如何从t检验模型生成的列表提取指标并整理为统一表格

问题描述 #

如何从统计模型（t检验）生成的列表中提取特定值？
我通过以下代码执行了多个t检验：

A<-lapply(merged_DF_final[2:6], function(x) t.test(x ~ merged_DF_final$Group))

我需要从每个子检验的结果中提取p值、t统计量、置信区间以及组均值，并将这些信息输出到一个统一表格中，请问该如何操作？

以下是存储在对象A中的内容：

$HC_HC_L_amygdala_baseline

    Welch Two Sample t-test

data:  x by merged_DF_final$Group 
t = 0.039543, df = 47.412, p-value = 0.9686
alternative hypothesis: true difference in means is not equal to 0 
95 percent confidence interval:  -0.4694404  0.4882694 
sample estimates:
mean in group CONN   mean in group HC 
         0.2954200          0.2860055 


$HC_HC_L_culmen_baseline

    Welch Two Sample t-test

data:  x by merged_DF_final$Group 
t = 0.81387, df = 53.695, p-value = 0.4193
alternative hypothesis: true difference in means is not equal to 0 
95 percent confidence interval:  -0.2970321  0.7028955 
sample estimates:
mean in group CONN   mean in group HC 
         0.4020883          0.1991566 


$HC_HC_L_fusiform_baseline

    Welch Two Sample t-test

data:  x by merged_DF_final$Group 
t = 0.024945, df = 53.851, p-value = 0.9802
alternative hypothesis: true difference in means is not equal to 0 
95 percent confidence interval:  -0.5768786  0.5914136 
sample estimates:
mean in group CONN   mean in group HC 
         0.5552184          0.5479509 


$HC_HC_L_insula_baseline

    Welch Two Sample t-test

data:  x by merged_DF_final$Group 
t = 0.79659, df = 52.141, p-value = 0.4293
alternative hypothesis: true difference in means is not equal to 0 
95 percent confidence interval:  -0.3000513  0.6951466 
sample estimates:
mean in group CONN   mean in group HC 
        0.12436946        -0.07317818 


$HC_HC_L_lingual_gyrus_baseline

    Welch Two Sample t-test

data:  x by merged_DF_final$Group 
t = -0.11033, df = 53.756, p-value = 0.9126
alternative hypothesis: true difference in means is not equal to 0 
95 percent confidence interval:  -0.5172863  0.4633268 
sample estimates:
mean in group CONN   mean in group HC 
         0.4395066          0.4664864

解决方案 #

每个t.test()返回的结果是一个包含各类统计值的列表，我们可以通过提取列表元素的方式获取所需信息，再合并成统一表格：

步骤1：定义提取函数 #

先写一个函数，从单个t检验结果中提取需要的统计量：

extract_tstats <- function(test_result) {
  data.frame(
    t_stat = test_result$statistic,       # t统计量
    df = test_result$parameter,           # 自由度
    p_value = test_result$p.value,        # p值
    ci_low = test_result$conf.int[1],     # 置信区间下限
    ci_high = test_result$conf.int[2],    # 置信区间上限
    mean_CONN = test_result$estimate[1],  # CONN组均值
    mean_HC = test_result$estimate[2]     # HC组均值
  )
}

步骤2：批量提取并合并表格 #

用lapply()遍历列表A，应用上面的提取函数，再用do.call(rbind, ...)合并成一个数据框：

# 批量提取每个检验的统计值
t_results <- lapply(A, extract_tstats)

# 合并为一个表格，保留变量名作为单独列
final_table <- do.call(rbind, t_results)
final_table$variable <- rownames(final_table)
rownames(final_table) <- NULL

# 调整列顺序（可选，让变量名在前）
final_table <- final_table[, c("variable", "t_stat", "df", "p_value", "ci_low", "ci_high", "mean_CONN", "mean_HC")]

最终结果示例 #

运行上述代码后，final_table会生成如下格式的表格（数值保留原精度）：

内容的提问来源于stack exchange，提问作者lotus