library(readxl)
## Warning: package 'readxl' was built under R version 4.3.3
library(dplyr)
##
## Attaching package: 'dplyr'
## The following objects are masked from 'package:stats':
##
## filter, lag
## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, union
library(tidyr)
library(ggplot2)
library(kableExtra)
##
## Attaching package: 'kableExtra'
## The following object is masked from 'package:dplyr':
##
## group_rows
# ディレクトリ設定とデータ読み込み
setwd("~/Library/CloudStorage/Dropbox/土居_関口/2.Analysis_Doi")
df <- read_excel("writing_results_scored_4criteria.xlsx")
# 条件名を英語化
df <- df %>%
mutate(
Condition = case_when(
条件 == "ai-wcf" ~ "AI",
条件 == "model text" ~ "Model",
条件 == "control" ~ "Control",
TRUE ~ 条件
)
)
# Conditionの順序をAI→Model→Controlに固定
df <- df %>%
mutate(Condition = factor(Condition, levels = c("AI", "Model", "Control")))
# 記述統計（Pre/Post別）
desc_total <- df %>%
group_by(Condition) %>%
summarise(
n = n(),
Mean_Pre = mean(Pre_Total, na.rm = TRUE),
SD_Pre = sd(Pre_Total, na.rm = TRUE),
Mean_Post = mean(Post_Total, na.rm = TRUE),
SD_Post = sd(Post_Total, na.rm = TRUE)
)
# 表として整形
desc_total %>%
mutate(
Mean_Pre = sprintf("%.2f", Mean_Pre),
SD_Pre = sprintf("%.2f", SD_Pre),
Mean_Post = sprintf("%.2f", Mean_Post),
SD_Post = sprintf("%.2f", SD_Post)
) %>%
kbl(
caption = "Table 1. Descriptive Statistics for Total Scores by Condition",
align = "lcccc",
col.names = c("Condition", "n", "M (Pre)", "SD (Pre)", "M (Post)", "SD (Post)")
) %>%
kable_classic(full_width = FALSE, html_font = "Times New Roman") %>%
footnote(
general = "Note. M = mean; SD = standard deviation; Scores are totals across four writing dimensions (0–36).",
general_title = " "
)
## Warning: 'xfun::attr()' is deprecated.
## Use 'xfun::attr2()' instead.
## See help("Deprecated")
Table 1. Descriptive Statistics for Total Scores by Condition
|
Condition
|
n
|
M (Pre)
|
SD (Pre)
|
M (Post)
|
SD (Post)
|
|
AI
|
20
|
14.80
|
3.40
|
17.15
|
2.80
|
|
Model
|
16
|
14.44
|
4.52
|
17.19
|
2.48
|
|
Control
|
19
|
14.21
|
3.74
|
16.00
|
2.92
|
|
|
|
Note. M = mean; SD = standard deviation; Scores are totals
across four writing dimensions (0–36).
|
# 縦長化
df_long_total <- df %>%
select(Condition, Pre_Total, Post_Total) %>%
pivot_longer(cols = starts_with("Pre_") | starts_with("Post_"),
names_to = "Time", values_to = "Score") %>%
mutate(Time = factor(Time, levels = c("Pre_Total", "Post_Total"),
labels = c("Pre-Test", "Post-Test")))
# 色設定
fill_colors <- c("Pre-Test" = "#1f78b4", "Post-Test" = "#e31a1c")
# プロット
ggplot(df_long_total, aes(x = Time, y = Score, fill = Time)) +
geom_violin(trim = FALSE, alpha = 0.65, width = 0.8, color = "gray40") +
geom_boxplot(width = 0.15, outlier.shape = NA, color = "gray20", alpha = 0.7) +
stat_summary(fun = mean, geom = "point", shape = 8, color = "yellow", size = 2) +
facet_wrap(~Condition, nrow = 1) +
scale_fill_manual(values = fill_colors) +
labs(
title = "Distribution of Total Scores by Condition",
x = "Test Time", y = "Total Score (0–36)", fill = "Test Type"
) +
theme_minimal(base_size = 15) +
theme(
plot.title = element_text(hjust = 0.5, face = "bold"),
axis.title = element_text(face = "bold"),
legend.position = "top",
strip.text = element_text(size = 13, face = "bold")
)
# 観点を一括処理
criteria <- c("TA", "CC", "LR", "GRA")
desc_list <- lapply(criteria, function(crit) {
df %>%
group_by(Condition) %>%
summarise(
Mean_Pre = mean(get(paste0("Pre_", crit)), na.rm = TRUE),
SD_Pre = sd(get(paste0("Pre_", crit)), na.rm = TRUE),
Mean_Post = mean(get(paste0("Post_", crit)), na.rm = TRUE),
SD_Post = sd(get(paste0("Post_", crit)), na.rm = TRUE)
) %>%
mutate(Criterion = crit)
})
desc_each <- bind_rows(desc_list)
# 出力
desc_each %>%
select(Criterion, Condition, Mean_Pre, SD_Pre, Mean_Post, SD_Post) %>%
mutate(across(starts_with("Mean"), ~ sprintf("%.2f", .)),
across(starts_with("SD"), ~ sprintf("%.2f", .))) %>%
kbl(
caption = "Table 2. Descriptive Statistics for Each Writing Criterion by Condition",
align = "lccccc"
) %>%
kable_classic(full_width = FALSE, html_font = "Times New Roman")
