Stekic et al. - Meta Analysis and Related Stuff
Alan Nielsen
October 30, 2017
First we’ll load in some previous data to make some nice graphs (and load in some packages we will need)
library(tidyverse)
library(ggthemes)
library(scales)
PrevData <- read.csv('F:/GitHub Repos/Stekic-et-al/Data/PrevWork.csv')
#Make a new graph theme
theme_alan <- function(base_size = 12 , base_family = "")
{
half_line <- base_size/2
colors <- ggthemes_data$few
gray <- colors$medium["gray"]
black <- colors$dark["black"]
theme(
line = element_line(colour = "black", size = 0.5, linetype = 1, lineend = "butt"),
rect = element_rect(fill = "white",
colour = "black", size = 0.5, linetype = 1),
text = element_text(family = base_family, face = "plain", colour = "black",
size = base_size, lineheight = 0.9, hjust = 0.5, vjust = 0.5,
angle = 0, margin = margin(), debug = FALSE),
axis.line = element_blank(),
axis.line.x = NULL,
axis.line.y = NULL,
axis.text = element_text(size = rel(0.8), colour = "grey30"),
axis.text.x = element_text(margin = margin(t = 0.8 * half_line/2), vjust = 1),
axis.text.x.top = element_text(margin = margin(b = 0.8 * half_line/2), vjust = 0),
axis.text.y = element_text(margin = margin(r = 0.8 * half_line/2), hjust = 1),
axis.text.y.right = element_text(margin = margin(l = 0.8 * half_line/2), hjust = 0),
axis.ticks = element_line(colour = "grey20"),
axis.ticks.length = unit(half_line/2, "pt"),
axis.title.x = element_text(margin = margin(t = half_line), vjust = 1),
axis.title.x.top = element_text(margin = margin(b = half_line), vjust = 0),
axis.title.y = element_text(angle = 90, margin = margin(r = half_line), vjust = 1),
axis.title.y.right = element_text(angle = -90, margin = margin(l = half_line), vjust = 0),
legend.background = element_rect(colour = NA),
legend.spacing = unit(0.4, "cm"),
legend.spacing.x = NULL,
legend.spacing.y = NULL,
legend.margin = margin(0.2, 0.2, 0.2, 0.2, "cm"),
legend.key = element_rect(fill = "white", colour = NA),
legend.key.size = unit(1.2, "lines"),
legend.key.height = NULL,
legend.key.width = NULL,
legend.text = element_text(size = rel(0.8)),
legend.text.align = NULL,
legend.title = element_text(hjust = 0),
legend.title.align = NULL,
legend.position = "right",
legend.direction = NULL,
legend.justification = "center",
legend.box = NULL,
legend.box.margin = margin(0, 0, 0, 0, "cm"),
legend.box.background = element_blank(),
legend.box.spacing = unit(0.4, "cm"),
panel.background = element_rect(fill = "white", colour = NA),
panel.border = element_rect(fill = NA, colour = "grey20"),
panel.grid.major = element_line(colour = "grey92"),
panel.grid.minor = element_line(colour = "grey92", size = 0.25),
panel.spacing = unit(half_line, "pt"),
panel.spacing.x = NULL,
panel.spacing.y = NULL,
panel.ontop = FALSE,
strip.background = element_rect(fill = "NA", colour = "NA"),
strip.text = element_text(colour = "grey10", size = rel(0.8)),
strip.text.x = element_text(margin = margin(t = half_line, b = half_line)),
strip.text.y = element_text(angle = 0, margin = margin(l = half_line, r = half_line)),
strip.placement = "inside",
strip.placement.x = NULL,
strip.placement.y = NULL,
strip.switch.pad.grid = unit(0.1, "cm"),
strip.switch.pad.wrap = unit(0.1, "cm"),
plot.background = element_rect(colour = "white"),
plot.title = element_text(size = rel(1.2), hjust = 0, vjust = 1, margin = margin(b = half_line * 1.2)),
plot.subtitle = element_text(size = rel(0.9), hjust = 0, vjust = 1, margin = margin(b = half_line * 0.9)),
plot.caption = element_text(size = rel(0.9), hjust = 1, vjust = 1, margin = margin(t = half_line * 0.9)),
plot.margin = margin(half_line, half_line, half_line, half_line),
complete = TRUE)
}Okay so the first sensible thing to do are some individual plots by experiment, just to make sure we understand those previous results:
L2007 <- subset(PrevData, Paper == "Lupyan et al. (2007)")
L2007Training <- subset(L2007, Train.vs..Test == "Training")
L2007Testing <- subset(L2007, Train.vs..Test == "Testing")
L2007Verification <- subset(L2007, Train.vs..Test == "Verification")
# Training - Performance
ggplot(data=L2007Training, aes(x=Label.Type, y=Performance, fill = SoundSym)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Label Type", y="Performance") +
facet_wrap(~Experiment) +
ggtitle("Lupyan 2007 - Training - Performance")
# Training - RT
ggplot(data=L2007Training, aes(x=Label.Type, y=RT, fill = SoundSym)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = RT - RT.SD,
ymax = RT + RT.SD,
width = 0.5)) +
labs(x="Label Type", y="Response Time") +
facet_wrap(~Experiment) +
ggtitle("Lupyan 2007 - Training - Response Time")## Warning: Removed 4 rows containing missing values (geom_bar).## Warning: Removed 4 rows containing missing values (geom_errorbar).
