##title: ‘Evenki Vowel Properties’ ##author: “Matt Fan” ##date: “27/12/2022” #########################################
library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.3 ✔ tidyr 1.3.1
## ✔ purrr 1.0.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errorslibrary(languageR)
library(broom)
library(broom.mixed) #install the package if you don't have it yet
library(lme4)## Loading required package: Matrix
##
## Attaching package: 'Matrix'
##
## The following objects are masked from 'package:tidyr':
##
## expand, pack, unpacklibrary(lmerTest)##
## Attaching package: 'lmerTest'
##
## The following object is masked from 'package:lme4':
##
## lmer
##
## The following object is masked from 'package:stats':
##
## steplibrary(arm)## Loading required package: MASS
##
## Attaching package: 'MASS'
##
## The following object is masked from 'package:dplyr':
##
## select
##
##
## arm (Version 1.14-4, built: 2024-4-1)
##
## Working directory is /Users/mfan/Documents/2022 Fall/FAN - FINAL PROJECT SUBMISSION/R analysislibrary("phonTools")##
## Attaching package: 'phonTools'
##
## The following object is masked from 'package:dplyr':
##
## slicelibrary(plotly)##
## Attaching package: 'plotly'
##
## The following object is masked from 'package:phonTools':
##
## slice
##
## The following object is masked from 'package:MASS':
##
## select
##
## The following object is masked from 'package:ggplot2':
##
## last_plot
##
## The following object is masked from 'package:stats':
##
## filter
##
## The following object is masked from 'package:graphics':
##
## layoutlibrary(reshape2)##
## Attaching package: 'reshape2'
##
## The following object is masked from 'package:tidyr':
##
## smithslibrary("vowels")##
## Attaching package: 'vowels'
##
## The following object is masked from 'package:phonTools':
##
## vowelplotlibrary("cowplot")##
## Attaching package: 'cowplot'
##
## The following object is masked from 'package:lubridate':
##
## stamplibrary(ggpubr)##
## Attaching package: 'ggpubr'
##
## The following object is masked from 'package:cowplot':
##
## get_legend
##
## The following object is masked from 'package:phonTools':
##
## rotatelibrary("RColorBrewer")
library(ggplot2)
library("ggeffects")##
## Attaching package: 'ggeffects'
##
## The following object is masked from 'package:cowplot':
##
## get_titleformant <- read.csv('formantmdpt.csv', stringsAsFactors = FALSE)
#formant <- filter(formant, speaker_id != "poligus khadonchina")formant_new <- formant %>% mutate(
VC= NULL,
Preceding.V = NULL,
Rounding = NULL
)
formant_new <- normalize(formant_new[,4:6], formant_new$speaker_id, formant_new$vowel_id, method = 'neareyE',
corners = NULL)
##formant_new$speaker <- as.character(formant_new$speaker)
##formant_new$vowel <-as.character(formant_new$vowel)
formant_new <- rename(formant_new,speaker_id=speaker)
formant_new <-rename(formant_new,vowel_id=vowel)
formant_new <- mutate(formant_new, F1_glide = NA, F2_glide = NA, F3_glide = NA, context =NA )
formant_new<-formant_new %>% relocate(context)
formant_new<-formant_new %>% relocate(vowel_id)
formant_new<-formant_new %>% relocate(speaker_id)
formant_new <-scalevowels(formant_new)
formant_new <- formant_new %>% mutate(
VH = formant$VC,
Preceding.V = formant$Preceding.V,
Rounding= formant$Rounding,
"F1_glide" =NULL,
"F2_glide" =NULL,
"F3_glide" =NULL,
context=NULL
)
formant_new <-rename(formant_new,speaker=speaker_id)
formant_new <-rename(formant_new,vowel=vowel_id)
str(formant_new)## 'data.frame': 1065 obs. of 8 variables:
## $ speaker : Factor w/ 5 levels "chirinda eldogir valentina",..: 1 1 1 1 1 1 1 1 1 1 ...
## $ vowel : Factor w/ 4 levels "a","e","o","u": 1 1 4 1 1 1 3 3 3 1 ...
## $ F1 : num 494 445 347 560 501 ...
## $ F2 : num 1630 1952 1868 1774 1585 ...
## $ F3 : num 5464 5484 5474 5319 5423 ...
## $ VH : chr "N" "Y" "N" "N" ...
## $ Preceding.V: chr NA "u" NA NA ...
## $ Rounding : chr "N" "Y" "Y" "N" ...
## - attr(*, "scaled.values")= logi TRUElevels(formant_new$vowel)## [1] "a" "e" "o" "u"formant_speaker1 <- filter(formant_new, speaker == "poligus khadonchina")
formant_speaker2 <- filter(formant_new, speaker == "chirinda eldogir valentina")
formant_speaker3 <- filter(formant_new, speaker == "iyengra kargapolovaalbinasergeyevna")
formant_speaker4 <- filter(formant_new, speaker == "strelka andreeva")
formant_speaker5 <- filter(formant_new, speaker == "mutoray yastrikova")formant_new$vowel <- relevel(formant_new$vowel, ref="a")
formant_F1 <- lm (F1 ~ vowel, data = formant_new)
formant_F2 <- lm (F2 ~ vowel, data = formant_new)
formant_F3 <- lm (F3 ~ vowel, data = formant_new)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -181.041 -30.546 1.422 29.584 239.721
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 510.279 2.637 193.50 <2e-16 ***
## vowele -82.916 3.588 -23.11 <2e-16 ***
## vowelo -66.335 4.525 -14.66 <2e-16 ***
## vowelu -138.414 5.071 -27.30 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.2 on 1061 degrees of freedom
## Multiple R-squared: 0.469, Adjusted R-squared: 0.4675
## F-statistic: 312.4 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -654.36 -137.19 5.47 131.28 785.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1755.27 10.57 166.134 < 2e-16 ***
## vowele 105.46 14.37 7.337 4.34e-13 ***
## vowelo -333.02 18.13 -18.370 < 2e-16 ***
## vowelu -290.49 20.32 -14.298 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 197.1 on 1061 degrees of freedom
## Multiple R-squared: 0.4413, Adjusted R-squared: 0.4397
## F-statistic: 279.3 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -734.55 -73.90 8.59 80.51 521.24
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5467.523 6.755 809.432 < 2e-16 ***
## vowele 55.794 9.190 6.071 1.76e-09 ***
## vowelo -19.486 11.590 -1.681 0.093 .
## vowelu 3.267 12.989 0.252 0.801
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 126 on 1061 degrees of freedom
## Multiple R-squared: 0.05477, Adjusted R-squared: 0.0521
## F-statistic: 20.49 on 3 and 1061 DF, p-value: 6.51e-13formant_new$vowel <- relevel(formant_new$vowel, ref="e")
formant_F1 <- lm (F1 ~ vowel, data = formant_new)
formant_F2 <- lm (F2 ~ vowel, data = formant_new)
formant_F3 <- lm (F3 ~ vowel, data = formant_new)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -181.041 -30.546 1.422 29.584 239.721
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 427.364 2.433 175.685 < 2e-16 ***
## vowela 82.916 3.588 23.111 < 2e-16 ***
## vowelo 16.581 4.409 3.761 0.000179 ***
## vowelu -55.498 4.968 -11.172 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.2 on 1061 degrees of freedom
## Multiple R-squared: 0.469, Adjusted R-squared: 0.4675
## F-statistic: 312.4 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -654.36 -137.19 5.47 131.28 785.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1860.734 9.746 190.928 < 2e-16 ***
## vowela -105.460 14.374 -7.337 4.34e-13 ***
## vowelo -438.476 17.664 -24.824 < 2e-16 ***
## vowelu -395.947 19.903 -19.894 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 197.1 on 1061 degrees of freedom
## Multiple R-squared: 0.4413, Adjusted R-squared: 0.4397
## F-statistic: 279.3 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -734.55 -73.90 8.59 80.51 521.24
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5523.318 6.231 886.465 < 2e-16 ***
## vowela -55.794 9.190 -6.071 1.76e-09 ***
## vowelo -75.280 11.293 -6.666 4.21e-11 ***
## vowelu -52.527 12.724 -4.128 3.94e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 126 on 1061 degrees of freedom
## Multiple R-squared: 0.05477, Adjusted R-squared: 0.0521
## F-statistic: 20.49 on 3 and 1061 DF, p-value: 6.51e-13formant_new$vowel <- relevel(formant_new$vowel, ref="o")
formant_F1 <- lm (F1 ~ vowel, data = formant_new)
formant_F2 <- lm (F2 ~ vowel, data = formant_new)
formant_F3 <- lm (F3 ~ vowel, data = formant_new)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -181.041 -30.546 1.422 29.584 239.721
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 443.944 3.677 120.734 < 2e-16 ***
## vowele -16.581 4.409 -3.761 0.000179 ***
## vowela 66.335 4.525 14.660 < 2e-16 ***
## vowelu -72.079 5.682 -12.686 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.2 on 1061 degrees of freedom
## Multiple R-squared: 0.469, Adjusted R-squared: 0.4675
## F-statistic: 312.4 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -654.36 -137.19 5.47 131.28 785.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1422.26 14.73 96.545 <2e-16 ***
## vowele 438.48 17.66 24.824 <2e-16 ***
## vowela 333.02 18.13 18.370 <2e-16 ***
## vowelu 42.53 22.76 1.868 0.062 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 197.1 on 1061 degrees of freedom
## Multiple R-squared: 0.4413, Adjusted R-squared: 0.4397
## F-statistic: 279.3 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -734.55 -73.90 8.59 80.51 521.24
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5448.037 9.418 578.451 < 2e-16 ***
## vowele 75.280 11.293 6.666 4.21e-11 ***
## vowela 19.486 11.590 1.681 0.093 .
## vowelu 22.753 14.553 1.563 0.118
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 126 on 1061 degrees of freedom
## Multiple R-squared: 0.05477, Adjusted R-squared: 0.0521
## F-statistic: 20.49 on 3 and 1061 DF, p-value: 6.51e-13formant_new$vowel <- relevel(formant_new$vowel, ref="u")
formant_F1 <- lm (F1 ~ vowel, data = formant_new)
formant_F2 <- lm (F2 ~ vowel, data = formant_new)
formant_F3 <- lm (F3 ~ vowel, data = formant_new)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -181.041 -30.546 1.422 29.584 239.721
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 371.866 4.331 85.85 <2e-16 ***
## vowelo 72.079 5.682 12.69 <2e-16 ***
## vowele 55.498 4.968 11.17 <2e-16 ***
## vowela 138.414 5.071 27.30 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.2 on 1061 degrees of freedom
## Multiple R-squared: 0.469, Adjusted R-squared: 0.4675
## F-statistic: 312.4 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -654.36 -137.19 5.47 131.28 785.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1464.79 17.35 84.410 <2e-16 ***
## vowelo -42.53 22.76 -1.868 0.062 .
## vowele 395.95 19.90 19.894 <2e-16 ***
## vowela 290.49 20.32 14.298 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 197.1 on 1061 degrees of freedom
## Multiple R-squared: 0.4413, Adjusted R-squared: 0.4397
## F-statistic: 279.3 on 3 and 1061 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_new)
##
## Residuals:
## Min 1Q Median 3Q Max
## -734.55 -73.90 8.59 80.51 521.24
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5470.791 11.094 493.111 < 2e-16 ***
## vowelo -22.753 14.553 -1.563 0.118
## vowele 52.527 12.724 4.128 3.94e-05 ***
## vowela -3.267 12.989 -0.252 0.801
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 126 on 1061 degrees of freedom
## Multiple R-squared: 0.05477, Adjusted R-squared: 0.0521
## F-statistic: 20.49 on 3 and 1061 DF, p-value: 6.51e-13#first speaker
formant_speaker1$vowel <- relevel(formant_speaker1$vowel, ref="a")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker1)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker1)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker1)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -190.295 -35.651 1.678 36.932 230.467
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 519.533 9.501 54.683 < 2e-16 ***
## vowele -112.524 13.649 -8.244 2.21e-13 ***
## vowelo -98.981 20.314 -4.873 3.35e-06 ***
## vowelu -159.865 19.778 -8.083 5.23e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 67.18 on 122 degrees of freedom
## Multiple R-squared: 0.4564, Adjusted R-squared: 0.4431
## F-statistic: 34.15 on 3 and 122 DF, p-value: 4.281e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -549.5 -119.7 14.0 122.3 402.0
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1788.54 27.16 65.861 < 2e-16 ***
## vowele 58.28 39.01 1.494 0.13779
## vowelo -381.45 58.06 -6.570 1.31e-09 ***
## vowelu -183.79 56.53 -3.251 0.00149 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 192 on 122 degrees of freedom
## Multiple R-squared: 0.355, Adjusted R-squared: 0.3391
## F-statistic: 22.38 on 3 and 122 DF, p-value: 1.3e-11summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -292.35 -92.42 -2.11 85.40 479.45
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5509.311 18.391 299.566 <2e-16 ***
## vowele -6.492 26.421 -0.246 0.8063
## vowelo -91.500 39.322 -2.327 0.0216 *
## vowelu -15.625 38.284 -0.408 0.6839
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130 on 122 degrees of freedom
## Multiple R-squared: 0.0447, Adjusted R-squared: 0.02121
## F-statistic: 1.903 on 3 and 122 DF, p-value: 0.1326formant_speaker1$vowel <- relevel(formant_speaker1$vowel, ref="e")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker1)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker1)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker1)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -190.295 -35.651 1.678 36.932 230.467
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 407.010 9.799 41.534 < 2e-16 ***
## vowela 112.524 13.649 8.244 2.21e-13 ***
## vowelo 13.543 20.455 0.662 0.509
## vowelu -47.341 19.923 -2.376 0.019 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 67.18 on 122 degrees of freedom
## Multiple R-squared: 0.4564, Adjusted R-squared: 0.4431
## F-statistic: 34.15 on 3 and 122 DF, p-value: 4.281e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -549.5 -119.7 14.0 122.3 402.0
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1846.82 28.01 65.935 < 2e-16 ***
## vowela -58.28 39.01 -1.494 0.138
## vowelo -439.73 58.47 -7.521 1.02e-11 ***
## vowelu -242.07 56.95 -4.251 4.19e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 192 on 122 degrees of freedom
## Multiple R-squared: 0.355, Adjusted R-squared: 0.3391
## F-statistic: 22.38 on 3 and 122 DF, p-value: 1.3e-11summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -292.35 -92.42 -2.11 85.40 479.45
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5502.819 18.969 290.098 <2e-16 ***
## vowela 6.492 26.421 0.246 0.8063
## vowelo -85.008 39.595 -2.147 0.0338 *
## vowelu -9.133 38.565 -0.237 0.8132
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130 on 122 degrees of freedom
## Multiple R-squared: 0.0447, Adjusted R-squared: 0.02121
## F-statistic: 1.903 on 3 and 122 DF, p-value: 0.1326formant_speaker1$vowel <- relevel(formant_speaker1$vowel, ref="o")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker1)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker1)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker1)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -190.295 -35.651 1.678 36.932 230.467
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 420.55 17.95 23.423 < 2e-16 ***
