# Libraries
pacman::p_load(knitr, # rendering of .Rmd file
lme4, # model specification/estimation
afex, # anova and deriving p-values via LRTs from lmer
broom.mixed, # extracting data from model fits
faux, # generate correlated values
tidyverse, # data wrangling and visualization
ggridges, # visualization
viridis, # color schemes for visualization
kableExtra, # helps with knitting the .Rmd file
cowplot, # visualization tool to include multiple plots
ggrain, # visualization tool for raincloud plots
dplyr, # data manipulation tools
here, # controlling directories/reproducibility
tidyr, # formatting package
readr, # data importation
Hmisc, # visualization
emmeans, # estimation of marginal means and contrasts
sjstats, # estimation of marginal means and contrasts
lmerTest, # deriving p-values from lmer
MuMIn, # ?
sjPlot, # APA formatting of numerical output
car, # data wrangling and visualization,
magrittr, # to use %>%
ggpubr, # ? (Rose had it in her script)
broom, # ? (Rose had it in her script)
Matrix, # model specification/estimation,
multcomp, # multiple comparison adjustments
forcats, # organizing visualizations
gtools, # data manipulation/wrangling
fuzzySim, # FDR correction,
ez, # anova
apaTables, # visualization of anova output
#renv # package versioning/reproducibility
)
##
## Your package installed
## Warning in pacman::p_load(knitr, lme4, afex, broom.mixed, faux, tidyverse, : Failed to install/load:
# Global theme set
set.seed(1234) # set random seed for reproducibility
knitr::opts_chunk$set(echo = TRUE, collapse = TRUE)
knitr::opts_chunk$set(tidy.opts=list(width.cutoff=60),tidy=TRUE, fig.width=10, fig.height=8, fig.fullwidth=TRUE)
plot_theme <-
theme_bw() +
theme(plot.title = element_text(hjust = 0.5, size = 18),
axis.text.x = element_text(size = 13),
axis.title.x = element_text(size = 13),
axis.text.y = element_text(size = 12),
axis.title.y = element_text(size = 13),
strip.background = element_rect(color="white", fill="white", linewidth = 1.5, linetype="solid"))
options(scipen=999)
Data Manipulations
# Load data
data <- read.csv(here("2025 S3 Data.csv"), stringsAsFactors=TRUE)
# Format data
data %<>%
mutate(Participant.ID = factor(Participant.ID),
lang = factor(Language),
lang.c = as.numeric(Language)-1.5,
social.percep = as.numeric(Social.Perception),
celf.c = as.numeric(scale(CELF.ENG, center = TRUE, scale = FALSE)),
lang.dev = factor(Lang.Dev.Status),
lang.dev.c = as.numeric(Lang.Dev.Status) - 1.5,
age.c = as.numeric(scale(Age, center = TRUE, scale = FALSE)),
child.ethnicity = factor(Ethnicity),
child.ethnicity.c = as.numeric(Ethnicity) - 1.5,
s.w = factor(Strength.Weakness),
#child.race = factor (child.race),
#free.red.lunch.c = as.numeric(free.or.reduced.lunch) - 1.5,
gender.c = as.numeric(Gender) - 1.5)
str(data)
## 'data.frame': 19 obs. of 24 variables:
## $ Participant.ID : Factor w/ 19 levels "030422_1M_11",..: 1 2 3 5 4 6 8 13 14 16 ...
## $ Gender : Factor w/ 2 levels "Female","Male ": 2 1 1 1 1 2 2 1 1 1 ...
## $ DOB : Factor w/ 19 levels "1/20/14","10/21/11",..: 6 13 17 12 9 3 15 18 16 5 ...
## $ DOT1 : Factor w/ 18 levels "10/3/23","3/10/22",..: 6 6 2 4 3 5 8 17 13 15 ...
## $ Age : num 11.34 9.71 10.59 10.74 11.84 ...
## $ Language : Factor w/ 2 levels "ENG","ENG/SPAN": 1 1 1 1 1 1 2 1 2 1 ...
## $ Social.Perception: int 1 1 0 1 1 0 0 0 0 0 ...
## $ Strength.Weakness: Factor w/ 2 levels "Strength","Weakness": 1 1 2 1 1 2 2 2 2 2 ...
## $ Statement : Factor w/ 19 levels "Because is hard to ask people if I could be their friend.",..: 16 19 18 14 9 15 12 6 5 7 ...
