-
Notifications
You must be signed in to change notification settings - Fork 0
/
basic_stats.R
303 lines (255 loc) · 12.8 KB
/
basic_stats.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
## LIBRARIES----------------------
source("subfunctions.R",encoding="UTF-8")
library(tidyverse)
library(R.matlab)
library(HDInterval)
theme_set(theme_void(base_family = "sans"))
## Load data ----
ilsup <- read.csv("data/superiority_data.csv")
ponzometa <- read.csv("data/ponzo.csv")
meanconf <- read.csv("data/meanconf.csv")
# Calculation
ponzometa$d_rel <- ponzometa$d1_1 - ponzometa$d1_0
ponzometa$c_rel <- ponzometa$c1_1 - ponzometa$c1_0
ponzometa$metac_rel <- ponzometa$metac_1 - ponzometa$metac_0
ponzometa$mceff_rel <- ponzometa$mceff_1 - ponzometa$mceff_0
ponzometa$ln_mceff_rel <- ponzometa$ln_mceff_1 - ponzometa$ln_mceff_0
ponzometa$ilsup <- ilsup$value
des <- psych::describe(ponzometa)
print(des,digits=3)
## t-tests ----
# Each condition
t.test(ponzometa$ilsup)
t.test(ponzometa$d1_0)
t.test(ponzometa$d1_0)$p.value
t.test(ponzometa$d1_1)
t.test(ponzometa$c1_0)
t.test(ponzometa$c1_1)
t.test(ponzometa$metac_0)
t.test(ponzometa$metac_1)
cohens_d(ponzometa$ilsup)
cohens_d(ponzometa$d1_0)
cohens_d(ponzometa$d1_1)
cohens_d(ponzometa$c1_0)
cohens_d(ponzometa$c1_1)
cohens_d(ponzometa$metac_0)
cohens_d(ponzometa$metac_1)
# Depth-Control
t.test(ponzometa$d_rel)
t.test(ponzometa$c_rel)
t.test(ponzometa$metac_rel)
cohens_d(ponzometa$d_rel)
cohens_d(ponzometa$c_rel)
cohens_d(ponzometa$metac_rel)
# MAP & HDI (Group-level): Log-Mratio
mcmc0 <- R.matlab::readMat("data/mcmc0.mat")
mcmc0 <- lapply(mcmc0, unlist, use.names=FALSE)
mcmc0 <- as.vector(mcmc0$mcmc0)
mc_density0 <- density(mcmc0,bw=bw_silverman(mcmc0),n=100)
mc_density0$x[match(max(mc_density0$y),mc_density0$y)] # find an index that gives max y
HDInterval::hdi(mcmc0)
mcmc1 <- R.matlab::readMat("data/mcmc1.mat")
mcmc1 <- lapply(mcmc1, unlist, use.names=FALSE)
mcmc1 <- as.vector(mcmc1$mcmc1)
mc_density1 <- density(mcmc1,bw=bw_silverman(mcmc1),n=100)
mc_density1$x[match(max(mc_density1$y),mc_density1$y)] # find an index that gives max y
HDInterval::hdi(mcmc1)
mcmcDiff <- mcmc1-mcmc0
mc_densityDiff <- density(mcmcDiff,bw=bw_silverman(mcmcDiff),n=100)
mc_densityDiff$x[match(max(mc_densityDiff$y),mc_densityDiff$y)] # find an index that gives max y
HDInterval::hdi(mcmcDiff)
# MAP & HDI (Group-level): Mratio
mcmc0exp <- exp(mcmc0)
mc_density0exp <- density(mcmc0exp,bw=bw_silverman(mcmc0exp),n=100)
mc_density0exp$x[match(max(mc_density0exp$y),mc_density0exp$y)] # find an index that gives max y
HDInterval::hdi(mcmc0exp)
mcmc1exp <- exp(mcmc1)
mc_density1exp <- density(mcmc1exp,bw=bw_silverman(mcmc1exp),n=100)
mc_density1exp$x[match(max(mc_density1exp$y),mc_density1exp$y)] # find an index that gives max y
HDInterval::hdi(mcmc1exp)
mcmcDiffexp <- mcmc1exp-mcmc0exp
mc_densityDiffexp <- density(mcmcDiffexp,bw=bw_silverman(mcmcDiffexp),n=100)
mc_densityDiffexp$x[match(max(mc_densityDiffexp$y),mc_densityDiffexp$y)] # find an index that gives max y
HDInterval::hdi(mcmcDiffexp)