## vowele -13.54 20.45 -0.662 0.5092
## vowela 98.98 20.31 4.873 3.35e-06 ***
## vowelu -60.88 24.97 -2.439 0.0162 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 67.18 on 122 degrees of freedom
## Multiple R-squared: 0.4564, Adjusted R-squared: 0.4431
## F-statistic: 34.15 on 3 and 122 DF, p-value: 4.281e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -549.5 -119.7 14.0 122.3 402.0
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1407.09 51.32 27.418 < 2e-16 ***
## vowele 439.73 58.47 7.521 1.02e-11 ***
## vowela 381.45 58.06 6.570 1.31e-09 ***
## vowelu 197.66 71.36 2.770 0.00648 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 192 on 122 degrees of freedom
## Multiple R-squared: 0.355, Adjusted R-squared: 0.3391
## F-statistic: 22.38 on 3 and 122 DF, p-value: 1.3e-11summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -292.35 -92.42 -2.11 85.40 479.45
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5417.81 34.76 155.883 <2e-16 ***
## vowele 85.01 39.60 2.147 0.0338 *
## vowela 91.50 39.32 2.327 0.0216 *
## vowelu 75.87 48.33 1.570 0.1190
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130 on 122 degrees of freedom
## Multiple R-squared: 0.0447, Adjusted R-squared: 0.02121
## F-statistic: 1.903 on 3 and 122 DF, p-value: 0.1326formant_speaker1$vowel <- relevel(formant_speaker1$vowel, ref="u")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker1)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker1)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker1)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -190.295 -35.651 1.678 36.932 230.467
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 359.67 17.35 20.735 < 2e-16 ***
## vowelo 60.88 24.97 2.439 0.0162 *
## vowele 47.34 19.92 2.376 0.0190 *
## vowela 159.86 19.78 8.083 5.23e-13 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 67.18 on 122 degrees of freedom
## Multiple R-squared: 0.4564, Adjusted R-squared: 0.4431
## F-statistic: 34.15 on 3 and 122 DF, p-value: 4.281e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -549.5 -119.7 14.0 122.3 402.0
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1604.75 49.58 32.367 < 2e-16 ***
## vowelo -197.66 71.36 -2.770 0.00648 **
## vowele 242.07 56.95 4.251 4.19e-05 ***
## vowela 183.79 56.53 3.251 0.00149 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 192 on 122 degrees of freedom
## Multiple R-squared: 0.355, Adjusted R-squared: 0.3391
## F-statistic: 22.38 on 3 and 122 DF, p-value: 1.3e-11summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -292.35 -92.42 -2.11 85.40 479.45
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5493.686 33.577 163.614 <2e-16 ***
## vowelo -75.875 48.326 -1.570 0.119
## vowele 9.133 38.565 0.237 0.813
## vowela 15.625 38.284 0.408 0.684
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130 on 122 degrees of freedom
## Multiple R-squared: 0.0447, Adjusted R-squared: 0.02121
## F-statistic: 1.903 on 3 and 122 DF, p-value: 0.1326#second speaker
formant_speaker2$vowel <- relevel(formant_speaker2$vowel, ref="a")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker2)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker2)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker2)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -84.602 -24.034 2.181 24.364 113.390
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 514.877 4.587 112.26 <2e-16 ***
## vowele -85.659 6.590 -13.00 <2e-16 ***
## vowelo -67.860 7.068 -9.60 <2e-16 ***
## vowelu -145.219 11.307 -12.84 <2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 37.26 on 185 degrees of freedom
## Multiple R-squared: 0.5913, Adjusted R-squared: 0.5847
## F-statistic: 89.23 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -427.38 -108.39 4.83 120.54 564.87
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1791.60 20.55 87.163 < 2e-16 ***
## vowele 151.75 29.53 5.138 7.01e-07 ***
## vowelo -369.02 31.68 -11.649 < 2e-16 ***
## vowelu -347.19 50.67 -6.852 1.05e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 167 on 185 degrees of freedom
## Multiple R-squared: 0.6267, Adjusted R-squared: 0.6207
## F-statistic: 103.5 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -689.17 -48.18 10.68 74.25 274.75
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5377.48 15.17 354.438 < 2e-16 ***
## vowele 100.45 21.80 4.608 7.56e-06 ***
## vowelo 7.28 23.38 0.311 0.7559
## vowelu 80.11 37.40 2.142 0.0335 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 123.3 on 185 degrees of freedom
## Multiple R-squared: 0.1258, Adjusted R-squared: 0.1116
## F-statistic: 8.874 on 3 and 185 DF, p-value: 1.59e-05formant_speaker2$vowel <- relevel(formant_speaker2$vowel, ref="e")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker2)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker2)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker2)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -84.602 -24.034 2.181 24.364 113.390
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 429.218 4.732 90.701 < 2e-16 ***
## vowela 85.659 6.590 12.998 < 2e-16 ***
## vowelo 17.800 7.164 2.485 0.0139 *
## vowelu -59.559 11.366 -5.240 4.35e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 37.26 on 185 degrees of freedom
## Multiple R-squared: 0.5913, Adjusted R-squared: 0.5847
## F-statistic: 89.23 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -427.38 -108.39 4.83 120.54 564.87
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1943.35 21.21 91.635 < 2e-16 ***
## vowela -151.75 29.53 -5.138 7.01e-07 ***
## vowelo -520.76 32.10 -16.221 < 2e-16 ***
## vowelu -498.93 50.94 -9.795 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 167 on 185 degrees of freedom
## Multiple R-squared: 0.6267, Adjusted R-squared: 0.6207
## F-statistic: 103.5 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -689.17 -48.18 10.68 74.25 274.75
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5477.93 15.65 349.946 < 2e-16 ***
## vowela -100.45 21.80 -4.608 7.56e-06 ***
## vowelo -93.17 23.70 -3.932 0.000119 ***
## vowelu -20.35 37.60 -0.541 0.589079
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 123.3 on 185 degrees of freedom
## Multiple R-squared: 0.1258, Adjusted R-squared: 0.1116
## F-statistic: 8.874 on 3 and 185 DF, p-value: 1.59e-05formant_speaker2$vowel <- relevel(formant_speaker2$vowel, ref="o")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker2)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker2)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker2)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -84.602 -24.034 2.181 24.364 113.390
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 447.017 5.378 83.116 < 2e-16 ***
## vowele -17.800 7.164 -2.485 0.0139 *
## vowela 67.860 7.068 9.600 < 2e-16 ***
## vowelu -77.359 11.650 -6.640 3.38e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 37.26 on 185 degrees of freedom
## Multiple R-squared: 0.5913, Adjusted R-squared: 0.5847
## F-statistic: 89.23 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -427.38 -108.39 4.83 120.54 564.87
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1422.59 24.10 59.022 <2e-16 ***
## vowele 520.76 32.10 16.221 <2e-16 ***
## vowela 369.02 31.68 11.649 <2e-16 ***
## vowelu 21.83 52.21 0.418 0.676
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 167 on 185 degrees of freedom
## Multiple R-squared: 0.6267, Adjusted R-squared: 0.6207
## F-statistic: 103.5 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -689.17 -48.18 10.68 74.25 274.75
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5384.76 17.79 302.675 < 2e-16 ***
## vowele 93.17 23.70 3.932 0.000119 ***
## vowela -7.28 23.38 -0.311 0.755895
## vowelu 72.83 38.54 1.890 0.060352 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 123.3 on 185 degrees of freedom
## Multiple R-squared: 0.1258, Adjusted R-squared: 0.1116
## F-statistic: 8.874 on 3 and 185 DF, p-value: 1.59e-05formant_speaker2$vowel <- relevel(formant_speaker2$vowel, ref="u")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker2)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker2)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker2)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -84.602 -24.034 2.181 24.364 113.390
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 369.66 10.33 35.77 < 2e-16 ***
## vowelo 77.36 11.65 6.64 3.38e-10 ***
## vowele 59.56 11.37 5.24 4.35e-07 ***
## vowela 145.22 11.31 12.84 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 37.26 on 185 degrees of freedom
## Multiple R-squared: 0.5913, Adjusted R-squared: 0.5847
## F-statistic: 89.23 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -427.38 -108.39 4.83 120.54 564.87
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1444.42 46.31 31.187 < 2e-16 ***
## vowelo -21.83 52.21 -0.418 0.676
## vowele 498.93 50.94 9.795 < 2e-16 ***
## vowela 347.19 50.67 6.852 1.05e-10 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 167 on 185 degrees of freedom
## Multiple R-squared: 0.6267, Adjusted R-squared: 0.6207
## F-statistic: 103.5 on 3 and 185 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -689.17 -48.18 10.68 74.25 274.75
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5457.59 34.19 159.647 <2e-16 ***
## vowelo -72.83 38.54 -1.890 0.0604 .
## vowele 20.35 37.60 0.541 0.5891
## vowela -80.11 37.40 -2.142 0.0335 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 123.3 on 185 degrees of freedom
## Multiple R-squared: 0.1258, Adjusted R-squared: 0.1116
## F-statistic: 8.874 on 3 and 185 DF, p-value: 1.59e-05#third speaker
formant_speaker3$vowel <- relevel(formant_speaker3$vowel, ref="a")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker3)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker3)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker3)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -96.528 -21.427 -4.423 22.329 148.091
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 478.829 6.239 76.747 < 2e-16 ***
## vowele -58.746 8.930 -6.578 1.20e-09 ***
## vowelo -59.101 10.290 -5.744 6.75e-08 ***
## vowelu -114.871 11.270 -10.192 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 40.91 on 124 degrees of freedom
## Multiple R-squared: 0.4803, Adjusted R-squared: 0.4677
## F-statistic: 38.2 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -483.08 -88.54 -3.86 117.94 464.59
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1776.61 26.54 66.929 < 2e-16 ***
## vowele 108.32 37.99 2.851 0.00511 **
## vowelo -362.91 43.78 -8.290 1.58e-13 ***
## vowelu -253.52 47.95 -5.287 5.42e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 174.1 on 124 degrees of freedom
## Multiple R-squared: 0.5335, Adjusted R-squared: 0.5223
## F-statistic: 47.28 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -457.69 -37.84 10.62 66.73 227.07
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5523.41 15.83 348.917 <2e-16 ***
## vowele 28.69 22.66 1.266 0.2078
## vowelo 46.16 26.11 1.768 0.0795 .
## vowelu -31.45 28.60 -1.100 0.2735
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 103.8 on 124 degrees of freedom
## Multiple R-squared: 0.05806, Adjusted R-squared: 0.03527
## F-statistic: 2.548 on 3 and 124 DF, p-value: 0.05894formant_speaker3$vowel <- relevel(formant_speaker3$vowel, ref="e")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker3)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker3)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker3)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -96.528 -21.427 -4.423 22.329 148.091
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 420.0830 6.3894 65.747 < 2e-16 ***
## vowela 58.7465 8.9303 6.578 1.20e-09 ***
## vowelo -0.3543 10.3815 -0.034 0.973
## vowelu -56.1244 11.3542 -4.943 2.44e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 40.91 on 124 degrees of freedom
## Multiple R-squared: 0.4803, Adjusted R-squared: 0.4677
## F-statistic: 38.2 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -483.08 -88.54 -3.86 117.94 464.59
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1884.92 27.18 69.338 < 2e-16 ***
## vowela -108.32 37.99 -2.851 0.00511 **
## vowelo -471.23 44.17 -10.669 < 2e-16 ***
## vowelu -361.84 48.31 -7.490 1.12e-11 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 174.1 on 124 degrees of freedom
## Multiple R-squared: 0.5335, Adjusted R-squared: 0.5223
## F-statistic: 47.28 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -457.69 -37.84 10.62 66.73 227.07
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5552.10 16.21 342.476 <2e-16 ***
## vowela -28.69 22.66 -1.266 0.2078
## vowelo 17.48 26.34 0.663 0.5083
## vowelu -60.14 28.81 -2.088 0.0389 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 103.8 on 124 degrees of freedom
## Multiple R-squared: 0.05806, Adjusted R-squared: 0.03527
## F-statistic: 2.548 on 3 and 124 DF, p-value: 0.05894formant_speaker3$vowel <- relevel(formant_speaker3$vowel, ref="o")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker3)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker3)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker3)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -96.528 -21.427 -4.423 22.329 148.091
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 419.7287 8.1824 51.296 < 2e-16 ***
## vowele 0.3543 10.3815 0.034 0.973
## vowela 59.1007 10.2897 5.744 6.75e-08 ***
## vowelu -55.7702 12.4518 -4.479 1.68e-05 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 40.91 on 124 degrees of freedom
## Multiple R-squared: 0.4803, Adjusted R-squared: 0.4677
## F-statistic: 38.2 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -483.08 -88.54 -3.86 117.94 464.59
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1413.69 34.81 40.608 < 2e-16 ***
## vowele 471.23 44.17 10.669 < 2e-16 ***
## vowela 362.91 43.78 8.290 1.58e-13 ***
## vowelu 109.39 52.98 2.065 0.041 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 174.1 on 124 degrees of freedom
## Multiple R-squared: 0.5335, Adjusted R-squared: 0.5223
## F-statistic: 47.28 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -457.69 -37.84 10.62 66.73 227.07
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5569.57 20.76 268.271 <2e-16 ***
