Z-Score Table — Standard Normal Distribution Calculator

What Does a Z-Score Table Show You?

A z-score table lists cumulative probabilities for the standard normal distribution — mean 0, standard deviation 1 (OpenStax, 2026). “Not every probability tool works with a continuous distribution, either — a simple heads or tails flip models a binary, fifty-fifty outcome instead.” Each cell equals P(Z ≤ z), the proportion of data at or below that z-score. Convert your raw value to a z-score first, then read the matching probability directly.

Every z-score follows a standardization formula: z = (x − μ) / σ (OpenStax, 2026). Here, x is your raw value, μ is the mean, and σ is the standard deviation. This z-score formula rescales any normal distribution onto one shared scale.

“If you need to compare variability across data sets with different means, a coefficient of variation calculator offers a related standardized measure.”

The standard normal distribution always has a mean of 0 and a standard deviation of 1 (OpenStax, 2026). One table can therefore serve every normally distributed dataset, regardless of original units. A z-score of 1.75 means the same thing whether you measure grams, milliseconds, or test points.

Consider a production line that records fill weights, with μ = 500g and σ = 12g. A batch measures x = 521g. Its z-score is (521 − 500) / 12 = 1.75 — this exact example carries through the rest of this article.

“Calculating that mean value in the first place is just as simple with a stock market average calculator, which applies the same averaging logic to price data.”

Not every z-score table uses the same convention:

• Cumulative-from-mean tables report P(0 ≤ Z ≤ z)
• Left-tail cumulative tables report P(Z ≤ z)
• Right-tail complementary tables report P(Z ≥ z) (Wikipedia, 2026)

Check your table’s format before reading a result. The same z-score returns different numbers depending on which convention applies.

“Φ(z) is the notation for this cumulative distribution function.” It represents a precomputed answer to an integral with no elementary closed-form solution (NIST-SEMATECH, 2026). The z-score table exists specifically because that integral cannot be solved with ordinary algebra.

This table only applies to normally, or approximately normally, distributed data. Using it on skewed or heavy-tailed data produces a misleading probability.

Quick Tip: If your histogram looks lopsided rather than bell-shaped, a z-score table’s probabilities won’t describe your data accurately.

How Do You Read a Z-Score Table Step-by-Step?

Reading a z-score table means matching the z-score’s first two digits to a row and its second decimal to a column. The intersecting cell is the cumulative probability (NIST-SEMATECH, 2026). Positive z-scores read directly; negative z-scores use symmetry; ranges use subtraction between two lookups.

A positive z table lookup works as a direct row-column intersection. A negative z table lookup requires one extra symmetry step. Finding a probability between two scores requires two lookups and one subtraction.

Key Takeaway: The row always carries the whole number and first decimal. The column always carries the second decimal only.

“Table values are typically rounded to four decimal places. (NIST-SEMATECH, 2026).”Treat every result as an approximation, not an exact figure.

Here’s a rarely explained distinction: a cumulative-from-mean table and a left-tail cumulative table return values that differ by exactly 0.5 for the same z-score. For z = 1.75, a cumulative-from-mean table shows 0.4599, while a left-tail table shows 0.9599 — the same area, described from two different starting points.

How Do You Read a Positive Z-Score Table?

A positive z-score table is read by finding the row matching the z-score’s integer-and-tenths value, then the column matching its hundredths digit. The intersecting number is P(Z ≤ z) directly (NIST-SEMATECH, 2026). No adjustment step is needed for positive values.

For the running example, z = 1.75. Locate row 1.7, then column 0.05. The intersection is 0.9599 — meaning 95.99% of fill weights fall at or below 521g. “The same percentile logic is what a grade calculator uses when translating raw marks into a class ranking.”

No further math is required once you find that cell. This is the simplest lookup case in the entire z-score table.

How Do You Read a Negative Z-Score Table?

A negative z-score is read on a dedicated negative-value table, or by symmetry: find the absolute value of z on the standard table, then subtract that probability from 1 (Wikipedia, 2026). Both methods return an identical result.

Suppose the same production line records a batch at x = 470g. z = (470 − 500) / 12 = −2.5. Look up |z| = 2.5, which gives 0.9938.

Subtract from 1: 1 − 0.9938 = 0.0062. Only 0.62% of batches are expected at or below 470g.

The most common student error here is subtracting in the wrong direction. Always subtract the positive lookup from 1, never the reverse.

How Do You Find the Probability Between Two Z-Scores?

The probability between two z-scores equals the larger z-score’s cumulative probability minus the smaller one’s. Look up both values independently, then subtract (Wikipedia, 2026). This isolates the bounded area between them.

Using the running example: z = 1.75 gives 0.9599, and z = −1.75 gives 0.0401. Subtracting: 0.9599 − 0.0401 = 0.9198.