# Mean confidence (metacognitive bias)
t.test(meanconf$depth,meanconf$control,paired=T)
cohens_d(meanconf$depth-meanconf$control)
## Create data frames for ggplot2 ----
# Data frame for ggplot2 (Depth vs Control)
sdt_raw <- ponzometa[,c(1:11)] %>% tidyr::pivot_longer(cols = -c("id"),names_to = c("index"), values_to = "val")
sdt_raw$index <- factor(sdt_raw$index,levels=colnames(ponzometa)[c(11:1)])
sdt_raw_means <- sdt_raw %>% group_by(index) %>% summarize(mean=mean(val), conf_lwr = mean(val) - qt(0.975, df=length(val)-1)*sd(val)/sqrt(length(val)), conf_upr = mean(val) + qt(0.975, df=length(val)-1)*sd(val)/sqrt(length(val)))
# Assign hBayesian estimates
sdt_raw_means[1,2] <- mc_density1$x[match(max(mc_density1$y),mc_density1$y)]
sdt_raw_means[1,3:4] <- t(HDInterval::hdi(mcmc1)[1:2])
sdt_raw_means[2,2] <- mc_density0$x[match(max(mc_density0$y),mc_density0$y)]
sdt_raw_means[2,3:4] <- t(HDInterval::hdi(mcmc0)[1:2])
sdt_raw_means[3,2] <- mc_density1exp$x[match(max(mc_density1exp$y),mc_density1exp$y)]
sdt_raw_means[3,3:4] <- t(HDInterval::hdi(mcmc1exp)[1:2])
sdt_raw_means[4,2] <- mc_density0exp$x[match(max(mc_density0exp$y),mc_density0exp$y)]
sdt_raw_means[4,3:4] <- t(HDInterval::hdi(mcmc0exp)[1:2])
sdt_raw_means # updated
# Data frame for ggplot2 (Diff)
sdt_rel <- ponzometa[,c(1,12:16)] %>% tidyr::pivot_longer(cols = -c("id"),names_to = c("index"), values_to = "val")
sdt_rel$index <- factor(sdt_rel$index,levels=colnames(ponzometa)[c(16:12)])
sdt_rel_means <- sdt_rel %>% group_by(index) %>% summarize(mean=mean(val), conf_lwr = mean(val)-qt(0.975, df=length(val)-1)*sd(val)/sqrt(length(val)), conf_upr = mean(val)+qt(0.975, df=length(val)-1)*sd(val)/sqrt(length(val)) )
# Assign hBayesian estimates
sdt_rel_means[1,2] <- mc_densityDiff$x[match(max(mc_densityDiff$y),mc_densityDiff$y)]
sdt_rel_means[1,3:4] <- t(HDInterval::hdi(mcmcDiff)[1:2])
sdt_rel_means[2,2] <- mc_densityDiffexp$x[match(max(mc_densityDiffexp$y),mc_densityDiffexp$y)]
sdt_rel_means[2,3:4] <- t(HDInterval::hdi(mcmcDiffexp))
sdt_rel_means # updated
## Correlations ----
# dprime (rel)
cor_spearman.test(ponzometa$d_rel,ponzometa$ilsup)
# Log M-ratio (rel)
cor_spearman.test(ponzometa$ln_mceff_rel,ponzometa$ilsup)
# dprime vs log M-ratio
cor_spearman.test(ponzometa$d_rel,ponzometa$ln_mceff_rel)
# Partial Correlations (each other)
pcor_ci_spearman.test(ponzometa$ilsup,ponzometa$d_rel,ponzometa$ln_mceff_rel)
pcor_ci_spearman.test(ponzometa$ilsup,ponzometa$ln_mceff_rel,ponzometa$d_rel)
#ppcor::pcor.test(ponzometa$ilsup,ponzometa$d_rel,ponzometa$ln_mceff_rel,"spearman")
#ppcor::pcor.test(ponzometa$ilsup,ponzometa$ln_mceff_rel,ponzometa$d_rel,"spearman")
# Partial Correlations (controlling for age)
pcor_ci_spearman.test(ponzometa$ilsup,ponzometa$d_rel,ilsup$age)
pcor_ci_spearman.test(ponzometa$ilsup,ponzometa$ln_mceff_rel,ilsup$age)
#ppcor::pcor.test(ponzometa$ilsup,ponzometa$d_rel,ilsup$age,"spearman")
#ppcor::pcor.test(ponzometa$ilsup,ponzometa$ln_mceff_rel,ilsup$age,"spearman")
# C (rel)
cor_spearman.test(ponzometa$c_rel,ponzometa$ilsup)
# Meta-C (rel)
cor_spearman.test(ponzometa$metac_rel,ponzometa$ilsup)