## vowele -17.48 26.34 -0.663 0.5083
## vowela -46.16 26.11 -1.768 0.0795 .
## vowelu -77.62 31.59 -2.457 0.0154 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 103.8 on 124 degrees of freedom
## Multiple R-squared: 0.05806, Adjusted R-squared: 0.03527
## F-statistic: 2.548 on 3 and 124 DF, p-value: 0.05894formant_speaker3$vowel <- relevel(formant_speaker3$vowel, ref="u")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker3)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker3)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker3)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -96.528 -21.427 -4.423 22.329 148.091
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 363.959 9.386 38.777 < 2e-16 ***
## vowelo 55.770 12.452 4.479 1.68e-05 ***
## vowele 56.124 11.354 4.943 2.44e-06 ***
## vowela 114.871 11.270 10.192 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 40.91 on 124 degrees of freedom
## Multiple R-squared: 0.4803, Adjusted R-squared: 0.4677
## F-statistic: 38.2 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -483.08 -88.54 -3.86 117.94 464.59
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1523.08 39.93 38.141 < 2e-16 ***
## vowelo -109.39 52.98 -2.065 0.041 *
## vowele 361.84 48.31 7.490 1.12e-11 ***
## vowela 253.52 47.95 5.287 5.42e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 174.1 on 124 degrees of freedom
## Multiple R-squared: 0.5335, Adjusted R-squared: 0.5223
## F-statistic: 47.28 on 3 and 124 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -457.69 -37.84 10.62 66.73 227.07
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5491.96 23.81 230.614 <2e-16 ***
## vowelo 77.62 31.59 2.457 0.0154 *
## vowele 60.14 28.81 2.088 0.0389 *
## vowela 31.45 28.60 1.100 0.2735
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 103.8 on 124 degrees of freedom
## Multiple R-squared: 0.05806, Adjusted R-squared: 0.03527
## F-statistic: 2.548 on 3 and 124 DF, p-value: 0.05894#fourth speaker
formant_speaker4$vowel <- relevel(formant_speaker4$vowel, ref="a")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker4)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker4)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker4)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -167.495 -30.402 2.555 29.938 140.407
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 515.238 4.340 118.707 < 2e-16 ***
## vowele -84.624 5.790 -14.614 < 2e-16 ***
## vowelo -60.336 7.288 -8.279 1.71e-15 ***
## vowelu -140.063 8.319 -16.837 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.68 on 416 degrees of freedom
## Multiple R-squared: 0.4636, Adjusted R-squared: 0.4598
## F-statistic: 119.9 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -611.44 -155.81 5.32 157.21 713.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1736.86 18.93 91.753 < 2e-16 ***
## vowele 80.94 25.25 3.205 0.00145 **
## vowelo -304.49 31.79 -9.579 < 2e-16 ***
## vowelu -298.42 36.28 -8.225 2.52e-15 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 216.7 on 416 degrees of freedom
## Multiple R-squared: 0.3579, Adjusted R-squared: 0.3533
## F-statistic: 77.3 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -566.53 -73.72 10.70 83.39 517.05
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5471.07 11.42 479.044 < 2e-16 ***
## vowele 70.17 15.24 4.605 5.48e-06 ***
## vowelo -29.09 19.18 -1.517 0.130
## vowelu -24.86 21.89 -1.136 0.257
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130.7 on 416 degrees of freedom
## Multiple R-squared: 0.09514, Adjusted R-squared: 0.08861
## F-statistic: 14.58 on 3 and 416 DF, p-value: 4.79e-09formant_speaker4$vowel <- relevel(formant_speaker4$vowel, ref="e")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker4)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker4)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker4)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -167.495 -30.402 2.555 29.938 140.407
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 430.614 3.833 112.351 < 2e-16 ***
## vowela 84.624 5.790 14.614 < 2e-16 ***
## vowelo 24.288 6.998 3.471 0.000573 ***
## vowelu -55.439 8.066 -6.873 2.3e-11 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.68 on 416 degrees of freedom
## Multiple R-squared: 0.4636, Adjusted R-squared: 0.4598
## F-statistic: 119.9 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -611.44 -155.81 5.32 157.21 713.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1817.81 16.72 108.748 < 2e-16 ***
## vowela -80.94 25.25 -3.205 0.00145 **
## vowelo -385.43 30.52 -12.629 < 2e-16 ***
## vowelu -379.37 35.18 -10.784 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 216.7 on 416 degrees of freedom
## Multiple R-squared: 0.3579, Adjusted R-squared: 0.3533
## F-statistic: 77.3 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -566.53 -73.72 10.70 83.39 517.05
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5541.24 10.09 549.451 < 2e-16 ***
## vowela -70.17 15.24 -4.605 5.48e-06 ***
## vowelo -99.26 18.41 -5.391 1.18e-07 ***
## vowelu -95.03 21.22 -4.478 9.76e-06 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130.7 on 416 degrees of freedom
## Multiple R-squared: 0.09514, Adjusted R-squared: 0.08861
## F-statistic: 14.58 on 3 and 416 DF, p-value: 4.79e-09formant_speaker4$vowel <- relevel(formant_speaker4$vowel, ref="o")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker4)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker4)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker4)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -167.495 -30.402 2.555 29.938 140.407
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 454.903 5.855 77.699 < 2e-16 ***
## vowele -24.288 6.998 -3.471 0.000573 ***
## vowela 60.336 7.288 8.279 1.71e-15 ***
## vowelu -79.727 9.200 -8.666 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.68 on 416 degrees of freedom
## Multiple R-squared: 0.4636, Adjusted R-squared: 0.4598
## F-statistic: 119.9 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -611.44 -155.81 5.32 157.21 713.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1432.378 25.534 56.097 <2e-16 ***
## vowele 385.431 30.519 12.629 <2e-16 ***
## vowela 304.487 31.786 9.579 <2e-16 ***
## vowelu 6.064 40.125 0.151 0.88
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 216.7 on 416 degrees of freedom
## Multiple R-squared: 0.3579, Adjusted R-squared: 0.3533
## F-statistic: 77.3 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -566.53 -73.72 10.70 83.39 517.05
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5441.98 15.40 353.257 < 2e-16 ***
## vowele 99.26 18.41 5.391 1.18e-07 ***
## vowela 29.09 19.18 1.517 0.130
## vowelu 4.23 24.21 0.175 0.861
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130.7 on 416 degrees of freedom
## Multiple R-squared: 0.09514, Adjusted R-squared: 0.08861
## F-statistic: 14.58 on 3 and 416 DF, p-value: 4.79e-09formant_speaker4$vowel <- relevel(formant_speaker4$vowel, ref="u")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker4)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker4)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker4)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -167.495 -30.402 2.555 29.938 140.407
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 375.175 7.097 52.865 < 2e-16 ***
## vowelo 79.727 9.200 8.666 < 2e-16 ***
## vowele 55.439 8.066 6.873 2.3e-11 ***
## vowela 140.063 8.319 16.837 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.68 on 416 degrees of freedom
## Multiple R-squared: 0.4636, Adjusted R-squared: 0.4598
## F-statistic: 119.9 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -611.44 -155.81 5.32 157.21 713.21
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1438.442 30.952 46.474 < 2e-16 ***
## vowelo -6.064 40.125 -0.151 0.88
## vowele 379.367 35.177 10.784 < 2e-16 ***
## vowela 298.422 36.282 8.225 2.52e-15 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 216.7 on 416 degrees of freedom
## Multiple R-squared: 0.3579, Adjusted R-squared: 0.3533
## F-statistic: 77.3 on 3 and 416 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -566.53 -73.72 10.70 83.39 517.05
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5446.21 18.67 291.648 < 2e-16 ***
## vowelo -4.23 24.21 -0.175 0.861
## vowele 95.03 21.22 4.478 9.76e-06 ***
## vowela 24.86 21.89 1.136 0.257
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130.7 on 416 degrees of freedom
## Multiple R-squared: 0.09514, Adjusted R-squared: 0.08861
## F-statistic: 14.58 on 3 and 416 DF, p-value: 4.79e-09#fifth speaker
formant_speaker5$vowel <- relevel(formant_speaker5$vowel, ref="a")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker5)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker5)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker5)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -139.680 -30.025 -2.002 28.681 132.692
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 509.186 5.978 85.170 < 2e-16 ***
## vowele -75.294 7.650 -9.842 < 2e-16 ***
## vowelo -65.422 11.807 -5.541 9.5e-08 ***
## vowelu -131.269 9.928 -13.222 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 45.53 on 198 degrees of freedom
## Multiple R-squared: 0.492, Adjusted R-squared: 0.4843
## F-statistic: 63.93 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -479.92 -119.19 4.68 130.01 835.26
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1711.02 24.08 71.069 < 2e-16 ***
## vowele 168.96 30.81 5.484 1.26e-07 ***
## vowelo -304.66 47.55 -6.408 1.05e-09 ***
## vowelu -296.28 39.98 -7.411 3.56e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 183.4 on 198 degrees of freedom
## Multiple R-squared: 0.5248, Adjusted R-squared: 0.5176
## F-statistic: 72.87 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -328.48 -56.89 -3.72 48.26 456.36
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5484.529 12.666 433.025 <2e-16 ***
## vowele 34.251 16.207 2.113 0.0358 *
## vowelo 6.428 25.013 0.257 0.7974
## vowelu 5.372 21.032 0.255 0.7987
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 96.46 on 198 degrees of freedom
## Multiple R-squared: 0.0267, Adjusted R-squared: 0.01196
## F-statistic: 1.811 on 3 and 198 DF, p-value: 0.1465formant_speaker5$vowel <- relevel(formant_speaker5$vowel, ref="e")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker5)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker5)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker5)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -139.680 -30.025 -2.002 28.681 132.692
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 433.892 4.773 90.907 < 2e-16 ***
## vowela 75.294 7.650 9.842 < 2e-16 ***
## vowelo 9.872 11.244 0.878 0.381
## vowelu -55.975 9.252 -6.050 7.12e-09 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 45.53 on 198 degrees of freedom
## Multiple R-squared: 0.492, Adjusted R-squared: 0.4843
## F-statistic: 63.93 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -479.92 -119.19 4.68 130.01 835.26
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1879.98 19.22 97.811 < 2e-16 ***
## vowela -168.96 30.81 -5.484 1.26e-07 ***
## vowelo -473.62 45.28 -10.460 < 2e-16 ***
## vowelu -465.24 37.26 -12.487 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 183.4 on 198 degrees of freedom
## Multiple R-squared: 0.5248, Adjusted R-squared: 0.5176
## F-statistic: 72.87 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -328.48 -56.89 -3.72 48.26 456.36
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5518.78 10.11 545.787 <2e-16 ***
## vowela -34.25 16.21 -2.113 0.0358 *
## vowelo -27.82 23.82 -1.168 0.2442
## vowelu -28.88 19.60 -1.473 0.1422
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 96.46 on 198 degrees of freedom
## Multiple R-squared: 0.0267, Adjusted R-squared: 0.01196
## F-statistic: 1.811 on 3 and 198 DF, p-value: 0.1465formant_speaker5$vowel <- relevel(formant_speaker5$vowel, ref="o")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker5)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker5)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker5)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -139.680 -30.025 -2.002 28.681 132.692
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 443.764 10.181 43.587 < 2e-16 ***
## vowele -9.872 11.244 -0.878 0.381
## vowela 65.422 11.807 5.541 9.50e-08 ***
## vowelu -65.847 12.902 -5.103 7.79e-07 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 45.53 on 198 degrees of freedom
## Multiple R-squared: 0.492, Adjusted R-squared: 0.4843
## F-statistic: 63.93 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -479.92 -119.19 4.68 130.01 835.26
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1406.359 40.999 34.302 < 2e-16 ***
## vowele 473.623 45.281 10.460 < 2e-16 ***
## vowela 304.665 47.545 6.408 1.05e-09 ***
## vowelu 8.385 51.958 0.161 0.872
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 183.4 on 198 degrees of freedom
## Multiple R-squared: 0.5248, Adjusted R-squared: 0.5176
## F-statistic: 72.87 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -328.48 -56.89 -3.72 48.26 456.36
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5490.957 21.569 254.579 <2e-16 ***
## vowele 27.823 23.821 1.168 0.244
## vowela -6.428 25.013 -0.257 0.797
## vowelu -1.057 27.334 -0.039 0.969
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 96.46 on 198 degrees of freedom
## Multiple R-squared: 0.0267, Adjusted R-squared: 0.01196
## F-statistic: 1.811 on 3 and 198 DF, p-value: 0.1465formant_speaker5$vowel <- relevel(formant_speaker5$vowel, ref="u")
formant_F1 <- lm (F1 ~ vowel, data = formant_speaker5)
formant_F2 <- lm (F2 ~ vowel, data = formant_speaker5)
formant_F3 <- lm (F3 ~ vowel, data = formant_speaker5)
summary(formant_F1)##
## Call:
## lm(formula = F1 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -139.680 -30.025 -2.002 28.681 132.692
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 377.917 7.926 47.681 < 2e-16 ***
## vowelo 65.847 12.902 5.103 7.79e-07 ***
## vowele 55.975 9.252 6.050 7.12e-09 ***
## vowela 131.269 9.928 13.222 < 2e-16 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 45.53 on 198 degrees of freedom
## Multiple R-squared: 0.492, Adjusted R-squared: 0.4843
## F-statistic: 63.93 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F2)##
## Call:
## lm(formula = F2 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -479.92 -119.19 4.68 130.01 835.26
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1414.744 31.918 44.325 < 2e-16 ***
## vowelo -8.385 51.958 -0.161 0.872
## vowele 465.238 37.258 12.487 < 2e-16 ***
## vowela 296.280 39.979 7.411 3.56e-12 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 183.4 on 198 degrees of freedom
## Multiple R-squared: 0.5248, Adjusted R-squared: 0.5176
## F-statistic: 72.87 on 3 and 198 DF, p-value: < 2.2e-16summary(formant_F3)##
## Call:
## lm(formula = F3 ~ vowel, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -328.48 -56.89 -3.72 48.26 456.36
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5489.900 16.791 326.950 <2e-16 ***
## vowelo 1.057 27.334 0.039 0.969
## vowele 28.880 19.601 1.473 0.142
## vowela -5.372 21.032 -0.255 0.799
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 96.46 on 198 degrees of freedom
## Multiple R-squared: 0.0267, Adjusted R-squared: 0.01196
## F-statistic: 1.811 on 3 and 198 DF, p-value: 0.1465formant_new$speaker<- as.character(formant_new$speaker)
formant_new$speaker <- replace(formant_new$speaker, formant_new$speaker == "poligus khadonchina","Speaker 1")
formant_new$speaker <-replace(formant_new$speaker, formant_new$speaker== "chirinda eldogir valentina","Speaker 2")
formant_new$speaker <-replace(formant_new$speaker, formant_new$speaker== "iyengra kargapolovaalbinasergeyevna","Speaker 3")
formant_new$speaker <-replace(formant_new$speaker, formant_new$speaker== "strelka andreeva","Speaker 4")
formant_new$speaker <-replace(formant_new$speaker, formant_new$speaker== "mutoray yastrikova","Speaker 5")
formant_new_1 <- formant_new %>%
group_by(vowel, speaker) %>%
summarise(F1.avg = mean(F1), F2.avg = mean(F2), F3.avg = mean(F3))## `summarise()` has grouped output by 'vowel'. You can override using the
## `.groups` argument.formant_new %>%
ggplot(aes(x = F2, y = F1, colour = vowel)) +
geom_density_2d(alpha = 0.5) +
scale_fill_distiller(palette = "Spectral", direction = -1)+
geom_label(data = formant_new_1, aes(x = F2.avg, y = F1.avg, label = vowel)) +
scale_x_reverse() +
scale_y_reverse() +
theme(legend.position = 'none')
levels(formant_new$vowel)## [1] "u" "o" "e" "a"formant_new %>%
ggplot(aes(x = F2, y = F1, colour = vowel)) +
stat_ellipse(aes(fill = vowel), geom='polygon', alpha = 0.2) +geom_point(alpha = 0.1) +
stat_ellipse(type = "t") +
stat_ellipse(type = "norm", linetype = 4)+
geom_label(data = formant_new_1, aes(x = F2.avg, y = F1.avg, label = vowel)) +
scale_x_reverse() +
scale_y_reverse() +
theme(legend.position = 'none') +
facet_wrap(~speaker)
formant_new %>%
ggplot(aes(x = F2, y = F1, colour = vowel)) +
geom_density_2d(alpha = 0.5) +
scale_fill_distiller(palette = "Spectral", direction = -1)+
geom_label(data = formant_new_1, aes(x = F2.avg, y = F1.avg, label = vowel)) +
scale_x_reverse() +
scale_y_reverse() +
theme(legend.position = 'none') +
facet_wrap(~speaker)
formant_new_2 <- formant_new %>%
group_by(vowel) %>%
summarise(F1.avg = mean(F1), F2.avg = mean(F2), F3.avg = mean(F3))
allplot_1 <- formant_new %>%
ggplot(aes(x = F2, y = F1, colour = vowel)) +
stat_ellipse(aes(fill = vowel), geom='polygon', alpha = 0.2) +geom_point(alpha = 0.1) +
stat_ellipse(type = "t") +
stat_ellipse(type = "norm", linetype = 4)+
geom_label(data = formant_new_2, aes(x = F2.avg, y = F1.avg, label = vowel)) +
scale_x_reverse() +
scale_y_reverse() +
theme(legend.position = 'none')
allplot_2 <- formant_new %>%
ggplot(aes(x = F2, y = F1, colour = vowel)) +
geom_density_2d(alpha = 0.5) +
scale_fill_distiller(palette = "Spectral", direction = -1)+
geom_label(data = formant_new_2, aes(x = F2.avg, y = F1.avg, label = vowel)) +
scale_x_reverse() +
scale_y_reverse() +
theme(legend.position = 'none')
ggarrange(allplot_1, allplot_2 + rremove("y.title")
,common.legend = FALSE,
ncol = 2, nrow = 1)
#vowel formants
boxplotv1 <- ggplot(formant_new, aes(x=vowel, y=F1, fill=vowel))+xlab("Vowel type")+
scale_fill_manual(values = c("#d8b365", "#5ab4ac","#F67280", "#6c5b7b"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=TRUE)
boxplotv2 <- ggplot(formant_new, aes(x=vowel, y=F2, fill=vowel))+xlab("Vowel type")+
scale_fill_manual(values = c("#d8b365", "#5ab4ac","#F67280", "#6c5b7b"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=TRUE)
boxplotv3 <- ggplot(formant_new, aes(x=vowel, y=F3, fill=vowel))+ylim(5150,5750)+xlab("Vowel type")+
scale_fill_manual(values = c("#d8b365", "#5ab4ac","#F67280", "#6c5b7b"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=TRUE)
ggarrange(boxplotv1, boxplotv2, boxplotv3
,common.legend = TRUE,
ncol = 3, nrow = 1)## Warning: Removed 35 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
formant_new$Preceding.V <-as.character(formant_new$Preceding.V)
formant_new[is.na(formant_new)] <- "base"
formant_new$Preceding.V <-as.factor(formant_new$Preceding.V)
formant_new$Preceding.V <- relevel(formant_new$Preceding.V, ref="base")
formant.aeonly <-formant_new %>%
filter(vowel=="a"| vowel=="e")
formant_F1 <- lmer (F1 ~ Preceding.V+vowel+(1+Preceding.V+vowel|speaker), data = formant.aeonly)## boundary (singular) fit: see help('isSingular')formant_F2 <- lmer (F2 ~ Preceding.V+vowel+(1+Preceding.V+vowel|speaker), data = formant.aeonly)## boundary (singular) fit: see help('isSingular')formant_F3 <- lmer (F3 ~ Preceding.V+vowel+(1+Preceding.V+vowel|speaker), data = formant.aeonly)## boundary (singular) fit: see help('isSingular')## Warning: Model failed to converge with 1 negative eigenvalue: -6.2e-01summary(formant_F1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ Preceding.V + vowel + (1 + Preceding.V + vowel | speaker)
## Data: formant.aeonly
##
## REML criterion at convergence: 8014.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.6500 -0.6130 0.0211 0.6231 4.6582
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 112.230 10.594
## Preceding.Va 4.223 2.055 -0.34
## Preceding.Ve 25.485 5.048 -0.75 0.87
## Preceding.Vo 245.125 15.656 0.14 0.87 0.55
## Preceding.Vu 76.088 8.723 -0.98 0.51 0.86 0.05
## vowela 383.016 19.571 -0.20 -0.84 -0.50 -1.00 0.01
## Residual 2379.020 48.775
## Number of obs: 757, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 434.226 6.192 3.500 70.127 1.23e-06 ***
## Preceding.Va -9.822 6.715 53.110 -1.463 0.14946
## Preceding.Ve -5.815 6.770 10.918 -0.859 0.40880
## Preceding.Vo -11.100 8.658 2.989 -1.282 0.29021
## Preceding.Vu -19.624 6.462 4.427 -3.037 0.03375 *
## vowela 83.135 9.755 3.327 8.522 0.00226 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Va Preceding.Ve Preceding.Vo Preceding.Vu
## Preceding.Va -0.184
## Preceding.Ve -0.553 0.175
## Preceding.Vo -0.113 0.274 0.334
## Preceding.Vu -0.762 0.265 0.445 0.221
## vowela -0.286 -0.236 -0.023 -0.736 0.032
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(formant_F2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F2 ~ Preceding.V + vowel + (1 + Preceding.V + vowel | speaker)
## Data: formant.aeonly
##
## REML criterion at convergence: 9956.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.6205 -0.6931 0.0409 0.7272 2.3938