## Warning: 'xfun::attr()' is deprecated.
## Use 'xfun::attr2()' instead.
## See help("Deprecated")
Table 2. Descriptive Statistics for Each Writing Criterion by Condition
|
Criterion
|
Condition
|
Mean_Pre
|
SD_Pre
|
Mean_Post
|
SD_Post
|
|
TA
|
AI
|
3.70
|
0.92
|
4.40
|
0.82
|
|
TA
|
Model
|
3.50
|
1.15
|
4.25
|
0.77
|
|
TA
|
Control
|
3.68
|
1.06
|
4.00
|
0.88
|
|
CC
|
AI
|
3.65
|
0.93
|
4.25
|
0.79
|
|
CC
|
Model
|
3.56
|
1.15
|
4.25
|
0.86
|
|
CC
|
Control
|
3.47
|
1.07
|
3.95
|
0.78
|
|
LR
|
AI
|
3.90
|
0.97
|
4.40
|
0.75
|
|
LR
|
Model
|
3.81
|
1.17
|
4.44
|
0.63
|
|
LR
|
Control
|
3.58
|
1.02
|
4.16
|
0.76
|
|
GRA
|
AI
|
3.55
|
0.76
|
4.10
|
0.72
|
|
GRA
|
Model
|
3.56
|
1.15
|
4.25
|
0.58
|
|
GRA
|
Control
|
3.47
|
0.90
|
3.89
|
0.74
|
# ======== 各観点別・条件別バイオリンプロット（Viewer表示用 完全版）======== #
library(dplyr)
library(tidyr)
library(ggplot2)
# --- データ縦長化 ---
df_long_criteria <- df %>%
select(
Condition,
starts_with("Pre_TA"), starts_with("Pre_CC"),
starts_with("Pre_LR"), starts_with("Pre_GRA"),
starts_with("Post_TA"), starts_with("Post_CC"),
starts_with("Post_LR"), starts_with("Post_GRA")
) %>%
pivot_longer(
cols = -Condition,
names_to = c("Time", "Criterion"),
names_pattern = "(Pre|Post)_(.*)",
values_to = "Score"
) %>%
mutate(
Time = factor(Time, levels = c("Pre", "Post")),
Criterion = factor(Criterion, levels = c("TA", "CC", "LR", "GRA")),
Condition = factor(Condition, levels = c("AI", "Model", "Control")) # ← 順序固定
)
# --- 色指定（Pre=青, Post=赤）---
fill_colors <- c("Pre" = "#1f78b4", "Post" = "#e31a1c")
# --- プロット ---
ggplot(df_long_criteria, aes(x = Time, y = Score, fill = Time)) +
geom_violin(trim = FALSE, alpha = 0.65, width = 0.8, color = "gray40") +
geom_boxplot(width = 0.15, outlier.shape = NA, color = "gray20", alpha = 0.7) +
stat_summary(fun = mean, geom = "point", shape = 8, color = "yellow", size = 2) +
facet_wrap(~ Criterion + Condition, ncol = 3, labeller = label_value) +
scale_fill_manual(values = fill_colors) +
labs(
title = "Distribution of Writing Scores by Criterion and Condition",
x = "Test Time", y = "Score (0–9)", fill = "Test Type"
) +
theme_minimal(base_size = 14) +
theme(
plot.title = element_text(hjust = 0.5, face = "bold", size = 16),
axis.title = element_text(size = 13, face = "bold"),
axis.text.x = element_text(size = 11),
axis.text.y = element_text(size = 11),
legend.position = "top",
legend.title = element_text(face = "bold"),
legend.text = element_text(size = 11),
strip.text = element_text(size = 12, face = "bold"),
panel.spacing = unit(1, "lines")
)