# Metacognitive bias (Mean confidence)
cor_spearman.test(meanconf$depth - meanconf$control,ponzometa$ilsup)
# dprime (Control)
cor_spearman.test(ponzometa$d1_0,ponzometa$ilsup)
# C (Control)
cor_spearman.test(ponzometa$c1_0,ponzometa$ilsup)
# Meta-C (Control)
cor_spearman.test(ponzometa$metac_0,ponzometa$ilsup)
# M-ratio (Control) # SAME AS Log M-ratio
#cor_spearman.test(ponzometa$mceff_0,ponzometa$ilsup)
# Log M-ratio (Control)
cor_spearman.test(ponzometa$ln_mceff_0,ponzometa$ilsup)
# Criterion and metacognitive performance
cor_spearman.test(ponzometa$c_rel,ponzometa$ln_mceff_rel)
cor_spearman.test(ponzometa$c_rel,meanconf$depth - meanconf$control)
# hBayes with SI as a covariate ----
simreg <- read.csv("data/reglogMratio.csv")
simreg$ln_mceff_rel <- simreg$ln_mceff_1 - simreg$ln_mceff_0
cor_spearman.test(simreg$ln_mceff_rel,ilsup$value)
## ggplots theme ----
# general theme
theme_set(theme_void(base_family = "sans"))
theme_update(
axis.text.x = element_text(color = "black", size = 16,
margin = margin(t = 6)),
axis.text.y = element_text(color = "black", size = 16, hjust = 1,
margin = margin(r = 6)),
axis.line.x = element_line(color = "black", linewidth = 1),
panel.grid.major.y = element_line(color = "grey90", linewidth = .6),
plot.margin = margin(rep(20, 4)),
axis.title.x = element_text(size=22, lineheight=.9, face = "bold"),
axis.title.y = element_text(size=22, lineheight=.9, face = "bold", angle = 90 )
)
# custom colors
my_pal <- c("#0041FF","#FF2800","#9A0079","#FAF500","#35A16B") #
## ggplot illusory superiority [Figure 2a]----
ils_raw <- ilsup[,c(1,4)] %>% tidyr::pivot_longer(cols = -c("id"),names_to = c("index"), values_to = "val")
ils_raw$index <- factor(ils_raw$index,levels=colnames(ponzometa)[16:12])
ils_raw_means <- ils_raw %>% group_by(index) %>% summarize(mean=mean(val), ci = qt(0.975, df=length(val)-1)*sd(val)/sqrt(length(val)))
cl <- wesanderson::wes_palette("Cavalcanti1")
g <- ggplot(ils_raw, aes(x = index, y = val, color = index)) +
geom_jitter(size = 3, alpha = 0.3, width = 0.2) +
geom_point(data = ils_raw_means, aes(x=index,y=mean,fill = index), size = 5, colour = "grey30", shape = 21) +
geom_errorbar(data = ils_raw_means, aes(x = index, y= NULL,ymin= mean - ci, ymax = mean + ci), width=.2) +
geom_hline(yintercept = 0, linetype = "dotted", colour = "grey30") +
coord_flip() +
scale_colour_manual(values = rep(c(cl[2]),5) ) +
scale_fill_manual(values = rep(c(cl[2]),5) ) +
scale_y_continuous("Value",breaks = seq(-1,1,by=0.5),limits = c(-1, 1)) +
scale_x_discrete(labels="Superiority") +
guides(colour = "none", fill = "none")
g
## ggplot SDT1 [Figure 2b]----
metaLabels <- c("ln Eff (Depth)","ln Eff (Control)",
"Efficiency (Depth)","Efficiency (Control)",
"Meta-C (Depth)","Meta-C (Control)",
"C (Depth)","C (Control)",
"d (Depth)","d (Control)")
cl <- wesanderson::wes_palette("Cavalcanti1")
g <- ggplot(sdt_raw, aes(x = index, y = val, color = index)) +
geom_jitter(size = 3, alpha = 0.3, width = 0.2) +
geom_point(data = sdt_raw_means, aes(x=index,y=mean,fill = index), size = 5, colour = "grey30", shape = 21) +