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 1651.7 40.64
## Preceding.Va 328.4 18.12 0.11
## Preceding.Ve 2457.4 49.57 0.71 0.78
## Preceding.Vo 1335.0 36.54 -0.92 0.30 -0.36
## Preceding.Vu 539.8 23.23 0.95 0.41 0.89 -0.74
## vowela 729.5 27.01 -0.38 -0.96 -0.92 -0.02 -0.65
## Residual 31678.6 177.98
## Number of obs: 757, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1898.750 23.092 3.735 82.226 3.18e-07 ***
## Preceding.Va -7.030 25.677 14.995 -0.274 0.78797
## Preceding.Ve -41.176 32.471 5.713 -1.268 0.25399
## Preceding.Vo -10.376 24.275 2.791 -0.427 0.69988
## Preceding.Vu -39.737 21.183 9.622 -1.876 0.09129 .
## vowela -124.717 19.452 3.375 -6.411 0.00533 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Va Preceding.Ve Preceding.Vo Preceding.Vu
## Preceding.Va -0.112
## Preceding.Ve 0.124 0.278
## Preceding.Vo -0.756 0.287 0.028
## Preceding.Vu 0.068 0.301 0.528 0.029
## vowela -0.437 -0.426 -0.238 -0.002 -0.158
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(formant_F3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ Preceding.V + vowel + (1 + Preceding.V + vowel | speaker)
## Data: formant.aeonly
##
## REML criterion at convergence: 9399.6
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.3919 -0.5442 0.0845 0.6386 3.8822
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 612.8 24.75
## Preceding.Va 312.3 17.67 0.69
## Preceding.Ve 706.2 26.57 -1.00 -0.62
## Preceding.Vo 970.3 31.15 0.80 0.11 -0.85
## Preceding.Vu 101.8 10.09 -0.47 -0.96 0.39 0.16
## vowela 1299.8 36.05 0.03 -0.71 -0.11 0.62 0.87
## Residual 15053.6 122.69
## Number of obs: 757, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5524.090 14.847 4.345 372.070 5.94e-11 ***
## Preceding.Va 9.566 18.605 10.199 0.514 0.6181
## Preceding.Ve -10.618 20.073 5.265 -0.529 0.6184
## Preceding.Vo 7.489 18.744 4.489 0.400 0.7078
## Preceding.Vu -7.432 13.464 14.735 -0.552 0.5892
## vowela -53.017 19.448 3.931 -2.726 0.0537 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Va Preceding.Ve Preceding.Vo Preceding.Vu
## Preceding.Va 0.101
## Preceding.Ve -0.769 -0.070
## Preceding.Vo 0.208 0.213 -0.199
## Preceding.Vu -0.478 0.091 0.344 0.294
## vowela -0.193 -0.417 0.099 0.397 0.293
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')#individual for a
formant_a.a <- formant_new %>%
filter(vowel=="a", Preceding.V=="a")
mean(formant_a.a$F1)## [1] 508.9103mean(formant_a.a$F2)## [1] 1756.225mean(formant_a.a$F3)## [1] 5485.566formant_a.a <- formant_new %>%
filter(vowel=="a", Preceding.V=="o")
mean(formant_a.a$F1)## [1] 510.9087mean(formant_a.a$F2)## [1] 1749.842mean(formant_a.a$F3)## [1] 5470.577formant_a.a <- formant_new %>%
filter(vowel=="a", Preceding.V=="u")
mean(formant_a.a$F1)## [1] 491.7216mean(formant_a.a$F2)## [1] 1719.372mean(formant_a.a$F3)## [1] 5469.995#individual for e
formant_e.e <- formant_new %>%
filter(vowel=="e", Preceding.V=="e")
mean(formant_e.e$F1)## [1] 429.6815mean(formant_e.e$F2)## [1] 1833.033mean(formant_e.e$F3)## [1] 5513.416formant_e.e <- formant_new %>%
filter(vowel=="e", Preceding.V=="o")
mean(formant_e.e$F1)## [1] 422.9433mean(formant_e.e$F2)## [1] 1897.809mean(formant_e.e$F3)## [1] 5514.976formant_e.e <- formant_new %>%
filter(vowel=="e", Preceding.V=="u")
mean(formant_e.e$F1)## [1] 421.9054mean(formant_e.e$F2)## [1] 1836.825mean(formant_e.e$F3)## [1] 5516.203##pairwise a
formant_a.a <- formant_new %>%
filter(vowel=="a", Preceding.V=="a" | Preceding.V== "base")
formant_F1 <- lmer (F1 ~ Preceding.V+(1|speaker), data = formant_a.a)
formant_F2 <- lmer (F2 ~ Preceding.V+(1|speaker), data = formant_a.a)
formant_F3 <- lmer (F3 ~ Preceding.V+(1|speaker), data = formant_a.a)
summary(formant_F1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2123.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.8220 -0.6296 -0.0464 0.6877 2.4001
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 337.7 18.38
## Residual 2587.8 50.87
## Number of obs: 199, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 521.493 9.695 4.350 53.790 2.61e-07 ***
## Preceding.Va -13.124 7.421 194.789 -1.768 0.0786 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.326summary(formant_F2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F2 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2539.3
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.1618 -0.6935 0.0944 0.6458 2.7182
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 349.9 18.7
## Residual 21946.8 148.1
## Number of obs: 199, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1780.895 16.500 5.556 107.931 1.92e-10 ***
## Preceding.Va -21.630 21.486 196.473 -1.007 0.315
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.536summary(formant_F3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2482.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.2432 -0.5915 0.1154 0.5928 2.5594
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 1720 41.47
## Residual 16088 126.84
## Number of obs: 199, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5453.023 22.521 4.645 242.126 1.05e-10 ***
## Preceding.Va 31.030 18.495 195.104 1.678 0.095 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.349formant_a.a <- formant_new %>%
filter(vowel=="a", Preceding.V=="o" | Preceding.V== "base")
formant_F1 <- lmer (F1 ~ Preceding.V+(1|speaker), data = formant_a.a)
formant_F2 <- lmer (F2 ~ Preceding.V+(1|speaker), data = formant_a.a)
formant_F3 <- lmer (F3 ~ Preceding.V+(1|speaker), data = formant_a.a)
summary(formant_F1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2176.2
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.5849 -0.5974 0.0353 0.5770 2.5794
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 219.2 14.80
## Residual 2598.3 50.97
## Number of obs: 204, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 519.404 8.347 4.641 62.228 5.83e-08 ***
## Preceding.Vo -10.010 7.633 196.236 -1.311 0.191
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.389summary(formant_F2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F2 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2627.7
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.47129 -0.60887 0.04162 0.61154 2.56136
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 667.9 25.84
## Residual 24633.5 156.95
## Number of obs: 204, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1784.61 19.14 3.61 93.251 3.09e-07 ***
## Preceding.Vo -37.91 23.04 160.32 -1.646 0.102
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.510summary(formant_F3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2551
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.8684 -0.6305 0.1015 0.6038 3.5092
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 4722 68.72
## Residual 16290 127.63
## Number of obs: 204, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5457.108 33.341 4.585 163.675 8.18e-10 ***
## Preceding.Vo 36.545 19.384 201.890 1.885 0.0608 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.249formant_a.a <- formant_new %>%
filter(vowel=="a", Preceding.V=="u" | Preceding.V== "base")
formant_F1 <- lmer (F1 ~ Preceding.V+(1|speaker), data = formant_a.a)
formant_F2 <- lmer (F2 ~ Preceding.V+(1|speaker), data = formant_a.a)
formant_F3 <- lmer (F3 ~ Preceding.V+(1|speaker), data = formant_a.a)
summary(formant_F1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 1991.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.9395 -0.6220 -0.0079 0.5590 4.1890
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 378.4 19.45
## Residual 3231.6 56.85
## Number of obs: 183, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 523.185 10.421 4.040 50.205 8.41e-07 ***
## Preceding.Vu -35.570 9.184 180.863 -3.873 0.00015 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.338summary(formant_F2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F2 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2368.6
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.39321 -0.68181 -0.05623 0.73020 2.51698
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 1482 38.49
## Residual 26260 162.05
## Number of obs: 183, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1787.63 23.58 5.40 75.811 2.19e-09 ***
## Preceding.Vu -65.23 25.96 178.28 -2.513 0.0129 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.416summary(formant_F3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ Preceding.V + (1 | speaker)
## Data: formant_a.a
##
## REML criterion at convergence: 2270.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.4718 -0.5551 0.1200 0.5528 2.7111
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 1444 37.99
## Residual 15186 123.23
## Number of obs: 183, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5458.07 21.03 4.65 259.542 7.41e-11 ***
## Preceding.Vu 13.08 19.86 180.48 0.658 0.511
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.361##pairwise e
formant_e.e <- formant_new %>%
filter(vowel=="e", Preceding.V=="e" | Preceding.V== "base")
formant_F1 <- lmer (F1 ~ Preceding.V+(1|speaker), data = formant_e.e)
formant_F2 <- lmer (F2 ~ Preceding.V+(1|speaker), data = formant_e.e)
formant_F3 <- lmer (F3 ~ Preceding.V+(1|speaker), data = formant_e.e)
summary(formant_F1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2164.6
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.03558 -0.61794 0.05908 0.56374 2.63811
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 80.17 8.954
## Residual 1817.64 42.634
## Number of obs: 210, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 430.763 5.746 6.114 74.963 2.7e-10 ***
## Preceding.Ve -3.690 5.936 204.643 -0.622 0.535
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.457summary(formant_F2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F2 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2797
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.14193 -0.77589 -0.00189 0.82217 2.01627
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 4031 63.49
## Residual 37559 193.80
## Number of obs: 210, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1888.714 34.142 5.279 55.320 1.68e-08 ***
## Preceding.Ve -37.295 26.990 204.426 -1.382 0.169
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.351summary(formant_F3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2587.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.5335 -0.5456 0.0491 0.5845 2.7194
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 97.28 9.863
## Residual 14109.38 118.783
## Number of obs: 210, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5542.90 11.97 11.78 462.89 <2e-16 ***
## Preceding.Ve -30.41 16.53 206.30 -1.84 0.0672 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.607formant_e.e <- formant_new %>%
filter(vowel=="e", Preceding.V=="o" | Preceding.V== "base")
formant_F1 <- lmer (F1 ~ Preceding.V+(1|speaker), data = formant_e.e)
formant_F2 <- lmer (F2 ~ Preceding.V+(1|speaker), data = formant_e.e)
formant_F3 <- lmer (F3 ~ Preceding.V+(1|speaker), data = formant_e.e)
summary(formant_F1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2315.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.70243 -0.59386 0.07007 0.60458 2.50228
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 140.3 11.85
## Residual 2248.7 47.42
## Number of obs: 220, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 430.827 7.018 5.001 61.387 2.16e-08 ***
## Preceding.Vo -11.024 6.511 217.610 -1.693 0.0919 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.415summary(formant_F2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F2 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2929.3
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.3549 -0.7343 0.1896 0.8059 1.6997
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 308.1 17.55
## Residual 38220.0 195.50
## Number of obs: 220, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1872.360 19.962 9.872 93.796 6.78e-16 ***
## Preceding.Vo 27.265 26.573 215.966 1.026 0.306
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.609summary(formant_F3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2737.3
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.5073 -0.5082 0.0204 0.6416 2.6919
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 1031 32.11
## Residual 15565 124.76
## Number of obs: 220, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5537.368 18.791 6.249 294.688 3.49e-14 ***
## Preceding.Vo -29.035 17.134 217.629 -1.695 0.0916 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.408formant_e.e <- formant_new %>%
filter(vowel=="e", Preceding.V=="u" | Preceding.V== "base")
formant_F1 <- lmer (F1 ~ Preceding.V+(1|speaker), data = formant_e.e)
formant_F2 <- lmer (F2 ~ Preceding.V+(1|speaker), data = formant_e.e)
formant_F3 <- lmer (F3 ~ Preceding.V+(1|speaker), data = formant_e.e)
summary(formant_F1)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2251.8
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.7027 -0.6098 0.0163 0.5698 4.1026
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 21.95 4.685
## Residual 2173.06 46.616
## Number of obs: 215, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 432.795 4.867 5.995 88.931 1.38e-10 ***
## Preceding.Vu -12.080 6.396 210.963 -1.889 0.0603 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.587summary(formant_F2)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F2 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2875.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.12812 -0.76171 0.08353 0.81888 1.95301
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 2785 52.77
## Residual 39986 199.97
## Number of obs: 215, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1885.217 30.581 5.586 61.648 3.94e-09 ***
## Preceding.Vu -30.919 27.494 210.316 -1.125 0.262
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.396summary(formant_F3)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ Preceding.V + (1 | speaker)
## Data: formant_e.e
##
## REML criterion at convergence: 2643.4
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -4.7544 -0.5185 0.0793 0.6411 2.8338
