T-score calculator
Convert a raw score to a T-score (mean 50, SD 10).
The T-score calculator rescales a raw score onto a standardized scale with a mean of 50 and a standard deviation of 10, using the group's actual mean and standard deviation — a common transformation for reporting test results in psychometrics and education.
How to use
- Enter the raw score, the group's mean, and its standard deviation.
- T-score = 50 + ((raw score − mean) ÷ SD) × 10.
Common use scenarios
- Reporting a standardized test result on the common T-score scale.
- Comparing a score against a group's distribution without exposing raw values.
- Converting a psychometric scale score for a report or research write-up.
- Double-checking a T-score a stats package or spreadsheet formula produced.
✓Content & verification reviewed: 2026-08-25
How we verify T-score calculator
How to confirm the T-score is calculated the way you expect.
Formula: T = 50 + ((raw score − mean) ÷ SD) × 10
The raw score is first converted to a Z-score (how many standard deviations from the mean it sits), then that Z-score is scaled by 10 and shifted by 50 — so a score exactly at the mean always produces a T-score of exactly 50.
A T-score is a rescaling, not a percentile
A T-score of 60 means "one standard deviation above the mean on this scale" — it doesn't mean "60th percentile." Don't read T-scores as percentile ranks; use the percentile rank calculator if that's what you actually need.
The mean and SD must describe the actual reference group
The result is only meaningful if the group mean and standard deviation you enter genuinely describe the population or sample the raw score belongs to — plugging in numbers from a different group produces a T-score that doesn't mean anything for your actual data.
Before you rely on the result
- The group mean and standard deviation you entered describe the same group the raw score came from.
- The standard deviation is greater than 0 (a required condition for the calculation to be defined).
- You're treating the result as a rescaled score, not as a percentile rank.
- If comparing across different tests, you've confirmed both used a comparable T-score definition.
Who this is for
- Anyone reporting a standardized test result on the T-score scale.
- People comparing a score against a group's distribution.
- Researchers converting a psychometric scale score for a write-up.
- Anyone double-checking a T-score a stats package produced.
Real-world examples
Standardized test reporting
Someone reporting results from a psychometric instrument converts each raw score to a T-score so results are comparable across different versions of the test.
Comparing to a norm group
A researcher with a participant's raw score and the published norm group's mean/SD converts it to a T-score to see how far above or below average that participant falls.
Verifying a spreadsheet formula
Someone who built a T-score formula in a spreadsheet enters the same raw score, mean, and SD here to confirm their formula produces a matching result.
Limitations
- This is a linear rescaling, not a percentile — a T-score of 60 doesn't mean "60th percentile".
Frequently asked questions
Does a T-score of 70 mean the 70th percentile?
No — a T-score of 70 means two standard deviations above the mean on this rescaled scale. It's not a percentile; use the percentile rank calculator if you need an actual percentile.
What if my standard deviation is 0?
The calculation is undefined when SD is 0, since it would require dividing by zero — the tool shows an error rather than a result in that case.
Can I use this for any test, or only IQ-style tests?
The T-score formula (mean 50, SD 10) is a general rescaling that works for any raw score as long as you have an accurate group mean and standard deviation — it's not specific to any one test.
Why is my T-score different from what my software reports?
Check that the mean and SD you entered match exactly what the software used — a mismatch in the reference group's mean or SD is the most common cause of a discrepancy.
🔒 Runs in your browser — nothing is uploaded. Results are for reference only; see the disclaimer. Last updated: 2026-08-25
Subscribe for updates
New tools and tips, straight to your inbox. No spam, unsubscribe anytime.