Raw Score Calculator
Turn a z-score or percentile back into the original score: X = μ + zσ. You can also solve for the z-score, the mean, or the standard deviation, and see where the score falls on the bell curve.
X = μ + z × σ
Enter the mean, the standard deviation, and the z-score. All three must describe the same data set.
—
—
Raw score (X)
—
—
Z-score
—
—
Percentile
—
share at or below X
Above X
—
if data are normal
Where the Score Falls
—
How it was calculated
Score summary
Notes on this result
Conversion Table for This Distribution
| Z-score | Raw score (X) | Percentile |
|---|
How to Use This Calculator
- Choose what to find. Raw score (X) is the default. The other three modes solve the same formula for the z-score, the mean, or the standard deviation.
- Enter the values you know, or pick a quick scale (standard normal, IQ, or T-score) to fill in the mean and standard deviation.
- Give the z-score directly, or switch to Percentile and enter a value between 0 and 100. The calculator converts it to a z-score for you.
- Press Calculate Raw Score. The result comes with the bell curve, the working, and a table of raw scores for z-scores from −3 to +3.
What Is a Raw Score?
A raw score calculator converts a standardized score into its original numerical value using the population mean and standard deviation. The raw score formula is x = μ + zσ, where x is the raw score, μ is the mean, z is the z-score, and σ is the standard deviation. Raw scores represent original, unstandardized values.
Examples include the number of correct answers on a test or a height in centimeters. A raw score has not been converted to a percentage, a percentile, or a standard score; in exams it is the number of marks earned before any scaling.
A raw score alone says little about performance. A score of 48 is high on a test with a mean of 40 and low on one with a mean of 60. A z-score fixes this by measuring how many standard deviations the raw score lies from the mean. This calculator goes the other way: from a z-score back to the raw score.
z < 0
The raw score is below the mean.
z = 0
The raw score equals the mean.
z > 0
The raw score is above the mean.
Definition reference: Wikipedia — Raw score.
Raw Score Formulas
Raw score
X = μ + z × σ
Also written X = zσ + μ.
Z-score
z = (X − μ) ÷ σ
Standard deviations from the mean.
Mean
μ = X − z × σ
Needs X, σ, and z.
Standard deviation
σ = (X − μ) ÷ z
Needs z ≠ 0. The result must be positive.
Percentile
P = Φ(z) × 100
Φ is the standard normal cumulative distribution function.
Other scales
T = 50 + 10z, IQ = 100 + 15z
Each is a z-score with a new mean and standard deviation.
Formula reference: Wikipedia — Standard score.
Worked Example 1: From a Z-Score to a Raw Score
A data set has a mean of 60 and a standard deviation of 3. What raw score has a z-score of −4?
- Formula: X = z × σ + μ = −4 × 3 + 60.
- Multiply: −4 × 3 = −12.
- Add the mean: −12 + 60 = 48.
- The sign of z matches the position: 48 is below the mean of 60, so z is negative.
Worked Example 2: From a Percentile to a Raw Score
On an IQ scale (mean 100, standard deviation 15), what score marks the 90th percentile?
- Percentile to z-score: the 90th percentile of the standard normal curve is z = 1.2816.
- Raw score: X = 100 + 1.2816 × 15 = 100 + 19.22 = 119.22.
Worked Example 3: Finding the Standard Deviation
A raw score of 48 has a z-score of −4 and the mean is 60. Then σ = (X − μ) ÷ z = (48 − 60) ÷ (−4) = −12 ÷ −4 = 3.
Common Standardized Scales
Many reported scores are z-scores rescaled to a convenient mean and standard deviation. The raw score is the number before that rescaling.
| Scale | Mean | Std. dev. | z = −2 | z = 0 | z = +2 |
|---|---|---|---|---|---|
| Z-score | 0 | 1 | −2 | 0 | 2 |
| T-score | 50 | 10 | 30 | 50 | 70 |
| IQ scale | 100 | 15 | 70 | 100 | 130 |
To recover the raw score from a T-score or IQ score, first convert it back to a z-score, or simply enter the scale's mean and standard deviation here. The quick-scale menu does the first step for you.