##
## Random effects:
## Groups Name Variance Std.Dev.
## speaker (Intercept) 510.8 22.6
## Residual 13522.7 116.3
## Number of obs: 215, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5541.617 15.092 5.581 367.186 1.9e-13 ***
## Preceding.Vu -30.056 15.978 210.433 -1.881 0.0613 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr)
## Preceding.V -0.468formant_new[formant_new=="base"] <- NAboxplot1a <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & vowel!="e" & Preceding.V !="base"), aes(x=Preceding.V, y=F1, fill=Preceding.V))+
xlab("Preceding vowel")+scale_fill_manual(values = c("#F67280","#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=TRUE, alpha=0.6)
boxplot2a <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & vowel!="e" & Preceding.V !="base"), aes(x=Preceding.V, y=F2, fill=Preceding.V))+
xlab("Preceding vowel")+scale_fill_manual(values = c("#F67280","#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=TRUE, alpha=0.6)
boxplot3a <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & vowel!="e" & Preceding.V !="base"), aes(x=Preceding.V, y=F3, fill=Preceding.V))+ylim(5150,5750)+
xlab("Preceding vowel")+scale_fill_manual(values = c("#F67280","#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=TRUE, alpha=0.6)
boxplot1e <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & vowel!="a" & Preceding.V !="base"), aes(x=Preceding.V, y=F1, fill=Preceding.V))+
xlab("Preceding vowel")+scale_fill_manual(values = c("#6c5b7b","#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=TRUE, alpha=0.6)
boxplot2e <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & vowel!="a" & Preceding.V !="base"), aes(x=Preceding.V, y=F2, fill=Preceding.V))+
xlab("Preceding vowel")+scale_fill_manual(values = c("#6c5b7b","#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=TRUE, alpha=0.6)
boxplot3e <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & vowel!="a" & Preceding.V !="base"), aes(x=Preceding.V, y=F3, fill=Preceding.V))+ylim(5150,5750)+
xlab("Preceding vowel")+scale_fill_manual(values = c("#6c5b7b","#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=TRUE, alpha=0.6)
boxplotae <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & Preceding.V !="base"), aes(x=Preceding.V, y=F3, fill=Preceding.V))+ylim(5150,5750)+
xlab("Preceding vowel")+scale_fill_manual(values = c("#F67280","#6c5b7b","#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=TRUE, alpha=0.6)+theme(
legend.direction = "horizontal")
ggarrange(boxplot1a+ rremove("x.title"), boxplot2a + rremove("x.title"), boxplot3a + rremove("x.title"),boxplot1e+ rremove("x.title"), boxplot2e + rremove("x.title"), boxplot3e + rremove("x.title"),legend.grob = get_legend(boxplotae),
ncol = 3, nrow = 2)## Warning: Removed 17 rows containing non-finite outside the scale range
## (`stat_boxplot()`).## Warning: Removed 6 rows containing non-finite outside the scale range
## (`stat_boxplot()`).## Warning: Removed 11 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
##vowel harmony participation charts
boxplot4 <- ggplot(subset(formant_new,vowel!="o" & vowel!="u"& Preceding.V!="a" &Preceding.V!="e"), aes(x=interaction(Preceding.V, vowel), y=F1, fill=VH))+ylim(300,650)+xlab("Preceding vowel.Vowel")+
scale_fill_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=TRUE)
boxplot5 <- ggplot(subset(formant_new,vowel!="o" & vowel!="u"& Preceding.V!="a" &Preceding.V!="e"), aes(x=interaction(Preceding.V, vowel), y=F2, fill=VH))+xlab("Preceding vowel.Vowel")+ylim(1250,2250)+
scale_fill_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=TRUE)
boxplot6 <- ggplot(subset(formant_new,vowel!="o" & vowel!="u"& Preceding.V!="a" &Preceding.V!="e"), aes(x=interaction(Preceding.V, vowel), y=F3, fill=VH))+ylim(5000,5900)+xlab("Preceding vowel.Vowel")+
scale_fill_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=TRUE)
boxplot7 <- ggplot(subset(formant_new,vowel!="o" & vowel!="u"& Preceding.V!="a" &Preceding.V!="e"), aes(x=interaction(Preceding.V, vowel), y=F1+F2, fill=VH))+xlab("Preceding vowel.Vowel")+
scale_fill_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#F67280", "#6c5b7b"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=TRUE)
rm_legend <- function(p){p + theme(legend.position = "none")}
plots <- ggarrange(boxplot4+ rremove("x.text")+ rremove("x.title")+rremove("legend"), boxplot5+rremove("x.text")+ rremove("x.title")+rremove("legend"), boxplot6+rremove("legend"), boxplot7+rremove("legend"),
labels = c("A", "B", "C", "D"),
ncol = 2, nrow = 2)## Warning: Removed 5 rows containing non-finite outside the scale range
## (`stat_boxplot()`).## Warning: Removed 3 rows containing non-finite outside the scale range
## (`stat_boxplot()`).legend_1 <- get_legend(boxplot4)## Warning: Removed 5 rows containing non-finite outside the scale range
## (`stat_boxplot()`).legend_2 <- get_legend(boxplot7)
legends <- ggarrange(legend_1, legend_2, nrow=1)
ggarrange(plots,legends, heights= c(0.84, 0.16),nrow=2)
boxplot8 <- ggplot(subset(formant_new,vowel!="o" & vowel!="u"& Preceding.V!="a" &Preceding.V!="e"), aes(x=Preceding.V, y=F1, fill=VH))+xlab("Preceding vowel")+ylim(300,600)+
scale_fill_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=FALSE)+
facet_wrap(.~speaker)+theme(
strip.background = element_blank(),
strip.text.x = element_blank()
)
boxplot9 <- ggplot(subset(formant_new,vowel!="o" & vowel!="u"& Preceding.V!="a" &Preceding.V!="e"), aes(x=Preceding.V, y=F2, fill=VH))+xlab("Preceding vowel")+
scale_fill_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=FALSE)+
facet_wrap(.~speaker)+theme(
strip.background = element_blank(),
strip.text.x = element_blank()
)
boxplot10 <- ggplot(subset(formant_new,vowel!="o" & vowel!="u"& Preceding.V!="a" &Preceding.V!="e"), aes(x=Preceding.V, y=F3, fill=VH))+xlab("Preceding vowel")+ylim(5000,5900)+
scale_fill_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=1, notch=FALSE)+
facet_wrap(.~speaker)+theme(
strip.background = element_blank(),
strip.text.x = element_blank()
)
ggarrange(boxplot8+ rremove("x.text")+ rremove("x.title"), boxplot9, boxplot10
,common.legend = TRUE,
ncol = 2, nrow = 2)## Warning: Removed 9 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
## Removed 9 rows containing non-finite outside the scale range
## (`stat_boxplot()`).## Warning: Removed 3 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
VC_participation1 <- ggplot(subset(formant_new,vowel =="a"& Preceding.V!="a" &Preceding.V!="e"), aes(x = F2, y = F1, colour = VH)) +
stat_ellipse(aes(color = VH), geom='polygon', alpha = 0.1) +geom_point(alpha = 0.3) +
stat_ellipse(type = "t") +
stat_ellipse(type = "norm", linetype = 4)+ggtitle("distibution of /a/")+
scale_color_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
scale_x_reverse() +
scale_y_reverse()
VC_participation2 <- ggplot(subset(formant_new,vowel =="e"& Preceding.V!="a" &Preceding.V!="e"), aes(x = F2, y = F1, colour = VH)) +
stat_ellipse(aes(color = VH), geom='polygon', alpha = 0.1) +geom_point(alpha = 0.3) +
stat_ellipse(type = "t") +
stat_ellipse(type = "norm", linetype = 4)+ggtitle("distibution of /e/")+
scale_color_manual(name = "Vowel Harmony", labels = c("No", "Yes"),values = c("#d8b365", "#5ab4ac"))+
scale_x_reverse() +
scale_y_reverse()
ggarrange(VC_participation1, VC_participation2,common.legend = TRUE,
ncol = 2)
boxplot8## Warning: Removed 9 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
boxplot <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & VH =="Y"), aes(x=Preceding.V, y=F2, fill=vowel))+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=FALSE)
boxplot
boxplot <- ggplot(subset(formant_new,vowel!="o" & vowel!="u" & VH =="Y"), aes(x=Preceding.V, y=F3, fill=vowel))+ylim(5150,5750)+
geom_boxplot(outlier.colour="black", outlier.shape=16,
outlier.size=2, notch=FALSE)
boxplot## Warning: Removed 7 rows containing non-finite outside the scale range
## (`stat_boxplot()`).
boxplot <- ggplot(subset(formant_new,vowel != "o" & vowel != "u"& VH =="Y"), aes(x=vowel, y=F2, fill=Preceding.V, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.5, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()## Warning: The `<scale>` argument of `guides()` cannot be `FALSE`. Use "none" instead as
## of ggplot2 3.3.4.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.boxplot
boxplot <- ggplot(subset(formant_new,vowel != "o" & vowel != "u"& VH =="Y"), aes(x=vowel, y=F1, fill=Preceding.V, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.5, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()
boxplot
boxplot <- ggplot(subset(formant_new,vowel == "a" ), aes(x=VH, y=F2, fill=Preceding.V, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.5, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()
boxplot
boxplot <- ggplot(subset(formant_new,vowel == "a" ), aes(x=VH, y=F1, fill=Preceding.V, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.5, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()
boxplot
boxplot1 <- ggplot(subset(formant_new,vowel == "a" ), aes(x=VH, y=F1, fill=speaker, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.6, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()
boxplot1
boxplot2 <- ggplot(subset(formant_new,vowel == "a" ), aes(x=VH, y=F2, fill=speaker, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.5, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()
boxplot2## Notch went outside hinges
## ℹ Do you want `notch = FALSE`?
## Notch went outside hinges
## ℹ Do you want `notch = FALSE`?
boxplot3 <- ggplot(subset(formant_new,vowel == "e"), aes(x=VH, y=F1, fill=speaker, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.6, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()
boxplot3## Notch went outside hinges
## ℹ Do you want `notch = FALSE`?
boxplot4 <- ggplot(subset(formant_new, vowel == "e" ), aes(x=VH, y=F2, fill=speaker, alpha=0.2))+
scale_fill_brewer(palette = "PuOr")+
geom_boxplot(outlier.shape=0.2, outlier.size=0.1, width=0.5, notch=TRUE) +
guides(alpha = FALSE)+
scale_y_reverse()+coord_flip()
boxplot4## Notch went outside hinges
## ℹ Do you want `notch = FALSE`?
formant_new[is.na(formant_new)] <- "base"
formant.ae <- filter(formant_new, vowel != "o" & vowel != "u" & Preceding.V != "a" & Preceding.V != "e") %>%
droplevels()
formant.ae$vowel <- relevel(formant.ae$vowel, ref="a")
F1.VT.ou <- glm(F1 ~ Preceding.V+vowel+VH, data = formant.ae)
F2.VT.ou <- glm(F2 ~ Preceding.V+vowel+VH, data = formant.ae)
summary(F1.VT.ou)##
## Call:
## glm(formula = F1 ~ Preceding.V + vowel + VH, data = formant.ae)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 518.648 3.978 130.367 < 2e-16 ***
## Preceding.Vo -7.652 5.794 -1.321 0.18717
## Preceding.Vu -16.778 5.765 -2.910 0.00375 **
## vowele -82.683 4.300 -19.231 < 2e-16 ***
## VHY -5.844 5.577 -1.048 0.29517
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 2665.05)
##
## Null deviance: 2609323 on 584 degrees of freedom
## Residual deviance: 1545729 on 580 degrees of freedom
## AIC: 6281.6
##
## Number of Fisher Scoring iterations: 2summary(F2.VT.ou)##
## Call:
## glm(formula = F2 ~ Preceding.V + vowel + VH, data = formant.ae)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1765.65 14.10 125.268 < 2e-16 ***
## Preceding.Vo 38.57 20.53 1.879 0.060742 .
## Preceding.Vu -15.68 20.43 -0.768 0.442954
## vowele 116.56 15.23 7.652 8.31e-14 ***
## VHY -73.32 19.76 -3.711 0.000226 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 33452.42)
##
## Null deviance: 21992707 on 584 degrees of freedom
## Residual deviance: 19402405 on 580 degrees of freedom
## AIC: 7761.6
##
## Number of Fisher Scoring iterations: 2F1.VT.ou.VH <- glm(F1 ~ Preceding.V+vowel+VH, data = formant.ae)
F2.VT.ou.VH <- glm(F2 ~ Preceding.V+vowel+VH, data = formant.ae)
F3.VT.ou.VH <- glm(F3 ~ Preceding.V+vowel+VH, data = formant.ae)
summary(F1.VT.ou.VH)##
## Call:
## glm(formula = F1 ~ Preceding.V + vowel + VH, data = formant.ae)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 518.648 3.978 130.367 < 2e-16 ***
## Preceding.Vo -7.652 5.794 -1.321 0.18717
## Preceding.Vu -16.778 5.765 -2.910 0.00375 **
## vowele -82.683 4.300 -19.231 < 2e-16 ***
## VHY -5.844 5.577 -1.048 0.29517
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 2665.05)
##
## Null deviance: 2609323 on 584 degrees of freedom
## Residual deviance: 1545729 on 580 degrees of freedom
## AIC: 6281.6
##
## Number of Fisher Scoring iterations: 2summary(F2.VT.ou.VH)##
## Call:
## glm(formula = F2 ~ Preceding.V + vowel + VH, data = formant.ae)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1765.65 14.10 125.268 < 2e-16 ***
## Preceding.Vo 38.57 20.53 1.879 0.060742 .
## Preceding.Vu -15.68 20.43 -0.768 0.442954
## vowele 116.56 15.23 7.652 8.31e-14 ***
## VHY -73.32 19.76 -3.711 0.000226 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 33452.42)
##
## Null deviance: 21992707 on 584 degrees of freedom
## Residual deviance: 19402405 on 580 degrees of freedom
## AIC: 7761.6
##
## Number of Fisher Scoring iterations: 2summary(F3.VT.ou.VH)##
## Call:
## glm(formula = F3 ~ Preceding.V + vowel + VH, data = formant.ae)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5465.681 10.010 545.995 < 2e-16 ***
## Preceding.Vo -2.094 14.579 -0.144 0.886
## Preceding.Vu -3.500 14.507 -0.241 0.809
## vowele 64.501 10.819 5.962 4.33e-09 ***
## VHY -7.825 14.033 -0.558 0.577
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 16873.5)
##
## Null deviance: 10393437 on 584 degrees of freedom
## Residual deviance: 9786630 on 580 degrees of freedom
## AIC: 7361.2
##
## Number of Fisher Scoring iterations: 2F1.VT.ou.rand <- lmer (F1 ~ Preceding.V*vowel+VH+(1+Preceding.V*vowel+VH|speaker), data = formant.ae, REML=FALSE)## boundary (singular) fit: see help('isSingular')F2.VT.ou.rand <- lmer (F2 ~ Preceding.V*vowel+VH+(1+Preceding.V*vowel+VH|speaker), data = formant.ae, REML=FALSE)## boundary (singular) fit: see help('isSingular')F3.VT.ou.rand <- lmer (F3 ~ Preceding.V*vowel+VH+(1+Preceding.V*vowel+VH|speaker), data = formant.ae, REML=FALSE)## boundary (singular) fit: see help('isSingular')summary(F1.VT.ou.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F1 ~ Preceding.V * vowel + VH + (1 + Preceding.V * vowel + VH |
## speaker)
## Data: formant.ae
##
## AIC BIC logLik deviance df.resid
## 6320.0 6477.3 -3124.0 6248.0 549
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.5061 -0.6012 0.0262 0.5785 4.6509
