Distribution Type
Z-Distribution
Standard Normal (σ known)
T-Distribution
Student's t (σ unknown)
Calculation Mode
Result
—
Step-by-Step Calculation
How to Use This Tool
Complete walkthrough for every calculation mode
Choose Distribution Type
Select Z-Distribution when the population standard deviation is known and your sample is large (n ≥ 30). Use T-Distribution when σ is unknown and you're working with a small sample.
Pick a Calculation Mode
Score → P-value converts a test statistic to a probability. P → Critical Score does the reverse. CI builds a confidence interval. Sample Stats lets you enter raw data.
Enter Your Values
Fill in the required fields shown for your selected mode. Each field has a short hint describing the expected value. Required fields are marked with a red asterisk (*).
Select Tail Type
Choose Two-tailed for ≠ hypotheses, Left-tailed for < hypotheses, and Right-tailed for > hypotheses. This affects how area under the curve is calculated.
Interpret Results
Your highlighted result appears with supporting stats. Scroll down to see the distribution curve, the interpretation statement, and the full step-by-step calculation breakdown.
Save & Compare
Every calculation is auto-saved to History. Click any history item to reload it. Use the Reset button to clear all fields and start fresh.
Frequently Asked Questions
When should I use Z vs T?
Use the Z-distribution when: (1) the population standard deviation σ is known, or (2) your sample size n ≥ 30. Use the T-distribution when σ is unknown and n < 30. For very large samples, both distributions converge, but the T is always the safer choice when σ is estimated.
What are degrees of freedom?
Degrees of freedom (df) represent the number of independent values in a calculation that are free to vary. For a one-sample t-test, df = n − 1. As df increases, the t-distribution approaches the standard normal (Z) distribution. More df = narrower, taller distribution tails.
What is a p-value?
A p-value is the probability of obtaining results at least as extreme as the observed results, assuming the null hypothesis is true. A small p-value (typically p < 0.05) suggests strong evidence against H₀. It does NOT measure the probability that H₀ is true.
What is a confidence interval?
A confidence interval (CI) is a range of values likely to contain the true population parameter. A 95% CI means: if you repeated the study 100 times, ~95 of the resulting intervals would contain the true parameter. It is NOT a 95% probability that the true value is in this specific interval.
How is the standard error calculated?
Standard Error (SE) = σ / √n (for Z-test) or s / √n (for T-test), where σ or s is the standard deviation and n is the sample size. SE measures the precision of the sample mean as an estimate of the population mean — smaller SE means more precise estimates.
Worked Examples
Click any row to load it into the calculator
Z-Distribution Examples
| Scenario | Inputs | Result | Action |
|---|---|---|---|
| IQ Score Is IQ 130 unusual? |
z = 2.0, Two-tailed |
p = 0.0455 | |
| Quality Control Defect rate test |
x̄=102, μ=100, σ=5, n=25 |
z = 2.0, p = 0.0228 | |
| Critical Z at α=0.05 Two-tailed test |
α = 0.05, Two-tailed |
z* = ±1.96 | |
| 95% Confidence Interval Sample mean estimation |
x̄=50, σ=10, n=100 |
[48.04, 51.96] |
T-Distribution Examples
| Scenario | Inputs | Result | Action |
|---|---|---|---|
| Drug Trial Before/after comparison |
t = 2.5, df = 14 |
p = 0.0253 | |
| Small Sample Mean Test n=10, unknown σ |
x̄=22, μ=20, s=3, n=10 |
t = 2.108, p = 0.064 | |
| Critical t at α=0.05 df = 20, Two-tailed |
α = 0.05, df = 20 |
t* = ±2.086 | |
| 95% CI for Small Sample n=15, unknown σ |
x̄=30, s=4, n=15 |
[27.78, 32.22] |
Common Critical Values Reference
| Confidence Level | α (two-tailed) | Z critical | t (df=10) | t (df=30) |
|---|---|---|---|---|
| 90% | 0.10 | ±1.645 | ±1.812 | ±1.697 |
| 95% | 0.05 | ±1.960 | ±2.228 | ±2.042 |
| 98% | 0.02 | ±2.326 | ±2.764 | ±2.457 |
| 99% | 0.01 | ±2.576 | ±3.169 | ±2.750 |
| 99.9% | 0.001 | ±3.291 | ±4.587 | ±3.646 |
Fascinating Facts
Key statistical insights about Z & T distributions
68%
of data falls within ±1σ of the mean in a normal distribution
95%
of data falls within ±1.96σ — the basis for 95% CI
99.7%
of data within ±3σ — the famous "three-sigma rule"
1908
Year William Gosset published the t-distribution (as "Student")
∞ df
At infinite df, the t-distribution becomes the Z-distribution
0.05
The conventional significance threshold α used in most sciences
±1.96
The Z critical values for a two-tailed test at α = 0.05
CLT
Central Limit Theorem: why the normal distribution is so universal
Did You Know?
Why "Student's" t-distribution?
William Sealy Gosset, while working as a statistician at Guinness Brewery, discovered the t-distribution in 1908. Guinness did not allow employees to publish research, so Gosset published under the pseudonym "Student" — hence the name "Student's t-distribution."
