在R中如何绘制参与者双调查得分的扁平斜率散点图？

推荐图表类型 #

• 有序斜率图（Ordered Slope Graph）：将参与者按得分变化幅度（Survey2 - Survey1）排序后绘制斜率图，避免线条交叉杂乱，既能展示个体得分变化，也能直观看到整体趋势。
• 横向差值连线图：以得分为横轴、按差值排序的参与者为纵轴，用线段连接同一参与者的两次得分，清晰呈现每个个体的变化方向与幅度。

R实现步骤（以有序斜率图为例） #

1. 加载所需包 #

library(tidyverse)

2. 准备示例数据 #

df <- tibble(
  participantNum = 1:6,
  survey1_score = c(-28, -14, 20, 6, -20, 22),
  survey2_score = c(4, 16, -10, 8, 26, 24)
)

3. 数据预处理 #

将宽格式转长格式，计算得分差值并按差值排序参与者：

df_processed <- df %>%
  mutate(score_diff = survey2_score - survey1_score) %>%
  arrange(score_diff) %>%
  mutate(participant_order = factor(participantNum, levels = participantNum)) %>%
  pivot_longer(
    cols = starts_with("survey"), 
    names_to = "survey", 
    values_to = "score",
    names_prefix = "survey"
  ) %>%
  mutate(survey = factor(survey, levels = c("1", "2")))

4. 绘制有序斜率图 #

ggplot(df_processed, aes(x = survey, y = score, group = participant_order)) +
  geom_line(aes(color = score_diff), alpha = 0.7, linewidth = 1) +
  geom_point(size = 3) +
  geom_text(aes(label = participantNum), vjust = -1, size = 3) +
  scale_color_gradient2(low = "blue", mid = "gray", high = "red", midpoint = 0) +
  labs(x = "调查轮次", y = "得分", color = "得分变化（Survey2 - Survey1）") +
  theme_minimal() +
  theme(axis.text.y = element_blank(), panel.grid.major.y = element_blank())

可选方案：横向差值连线图 #

更聚焦个体变化的横向对比：

df_diff <- df %>%
  mutate(score_diff = survey2_score - survey1_score) %>%
  arrange(score_diff) %>%
  mutate(participant_order = factor(participantNum, levels = participantNum))

ggplot(df_diff) +
  geom_segment(
    aes(x = survey1_score, xend = survey2_score, 
        y = participant_order, yend = participant_order,
        color = score_diff), 
    linewidth = 1
  ) +
  geom_point(aes(x = survey1_score, y = participant_order), size = 3, color = "darkblue") +
  geom_point(aes(x = survey2_score, y = participant_order), size = 3, color = "darkred") +
  scale_color_gradient2(low = "blue", mid = "gray", high = "red", midpoint = 0) +
  labs(
    x = "得分", y = "参与者（按得分变化排序）", 
    color = "得分变化（Survey2 - Survey1）",
    title = "两次调查得分变化对比"
  ) +
  theme_minimal()

内容的提问来源于stack exchange，提问作者fourteen14