如何在R中统计问卷不同答案？寻求列名修改代码优化方案

Great question—hardcoding column numbers is definitely brittle, especially if your questionnaire ever adds or removes questions later. Let’s look at a couple of concise, flexible approaches that work regardless of how many Q_* columns you have in your data frame DS.

Method 1: Tidyverse (dplyr + tidyr) – Most Readable #

This approach reshapes your data to long format first, then counts responses across all questions in one go. It’s intuitive and easy to extend:

library(dplyr)
library(tidyr)

# Pivot all Q_* columns to long format, then count each answer
response_counts <- DS %>%
  select(starts_with("Q_")) %>%  # Auto-grab all question columns (no hardcoding!)
  pivot_longer(everything(), names_to = "question", values_to = "answer") %>%
  count(question, answer, name = "count") %>%
  # Optional: Map numeric codes to labels if needed
  mutate(answer_label = case_when(
    answer == 1 ~ "a",
    answer == 2 ~ "b",
    # Add more mappings as required
    TRUE ~ as.character(answer)
  ))

Method 2: Base R – No External Packages #

If you prefer sticking to base R, use lapply to iterate over question columns and combine results cleanly:

# Get all question columns using regex
question_cols <- grep("^Q_", colnames(DS), value = TRUE)

# Calculate counts for each column and merge into a single data frame
response_counts <- do.call(rbind, lapply(question_cols, function(col) {
  tab <- table(DS[[col]])
  data.frame(
    question = col,
    answer = as.integer(names(tab)),
    count = as.integer(tab)
  )
}))

# Optional: Add answer labels
response_counts$answer_label <- factor(response_counts$answer, levels = c(1,2), labels = c("a","b"))

Why These Are Better Than Fixed Column Numbers #

• Flexibility: Both methods automatically detect all columns starting with Q_, so you won’t need to update code if you add Q_097 or remove Q_093 later.
• Readability: Anyone reviewing your code can immediately understand what’s happening, instead of guessing what arbitrary column indices like 5:9 refer to.
• Scalability: Works seamlessly for 5 questions or 500 without changing a single line of code.

If you had a specific conta_respo... code snippet you were working with, feel free to share it and we can refine these approaches even further!

内容的提问来源于stack exchange，提问作者rmaranhao