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 338.8 18.41
## Preceding.Vo 232.7 15.25 -0.38
## Preceding.Vu 161.4 12.71 -0.39 -0.45
## vowele 291.7 17.08 -0.84 0.82 0.03
## VHY 135.2 11.63 -0.76 -0.30 0.79 0.30
## Preceding.Vo:vowele 148.9 12.20 -0.44 -0.39 0.13 -0.01 0.64
## Preceding.Vu:vowele 194.3 13.94 0.45 0.17 -0.95 -0.25 -0.68
## Residual 2468.0 49.68
##
##
##
##
##
##
##
## 0.12
##
## Number of obs: 585, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 521.852 9.655 3.608 54.051 2.22e-06 ***
## Preceding.Vo -9.487 10.667 4.561 -0.889 0.41827
## Preceding.Vu -31.579 10.054 6.568 -3.141 0.01778 *
## vowele -91.268 10.261 3.001 -8.895 0.00299 **
## VHY -5.504 7.610 4.270 -0.723 0.50717
## Preceding.Vo:vowele 3.683 11.465 7.063 0.321 0.75734
## Preceding.Vu:vowele 24.299 12.392 6.921 1.961 0.09118 .
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Vo Preceding.Vu vowele VHY
## Preceding.Vo -0.445
## Preceding.Vu -0.449 0.119
## vowele -0.783 0.616 0.255
## VHY -0.459 -0.352 0.195 0.159
## Preceding.Vo:vwl 0.027 -0.574 -0.186 -0.383 0.240
## Preceding.Vu:vwl 0.410 -0.113 -0.791 -0.463 -0.290
## Preceding.Vo:vwl
## Preceding.Vo
## Preceding.Vu
## vowele
## VHY
## Preceding.Vo:vwl
## Preceding.Vu:vwl 0.351
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(F2.VT.ou.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F2 ~ Preceding.V * vowel + VH + (1 + Preceding.V * vowel + VH |
## speaker)
## Data: formant.ae
##
## AIC BIC logLik deviance df.resid
## 7805.8 7963.2 -3866.9 7733.8 549
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.5156 -0.7249 0.0514 0.7088 2.2826
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 393.1 19.83
## Preceding.Vo 499.8 22.36 -0.01
## Preceding.Vu 1308.3 36.17 0.92 -0.39
## vowele 1335.9 36.55 0.49 -0.87 0.79
## VHY 329.6 18.15 0.80 -0.60 0.97 0.92
## Preceding.Vo:vowele 3214.4 56.70 -1.00 0.03 -0.94 -0.52 -0.82
## Preceding.Vu:vowele 3315.3 57.58 -0.99 0.15 -0.97 -0.62 -0.88
## Residual 31621.0 177.82
##
##
##
##
##
##
##
## 0.99
##
## Number of obs: 585, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1784.855 18.757 8.429 95.156 4.47e-14 ***
## Preceding.Vo -5.328 29.818 14.528 -0.179 0.86066
## Preceding.Vu -33.878 32.963 11.083 -1.028 0.32597
## vowele 103.958 28.630 8.062 3.631 0.00659 **
## VHY -63.435 21.130 18.047 -3.002 0.00763 **
## Preceding.Vo:vowele 44.358 43.997 6.102 1.008 0.35165
## Preceding.Vu:vowele 28.389 45.409 7.785 0.625 0.54973
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Vo Preceding.Vu vowele VHY
## Preceding.Vo -0.486
## Preceding.Vu -0.223 0.285
## vowele -0.363 0.138 0.510
## VHY 0.160 -0.419 0.000 0.209
## Preceding.Vo:vwl 0.037 -0.517 -0.461 -0.600 -0.155
## Preceding.Vu:vwl 0.042 -0.149 -0.783 -0.611 -0.242
## Preceding.Vo:vwl
## Preceding.Vo
## Preceding.Vu
## vowele
## VHY
## Preceding.Vo:vwl
## Preceding.Vu:vwl 0.598
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(F3.VT.ou.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F3 ~ Preceding.V * vowel + VH + (1 + Preceding.V * vowel + VH |
## speaker)
## Data: formant.ae
##
## AIC BIC logLik deviance df.resid
## 7364.6 7522.0 -3646.3 7292.6 549
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.3985 -0.5358 0.0607 0.6211 3.6926
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 1916.710 43.780
## Preceding.Vo 1261.969 35.524 0.90
## Preceding.Vu 390.452 19.760 -0.98 -0.97
## vowele 2175.358 46.641 -0.87 -0.56 0.76
## VHY 1.546 1.243 -0.88 -0.57 0.76 1.00
## Preceding.Vo:vowele 261.503 16.171 -0.30 -0.69 0.48 -0.21 -0.20
## Preceding.Vu:vowele 2660.585 51.581 0.98 0.80 -0.93 -0.95 -0.95
## Residual 14754.492 121.468
##
##
##
##
##
##
##
## -0.11
##
## Number of obs: 585, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5455.656 22.815 4.620 239.129 1.24e-10 ***
## Preceding.Vo 52.561 25.605 8.120 2.053 0.0737 .
## Preceding.Vu 19.490 21.625 12.450 0.901 0.3846
## vowele 82.736 26.790 5.079 3.088 0.0267 *
## VHY -14.962 13.278 406.633 -1.127 0.2605
## Preceding.Vo:vowele -67.869 25.740 31.555 -2.637 0.0129 *
## Preceding.Vu:vowele -38.436 34.808 5.718 -1.104 0.3138
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Vo Preceding.Vu vowele VHY
## Preceding.Vo 0.249
## Preceding.Vu -0.621 0.064
## vowele -0.814 -0.071 0.483
## VHY -0.026 -0.329 -0.202 0.027
## Preceding.Vo:vwl 0.165 -0.647 -0.209 -0.439 0.070
## Preceding.Vu:vwl 0.744 0.176 -0.744 -0.795 -0.061
## Preceding.Vo:vwl
## Preceding.Vo
## Preceding.Vu
## vowele
## VHY
## Preceding.Vo:vwl
## Preceding.Vu:vwl 0.289
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')formant.all <- filter(formant_new, vowel != "o" & vowel != "u") %>%
droplevels()
formant.all[formant.all=="base"] <- NA
F1.VT.all <- glm(F1 ~ Preceding.V+vowel+VH, data = formant.all)
F2.VT.all <- glm(F2 ~ Preceding.V+vowel+VH, data = formant.all)
summary(F1.VT.all)##
## Call:
## glm(formula = F1 ~ Preceding.V + vowel + VH, data = formant.all)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 434.9695 9.4025 46.261 <2e-16 ***
## Preceding.Ve 0.5359 9.3178 0.058 0.9542
## Preceding.Vo -5.3400 7.7212 -0.692 0.4895
## Preceding.Vu -14.6314 8.1658 -1.792 0.0738 .
## vowela 79.7647 5.3936 14.789 <2e-16 ***
## VHY -5.8239 5.3696 -1.085 0.2786
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 2470.322)
##
## Null deviance: 2107190 on 519 degrees of freedom
## Residual deviance: 1269746 on 514 degrees of freedom
## (237 observations deleted due to missingness)
## AIC: 5546
##
## Number of Fisher Scoring iterations: 2summary(F2.VT.all)##
## Call:
## glm(formula = F2 ~ Preceding.V + vowel + VH, data = formant.all)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1963.04 34.48 56.933 < 2e-16 ***
## Preceding.Ve -56.80 34.17 -1.662 0.097043 .
## Preceding.Vo -34.57 28.31 -1.221 0.222672
## Preceding.Vu -89.79 29.94 -2.998 0.002846 **
## vowela -133.61 19.78 -6.755 3.87e-11 ***
## VHY -73.21 19.69 -3.718 0.000223 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 33219.62)
##
## Null deviance: 19479625 on 519 degrees of freedom
## Residual deviance: 17074887 on 514 degrees of freedom
## (237 observations deleted due to missingness)
## AIC: 6897.3
##
## Number of Fisher Scoring iterations: 2F1.VT.all.VH <- glm(F1 ~ Preceding.V+vowel+VH, data = formant.all)
F2.VT.all.VH <- glm(F2 ~ Preceding.V+vowel+VH, data = formant.all)
summary(F1.VT.all.VH)##
## Call:
## glm(formula = F1 ~ Preceding.V + vowel + VH, data = formant.all)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 434.9695 9.4025 46.261 <2e-16 ***
## Preceding.Ve 0.5359 9.3178 0.058 0.9542
## Preceding.Vo -5.3400 7.7212 -0.692 0.4895
## Preceding.Vu -14.6314 8.1658 -1.792 0.0738 .
## vowela 79.7647 5.3936 14.789 <2e-16 ***
## VHY -5.8239 5.3696 -1.085 0.2786
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 2470.322)
##
## Null deviance: 2107190 on 519 degrees of freedom
## Residual deviance: 1269746 on 514 degrees of freedom
## (237 observations deleted due to missingness)
## AIC: 5546
##
## Number of Fisher Scoring iterations: 2summary(F2.VT.all.VH)##
## Call:
## glm(formula = F2 ~ Preceding.V + vowel + VH, data = formant.all)
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1963.04 34.48 56.933 < 2e-16 ***
## Preceding.Ve -56.80 34.17 -1.662 0.097043 .
## Preceding.Vo -34.57 28.31 -1.221 0.222672
## Preceding.Vu -89.79 29.94 -2.998 0.002846 **
## vowela -133.61 19.78 -6.755 3.87e-11 ***
## VHY -73.21 19.69 -3.718 0.000223 ***
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for gaussian family taken to be 33219.62)
##
## Null deviance: 19479625 on 519 degrees of freedom
## Residual deviance: 17074887 on 514 degrees of freedom
## (237 observations deleted due to missingness)
## AIC: 6897.3
##
## Number of Fisher Scoring iterations: 2F1.VT.all.rand <- lmer (F1 ~ Preceding.V*vowel+VH+(1+Preceding.V*vowel+VH|speaker), data = formant.all, REML=FALSE)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients## boundary (singular) fit: see help('isSingular')## Warning: Model failed to converge with 2 negative eigenvalues: -1.7e-06
## -6.0e-03## Warning in as_lmerModLT(model, devfun): Model may not have converged with 7
## eigenvalues close to zero: 1.9e-13 4.2e-14 4.8e-15 6.2e-16 -2.6e-16 -1.6e-14
## -5.0e-14F2.VT.all.rand <- lmer (F2 ~ Preceding.V*vowel+VH+(1+Preceding.V*vowel+VH|speaker), data = formant.all, REML=FALSE)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## boundary (singular) fit: see help('isSingular')## Warning: Model failed to converge with 1 negative eigenvalue: -3.0e-02## Warning in as_lmerModLT(model, devfun): Model may not have converged with 7
## eigenvalues close to zero: 6.2e-14 6.5e-15 2.7e-15 -5.3e-20 -3.0e-14 -4.5e-14
## -8.1e-14F3.VT.all.rand <- lmer (F3 ~ Preceding.V*vowel+VH+(1+Preceding.V*vowel+VH|speaker),data = formant.all, REML=FALSE)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## boundary (singular) fit: see help('isSingular')## Warning in as_lmerModLT(model, devfun): Model may not have converged with 7
## eigenvalues close to zero: 1.7e-15 7.4e-18 -1.4e-14 -3.1e-14 -5.2e-14 -6.9e-14
## -7.0e-14summary(F1.VT.all.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F1 ~ Preceding.V * vowel + VH + (1 + Preceding.V * vowel + VH |
## speaker)
## Data: formant.all
##
## AIC BIC logLik deviance df.resid
## 5617.9 5843.4 -2756.0 5511.9 467
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.6564 -0.6226 0.0539 0.6196 4.8461
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 325.77 18.049
## Preceding.Ve 24.54 4.954 -0.78
## Preceding.Vo 187.17 13.681 -0.36 0.13
## Preceding.Vu 141.49 11.895 -0.99 0.85 0.31
## vowela 359.93 18.972 -0.24 0.53 -0.77 0.31
## VHY 91.93 9.588 -0.90 0.49 0.65 0.85 -0.20
## Preceding.Ve:vowela 1789.46 42.302 0.09 -0.03 -0.11 -0.08 0.07
## Preceding.Vo:vowela 42.27 6.501 0.89 -0.83 -0.62 -0.91 -0.02
## Preceding.Vu:vowela 36.72 6.059 -0.69 0.14 0.07 0.61 0.09
## Residual 2282.38 47.774
##
##
##
##
##
##
##
## -0.12
## -0.84 0.10
## 0.73 -0.08 -0.34
##
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 445.559 15.242 4.424 29.233 3.12e-06 ***
## Preceding.Ve -11.859 11.364 15.930 -1.043 0.31230
## Preceding.Vo -20.538 13.209 5.811 -1.555 0.17256
## Preceding.Vu -22.526 11.574 5.283 -1.946 0.10611
## vowela 66.419 12.073 4.938 5.501 0.00282 **
## VHY -4.869 6.852 4.729 -0.711 0.51087
## Preceding.Vo:vowela 18.649 11.670 10.851 1.598 0.13873
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Ve Preceding.Vo Preceding.Vu vowela VHY
## Preceding.Ve -0.764
## Preceding.Vo -0.756 0.689
## Preceding.Vu -0.930 0.733 0.697
## vowela -0.580 0.593 0.202 0.519
## VHY -0.698 0.224 0.487 0.602 0.065
## Prcdng.V:vw 0.646 -0.587 -0.723 -0.553 -0.521 -0.328
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(F2.VT.all.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F2 ~ Preceding.V * vowel + VH + (1 + Preceding.V * vowel + VH |
## speaker)
## Data: formant.all
##
## AIC BIC logLik deviance df.resid
## 6967.6 7193.1 -3430.8 6861.6 467
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.87241 -0.74935 0.02322 0.71296 2.38695
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 82.38 9.076
## Preceding.Ve 3446.32 58.705 -0.32
## Preceding.Vo 647.07 25.438 -0.46 0.32
## Preceding.Vu 1659.02 40.731 -0.37 0.99 0.22
## vowela 139.26 11.801 -0.79 0.59 0.00 0.69
## VHY 357.10 18.897 -0.08 0.75 -0.38 0.82 0.65
## Preceding.Ve:vowela 34242.11 185.046 -0.10 -0.11 -0.12 -0.08 0.11
## Preceding.Vo:vowela 2681.12 51.780 0.32 -0.56 -0.95 -0.44 -0.02
## Preceding.Vu:vowela 1773.78 42.116 0.36 -0.60 -0.94 -0.49 -0.08
## Residual 30678.22 175.152
##
##
##
##
##
##
##
## 0.00
## 0.13 0.17
## 0.09 0.17 1.00
##
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1963.878 44.641 12.396 43.993 5.44e-15 ***
