BY: Statistics Fundamentals Team
Reviewed By: Minsa A (Senior Statistics Editor)

Free Chebyshev's Theorem Calculator

Calculate the minimum probability that any dataset falls within k standard deviations of the mean, for any distribution shape. Enter k (and optionally your mean and standard deviation) to get the Chebyshev bound, the data interval, and a full step-by-step solution — no signup required.

Chebyshev's Theorem Calculator

Formula P(|X − μ| < kσ) ≥ 1 − 1/k² Requires k > 1
Must be greater than 1
Interval (μ − kσ, μ + kσ) Min. coverage 1 − 1/k²

The table below compares Chebyshev's guaranteed lower bounds with the Empirical Rule's exact percentages (which apply only to normally distributed data). Chebyshev's bounds are always more conservative.

k (std deviations) Chebyshev Min. Bound Empirical Rule (Normal only) Gap (Conservative margin)
k = 1 No guarantee (k must be > 1) 68%
k = 1.41 At least 50% ≈ 84% (normal) 34% gap
k = 2 At least 75% 95% (Empirical Rule) 20% gap
k = 3 At least 88.9% 99.7% (Empirical Rule) 10.8% gap
k = 4 At least 93.75% ≈ 99.994% (normal) 6.2% gap
k = 5 At least 96% ≈ 99.99994% (normal) 4% gap
k = 10 At least 99% ≈ 100% (normal) 1% gap

The gap closes as k increases. For large k, even Chebyshev's bound becomes a very strong guarantee. The Empirical Rule is exact only for normal distributions; Chebyshev's result is universal.

Run a calculation in the Basic or Interval tab first, then return here to see the full step-by-step solution.

No data yet — enter a k value in the Basic or Interval tab first.

What is Chebyshev's Theorem? Chebyshev's Theorem states that for any dataset with a finite mean and standard deviation, at least 1 − 1/k² of the data values lie within k standard deviations of the mean, where k > 1. This holds for every probability distribution, regardless of shape. For example, at k = 2, at least 75% of any dataset falls within 2 standard deviations of the mean; at k = 3, at least 88.9% does.

What Is Chebyshev's Theorem?

Chebyshev's Theorem (also called Chebyshev's Inequality or Bienaymé–Chebyshev Inequality) is one of the most useful results in probability theory. It gives a guaranteed lower bound on the proportion of data that must fall within any specified number of standard deviations from the mean — no matter what shape the distribution takes.

The theorem was first stated by the French mathematician Irénée-Jules Bienaymé in 1853 and later refined and proved by Russian mathematician Pafnuty Chebyshev in 1867. It is studied in AP Statistics, introductory college statistics, and data science courses precisely because it applies universally, not just to bell-shaped distributions.

The Chebyshev's Theorem Formula

For a random variable X with finite mean μ and standard deviation σ, and for any real number k > 1:

Core Inequality

P(|X − μ| < kσ) ≥ 1 − 1/k²

Outside Probability

P(|X − μ| ≥ kσ) ≤ 1/k²

Chebyshev Interval

(μ − kσ, μ + kσ)

Minimum % Coverage

(1 − 1/k²) × 100%

The critical point: 1 − 1/k² is a lower bound, not an exact probability. Any real dataset may have more than this proportion within k standard deviations. Chebyshev's Theorem guarantees the minimum; the actual percentage can only be higher.

Quick Reference: Minimum Coverage by k Value

k=2
At least 75%
1 − 1/4
k=3
At least 88.9%
1 − 1/9
k=4
At least 93.75%
1 − 1/16
k=5
At least 96%
1 − 1/25

How to Use This Chebyshev's Theorem Calculator

1
Choose your calculator mode

Use the Basic (k only) tab to find the minimum probability bound from k alone. Use the Interval tab when you also know the mean (μ) and standard deviation (σ) and want the exact data interval.

2
Enter k (required for both modes)

k is the number of standard deviations from the mean. It must be greater than 1. Common exam values are k = 2, k = 3, and k = 4.