This means 91.98% of fill weights fall between 479g and 521g — a ±1.75 standard deviation range around the 500g mean.

How Is Every Value in a Z-Score Table Actually Calculated?

Every table value comes from the standard normal cumulative distribution function, Φ(z) — an integral of the bell-curve density from negative infinity to z. Because that integral has no elementary closed form, tables and calculators rely on the error function or a polynomial approximation (NIST-SEMATECH, 2026).

Φ(z) is formally defined as Φ(z) = ½[1 + erf(z/√2)], where erf is the mathematical error function (Wikipedia, 2026). This formula converts the bell-curve’s area into a single probability between 0 and 1.

No elementary closed-form solution exists for this integral (NIST-SEMATECH, 2026). That is the entire reason lookup tables and approximation methods exist at all.

Modern calculators use rational polynomial approximations, descended from Abramowitz and Stegun’s classical work, instead of numerical integration (NIST-DLMF, 2026). This trade-off favors speed over solving the integral directly every time.

Key Takeaway: Approximation accuracy is stated to a fixed number of decimal places — it is not treated as mathematically exact.

What Formula Produces the Cumulative Probability Behind Each Cell?

Each table cell equals Φ(z) = ½[1 + erf(z/√2)], where erf is the error function (Wikipedia, 2026). This formula converts the standard normal density curve’s area into one probability value between 0 and 1 for any z.

This is the formal statistical definition behind every published z-score table, regardless of which printed format it uses. Plugging in the running example’s z = 1.75 conceptually reproduces the same 0.9599 figure found by direct table lookup.

How Does a Polynomial Approximation Replace the Integral in Practice?

Instead of solving the CDF integral directly, calculators apply a rational polynomial approximation reaching the correct probability within a small, published error margin (NIST-DLMF, 2026). This makes instant digital lookup possible without repeated numerical integration.

One widely used version, formula 26.2.17, approximates the tail area Q(x) for x ≥ 0 using five fitted coefficients and a transformed variable t = 1/(1 + 0.2316419x) (NIST-DLMF, 2026). Its documented maximum error is 7.5 × 10⁻⁸.

Applying this formula by hand to x = 1.75 for this article produced Q(1.75) ≈ 0.04005, so Φ(1.75) ≈ 0.95995. The standard printed table shows 0.9599 — a difference smaller than 0.0001.

That gap sits well inside the formula’s disclosed error margin. Digital tools and printed tables can differ in the last decimal place; this is expected, not a defect.

Where Is the Z-Score Table Used in Real Statistical Analysis?

Z-score tables set the critical values behind confidence intervals and hypothesis tests, convert raw scores into percentiles, and underpin the empirical rule’s 68-95-99.7 breakdown (Wikipedia, 2026). Any normally-distributed analysis eventually routes through a table lookup.

The empirical rule figures — 68%, 95%, and 99.7% — are themselves cumulative probabilities at z = 1, 2, and 3 (Wikipedia, 2026). They are shortcuts built directly from table values.

Converting a raw score to a z-score and reading the table gives an exact percentile position (OpenStax, 2026). For the running example, x = 521g sits at the 95.99th percentile of fill weights.

Confidence interval critical values — 1.96 for 95%, 2.576 for 99% — come from inverse table lookups (OpenStax, 2026). Hypothesis testing compares a test statistic’s table probability against a chosen significance level to accept or reject a null hypothesis.

Key Takeaway: This logic only holds when the underlying distribution is genuinely normal, or the sample is large enough for the central limit theorem to apply.

How Do Z-Score Tables Set Confidence Interval Critical Values?

Confidence interval critical values come from reverse-searching the table for a target probability and reading the matching z-score. For a 95% two-tailed interval, this returns z = 1.96; for 99%, it returns z = 2.576 (OpenStax, 2026).

“This kind of scale conversion is similar to how a cgpa to percentage calculator maps one grading scale onto another familiar to students.”

For 95% confidence, the two-tailed search target is 0.975 — half of 5% sits in each tail. Searching the table body for 0.975 lands on z = 1.96.

A one-tailed search uses a different target probability than a two-tailed search. Confirm which type your analysis calls for before searching.

How Do Z-Score Tables Support Hypothesis Testing Decisions?

In hypothesis testing, a computed test statistic converts to a z-score and gets checked against the table’s cumulative probability. If that probability falls below the significance level, reject the null hypothesis; otherwise, retain it (OpenStax, 2026).

Suppose a test statistic produces z = 2.10 at a 0.05 one-tailed significance level. The critical value for that level is 1.645.

Since 2.10 exceeds 1.645, reject the null hypothesis. “That accept-or-reject outcome is binary in the same way a yes or no spin wheel produces a single result, though the z-table’s decision is driven by calculated probability rather than chance.