## Preceding.Ve -44.553 44.689 11.887 -0.997 0.33864
## Preceding.Vo -35.855 41.781 43.465 -0.858 0.39551
## Preceding.Vu -78.430 37.345 10.015 -2.100 0.06202 .
## vowela -135.064 34.733 7.657 -3.889 0.00503 **
## VHY -62.123 21.034 17.533 -2.953 0.00868 **
## Preceding.Vo:vowela -13.716 39.582 63.966 -0.347 0.73010
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Ve Preceding.Vo Preceding.Vu vowela VHY
## Preceding.Ve -0.658
## Preceding.Vo -0.867 0.630
## Preceding.Vu -0.804 0.764 0.680
## vowela -0.787 0.614 0.686 0.596
## VHY -0.453 0.239 0.237 0.494 0.119
## Prcdng.V:vw 0.462 -0.439 -0.694 -0.319 -0.533 -0.068
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(F3.VT.all.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F3 ~ Preceding.V * vowel + VH + (1 + Preceding.V * vowel + VH |
## speaker)
## Data: formant.all
##
## AIC BIC logLik deviance df.resid
## 6586.6 6812.1 -3240.3 6480.6 467
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.3921 -0.5539 0.0589 0.6613 3.7824
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 1742.492 41.743
## Preceding.Ve 1753.237 41.872 -1.00
## Preceding.Vo 606.294 24.623 -0.41 0.41
## Preceding.Vu 526.891 22.954 -0.54 0.54 0.95
## vowela 563.815 23.745 -0.35 0.35 0.98 0.98
## VHY 2.208 1.486 -0.73 0.73 0.03 0.00 -0.14
## Preceding.Ve:vowela 13733.114 117.188 0.05 -0.05 -0.22 -0.21 -0.23
## Preceding.Vo:vowela 1795.392 42.372 0.95 -0.95 -0.53 -0.56 -0.40
## Preceding.Vu:vowela 984.670 31.379 -0.10 0.10 -0.86 -0.76 -0.88
## Residual 14766.527 121.518
##
##
##
##
##
##
##
## 0.05
## -0.82 0.07
## 0.40 0.21 0.04
##
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5539.690 37.370 5.997 148.237 6.43e-12 ***
## Preceding.Ve -10.527 36.742 5.352 -0.287 0.7853
## Preceding.Vo -12.385 29.077 30.476 -0.426 0.6731
## Preceding.Vu -21.338 24.056 12.899 -0.887 0.3913
## vowela -43.400 20.926 16.234 -2.074 0.0543 .
## VHY -15.669 13.292 331.846 -1.179 0.2393
## Preceding.Vo:vowela 22.020 35.664 6.175 0.617 0.5590
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) Preceding.Ve Preceding.Vo Preceding.Vu vowela VHY
## Preceding.Ve -0.877
## Preceding.Vo -0.803 0.713
## Preceding.Vu -0.854 0.739 0.791
## vowela -0.734 0.724 0.733 0.616
## VHY -0.414 0.060 0.296 0.371 0.074
## Prcdng.V:vw 0.733 -0.715 -0.689 -0.518 -0.702 -0.117
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')formant.all <- na.omit(formant.all)
vowel_matrix <-model.matrix(~Preceding.V, formant.all)
formant.all$v12 <- vowel_matrix[,2]
formant.all$v13 <- vowel_matrix[,3]
formant.all$v14 <- vowel_matrix[,4]
formant.all <- formant.all %>% mutate(
VH_Y = rescale(VH),
vowel_a = rescale(vowel),
)
F1.VT.all.rand <- lmer (F1 ~ (v12+v13+v14)*VH_Y+vowel_a+(1+(v12+v13+v14)+VH_Y+vowel_a|speaker), data = formant.all)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients## boundary (singular) fit: see help('isSingular')F2.VT.all.rand <- lmer (F2 ~ (v12+v13+v14)*VH_Y+vowel_a+(1+(v12+v13+v14)+VH_Y+vowel_a|speaker), data = formant.all, REML=FALSE)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## boundary (singular) fit: see help('isSingular')F3.VT.all.rand <- lmer (F3 ~ (v12+v13+v14)*VH_Y+vowel_a+(1+(v12+v13+v14)+VH_Y+vowel_a|speaker), data = formant.all, REML=FALSE)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## boundary (singular) fit: see help('isSingular')summary(F1.VT.all.rand)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ (v12 + v13 + v14) * VH_Y + vowel_a + (1 + (v12 + v13 + v14) +
## VH_Y + vowel_a | speaker)
## Data: formant.all
##
## REML criterion at convergence: 5476.9
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.6482 -0.6101 0.0544 0.6362 4.7549
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 55.82 7.472
## v12 196.50 14.018 -0.50
## v13 136.20 11.671 -0.18 0.94
## v14 334.11 18.279 -0.93 0.78 0.53
## VH_Y 116.75 10.805 -0.04 -0.85 -0.98 -0.33
## vowel_a 562.22 23.711 0.97 -0.27 0.06 -0.81 -0.28
## Residual 2325.59 48.224
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 450.211 5.421 5.241 83.052 2.22e-09 ***
## v12 15.034 11.008 5.719 1.366 0.22332
## v13 13.222 9.596 6.400 1.378 0.21447
## v14 8.696 10.097 4.190 0.861 0.43560
## VH_Y -2.489 9.215 9.898 -0.270 0.79260
## vowel_a 76.371 11.991 3.553 6.369 0.00462 **
## v14:VH_Y -2.778 10.692 448.430 -0.260 0.79514
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) v12 v13 v14 VH_Y vowel_
## v12 -0.595
## v13 -0.493 0.588
## v14 -0.808 0.605 0.469
## VH_Y 0.278 -0.645 -0.679 -0.309
## vowel_a 0.575 -0.298 0.102 -0.629 -0.089
## v14:VH_Y -0.251 0.346 0.330 0.188 -0.622 -0.057
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(F2.VT.all.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F2 ~ (v12 + v13 + v14) * VH_Y + vowel_a + (1 + (v12 + v13 + v14) +
## VH_Y + vowel_a | speaker)
## Data: formant.all
##
## AIC BIC logLik deviance df.resid
## 6919.8 7043.1 -3430.9 6861.8 491
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -2.89029 -0.74748 0.02488 0.70867 2.38639
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 2508.7 50.09
## v12 1879.0 43.35 -0.86
## v13 352.3 18.77 0.59 -0.09
## v14 1807.2 42.51 -1.00 0.84 -0.61
## VH_Y 359.7 18.97 1.00 -0.88 0.56 -1.00
## vowel_a 2010.7 44.84 0.24 -0.70 -0.65 -0.20 0.27
## Residual 30709.4 175.24
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 1785.512 27.562 5.656 64.783 2.44e-09 ***
## v12 83.532 37.761 10.839 2.212 0.04938 *
## v13 32.273 30.326 35.139 1.064 0.29449
## v14 37.701 28.186 7.623 1.338 0.21956
## VH_Y -63.876 29.771 59.732 -2.146 0.03598 *
## vowel_a -141.951 28.428 4.094 -4.993 0.00708 **
## v14:VH_Y 2.044 38.451 502.054 0.053 0.95762
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) v12 v13 v14 VH_Y vowel_
## v12 -0.683
## v13 -0.221 0.329
## v14 -0.878 0.612 0.229
## VH_Y 0.479 -0.587 -0.464 -0.434
## vowel_a 0.175 -0.509 0.010 -0.146 0.116
## v14:VH_Y -0.187 0.365 0.380 0.266 -0.709 -0.074
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(F3.VT.all.rand)## Linear mixed model fit by maximum likelihood . t-tests use Satterthwaite's
## method [lmerModLmerTest]
## Formula: F3 ~ (v12 + v13 + v14) * VH_Y + vowel_a + (1 + (v12 + v13 + v14) +
## VH_Y + vowel_a | speaker)
## Data: formant.all
##
## AIC BIC logLik deviance df.resid
## 6542.0 6665.3 -3242.0 6484.0 491
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.2838 -0.5253 0.0558 0.6655 3.9194
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 8.588e+02 29.3050
## v12 1.575e-02 0.1255 1.00
## v13 5.758e+02 23.9960 -1.00 -1.00
## v14 6.835e+02 26.1443 1.00 1.00 -1.00
## VH_Y 6.694e-01 0.8182 1.00 1.00 -1.00 1.00
## vowel_a 2.754e+02 16.5951 1.00 1.00 -1.00 1.00 1.00
## Residual 1.488e+04 121.9984
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5490.253 16.964 4.417 323.635 7.83e-11 ***
## v12 17.591 22.222 506.341 0.792 0.4290
## v13 13.040 22.991 9.202 0.567 0.5842
## v14 18.061 18.708 7.237 0.965 0.3655
## VH_Y -16.798 19.622 485.884 -0.856 0.3924
## vowel_a -37.112 15.444 11.058 -2.403 0.0349 *
## v14:VH_Y 7.754 26.664 512.327 0.291 0.7713
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) v12 v13 v14 VH_Y vowel_
## v12 -0.399
## v13 -0.730 0.356
## v14 0.129 0.370 0.031
## VH_Y 0.286 -0.545 -0.498 -0.232
## vowel_a 0.412 -0.356 -0.045 0.259 0.070
## v14:VH_Y -0.201 0.426 0.346 0.259 -0.739 -0.098
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')F1.VT.all.rand <- lmer (F1 ~ (v12+v13+v14)*VH_Y+vowel_a+(1+VH_Y|speaker), data = formant.all)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## boundary (singular) fit: see help('isSingular')F2.VT.all.rand <- lmer (F2 ~ (v12+v13+v14)*VH_Y+vowel_a+(1+VH_Y|speaker), data = formant.all)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## boundary (singular) fit: see help('isSingular')F3.VT.all.rand <- lmer (F3 ~ (v12+v13+v14)*VH_Y+vowel_a+(1+VH_Y|speaker), data = formant.all)## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## boundary (singular) fit: see help('isSingular')summary(F1.VT.all.rand)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F1 ~ (v12 + v13 + v14) * VH_Y + vowel_a + (1 + VH_Y | speaker)
## Data: formant.all
##
## REML criterion at convergence: 5494.2
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -3.4689 -0.5562 0.0389 0.6148 5.1253
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 29.50 5.432
## VH_Y 23.16 4.813 -1.00
## Residual 2452.38 49.522
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 451.417 5.003 15.713 90.228 <2e-16 ***
## v12 12.855 9.003 510.287 1.428 0.1539
## v13 14.172 8.201 509.763 1.728 0.0846 .
## v14 7.538 5.786 504.303 1.303 0.1932
## VH_Y -2.753 8.256 49.593 -0.333 0.7402
## vowel_a 80.570 5.479 499.582 14.704 <2e-16 ***
## v14:VH_Y -3.129 10.839 498.294 -0.289 0.7729
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) v12 v13 v14 VH_Y vowel_
## v12 -0.553
## v13 -0.569 0.413
## v14 -0.649 0.492 0.483
## VH_Y 0.224 -0.530 -0.532 -0.313
## vowel_a 0.055 -0.404 0.225 -0.082 0.075
## v14:VH_Y -0.276 0.424 0.394 0.342 -0.710 -0.108
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')summary(F3.VT.all.rand)## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: F3 ~ (v12 + v13 + v14) * VH_Y + vowel_a + (1 + VH_Y | speaker)
## Data: formant.all
##
## REML criterion at convergence: 6444.1
##
## Scaled residuals:
## Min 1Q Median 3Q Max
## -5.3438 -0.5482 0.0770 0.6615 3.9644
##
## Random effects:
## Groups Name Variance Std.Dev. Corr
## speaker (Intercept) 1889.35 43.47
## VH_Y 91.58 9.57 -1.00
## Residual 15422.14 124.19
## Number of obs: 520, groups: speaker, 5
##
## Fixed effects:
## Estimate Std. Error df t value Pr(>|t|)
## (Intercept) 5487.900 22.385 5.734 245.156 9.47e-13 ***
## v12 22.041 22.581 509.163 0.976 0.32949
## v13 11.165 20.590 509.342 0.542 0.58789
## v14 16.089 14.630 511.390 1.100 0.27196
## VH_Y -19.710 20.433 98.912 -0.965 0.33709
## vowel_a -39.045 13.801 508.247 -2.829 0.00485 **
## v14:VH_Y 10.377 27.133 508.473 0.382 0.70230
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Correlation of Fixed Effects:
## (Intr) v12 v13 v14 VH_Y vowel_
## v12 -0.309
## v13 -0.321 0.411
## v14 -0.364 0.487 0.482
## VH_Y 0.022 -0.535 -0.542 -0.318
## vowel_a 0.025 -0.403 0.228 -0.073 0.067
## v14:VH_Y -0.156 0.425 0.395 0.344 -0.722 -0.106
## fit warnings:
## fixed-effect model matrix is rank deficient so dropping 2 columns / coefficients
## optimizer (nloptwrap) convergence code: 0 (OK)
## boundary (singular) fit: see help('isSingular')F1.VT.all.rand <- lm (F1 ~ (v12+v13+v14)*VH_Y+vowel_a, data = formant.all)
F2.VT.all.rand <- lm (F2 ~ (v12+v13+v14)*VH_Y+vowel_a, data = formant.all)
F3.VT.all.rand <- lm (F3 ~ (v12+v13+v14)*VH_Y+vowel_a, data = formant.all)
