You recode in SPSS with Transform > Recode into Different Variables. Move the variable in, name a new output variable and click Change, then open Old and New Values to map each old code to a new one, click Add for every pair, and press Continue then OK. This route writes to a new column and leaves your raw data intact, which is why it is the safe choice for reverse scoring, binning and collapsing categories.
Recoding means turning the values of a variable into new values: flipping a reversed scale, grouping ages into brackets, or merging several categories into fewer. SPSS gives you two routes, and the one you pick decides whether your original data survives. This guide is part of our statistics homework help. It gives the exact menu path, the Old-and-New-Values mapping for three real tasks, and the mistakes that quietly break a scale score.
The golden rule. Always choose Recode into Different Variables, never Recode into Same Variables. The "same" route overwrites the original column, so the raw data is gone and a mistake cannot be undone. The "different" route writes the result to a new variable and leaves the original in place, so you can check the recode against the source and rerun it if a rule was wrong.
Every recode below uses the same dialog. Learn this once and the three tasks are only different mappings.
Two habits save marks. First, give the output variable a name you will recognise later, such as the source name with a suffix like _R for reversed or _grp for grouped. Second, after the recode, open Variable View and add value labels to the new variable, because Recode writes numbers only and the new column has no labels until you add them. Without labels your tables show bare codes instead of the group names.
Inside Old and New Values you also have Range options for continuous data: Range, LOWEST through value, Range, value through HIGHEST, and a Range from one number through another. The All other values option at the bottom lets you send everything you did not list to a single code or to system-missing, which is a clean way to handle leftovers.
The exact Old-and-New-Values mapping for the jobs you actually get set: reverse scoring, binning and collapsing.
A negatively worded item points the wrong way on the scale. You flip it before you average items or run reliability, or the scale score and Cronbach's alpha come out wrong.
Say a job-satisfaction scale is measured on five points, where one is strongly disagree and five is strongly agree, and the item "I often think about quitting" is negatively worded. A person who is satisfied should score high on the scale, but on this item they answer low, so its direction has to be reversed before it joins the others. For a 5-point item you swap one with five, two with four, and leave three where it is.
| Old Value | New Value | Meaning |
|---|---|---|
| 1 | 5 | strongly disagree becomes strongly agree |
| 2 | 4 | disagree becomes agree |
| 3 | 3 | neutral stays neutral |
| 4 | 2 | agree becomes disagree |
| 5 | 1 | strongly agree becomes strongly disagree |
Worked check. A respondent who scores two on the raw item now scores four on quit_think_R, which lines up with the positively worded items. Run a quick Analyze > Descriptive Statistics > Frequencies on both columns side by side: the counts mirror each other, one against five, two against four. That mirror is the proof the recode worked.
Common mistakes
Not sure which items on your scale are reversed? Send the questionnaire and data and a statistician will reverse the right ones and rebuild the scale. Get a quote →
Turns a continuous measure such as age into ordered brackets, for a grouped comparison or a cleaner table.
Here you keep age as a continuous variable and add a grouped copy, so you can still use exact age elsewhere. The trick is the Range controls in Old and New Values. Use LOWEST through for the bottom band and through HIGHEST for the top band so no real value falls outside your rules.
| Old Value (Range) | New Value | Group label to add later |
|---|---|---|
| Lowest through 29 | 1 | Under 30 |
| 30 through 44 | 2 | 30 to 44 |
| 45 through 59 | 3 | 45 to 59 |
| 60 through Highest | 4 | 60 and over |
Then label it. In Variable View, click the Values cell for age_grp and add one for Under 30, two for 30 to 44, three for 45 to 59, four for 60 and over. Set the Measure to Ordinal, since the groups have an order. Run Frequencies on age_grp to confirm the counts add up to your sample size, which tells you no case was left unmapped.
Common mistakes
Merges several categories of a nominal variable into a smaller set, for example when thin categories make a table unstable.