# Testing - Performance
ggplot(data=L2007Testing, aes(x=Label.Type, y=Performance, fill = SoundSym)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Label Type", y="Performance") +
facet_wrap(~Experiment) +
ggtitle("Lupyan 2007 - Testing - Performance")
# Testing - RT
ggplot(data=L2007Testing, aes(x=Label.Type, y=RT, fill = SoundSym)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = RT - RT.SD,
ymax = RT + RT.SD,
width = 0.5)) +
labs(x="Label Type", y="Response Time") +
facet_wrap(~Experiment) +
ggtitle("Lupyan 2007 - Testing - Response Time")
# Verification - Performance
ggplot(data=L2007Verification, aes(x=Presentation, y=Performance, fill = Label.Type)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Label Type", y="Performance") +
facet_wrap(~Experiment) +
ggtitle("Lupyan 2007 - Verification - Performance")
K2010 <- subset(PrevData, Paper == "Kovic et al. (2010)")
K2010Testing <- subset(K2010, Train.vs..Test == "Testing")
# Testing - Performance
ggplot(data=K2010Testing, aes(x=Congruency, y=Performance, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Label Type", y="Performance") +
facet_wrap(~Experiment) +
ggtitle("Kovic et al. 2010 - Testing - Performance")
# Testing - RT
ggplot(data=K2010Testing, aes(x=Congruency, y=RT, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = RT - RT.SD,
ymax = RT + RT.SD,
width = 0.5)) +
labs(x="Label Type", y="Response Time") +
facet_wrap(~Experiment) +
ggtitle("Kovic et al. 2010 - Testing - Response Time")
L2014 <- subset(PrevData, Paper == "Lupyan et al. (2014)")
# Training - Performance
ggplot(data=L2014, aes(x=SoundSym, y=Performance, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Label Type", y="Performance") +
facet_wrap(~Experiment) +
ggtitle("Lupyan et al. 2014 - Training - Performance")
# Training - RT
ggplot(data=L2014, aes(x=SoundSym, y=RT, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = RT - RT.SD,
ymax = RT + RT.SD,
width = 0.5)) +
labs(x="Label Type", y="Response Time") +
facet_wrap(~Experiment) +
ggtitle("Lupyan et al. 201 - Training - Response Time")## Warning: Removed 1 rows containing missing values (geom_bar).## Warning: Removed 1 rows containing missing values (geom_errorbar).