## $ CELF.ENG : int 108 121 123 123 111 115 132 111 99 123 ...
## $ CELF.SPAN : int NA NA NA NA NA NA 102 NA 75 NA ...
## $ Lang.Dev.Status : Factor w/ 2 levels "DLD","TD": 2 2 2 2 2 2 2 2 2 2 ...
## $ Ethnicity : Factor w/ 3 levels "","Hispanic/Latino/Spanish",..: 3 3 3 1 3 3 2 3 2 2 ...
## $ lang : Factor w/ 2 levels "ENG","ENG/SPAN": 1 1 1 1 1 1 2 1 2 1 ...
## $ lang.c : num -0.5 -0.5 -0.5 -0.5 -0.5 -0.5 0.5 -0.5 0.5 -0.5 ...
## $ social.percep : num 1 1 0 1 1 0 0 0 0 0 ...
## $ celf.c : num 8.11 21.11 23.11 23.11 11.11 ...
## $ lang.dev : Factor w/ 2 levels "DLD","TD": 2 2 2 2 2 2 2 2 2 2 ...
## $ lang.dev.c : num 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 ...
## $ age.c : num 0.5844 -1.043 -0.1608 -0.0128 1.0913 ...
## $ child.ethnicity : Factor w/ 3 levels "","Hispanic/Latino/Spanish",..: 3 3 3 1 3 3 2 3 2 2 ...
## $ child.ethnicity.c: num 1.5 1.5 1.5 -0.5 1.5 1.5 0.5 1.5 0.5 0.5 ...
## $ s.w : Factor w/ 2 levels "Strength","Weakness": 1 1 2 1 1 2 2 2 2 2 ...
## $ gender.c : num 0.5 -0.5 -0.5 -0.5 -0.5 0.5 0.5 -0.5 -0.5 -0.5 ...
# Notes:
# For gender, female = -0.5 (after centering), male = 0.5 (after centering)
# For language, English = -0.5 (after centering), English/Spanish = 0.5 (after centering)
# For language development status, DLD = -0.5 (after centering), TD = 0.5 (after centering)
# For ethnicity, Hispanic/Latino/Spanish = -0.5 (after centering), Not Hispanic/Latino/Spanish = 0.5 (after centering)
# For social perception, weakness = 0, strength = 1
Participants
Demographics
# Age
data %>%
distinct(Participant.ID, Age)%>%
summarise_at('Age',
list(~mean(., na.rm=T),
~sd(., na.rm=T),
~median(., na.rm=T),
~min(., na.rm=T),
~max(., na.rm=T),
~sum(!is.na(.))))%>%
dplyr::rename("n" = "sum")%>%
dplyr::select(n, mean, sd, median, min, max)%>%
mutate(total.n = sum(n))%>%
knitr::kable()
| 19 |
10.75256 |
1.025912 |
10.7863 |
9.120548 |
12.7863 |
19 |
# Gender
data %>%
distinct(Participant.ID, Gender)%>%
group_by(Gender)%>%
summarise(n = n())%>%
mutate(total.Gender.n = sum(n))%>%
knitr::kable()
# Language
data %>%
distinct(Participant.ID, lang)%>%
group_by(lang)%>%
summarise(n = n())%>%
mutate(lang.n = sum(n))%>%
knitr::kable()
# Language Development Status
data %>%
distinct(Participant.ID, Lang.Dev.Status)%>%
group_by(Lang.Dev.Status)%>%
summarise(n = n())%>%
mutate(Lang.Dev.Status.n = sum(n))%>%
knitr::kable()
# # Race
# data %>%
# distinct(Participant.ID, child.race)%>%
# group_by(child.race)%>%
# summarise(n = n())%>%
# mutate(total.child.race.n = sum(n))%>%
# knitr::kable()
# Ethnicity
data %>%
distinct(Participant.ID, child.ethnicity)%>%
group_by(child.ethnicity)%>%
summarise(n = n())%>%
mutate(total.child.ethnicity.n = sum(n))%>%
knitr::kable()
|
1 |
19 |
| Hispanic/Latino/Spanish |
7 |
19 |
| Not Hispanic/Latino/Spanish |
11 |
19 |
Descriptives
# CELF-4 standard scores
data %>%
distinct(Participant.ID, CELF.ENG)%>%
summarise_at('CELF.ENG',
list(~mean(., na.rm=T),
~sd(., na.rm=T),
~median(., na.rm=T),
~min(., na.rm=T),
~max(., na.rm=T),
~sum(!is.na(.))))%>%
dplyr::rename("n" = "sum")%>%
dplyr::select(n, mean, sd, median, min, max)%>%
mutate(total.n = sum(n))%>%
knitr::kable()
| 19 |
99.89474 |
24.50373 |
108 |
46 |
132 |
19 |
Visualizations
ggplot(data, aes(x = factor(s.w), y = CELF.ENG)) +
geom_boxplot(color = "black", outlier.shape = NA, alpha = 0.3) + # Single black boxplot per category
geom_jitter(aes(color = factor(s.w), shape = factor(lang)), width = 0.2, alpha = 0.7) + # Keep color/shape mapping for points
