Survey data preparation

Survey label mapper

Published by crosstabs.com · Last updated ·
How results are checkedNamed statistical review pending. Selected automated reference cases are published on the validation page.

简体中文版

Codes 4–5 → “Agree”

210of 600

Brand statement 1 reads as 35.0% agree once 1–5 codes are labelled

One mapping, applied to the whole 8-item grid

Synthetic sample · 600 respondents · same data as “Use sample data”

Value labels translate the numeric codes a survey platform exports (1, 2, 3…) into the answers respondents actually saw. This tool lists each code with its count, lets you type or paste a label for every code (merging codes that share a label), applies the same mapping to a whole grid in one step, and exports a labeled CSV — free, with no account, and the file never leaves your browser.

Map codes to labels online

Open a data file (or the synthetic sample), choose a column and type a label next to each code, or paste a block such as 1=Male — comma and tab separators work too. If the file is an SPSS .sav that already carries value labels, they are filled in for you. Press Apply mapping for each column (tick other columns to reuse the same labels across a Likert grid), check the before and after counts, then download the relabeled file.

CSV, Excel, SPSS, Stata, SAS, R, Parquet, SQLite or JSON. The file is read in this browser tab and never uploaded.

How the mapping works

Each distinct value in the column is a code. A code with a label is replaced by that label; a code without one is either kept as it was or set to blank, depending on the option you choose. Blank cells stay blank. Numeric codes match regardless of spelling, so a mapping for 1 also applies to cells stored as 1.0.

Giving two codes the same label merges them — the usual way to build top-2-box or bottom-2-box groupings. On export you can replace the codes in place or keep them and add a <column>_label column next to each mapped column, which is safer when you still need the numeric values for means.

The method

Worked example

Worked example

In the synthetic sample (600 rows), brand_grid_1 is a 1–5 agreement item with these counts: 1 → 61, 2 → 124, 3 → 205, 4 → 153, 5 → 57.

Labelling 1 = Strongly disagree … 5 = Strongly agree keeps the five counts and only changes their names. Mapping 1 and 2 to Bottom-2 box, 3 to Neutral, and 4 and 5 to Top-2 box merges them into three groups: Top-2 box 210 (153 + 57), Neutral 205, Bottom-2 box 185 (61 + 124) — still 600 rows in total. Ticking brand_grid_2…brand_grid_8 applies the same labels to the rest of the grid.

Using it well

When to use it

Use it when

  • Use: A CSV or Excel export contains numeric answer codes and you have the codebook or questionnaire.
  • Use: You want readable answers in banner tables and charts instead of 1–5 codes.
  • Use: You need to collapse a scale into top-2 / bottom-2 boxes or merge small categories.

Not the right tool when

  • Avoid: You still need to compute means on the scale — keep the codes (use the add-a-column option) rather than replacing them.
  • Avoid: The column is free text or an ID with thousands of distinct values; labels only make sense for coded answers.

How to interpret it

Rule of thumb

Check the after-counts against the before-counts: merged labels should equal the sum of their codes, and any value listed as unlabeled is a code your codebook did not cover — often a “don't know” or missing-value code worth handling explicitly.

Next steps

Screen the file for speeders and straight-liners with the Data Cleaner, then open the labeled CSV in the Crosstabs workspace to build tables. SPSS users can read value labels directly — see open a .sav file without SPSS.

Questions

Frequently asked questions

What formats can I paste a mapping in?

One code per line, separated from its label by an equals sign (1=Male), a comma (1,Male) or a tab — so two columns copied from a spreadsheet codebook paste directly. Only the first separator on a line splits it, so labels may contain commas. Blank lines are ignored and unreadable lines are reported by line number.

How do I merge two answer codes into one category?

Type the same label for both codes. The after-counts table marks the merged label and shows the combined count, for example 4 and 5 both labelled Top-2 box.

Can I label a whole Likert grid at once?

Yes. Label one item, tick the other grid columns under 'Also apply this mapping to', and press Apply mapping. Each ticked column receives the same labels and appears in the applied-mappings list, where it can be edited or removed individually.

What happens to codes I do not label?

You choose per column: keep the original code (useful when only some codes need names) or set it to blank (useful for dropping a 'refused' code from tables). Cells that were already blank stay blank either way.

Does it read SPSS value labels?

When you open an SPSS .sav file, the value labels stored in the file are prefilled for each column, so you can review, edit or merge them before exporting a CSV that carries the labels as text.

Sources and next steps

References & further reading

  1. [1]IBM SPSS Statistics Command Syntax Reference — VALUE LABELS and ADD VALUE LABELS commands.
  2. [2]IBM SPSS Statistics Command Syntax Reference — RECODE command (collapsing codes into new categories).

Try it on your own data — free, no signup

Upload a CSV, Excel, SPSS, Stata, SAS, R, Parquet or JSON file. Everything runs in your browser; your file never leaves your device.

Open the workspace →

← All calculators & guides