel_lengthmono_diph -1.3709 0.7670 -1.787 0.0739 . ## vowel_lengthj_w -2.6248 1.3999 -1.875 0.0608 . ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) vwl_lngthm_ ## vwl_lngthm_ -0.801 ## vwl_lngthj_ -0.642 0.603
dat_phonological_s2 %>% mutate(vowel_length = if_else(vowel_length == "monophthong", "monophthong", "diphthong")) %>% ggplot(aes(vowel_length,fill=Sex)) + geom_bar(position="fill") + xlab("Vowel Length") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Vowel Length") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
dat_phonological_s2 %>% filter(vowel_length != "monophthong") %>% ggplot(aes(vowel_length,fill=Sex)) + geom_bar(position="fill") + xlab("Diphthong on/off-glide") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Diphthong Type") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Stops lm_s2_stops = glmer(Sex~stops+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_stops)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ stops + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1989.1 2011.4 -990.6 1981.1 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.5435 -0.6413 -0.1409 0.5703 4.2388 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.005 2.647 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 5.711 19.395 0.294 0.768 ## stopszero -9.538 29.091 -0.328 0.743 ## stopsone_two -14.030 58.197 -0.241 0.809 ## ## Correlation of Fixed Effects: ## (Intr) stpszr ## stopszero -1.000 ## stopsone_tw -1.000 1.000
dat_phonological_s2 %>% ggplot(aes(stops,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Stops") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Stop Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Sonorants lm_s2_sonorants = glmer(Sex~sonorants+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_sonorants)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ sonorants + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1994.9 2017.2 -993.4 1986.9 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.7538 -0.6472 -0.1074 0.5670 4.1308 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.932 2.633 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -0.7533 0.3450 -2.184 0.028994 * ## sonorantszero 1.4057 0.6520 2.156 0.031086 * ## sonorantsone_two 3.1567 0.9463 3.336 0.000851 *** ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) snrnts ## sonorantszr -0.343 ## sonrntsn_tw -0.643 0.518
dat_phonological_s2 %>% ggplot(aes(sonorants,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Sonorants") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Sonorant Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Lenis Obstruents lm_s2_plain = glmer(Sex~plain+(1|syllable_2) , data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_plain)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ plain + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1979.9 2002.2 -985.9 1971.9 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.4389 -0.6505 -0.1449 0.5730 4.2454 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 6.069 2.463 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 5.966 18.312 0.326 0.745 ## plainzero -10.437 27.464 -0.380 0.704 ## plainone_two -17.965 54.931 -0.327 0.744 ## ## Correlation of Fixed Effects: ## (Intr) planzr ## plainzero -1.000 ## plainone_tw -1.000 1.000
dat_phonological_s2 %>% ggplot(aes(plain,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Lenis Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Fortis Obstruents lm_2s_tense = glmer(Sex~tense+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_2s_tense)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ tense + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1774.6 1796.9 -883.3 1766.6 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.6473 -0.3997 -0.0343 0.3751 2.6038 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.288 2.700 ## syllable_1 (Intercept) 3.175 1.782 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -8.699 21.116 -0.412 0.68 ## tense1 -17.969 42.229 -0.426 0.67 ## ## Correlation of Fixed Effects: ## (Intr) ## tense1 1.000
dat_phonological_full %>% ggplot(aes(tense,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Fortis Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
# Number of Aspirated Obstruents lm_s2_aspirated = glmer(Sex~aspirated + (1|syllable_2), data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(lm_s2_aspirated)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ aspirated + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2003.6 2020.3 -998.8 1997.6 1941 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -5.0461 -0.6451 -0.1480 0.5678 4.1797 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 7.274 2.697 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.6801 0.6293 1.081 0.280 ## aspirated1 1.5205 1.2594 1.207 0.227 ## ## Correlation of Fixed Effects: ## (Intr) ## aspirated1 0.912
dat_phonological_s2 %>% ggplot(aes(aspirated,fill=Sex)) + geom_bar(position="fill") + xlab("Number of Obstruents") + ylab("Proportion") + theme_bw() + guides(fill=guide_legend(title="Sex")) + ggtitle("Aspirated Obstruent Count") + scale_fill_manual(values=c("#663366","#FFCCCC")) + theme(legend.position="bottom")