summary(F1.VT.all.rand)##
## Call:
## lm(formula = F1 ~ (v12 + v13 + v14) * VH_Y + vowel_a, data = formant.all)
##
## Residuals:
## Min 1Q Median 3Q Max
## -173.562 -29.543 2.699 30.647 250.654
##
## Coefficients: (2 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 452.241 4.345 104.081 <2e-16 ***
## v12 13.234 9.029 1.466 0.1433
## v13 13.979 8.226 1.699 0.0898 .
## v14 8.545 5.760 1.484 0.1385
## VH_Y -3.669 7.993 -0.459 0.6464
## vowel_a 79.974 5.429 14.732 <2e-16 ***
## v12:VH_Y NA NA NA NA
## v13:VH_Y NA NA NA NA
## v14:VH_Y -3.953 10.854 -0.364 0.7158
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.74 on 513 degrees of freedom
## Multiple R-squared: 0.3976, Adjusted R-squared: 0.3905
## F-statistic: 56.43 on 6 and 513 DF, p-value: < 2.2e-16summary(F2.VT.all.rand)##
## Call:
## lm(formula = F2 ~ (v12 + v13 + v14) * VH_Y + vowel_a, data = formant.all)
##
## Residuals:
## Min 1Q Median 3Q Max
## -499.08 -131.29 6.55 137.47 380.93
##
## Coefficients: (2 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1765.45 15.93 110.803 < 2e-16 ***
## v12 95.54 33.11 2.886 0.00407 **
## v13 37.88 30.16 1.256 0.20977
## v14 58.29 21.12 2.760 0.00599 **
## VH_Y -82.09 29.31 -2.801 0.00529 **
## vowel_a -134.47 19.91 -6.755 3.88e-11 ***
## v12:VH_Y NA NA NA NA
## v13:VH_Y NA NA NA NA
## v14:VH_Y 16.28 39.80 0.409 0.68259
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 182.4 on 513 degrees of freedom
## Multiple R-squared: 0.1237, Adjusted R-squared: 0.1135
## F-statistic: 12.07 on 6 and 513 DF, p-value: 1.04e-12summary(F3.VT.all.rand)##
## Call:
## lm(formula = F3 ~ (v12 + v13 + v14) * VH_Y + vowel_a, data = formant.all)
##
## Residuals:
## Min 1Q Median 3Q Max
## -725.04 -75.02 10.65 88.19 517.07
##
## Coefficients: (2 not defined because of singularities)
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5492.831 11.290 486.533 < 2e-16 ***
## v12 23.049 23.460 0.982 0.32632
## v13 5.221 21.373 0.244 0.80710
## v14 2.138 14.965 0.143 0.88645
## VH_Y -13.214 20.768 -0.636 0.52488
## vowel_a -45.677 14.105 -3.238 0.00128 **
## v12:VH_Y NA NA NA NA
## v13:VH_Y NA NA NA NA
## v14:VH_Y 9.646 28.201 0.342 0.73245
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 129.2 on 513 degrees of freedom
## Multiple R-squared: 0.02471, Adjusted R-squared: 0.0133
## F-statistic: 2.166 on 6 and 513 DF, p-value: 0.04491formant_speaker1 <- filter(formant.ae, speaker == "Speaker 1")
formant_speaker2 <- filter(formant.ae, speaker == "Speaker 2")
formant_speaker3 <- filter(formant.ae, speaker == "Speaker 3")
formant_speaker4 <- filter(formant.ae, speaker == "Speaker 4")
formant_speaker5 <- filter(formant.ae, speaker == "Speaker 5")
##first speaker
F1.VT.ou.rand <- lm (F1 ~ Preceding.V*vowel+VH, data = formant_speaker1)
F2.VT.ou.rand <- lm (F2 ~ Preceding.V*vowel+VH, data = formant_speaker1)
F3.VT.ou.rand <- lm (F3 ~ Preceding.V*vowel+VH, data = formant_speaker1)
summary(F1.VT.ou.rand)##
## Call:
## lm(formula = F1 ~ Preceding.V * vowel + VH, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -177.017 -51.226 -6.375 45.661 228.216
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 556.674 31.883 17.460 < 2e-16 ***
## Preceding.Vo -53.908 39.505 -1.365 0.177089
## Preceding.Vu -34.890 37.681 -0.926 0.357900
## vowele -139.917 38.108 -3.672 0.000489 ***
## VHY -3.361 22.689 -0.148 0.882681
## Preceding.Vo:vowele 20.385 48.669 0.419 0.676713
## Preceding.Vu:vowele 36.281 52.172 0.695 0.489279
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 78.1 on 65 degrees of freedom
## Multiple R-squared: 0.3921, Adjusted R-squared: 0.336
## F-statistic: 6.987 on 6 and 65 DF, p-value: 9.195e-06summary(F2.VT.ou.rand)##
## Call:
## lm(formula = F2 ~ Preceding.V * vowel + VH, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -402.96 -103.41 11.81 119.70 308.03
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1782.836 73.042 24.408 <2e-16 ***
## Preceding.Vo 79.532 90.503 0.879 0.383
## Preceding.Vu -6.248 86.324 -0.072 0.943
## vowele 82.452 87.302 0.944 0.348
## VHY -44.822 51.979 -0.862 0.392
## Preceding.Vo:vowele -51.655 111.498 -0.463 0.645
## Preceding.Vu:vowele -31.465 119.522 -0.263 0.793
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 178.9 on 65 degrees of freedom
## Multiple R-squared: 0.07211, Adjusted R-squared: -0.01354
## F-statistic: 0.8419 on 6 and 65 DF, p-value: 0.5423summary(F3.VT.ou.rand)##
## Call:
## lm(formula = F3 ~ Preceding.V * vowel + VH, data = formant_speaker1)
##
## Residuals:
## Min 1Q Median 3Q Max
## -351.17 -96.96 -1.67 89.13 420.63
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5517.476 58.165 94.859 <2e-16 ***
## Preceding.Vo 50.660 72.069 0.703 0.485
## Preceding.Vu -15.550 68.742 -0.226 0.822
## vowele 12.713 69.520 0.183 0.855
## VHY -39.100 41.392 -0.945 0.348
## Preceding.Vo:vowele -83.113 88.788 -0.936 0.353
## Preceding.Vu:vowele -7.273 95.178 -0.076 0.939
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 142.5 on 65 degrees of freedom
## Multiple R-squared: 0.04591, Adjusted R-squared: -0.04216
## F-statistic: 0.5213 on 6 and 65 DF, p-value: 0.7901##second speaker
F1.VT.ou.rand <- lm (F1 ~ Preceding.V*vowel+VH, data = formant_speaker2)
F2.VT.ou.rand <- lm (F2 ~ Preceding.V*vowel+VH, data = formant_speaker2)
F3.VT.ou.rand <- lm (F3 ~ Preceding.V*vowel+VH, data = formant_speaker2)
summary(F1.VT.ou.rand)##
## Call:
## lm(formula = F1 ~ Preceding.V * vowel + VH, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -71.739 -21.254 0.998 22.992 115.417
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 527.1072 7.8747 66.937 < 2e-16 ***
## Preceding.Vo 0.7437 10.7834 0.069 0.94515
## Preceding.Vu -40.3469 13.4032 -3.010 0.00330 **
## vowele -99.9161 11.4793 -8.704 6.79e-14 ***
## VHY -22.5200 8.1753 -2.755 0.00698 **
## Preceding.Vo:vowele 16.3270 14.7727 1.105 0.27172
## Preceding.Vu:vowele 40.9438 18.5957 2.202 0.02998 *
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 33.41 on 100 degrees of freedom
## Multiple R-squared: 0.6588, Adjusted R-squared: 0.6383
## F-statistic: 32.18 on 6 and 100 DF, p-value: < 2.2e-16summary(F2.VT.ou.rand)##
## Call:
## lm(formula = F2 ~ Preceding.V * vowel + VH, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -377.68 -78.72 2.72 114.85 325.13
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1815.85 37.54 48.369 < 2e-16 ***
## Preceding.Vo -39.02 51.41 -0.759 0.449632
## Preceding.Vu 38.69 63.90 0.605 0.546237
## vowele 185.88 54.73 3.397 0.000981 ***
## VHY -51.33 38.97 -1.317 0.190847
## Preceding.Vo:vowele -32.59 70.43 -0.463 0.644583
## Preceding.Vu:vowele -104.08 88.65 -1.174 0.243178
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 159.3 on 100 degrees of freedom
## Multiple R-squared: 0.2448, Adjusted R-squared: 0.1994
## F-statistic: 5.401 on 6 and 100 DF, p-value: 7.187e-05summary(F3.VT.ou.rand)##
## Call:
## lm(formula = F3 ~ Preceding.V * vowel + VH, data = formant_speaker2)
##
## Residuals:
## Min 1Q Median 3Q Max
## -652.50 -53.66 11.37 79.20 290.61
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5374.042 33.258 161.586 < 2e-16 ***
## Preceding.Vo -17.377 45.543 -0.382 0.70360
## Preceding.Vu 60.736 56.607 1.073 0.28589
## vowele 145.358 48.482 2.998 0.00343 **
## VHY -8.991 34.528 -0.260 0.79509
## Preceding.Vo:vowele -51.770 62.391 -0.830 0.40865
## Preceding.Vu:vowele -129.861 78.537 -1.654 0.10136
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 141.1 on 100 degrees of freedom
## Multiple R-squared: 0.15, Adjusted R-squared: 0.09903
## F-statistic: 2.942 on 6 and 100 DF, p-value: 0.01099##third speaker
F1.VT.ou.rand <- lm (F1 ~ Preceding.V*vowel+VH, data = formant_speaker3)
F2.VT.ou.rand <- lm (F2 ~ Preceding.V*vowel+VH, data = formant_speaker3)
F3.VT.ou.rand <- lm (F3 ~ Preceding.V*vowel+VH, data = formant_speaker3)
summary(F1.VT.ou.rand)##
## Call:
## lm(formula = F1 ~ Preceding.V * vowel + VH, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -98.437 -30.926 -1.058 22.915 164.684
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 485.72 12.25 39.661 < 2e-16 ***
## Preceding.Vo -17.40 23.12 -0.753 0.454892
## Preceding.Vu -23.48 24.17 -0.972 0.335333
## vowele -74.95 19.04 -3.937 0.000231 ***
## VHY 22.48 18.10 1.242 0.219299
## Preceding.Vo:vowele 33.61 32.74 1.027 0.308999
## Preceding.Vu:vowele 17.92 32.07 0.559 0.578625
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 50.49 on 56 degrees of freedom
## Multiple R-squared: 0.3255, Adjusted R-squared: 0.2532
## F-statistic: 4.504 on 6 and 56 DF, p-value: 0.0008648summary(F2.VT.ou.rand)##
## Call:
## lm(formula = F2 ~ Preceding.V * vowel + VH, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -514.35 -70.34 11.59 106.82 244.90
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1796.73 37.01 48.554 <2e-16 ***
## Preceding.Vo -68.66 69.85 -0.983 0.330
## Preceding.Vu -44.91 73.02 -0.615 0.541
## vowele 93.96 57.53 1.633 0.108
## VHY -18.08 54.68 -0.331 0.742
## Preceding.Vo:vowele 18.15 98.92 0.184 0.855
## Preceding.Vu:vowele 70.41 96.92 0.727 0.471
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 152.6 on 56 degrees of freedom
## Multiple R-squared: 0.1831, Adjusted R-squared: 0.09559
## F-statistic: 2.092 on 6 and 56 DF, p-value: 0.06854summary(F3.VT.ou.rand)##
## Call:
## lm(formula = F3 ~ Preceding.V * vowel + VH, data = formant_speaker3)
##
## Residuals:
## Min 1Q Median 3Q Max
## -377.20 -35.02 13.10 48.53 101.01
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5504.07 20.45 269.104 <2e-16 ***
## Preceding.Vo 42.16 38.61 1.092 0.2794
## Preceding.Vu 12.64 40.36 0.313 0.7553
## vowele 62.53 31.80 1.967 0.0542 .
## VHY 25.91 30.22 0.857 0.3950
## Preceding.Vo:vowele -62.09 54.67 -1.136 0.2609
## Preceding.Vu:vowele -26.75 53.57 -0.499 0.6195
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 84.33 on 56 degrees of freedom
## Multiple R-squared: 0.1157, Adjusted R-squared: 0.021
## F-statistic: 1.222 on 6 and 56 DF, p-value: 0.3091##fourth speaker
F1.VT.ou.rand <- lm (F1 ~ Preceding.V*vowel+VH, data = formant_speaker4)
F2.VT.ou.rand <- lm (F2 ~ Preceding.V*vowel+VH, data = formant_speaker4)
F3.VT.ou.rand <- lm (F3 ~ Preceding.V*vowel+VH, data = formant_speaker4)
summary(F1.VT.ou.rand)##
## Call:
## lm(formula = F1 ~ Preceding.V * vowel + VH, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -168.312 -28.369 5.452 32.076 110.750
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 518.7201 6.8350 75.892 < 2e-16 ***
## Preceding.Vo 6.0589 12.6810 0.478 0.633
## Preceding.Vu -17.8136 14.2444 -1.251 0.212
## vowele -76.1768 10.0271 -7.597 7.77e-13 ***
## VHY -8.7243 8.6330 -1.011 0.313
## Preceding.Vo:vowele -21.0456 15.5384 -1.354 0.177
## Preceding.Vu:vowele 0.6214 17.2400 0.036 0.971
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 49.76 on 228 degrees of freedom
## Multiple R-squared: 0.4389, Adjusted R-squared: 0.4241
## F-statistic: 29.72 on 6 and 228 DF, p-value: < 2.2e-16summary(F2.VT.ou.rand)##
## Call:
## lm(formula = F2 ~ Preceding.V * vowel + VH, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -624.70 -163.72 29.82 160.69 407.41
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1777.2529 27.6725 64.225 < 2e-16 ***
## Preceding.Vo 0.4745 51.3412 0.009 0.99263
## Preceding.Vu -70.7122 57.6709 -1.226 0.22141
## vowele 53.8206 40.5963 1.326 0.18625
## VHY -96.3492 34.9522 -2.757 0.00631 **
## Preceding.Vo:vowele 102.8125 62.9096 1.634 0.10358
## Preceding.Vu:vowele 66.8443 69.7991 0.958 0.33925
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 201.5 on 228 degrees of freedom
## Multiple R-squared: 0.1276, Adjusted R-squared: 0.1046
## F-statistic: 5.557 on 6 and 228 DF, p-value: 2.135e-05summary(F3.VT.ou.rand)##
## Call:
## lm(formula = F3 ~ Preceding.V * vowel + VH, data = formant_speaker4)
##
## Residuals:
## Min 1Q Median 3Q Max
## -549.84 -86.48 12.51 81.38 334.27
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5442.30 17.85 304.865 < 2e-16 ***
## Preceding.Vo 77.05 33.12 2.326 0.0209 *
## Preceding.Vu 13.34 37.20 0.358 0.7203
## vowele 134.46 26.19 5.134 6.07e-07 ***
## VHY -17.28 22.55 -0.766 0.4443
## Preceding.Vo:vowele -97.89 40.58 -2.412 0.0167 *
## Preceding.Vu:vowele -65.55 45.03 -1.456 0.1468
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 130 on 228 degrees of freedom
## Multiple R-squared: 0.1349, Adjusted R-squared: 0.1121
## F-statistic: 5.925 on 6 and 228 DF, p-value: 9.083e-06##fifth speaker
F1.VT.ou.rand <- lm (F1 ~ Preceding.V*vowel+VH, data = formant_speaker5)
F2.VT.ou.rand <- lm (F2 ~ Preceding.V*vowel+VH, data = formant_speaker5)
F3.VT.ou.rand <- lm (F3 ~ Preceding.V*vowel+VH, data = formant_speaker5)
summary(F1.VT.ou.rand)##
## Call:
## lm(formula = F1 ~ Preceding.V * vowel + VH, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -101.959 -33.933 1.828 26.907 100.239
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 535.582 8.662 61.832 < 2e-16 ***
## Preceding.Vo -9.296 29.302 -0.317 0.75170
## Preceding.Vu -63.399 15.307 -4.142 7.16e-05 ***
## vowele -94.245 11.728 -8.036 1.80e-12 ***
## VHY -13.863 12.583 -1.102 0.27321
## Preceding.Vo:vowele 8.244 30.504 0.270 0.78750
## Preceding.Vu:vowele 55.104 19.127 2.881 0.00484 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 43.31 on 101 degrees of freedom
## Multiple R-squared: 0.5148, Adjusted R-squared: 0.486
## F-statistic: 17.86 on 6 and 101 DF, p-value: 5.13e-14summary(F2.VT.ou.rand)##
## Call:
## lm(formula = F2 ~ Preceding.V * vowel + VH, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -375.25 -121.94 -6.59 116.43 373.61
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 1737.587 33.571 51.759 <2e-16 ***
## Preceding.Vo -2.028 113.566 -0.018 0.9858
## Preceding.Vu -73.373 59.325 -1.237 0.2190
## vowele 115.515 45.455 2.541 0.0126 *
## VHY -76.923 48.769 -1.577 0.1179
## Preceding.Vo:vowele 146.725 118.225 1.241 0.2175
## Preceding.Vu:vowele 119.788 74.131 1.616 0.1092
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 167.9 on 101 degrees of freedom
## Multiple R-squared: 0.2634, Adjusted R-squared: 0.2197
## F-statistic: 6.021 on 6 and 101 DF, p-value: 2.056e-05summary(F3.VT.ou.rand)##
## Call:
## lm(formula = F3 ~ Preceding.V * vowel + VH, data = formant_speaker5)
##
## Residuals:
## Min 1Q Median 3Q Max
## -314.376 -50.227 -0.513 51.399 250.645
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 5477.02 18.42 297.405 <2e-16 ***
## Preceding.Vo 75.98 62.30 1.220 0.225
## Preceding.Vu 10.25 32.54 0.315 0.753
## vowele 27.66 24.94 1.109 0.270
## VHY -21.24 26.75 -0.794 0.429
## Preceding.Vo:vowele -38.93 64.85 -0.600 0.550
## Preceding.Vu:vowele 23.82 40.67 0.586 0.559
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 92.08 on 101 degrees of freedom
## Multiple R-squared: 0.06012, Adjusted R-squared: 0.004284
## F-statistic: 1.077 on 6 and 101 DF, p-value: 0.3815formant.ae <- formant.ae %>% mutate(
VH_Y = rescale(VH),
)
m3 = lmer (F2 ~ Preceding.V*vowel+VH_Y+(1+Preceding.V*vowel+VH_Y|speaker), data = formant.ae, REML=FALSE)## boundary (singular) fit: see help('isSingular')ggpredict(m3, terms = c("VH_Y", "speaker"), type = "re",interactive=TRUE, ci.lvl = NA) %>%
plot()+ labs(x = "Vowel harmony", y = "F2", title = "Predicted values of F2") + scale_x_continuous(breaks=c(-0.2752137,0.7247863), labels=c("NO", "YES"))+
theme_minimal()## Warning: Argument `ci.lvl` is deprecated and will be removed in the future.
## Please use `ci_level` instead.
ggpredict(m3, terms = c("VH_Y", "speaker"), type = "re",interactive=TRUE) %>%
plot()+ labs(x = "Vowel harmony", y = "F2") + scale_x_continuous(breaks=c(-0.2752137,0.7247863), labels=c("NO", "YES"))+
theme_minimal()