K2017 <- subset(PrevData, Paper == "Kovic et al. (2017)")
K2017$Trial.Type <- factor(K2017$Trial.Type, c("Label", "No Label", "Identical", "Mismatch"))
# Performance
ggplot(data=K2017, aes(x=Label.Type, y=Performance, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Label Type", y="Performance") +
facet_grid(Trial.Type ~ Train.vs..Test) +
ggtitle("Kovic et al. 2017 - Performance")
# Response Time
ggplot(data=K2017, aes(x=Label.Type, y=RT, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = RT - RT.SD,
ymax = RT + RT.SD,
width = 0.5)) +
labs(x="Label Type", y="Response Time") +
facet_grid(Trial.Type ~ Train.vs..Test) +
ggtitle("Kovic et al. 2017 - Response Time")
N2017 <- subset(PrevData, Paper == "Nielsen et al. (2017)")
# Performance
ggplot(data=N2017, aes(x=SoundSym, y=Performance, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Label Type", y="Performance") +
ggtitle("Nielsen et al. 2017 - Performance")
# Response Time
ggplot(data=N2017, aes(x=SoundSym, y=RT, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(aes(ymin = RT - RT.SD,
ymax = RT + RT.SD,
width = 0.5)) +
labs(x="Label Type", y="Response Time") +
ggtitle("Nielsen et al. 2017 - Response Time")
Okay those graphs are all pretty fucky and not well formatted, but we want to look at everything together anyways, so lets try that
First we will look at Performance
# First we get rid of some stuff not super relevant for what we care about here
PData <- subset(PrevData, Performance != 'NA')
PData <- subset(PData, Congruency != "Location")
PData <- subset(PData, Congruency != "Mixed")
PData <- subset(PData, SoundSym != "Real")
PData <- subset(PData, Train.vs..Test != "Verification")
dataagg1 <- aggregate(Performance ~ Train.vs..Test + Congruency + Paper + Label.Type, data = PData, mean)
dataagg2 <- aggregate(Performance.SD ~ Train.vs..Test + Congruency + Paper + Label.Type, data = PData, mean)
PDataAgg <- dataagg1
PDataAgg$Performance.SD <- dataagg2$Performance.SD
PDataAgg <-droplevels(PDataAgg)
PDataAgg$Label.Type <- factor(PDataAgg$Label.Type, c("Item", "Category"))
PDataAgg$Paper <- factor(PDataAgg$Paper,
c("Lupyan et al. (2007)", "Kovic et al. (2010)", "Lupyan et al. (2014)",
"Kovic et al. (2017)", "Nielsen et al. (2017)"))
PDataAgg$Congruency <- factor(PDataAgg$Congruency, c("Congruent", "Incongruent", "Conventional", "None"))
ggplot(data=PDataAgg, aes(x=Train.vs..Test, y=Performance, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity") +
geom_errorbar(position= position_dodge(1),
aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Trial Type", y="Performance") +
ggtitle("Previous Studies - Training - Performance") +
facet_grid(Paper ~ Label.Type) +
theme(strip.text.y = element_text(angle = 0))+
theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust = 1)) 
So that’s not perfect, mostly because it collapses bars into different sizes when you ask it to dodge, instead of just pretending there are all possible values- so we’ll need to manipulate our data frame to put those back
Papers <- c("Lupyan et al. (2007)", "Kovic et al. (2010)", "Lupyan et al. (2014)",
"Kovic et al. (2017)", "Nielsen et al. (2017)")
Congruency <- c("Congruent", "Incongruent", "Conventional", "None")
Train.vs..Test <- c("Training", "Testing")
Label.Type <- c("Item", "Category")
PDataNew <- expand.grid(Papers, Congruency, Train.vs..Test, Label.Type)
colnames(PDataNew) <- c("Paper", "Congruency", "Train.vs..Test", "Label.Type")
PDataNew <- unite(PDataNew, "ID", c("Paper", "Congruency", "Train.vs..Test", "Label.Type"),
sep = '-', remove= FALSE)
dataagg1 <- aggregate(Performance ~ Train.vs..Test + Congruency + Paper + Label.Type, data = PData, mean)
dataagg2 <- aggregate(Performance.SD ~ Train.vs..Test + Congruency + Paper + Label.Type, data = PData, mean)
PDataAgg <- dataagg1
PDataAgg$Performance.SD <- dataagg2$Performance.SD
PDataAgg <-droplevels(PDataAgg)
PDataAgg <- unite(PDataAgg, "ID", c("Paper", "Congruency", "Train.vs..Test", "Label.Type"),
sep = '-', remove= FALSE)
PDataNew$Performance <- PDataAgg[match(PDataNew$ID, PDataAgg$ID),]$Performance
PDataNew$Performance.SD <- PDataAgg[match(PDataNew$ID, PDataAgg$ID),]$Performance.SD
ggplot(data=PDataNew, aes(x=Train.vs..Test, y=Performance, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity", width = .75) +
geom_errorbar(position= position_dodge(.75),
aes(ymin = Performance - Performance.SD,
ymax = Performance + Performance.SD,
width = 0.5)) +
labs(x="Trial Type", y="Performance") +
ggtitle("Previous Studies - Performance") +
facet_grid(Paper ~ Label.Type) +
theme_alan() +
scale_y_continuous(breaks = c(0,0.5,1), minor_breaks = c(0.25, 0.75)) +
theme(strip.text.y = element_text(hjust= 0)) +
theme(panel.spacing.y = unit(10, "pt"))## Warning: Removed 60 rows containing missing values (geom_bar).## Warning: Removed 60 rows containing missing values (geom_errorbar).
