数据分析：多组ANOVA和后置检验

十三画者2024-05-16 8:52

介绍

单因素多个levels数据在做方差分析后，需要进一步做如t检验等置后检验，判断两两组间差异。

加载R包和准备数据

{r} 复制代码

# Load required R packages
library(tidyverse)
library(rstatix)
library(ggpubr)
  
data("PlantGrowth")
set.seed(1234)
PlantGrowth %>% sample_n_by(group, size = 1)

查看统计量

{r} 复制代码

PlantGrowth %>%
  group_by(group) %>%
  get_summary_stats(weight, type = "mean_sd")

使用ANOVA test比较多组的平均值

{r} 复制代码

res.aov <- PlantGrowth %>% anova_test(weight ~ group)
res.aov

组间t检验

{r} 复制代码

pwc <- PlantGrowth %>%
  pairwise_t_test(weight ~ group, p.adjust.method = "bonferroni")
pwc

p值的坐标轴参数

{r} 复制代码

pwc <- pwc %>% add_xy_position(x = "group")
pwc

可视化

{r} 复制代码

ggboxplot(PlantGrowth, x = "group", y = "weight") +
  stat_pvalue_manual(pwc, label = "p.adj", tip.length = 0, step.increase = 0.1) +
  labs(
    subtitle = get_test_label(res.aov, detailed = TRUE),
    caption = get_pwc_label(pwc))

Reference

HOW TO PERFORM T-TEST FOR MULTIPLE GROUPS IN R