Z-Scores and Percentiles
For normally distributed data, each z-score matches a fixed share of values at or below it.
| Z-score | −3 | −2 | −1 | 0 | 1 | 2 | 3 |
|---|---|---|---|---|---|---|---|
| Percentile | 0.13 | 2.28 | 15.87 | 50 | 84.13 | 97.72 | 99.87 |
About 68% of normal data lies within one standard deviation of the mean, about 95% within two, and about 99.7% within three. A z-score beyond ±3 is rare, so check that the inputs belong to the same data set.
Empirical rule reference: Penn State STAT 200 — Empirical Rule.
Frequently Asked Questions
What is a raw score?
A raw score is the original value of a measurement before it is transformed or standardized. For an exam score, it is the number of earned points out of the maximum score, such as 42 of 50 total points, before any percentage or grade is worked out. On its own it does not show how good the result is; you need the mean and standard deviation for that.
How do I find a raw score from a z-score?
Multiply the z-score by the standard deviation and add the mean (the average of the dataset): X = z × σ + μ. With z = −4, σ = 3, and μ = 60, the raw score is −4 × 3 + 60 = 48. The input values must come from the same dataset, and the output is in the original units of that data.
Can I find a raw score from a percentile?
Yes, if the data are roughly normal. First convert the percentile to a z-score with the inverse of the standard normal distribution, then apply X = μ + zσ. The percentile option in this calculator does both steps. For the 90th percentile, z is about 1.2816. Because a percentile is a ranking, this also shows what score would put a student in the top 10%.
Why must the sign of the z-score match the raw score?
The sign shows which side of the mean the score is on. A raw score below the mean always has a negative z-score, and one above the mean has a positive z-score. A negative z-score only describes position, not good or bad performance. If you are solving for the standard deviation and the result comes out negative, the signs of X − μ and z do not agree, so recheck the observation you started from.
What is the difference between a raw score and a standard score?
A raw score is in the original units of the data. A standard score, such as a z-score, T-score, or IQ score, expresses the same result relative to the mean and standard deviation, so scores from different tests can be compared. Standard scores are sometimes loosely called normalized scores. The conversion is a straight-line formula that works in either direction, and it lets you put a quiz and an exam on the same scale.
How do I find the standard deviation from a raw score and z-score?
Rearrange the formula to σ = (X − μ) ÷ z. For X = 48, μ = 60, and z = −4, σ = (48 − 60) ÷ (−4) = 3. This only works when z is not zero, because a z-score of 0 means X equals the mean for any standard deviation. It is a common back-calculation in statistics when a report gives the score and z-score but not the spread.
Can a raw score be negative?
Yes. The formula does not limit X. A negative raw score is possible when the mean is small compared with the spread, as with temperatures below zero or profit and loss figures. For an assessment, quiz, or test score on a 0 to 100 scale, though, a negative result means the mean, standard deviation, or z-score is wrong.
Do the mean and standard deviation have to come from the same data set?
Yes. The z-score only means something relative to the mean and standard deviation it was calculated from. Mixing values from different groups or tests gives a raw score that fits neither, and any evaluation built on it will be off. Use the population values if you know them, or the sample values if you do not.
Other Useful Calculators
Empirical Rule Calculator
The 68-95-99.7 rule for normal data.
Mean Absolute Deviation Calculator
Average distance of each value from the mean.
Residual Calculator
Observed minus predicted for a regression line.
Correlation Coefficient Calculator
Pearson, Spearman, and Kendall with a regression line.
Lognormal Distribution Calculator
Probabilities and percentiles for skewed data.
Process Capability Index Calculator
Cp, Cpk, and PPM against tolerance limits.