scale_color_manual(values = c("Weakness" = "red", "Strength" = "blue")) + # Red for weakness, blue for strength
geom_hline(yintercept = 80, linetype = "dashed", color = "gray") + # Dashed gray line at y = 80
scale_shape_manual(values = c("ENG" = 16, "ENG/SPAN" = 17),
labels = c("ENG" = "Monolingual", "ENG/SPAN" = "Bilingual")) + # Shape and labels for language
scale_y_continuous(breaks = seq(floor(min(data$CELF.ENG)-6), ceiling(max(data$CELF.ENG)+6), by = 10)) + # Ensure labels at every gridline
labs(x = "Self-Perception of Ability to Make Friends", y = "CELF-4 Standard Score",
color = "Self-Perception", shape = "Language Group") +
theme_minimal()+
theme(
axis.title.x = element_text(size = 16, face = "bold", margin = margin(t = 10)), # Add space between x-axis title and labels
axis.title.y = element_text(size = 16, face = "bold"), # Make y-axis title bold and larger
axis.text = element_text(size = 14), # Increase font size of axis values
legend.text = element_text(size = 14), # Increase font size of legend text
legend.title = element_text(size = 16, face = "bold") # Make legend title bold and larger
)
#legend <- cowplot::get_legend(plot)
#grid.newpage()
#grid.draw(legend)
Analyses
data.m1 <- glm(social.percep ~ celf.c + gender.c + age.c + lang.c + lang.dev.c + child.ethnicity.c, data=data) #converges and all that jazz
tab_model(data.m1, show.se = TRUE,show.ci = FALSE, transform = "exp") #,show.stat = TRUE, show.df = TRUE)
|
|
social.percep
|
|
Predictors
|
Estimates
|
std. Error
|
p
|
|
(Intercept)
|
1.69
|
0.41
|
0.033
|
|
celf c
|
0.98
|
0.01
|
0.067
|
|
gender c
|
1.22
|
0.29
|
0.404
|
|
age c
|
0.89
|
0.10
|
0.302
|
|
lang c
|
0.61
|
0.19
|
0.119
|
|
lang dev c
|
2.55
|
1.43
|
0.095
|
|
child ethnicity c
|
0.75
|
0.16
|
0.178
|
|
Observations
|
19
|
|
R2
|
0.407
|
# data.m2 <- lm(celf.c ~ s.w + gender.c + age.c + lang.c + lang.dev.c + child.ethnicity.c, data=data) #converges and all that jazz
# tab_model(data.m2, show.se = TRUE,show.ci = FALSE, transform = "exp") #,show.stat = TRUE, show.df = TRUE)
# Notes:
# For gender, female = -0.5 (after centering), male = 0.5 (after centering)
# For language development status, DLD = -0.5 (after centering), TD = 0.5 (after centering)
# For ethnicity, Hispanic/Latino/Spanish = -0.5 (after centering), Not Hispanic/Latino/Spanish = 0.5 (after centering)
# For social perception, weakness = 0, strength = 1
Summary of the results:
There are no within-subject or within-item measures (only a single
summary data point from each participant for each measure), so a LMER is
not appropriate; a simple (logistic) linear regression is. When
controlling for the other variables (i.e., gender, age, language spoken,
language development status, ethnicity), CELF-4 Core Language Scores do
not significantly predict the likelihood a child will perceive
their social abilities as a strength. The relationship does trend toward
significance, suggesting that if the same pattern held with a larger
sample size, we might find that higher CELF-4 CLS scores actually
predict that a child is more likely to perceive their social abilities
as a weakness (not a strength). The value of 0.98 means means
that for each one-point increase in CELF-4 CLS, the odds of a child
perceiving their social abilities as a strength decrease by about 1.8%.
However, this is a relatively small effect. Thus, language ability alone
does not appear to be a strong predictor of how social abilities are
perceived.