Suppose marital status is coded one married, two widowed, three divorced, four separated, five never married. For a simple partnered against not-partnered comparison you collapse the five into two. Every source code needs a rule, so nothing is dropped.
| Old Value | New Value | New group |
|---|---|---|
| 1 (married) | 1 | Partnered |
| 2 (widowed) | 2 | Not partnered |
| 3 (divorced) | 2 | Not partnered |
| 4 (separated) | 2 | Not partnered |
| 5 (never married) | 2 | Not partnered |
Handle the leftovers. If a variable has many codes and you only care about a few, map the ones you want and set All other values to a catch-all code or to system-missing on purpose, rather than leaving them to become accidental blanks. Then label partnered one for Partnered, two for Not partnered, and set its Measure to Nominal.
Common mistakes
Three tools change values, and each suits a different job. Pick by the shape of the mapping, not by habit.
| Tool | Menu path | Best for |
|---|---|---|
| Recode into Different Variables | Transform > Recode into Different Variables | Any value-to-value or range-to-value mapping: reverse scoring, binning, collapsing. Keeps the original. |
| Compute Variable | Transform > Compute Variable | A formula across the whole scale, such as new = 6 - old to reverse a 5-point item, or building a mean of items. |
| Visual Binning | Transform > Visual Binning | Cutting a continuous variable into equal-width or equal-count bands, with cut points and labels written for you. |
Procedure on this page was checked against the IBM SPSS Statistics documentation, the Laerd Statistics recoding guides, the UCLA OARC SPSS resources and the Kent State SPSS tutorials.
When a whole questionnaire needs reversing, the brackets are fussy, or a recode keeps writing blanks, a specialist builds it on your file and walks you through the mapping so you can defend it.
Upload your data file, the questionnaire or brief and your deadline, then get a free quote from support.
Approve the price and a statistician who works in SPSS every day sets up the recode right away.
Get the recoded file, the syntax and a short explanation in your account, with free revisions if you need them.
Your work goes to a specialist who passed a subject-specific statistics test and a background check, and who works in SPSS every day. What we deliver is a model answer and study aid for reference, and many students use it to learn the method and check their own recode.
Everything stays private. Your data and details are confidential, we never contact your school, payments are secure, and the work is original and plagiarism-checked.
Get SPSS help you can trust →"If there was a 6 rating I would choose it. First of all customer service was better than I even expected and I expected top notch customer service."
"This company is wonderful. I was able to depend on them all semester for my Statistics assignments. Dependable with competitive prices. Would definitely use again!"
Recode into Same Variables overwrites the original column, so the raw data is lost. Recode into Different Variables writes the result to a new column and leaves the original untouched. Always use Recode into Different Variables so you can check your work, undo a mistake and keep the original codes.
Open Transform > Recode into Different Variables, move the item in, name the output variable and click Change, then open Old and New Values. For a 5-point item map one to five, two to four, three to three, four to two and five to one, clicking Add after each pair, then Continue and OK. Do this before you compute a scale mean or run reliability.
Use Transform > Recode into Different Variables, name a new grouped variable and click Change, then in Old and New Values use the Range options. Map Lowest through 29 to one, 30 through 44 to two, 45 through 59 to three and 60 through Highest to four, then Continue and OK. Add value labels so the groups read as names.
Yes. Recode writes numbers only, so the new variable has no labels until you add them in Variable View under the Values column. Without labels your output shows bare codes such as one and two instead of the group names.
Both work. Recode maps each old value to a new one through a dialog. Compute reverses a scale with one formula, such as new = 6 minus old for a 5-point item, where the constant is the maximum plus one. Compute is faster for a whole set of items on the same scale; Recode is clearer when the mapping is irregular.
A blank means an old value had no rule, so nothing was written for that case. Check that every value is covered, that your ranges leave no gaps at the boundaries, and that you set All other values on purpose. Off-by-one range limits are the usual cause.
Hand it to a statistician and get reversed, binned and collapsed variables that keep your original data and pass a check before your deadline.
Get my free quote →