geom_errorbar(data = sdt_raw_means, aes(x = index, y= NULL,ymin= conf_lwr, ymax = conf_upr), width=.2) +
geom_hline(yintercept = 0, linetype = "dotted", colour = "grey30") +
coord_flip() +
scale_colour_manual(values = rep(c(cl[5],cl[1]),5) ) +
scale_fill_manual(values = rep(c(cl[5],cl[1]),5) ) +
scale_y_continuous("Value",breaks = seq(-6,6,by=2),limits = c(-6, 6),expand=c(0,0)) +
scale_x_discrete(labels=metaLabels) +
guides(colour = "none", fill = "none")
g
## ggplot SDT2 Diff [Figure 2c]----
metaLabels2 <- c("ln Eff (Control)",
"Efficiency (Control)",
"Meta-C",
"C",
"d")
cl <- wesanderson::wes_palette("Cavalcanti1")
g <- ggplot(sdt_rel, aes(x = index, y = val, color = index)) +
geom_jitter(size = 3, alpha = 0.3, width = 0.2) +
geom_point(data = sdt_rel_means, aes(x=index,y=mean,fill = index), size = 5, colour = "grey30", shape = 21) +
geom_errorbar(data = sdt_rel_means, aes(x = index, y= NULL,ymin= conf_lwr, ymax = conf_upr), width=.2) +
geom_hline(yintercept = 0, linetype = "dotted", colour = "grey30") +
coord_flip() +
scale_colour_manual(values = rep(c(cl[2]),5) ) +
scale_fill_manual(values = rep(c(cl[2]),5) ) +
scale_y_continuous("Value",breaks = seq(-6,6,by=2),limits = c(-6, 6),expand=c(0,0)) +
scale_x_discrete(labels=metaLabels2) +
guides(colour = "none", fill = "none")
g
## ggscatter plots (relative) [Figure 3a,b] ----
scaleFUN <- function(x) sprintf("%.1f", x)
p <- ggpubr::ggscatter(ponzometa, x = "ilsup", y = "d_rel",
add = "reg.line", # Add regression line
conf.int = TRUE, # Add confidence interval
ylim = c(-5.8, 0),
#ellipse = TRUE,
add.params = list(color = "blue",
fill = "lightgray")
) +
scale_y_continuous(labels=scaleFUN) +
ggpubr::stat_cor(
method = "spearman", label.x = 0.0, label.y = 0, r.digits = 7) # Add correlation coefficient
ggExtra::ggMarginal(p, type = "density")
cor_spearman.test(ponzometa$ilsup,ponzometa$d_rel)
p <- ggpubr::ggscatter(ponzometa, x = "ilsup", y = "ln_mceff_rel",
add = "reg.line", # Add regression line
conf.int = TRUE, # Add confidence interval
ylim = c(-0.6, 0.3),
#ellipse = TRUE,
add.params = list(color = "blue",
fill = "lightgray")
) +
ggpubr::stat_cor(
method = "spearman", label.x = -0.2, label.y = 0.3, r.digits = 7) # Add correlation coefficient
ggExtra::ggMarginal(p, type = "density")
cor_spearman.test(ponzometa$ilsup,ponzometa$ln_mceff_rel)
p <- ggpubr::ggscatter(ponzometa, x = "d_rel", y = "ln_mceff_rel",
add = "reg.line", # Add regression line
conf.int = TRUE, # Add confidence interval
ylim = c(-0.6, 0.3),
#ellipse = TRUE,
add.params = list(color = "blue",
fill = "lightgray")
) +
scale_x_continuous(labels=scaleFUN) +
scale_y_continuous(labels=scaleFUN) +
ggpubr::stat_cor(
method = "spearman", label.x = -4, label.y = 0.1, r.digits = 7) # Add correlation coefficient
ggExtra::ggMarginal(p, type = "density")
cor_spearman.test(ponzometa$d_rel,ponzometa$ln_mceff_rel)
# remove 2 outliers in log M-ratio
cor_spearman.test(ponzometa$d_rel[abs(ponzometa$ln_mceff_rel)< 0.2],ponzometa$ln_mceff_rel[abs(ponzometa$ln_mceff_rel)< 0.2])