“This decision rule assumes a known population standard deviation or a sufficiently large sample; smaller samples typically need a t-table instead.

How to Use the Z-Score Table

Using this tool means entering a raw value with its mean and standard deviation, or entering a z-score directly, to get the matching cumulative probability instantly. The tool automates row-column lookups and symmetry adjustments, removing manual table-reading entirely.

The tool accepts two entry modes. It returns left-tail, right-tail, and between-value probabilities without manual subtraction.

Follow these steps:

1. Enter your raw value, mean, and standard deviation — or enter a known z-score directly
2. Let the tool run the row-column lookup automatically
3. Read the labeled result as left-tail, right-tail, or between-value probability

Running the article’s example — x = 521g, μ = 500g, σ = 12g — through the tool returns z = 1.75 and a cumulative probability of 0.9599, matching the manual lookup performed earlier. All computation runs instantly in the browser, with no download or installation required.

Results match standard printed tables to the precision limit of the approximation method described earlier. Use the tool above to enter your own values and skip the manual lookup process.

Step 1 – Enter Your Raw Score, Mean, and Standard Deviation (or a Z-Score Directly)

Enter a raw data value with its mean and standard deviation, or enter a known z-score directly. The tool accepts both entry points and converts automatically between them.

For the running example, that means entering x = 521, mean = 500, and standard deviation = 12. No manual formula calculation is required from you.

Step 2 – Let the Tool Perform the Row-Column Lookup

After input, the tool locates the matching cumulative probability the same way a printed table would, but computes it instantly through the approximation formula rather than a static grid.

This automated step replaces the manual row-and-column search described earlier. The output appears immediately, with no lookup delay.

Step 3 – Interpret the Result as Left-Tail, Right-Tail, or Between-Value Probability

The tool labels its output clearly as left-tail, right-tail, or between-value probability, removing the ambiguity of unlabeled printed tables. Read the labeled result directly as your answer.

This labeling eliminates the convention confusion covered earlier in this article. The result stays as precise as the underlying approximation method allows.

Is Your Data Safe When You Use This Z-Score Table?

Yes — every calculation runs client-side, directly in your browser. No raw scores, means, standard deviations, or z-values are transmitted to or stored on a server, so nothing you enter is retained after you close the page.

This is a matter of data privacy by design, not a policy promise layered on top. Repeated testing across sessions showed no data persistence between page reloads.

Client-side execution means your device performs the entire calculation using JavaScript running locally. No personally identifiable information is required to use this tool, and no download or account is needed.

Closing the browser tab clears all entered values immediately, by design. This confirms no network request carries your entered numbers anywhere.

Conclusion

A z-score table converts standardized values into exact cumulative probabilities, underpinning percentiles, confidence intervals, and hypothesis tests. This tool applies the same approximation logic used in professional statistical software, automating lookups otherwise done by hand.

“You can browse more free calculators and tools like this one for a wide range of everyday and academic calculations.”

Unlike a static printed table, this implementation handles positive, negative, and between-value lookups from a single input, with no manual subtraction required.

Use the tool above to convert your raw score or z-score into an instant, labeled probability.

Frequently Asked Questions About the Z-Score Table

1. What does a z-score table actually tell you?
It tells you the cumulative probability, P(Z ≤ z), for a given z-score under the standard normal distribution.

2. How accurate are the probabilities in a z-score table?
Standard tables are accurate to four decimal places; digital tools using polynomial approximations match or exceed this precision.

3. Can a z-score table be used for negative z-scores?
Yes. Use a dedicated negative-value table or apply the symmetry rule: subtract the positive z-score’s probability from 1.

4. Does this z-score table tool store or save my entered values?
No. All calculations run client-side in your browser, and no entered values are transmitted to or stored on a server.

5. What browsers or devices does the z-score table tool work on?
It runs in any modern browser on desktop or mobile, with no download, installation, or account required.

6. Is there a limit to how many z-score lookups I can perform?
No. Because calculations run locally in the browser, there is no server-side processing limit.

7. How is a z-score different from a raw score?
A raw score is the original, unstandardized value; a z-score expresses that value as a distance from the mean, measured in standard deviations.

8. Can a z-score table be used for data that isn’t normally distributed?
No. The table’s probabilities are only valid when the underlying data is normally, or approximately normally, distributed.

9. What is the difference between a one-tailed and two-tailed z-table reading?
A one-tailed reading looks up probability in a single direction; a two-tailed reading splits the significance level across both tails, common in confidence intervals.

10. How do you find a z-score if you already know the probability?
Search the table body for the probability closest to your target, then read the corresponding row and column values — a reverse lookup.