########################################## ######### MULTIVARIATE ANALYSIS ########## ########################################## ########### Full name glm_2s_trimmed = glmer(Sex~open_syll_cnt+stops+round_v_cnt+w_cnt+tense+aspirated+j_cnt+dark_v_cnt+ (1|syllable_1) + (1|syllable_2), data=dat_phonological_full, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(glm_2s_trimmed)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ open_syll_cnt + stops + round_v_cnt + w_cnt + tense + aspirated + ## j_cnt + dark_v_cnt + (1 | syllable_1) + (1 | syllable_2) ## Data: dat_phonological_full ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1720.0 1809.1 -844.0 1688.0 1928 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -3.5436 -0.3887 -0.0235 0.3618 3.0813 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 4.739 2.177 ## syllable_1 (Intercept) 2.167 1.472 ## Number of obs: 1944, groups: syllable_2, 145; syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) -6.5538 23.2781 -0.282 0.778294 ## open_syll_cntzero 2.3935 0.4762 5.027 4.99e-07 *** ## open_syll_cntone_two 1.6800 0.3719 4.518 6.25e-06 *** ## stopszero -2.4025 0.5581 -4.305 1.67e-05 *** ## stopsone_two -1.4556 0.5966 -2.440 0.014689 * ## round_v_cntzero -1.7540 0.5239 -3.348 0.000813 *** ## round_v_cntone_two -0.9484 0.5138 -1.846 0.064903 . ## w_cnt1 1.8416 0.8427 2.185 0.028855 * ## tense1 -17.7421 46.5473 -0.381 0.703083 ## aspirated1 1.3393 0.6467 2.071 0.038350 * ## j_cntzero 1.6435 0.6024 2.728 0.006368 ** ## j_cntone_two 0.8807 0.5856 1.504 0.132640 ## dark_v_cntzero -1.2708 0.5004 -2.540 0.011100 * ## dark_v_cntone_two -0.7976 0.3787 -2.106 0.035182 * ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
########### 1st Syllable glm_s1_trimmed = glmer(Sex~syll_1_type+stops + (1|syllable_1), data=dat_phonological_s1, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(glm_s1_trimmed)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ syll_1_type + stops + (1 | syllable_1) ## Data: dat_phonological_s1 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 2352.0 2374.3 -1172.0 2344.0 1940 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.2567 -0.7768 -0.2100 0.7350 3.2090 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_1 (Intercept) 1.993 1.412 ## Number of obs: 1944, groups: syllable_1, 119 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 0.1962 0.1921 1.021 0.3071 ## syll_1_typeOpen -1.3023 0.3269 -3.983 6.8e-05 *** ## stops1 0.7880 0.3829 2.058 0.0396 * ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) sy_1_O ## syll_1_typO -0.055 ## stops1 0.531 0.045
########## 2nd Syllable glm_s2_trimmed = glmer(Sex~plain+sonorants+round_v_cnt+syll_2_type+vowel_length+aspirated + (1|syllable_2), data=dat_phonological_s2, family="binomial", control = glmerControl(optimizer = "bobyqa", optCtrl=list(maxfun=2e6))) summary(glm_s2_trimmed)
## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [glmerMod ## ] ## Family: binomial ( logit ) ## Formula: Sex ~ plain + sonorants + round_v_cnt + syll_2_type + vowel_length + ## aspirated + (1 | syllable_2) ## Data: dat_phonological_s2 ## Control: glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 2e+06)) ## ## AIC BIC logLik deviance df.resid ## 1957.9 2019.2 -968.0 1935.9 1933 ## ## Scaled residuals: ## Min 1Q Median 3Q Max ## -4.7013 -0.6553 -0.1237 0.5739 4.4107 ## ## Random effects: ## Groups Name Variance Std.Dev. ## syllable_2 (Intercept) 4.328 2.08 ## Number of obs: 1944, groups: syllable_2, 145 ## ## Fixed effects: ## Estimate Std. Error z value Pr(>|z|) ## (Intercept) 7.8786 11.8318 0.666 0.505486 ## plainzero -9.9206 17.7280 -0.560 0.575749 ## plainone_two -17.7709 35.4600 -0.501 0.616263 ## sonorantszero 1.3892 1.0910 1.273 0.202929 ## sonorantsone_two 2.0493 0.8653 2.368 0.017869 * ## round_v_cnt1 1.7882 0.4914 3.639 0.000273 *** ## syll_2_typeOpen -2.0040 0.8818 -2.273 0.023047 * ## vowel_lengthmono_diph -1.1350 0.6386 -1.777 0.075518 . ## vowel_lengthj_w -2.5526 1.1420 -2.235 0.025398 * ## aspirated1 1.9155 1.1667 1.642 0.100636 ## --- ## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1 ## ## Correlation of Fixed Effects: ## (Intr) planzr plnn_t snrnts snrnt_ rnd__1 sy_2_O vwl_lngthm_ vwl_lngthj_ ## plainzero -0.997 ## plainone_tw -0.997 1.000 ## sonorantszr -0.019 -0.007 -0.022 ## sonrntsn_tw -0.019 -0.005 -0.015 0.533 ## rond_v_cnt1 0.031 0.005 0.008 -0.153 -0.091 ## syll_2_typO 0.013 0.007 0.018 -0.820 -0.276 0.034 ## vwl_lngthm_ -0.022 -0.014 -0.017 0.147 0.102 -0.107 -0.113 ## vwl_lngthj_ -0.012 -0.013 -0.014 0.012 0.013 -0.099 0.025 0.591 ## aspirated1 0.040 0.008 0.018 -0.328 -0.170 0.112 0.249 -0.136 -0.033