The normal distribution in nature
The normal distribution appears everywhere in nature: heights of people, measurement errors, SAT scores, IQ scores, blood pressure readings, and even the thermal noise in electronics. This ubiquity is explained by the Central Limit Theorem, which states that averages of large numbers of independent random variables tend toward normality regardless of the underlying distribution.
Heavier tails of the t-distribution
The t-distribution has heavier tails than the normal distribution. This reflects greater uncertainty when estimating population parameters from small samples. As sample size increases (df → ∞), the t-distribution converges to the standard normal. This is why t-tests are more conservative than z-tests — they require more extreme test statistics to achieve the same p-value.
Expert Tips
Get the most accurate and meaningful results
- Check normality first. Both Z and T tests assume the underlying population is approximately normal. For small samples, verify normality with a Shapiro-Wilk test or Q-Q plot before proceeding.
- Always specify your tail direction before collecting data. Choosing one-tailed vs two-tailed after seeing your data (p-hacking) invalidates your results. Decide based on your research hypothesis.
- Sample size matters enormously. With n ≥ 30, the Z and T results are nearly identical. For n < 10, the t-distribution's heavier tails can significantly affect your p-values and CI width.
- Statistical significance ≠ practical significance. A very large sample can yield a statistically significant result (p < 0.05) even for a trivially small effect. Always compute effect size (Cohen's d) alongside your p-value.
- Use the population σ (not sample s) for Z-tests. Using sample standard deviation in a Z-test underestimates variability and inflates your test statistic. If σ is unknown, always use the T-test.
- Wider confidence intervals are more honest. A 99% CI is wider than a 95% CI because we're more confident. Don't default to 95% — choose the confidence level appropriate for your decision's stakes.
- Report exact p-values, not just "significant/not." Reporting p = 0.032 is more informative than "p < 0.05". Readers can apply their own significance thresholds, and effect size is easier to judge from exact values.
- Use the History tab to compare calculations. When testing multiple hypotheses, save each calculation to history and compare p-values. Remember to apply Bonferroni correction when making multiple comparisons.
Real-World Use Cases
Where Z and T distributions are applied professionally
01
Clinical Trials
Pharmaceutical researchers use t-tests to compare treatment and control groups when sample sizes are small. The t-distribution's heavier tails account for the added uncertainty of small-sample drug efficacy studies.
02
Quality Control
Manufacturing engineers apply Z-tests to monitor production processes. When a product dimension falls outside ±3σ (Z > 3), the machine may need recalibration — a direct application of the three-sigma rule.
03
Psychology Research
Behavioral scientists use t-tests to compare group means (e.g., test vs. control in a cognitive study). Small samples are common in psychology, making the t-distribution the standard tool.
04
Finance & Risk
Risk analysts use Z-scores to detect anomalous returns and identify outliers. The Z-score of a stock return tells you how many standard deviations it is from historical mean — crucial for VaR calculations.
05
Education Testing
Standardized test designers use Z-scores to normalize scores across test versions. IQ scores (mean 100, SD 15) are Z-scores transformed for interpretability — Z = 2.0 corresponds to IQ ≈ 130.
06
A/B Testing
Product teams use Z-tests (large-sample) or t-tests (small-sample) to determine whether version A or B of a feature performs significantly better. The p-value guides rollout decisions.
07
Environmental Science
Scientists use t-tests to compare pollution levels before and after an intervention. With limited monitoring stations, small-sample t-tests provide statistically valid inferences about environmental change.
08
Economics & Surveys
Economists use confidence intervals (Z or T) to report survey-based estimates like unemployment rates or consumer sentiment indices. The CI communicates the inherent uncertainty in the estimate.
About This Tool
What it is, how it works, and what makes it special
Professional-Grade Statistics, Free & Instant
This tool provides a complete environment for working with the two most fundamental probability distributions in inferential statistics — the Z (standard normal) and Student's T distributions. It is designed for researchers, students, analysts, and engineers who need fast, accurate, and explainable statistical calculations without needing to install specialized software.
100% browser-based — no data leaves your device
Four calculation modes: Score↔P-value, Critical Values, Confidence Intervals, Sample Stats
Interactive SVG distribution curves with shaded regions
Full step-by-step calculation breakdowns for educational transparency
Calculation history — compare and revisit past results
High-precision algorithms using Horner's method & iterative t-approximation
Algorithm Precision
Uses Abramowitz & Stegun polynomial approximations for the normal CDF (max error < 7.5×10⁻⁸) and a regularized incomplete beta function for the t-distribution CDF. Results are consistent with standard statistical tables and R/Python outputs.
Privacy First
All calculations run entirely in your browser using JavaScript. No data is transmitted to any server. No cookies are set. Your statistical inputs remain completely private and confidential.
Fully Responsive
Optimized for desktop, tablet, and mobile devices. The adaptive grid layout and fluid typography ensure a great experience on any screen size, from a 320px phone to a 4K monitor.
Calculation History
Your recent calculations — click any item to reload
Recent Calculations
No calculations yet.
Run a calculation to see it here.
Run a calculation to see it here.
| Feature | Details |
|---|---|
| Price | Free |
| Rendering | Client-Side Rendering |
| Language | JavaScript |
| Paywall | No |
0 Comments