如何一次性批量重编码多个Likert量表列？

问题场景 #

给定如下示例数据集：

# Example Dataset
Q1 <- c("Agree", "Disagree", "Strongly Agree", "Strongly Disagree", "Strongly Disagree")
Q2 <- c("Disagree", "Disagree", "Agree", "Disagree", "Strongly Disagree")
Q3 <- c("Agree", "Disagree", "Strongly Agree", "Strongly Agree", "Strongly Agree")
Q4 <- c("Disagree", "Disagree", "Disagree", "Strongly Disagree", "Strongly Disagree")

data <- data.frame(Q1, Q2, Q3, Q4)

原处理方式为逐个列使用car包的recode函数，代码冗余：

library(car)

data$Q1 <- car::recode(data$Q1, "
                       'Strongly Disagree' = 1; 
                       'Disagree' = 2; 
                       'Agree' = 3; 
                       'Strongly Agree' = 4")
data$Q2 <- car::recode(data$Q2, "
                       'Strongly Disagree' = 1; 
                       'Disagree' = 2; 
                       'Agree' = 3; 
                       'Strongly Agree' = 4")
data$Q3 <- car::recode(data$Q3, "
                       'Strongly Disagree' = 1; 
                       'Disagree' = 2; 
                       'Agree' = 3; 
                       'Strongly Agree' = 4")
data$Q4 <- car::recode(data$Q4, "
                       'Strongly Disagree' = 1; 
                       'Disagree' = 2; 
                       'Agree' = 3; 
                       'Strongly Agree' = 4")

尝试直接指定列范围批量处理但失败：

data <- recode(data$[1:4], "
                  'Strongly Disagree' = 1; 
                  'Disagree' = 2; 
                  'Agree' = 3; 
                  'Strongly Agree' = 4")

需求：实现批量按列重编码，简化代码。

解决方案 #

方案1：Base R 结合 lapply 批量处理 #

利用lapply遍历目标列，统一应用recode规则，直接替换原数据列：

library(car)

# 可指定列名或列索引，这里处理所有列
target_cols <- names(data)
data[target_cols] <- lapply(data[target_cols], function(col) {
  car::recode(col, "'Strongly Disagree' = 1; 'Disagree' = 2; 'Agree' = 3; 'Strongly Agree' = 4")
})

方案2：tidyverse 批量操作（dplyr） #

使用dplyr的mutate_all（处理所有列）或mutate_at（指定列）实现简洁批量处理：

library(dplyr)
library(car)

# 处理所有列
data <- data %>%
  mutate_all(~ car::recode(., "'Strongly Disagree' = 1; 'Disagree' = 2; 'Agree' = 3; 'Strongly Agree' = 4"))

# 仅处理第1-4列（或指定列名，如vars(Q1:Q4)）
data <- data %>%
  mutate_at(vars(1:4), ~ car::recode(., "'Strongly Disagree' = 1; 'Disagree' = 2; 'Agree' = 3; 'Strongly Agree' = 4"))

方案3：因子转换（无需car包） #

由于量表类别是有序的，可直接将字符列转为有序因子，再转换为数值，轻量且直观：

# 定义类别顺序（必须与编码规则对应）
level_order <- c("Strongly Disagree", "Disagree", "Agree", "Strongly Agree")

# 批量转换所有列
data[] <- lapply(data, function(col) {
  as.integer(factor(col, levels = level_order, ordered = TRUE))
})

内容的提问来源于stack exchange，提问作者HelplessStatistician