Now we just need the same for Response time
RTData <- subset(PrevData, RT != 'NA')
RTData <- subset(RTData, Congruency != "Location")
RTData <- subset(RTData, Congruency != "Mixed")
RTData <- subset(RTData, SoundSym != "Real")
RTData <- subset(RTData, Train.vs..Test != "Verification")
dataagg3 <- aggregate(RT ~ Train.vs..Test + Congruency + Paper + Label.Type, data = RTData, mean)
dataagg4 <- aggregate(RT.SD ~ Train.vs..Test + Congruency + Paper + Label.Type, data = RTData, mean)
RTDataAgg <- dataagg3
RTDataAgg$RT.SD <- dataagg4$RT.SD
RTDataAgg <-droplevels(RTDataAgg)
RTDataAgg$Label.Type <- factor(RTDataAgg$Label.Type, c("Item", "Category"))
RTDataAgg$Paper <- factor(RTDataAgg$Paper,
c("Lupyan et al. (2007)", "Kovic et al. (2010)", "Lupyan et al. (2014)",
"Kovic et al. (2017)", "Nielsen et al. (2017)"))
RTDataAgg$Congruency <- factor(RTDataAgg$Congruency, c("Congruent", "Incongruent", "Conventional", "None"))
RTDataAgg <- unite(RTDataAgg, "ID", c("Paper", "Congruency", "Train.vs..Test", "Label.Type"),
sep = '-', remove= FALSE)
Papers <- c("Lupyan et al. (2007)", "Kovic et al. (2010)", "Lupyan et al. (2014)",
"Kovic et al. (2017)", "Nielsen et al. (2017)")
Congruency <- c("Congruent", "Incongruent", "Conventional", "None")
Train.vs..Test <- c("Training", "Testing")
Label.Type <- c("Item", "Category")
RTDataNew <- expand.grid(Papers, Congruency, Train.vs..Test, Label.Type)
colnames(RTDataNew) <- c("Paper", "Congruency", "Train.vs..Test", "Label.Type")
RTDataNew <- unite(RTDataNew, "ID", c("Paper", "Congruency", "Train.vs..Test", "Label.Type"),
sep = '-', remove= FALSE)
RTDataNew$RT <- RTDataAgg[match(RTDataNew$ID, RTDataAgg$ID),]$RT
RTDataNew$RT.SD <- RTDataAgg[match(RTDataNew$ID, RTDataAgg$ID),]$RT.SD
ggplot(data=RTDataNew, aes(x=Train.vs..Test, y=RT, fill = Congruency)) +
geom_bar(position= position_dodge(), stat = "identity", width = 0.75) +
geom_errorbar(position= position_dodge(0.75),
aes(ymin = RT - RT.SD,
ymax = RT + RT.SD,
width = 0.5)) +
labs(x="Trial Type", y="ResponseTime") +
ggtitle("Previous Studies - Response Times") +
facet_grid(Paper ~ Label.Type, scales = "free_y") +
theme_alan() +
theme(strip.text.y = element_text(hjust= 0)) +
theme(panel.spacing.y = unit(10, "pt")) +
scale_y_continuous(breaks = scales::pretty_breaks(n = 2)) ## Warning: Removed 60 rows containing missing values (geom_bar).## Warning: Removed 60 rows containing missing values (geom_errorbar).