3
Enter μ and σ for the Interval mode

If you know the mean and standard deviation of your dataset, the Interval tab computes the actual data range (μ − kσ, μ + kσ) and draws a number line diagram.

4
Read the minimum bound and interpret it

The result tells you the guaranteed minimum proportion. For example, k = 2 gives 75%: at least 75% of your data — regardless of distribution shape — falls within 2 standard deviations of the mean.

5
Check the Step-by-Step tab for the full solution

After calculating, switch to the Step-by-Step tab to see every calculation laid out in order — useful for verifying homework or understanding the method before an exam.

Worked Example: Step-by-Step Calculation

Scenario: A factory produces bolts. The mean diameter is μ = 50 mm with a standard deviation of σ = 2 mm. What minimum percentage of bolts have a diameter within 3 standard deviations of the mean? What is the corresponding interval?
Identify the given values

μ = 50 mm, σ = 2 mm, k = 3

Calculate k²

k² = 3² = 9

Apply the Chebyshev formula

Min. proportion = 1 − 1/k² = 1 − 1/9 = 8/9 ≈ 0.8889 = 88.89%

Calculate the Chebyshev interval

Lower = μ − kσ = 50 − 3×2 = 44 mm
Upper = μ + kσ = 50 + 3×2 = 56 mm

Interpret the result

At least 88.89% of all bolts produced have a diameter between 44 mm and 56 mm, regardless of how the diameters are distributed. This holds even if the diameter distribution is skewed or irregular.

Result: Using k = 3, μ = 50, σ = 2: at least 88.89% of bolts fall in the interval (44, 56) mm. You can verify this result using the Interval tab of the calculator above.

🧠 The GMC Framework: Understanding Chebyshev's Theorem Without Heavy Math

The GMC Framework is a memory structure for the three things you need to remember about Chebyshev's Theorem. It is designed for students, data analysts, and anyone who needs to apply the theorem correctly under exam or workplace conditions.

G
Guaranteed
The bound 1 − 1/k² is guaranteed for every distribution with finite mean and variance. It does not depend on any shape assumption. This is Chebyshev's power.
M
Minimum
The bound is a minimum, not the exact answer. Real data may contain a higher percentage within the interval. Chebyshev's result is a floor, not a ceiling.
C
Conservative
Compared to the Empirical Rule, Chebyshev's bounds are conservative. The trade-off: they apply universally, while the Empirical Rule requires a normal distribution.
Intuitive Analogy: Think of Chebyshev's Theorem like a legal minimum wage. The law guarantees a floor; actual pay at any given company may be higher. Chebyshev guarantees a floor for data concentration around the mean; the actual concentration in your dataset may — and usually does — exceed this floor.

Chebyshev's Theorem vs. the Empirical Rule

The Empirical Rule (68-95-99.7 Rule) and Chebyshev's Theorem both describe how data clusters around the mean, but they serve different purposes and apply under different conditions. Understanding when to use each is a key skill in statistics.

Table: Chebyshev's Theorem vs. Empirical Rule — Complete Comparison

PropertyChebyshev's TheoremEmpirical Rule (68-95-99.7)
Distribution requiredAny distribution (normal or not)Normal (bell-shaped) only
At k = 1 (±1σ)No guarantee (k must be > 1)Exactly 68%
At k = 2 (±2σ)At least 75%Exactly 95%
At k = 3 (±3σ)At least 88.9%Exactly 99.7%
Type of resultLower bound (minimum guarantee)Exact percentage
Best use caseUnknown or skewed distributionsConfirmed normal data
AP StatisticsRequired topicRequired topic

The rule of thumb: use the Empirical Rule when you have confirmed your data is approximately normal (e.g., via a histogram, Q-Q plot, or normality test). Use Chebyshev's Theorem when the distribution is unknown or clearly non-normal. The Empirical Rule guide on Statistics Fundamentals covers the normal-distribution case in depth.

📊 Worked Case Studies

Case Study 1 — Manufacturing Quality Control

Scenario: A machine fills bags of flour. Mean weight μ = 1,000 g, standard deviation σ = 8 g. The distribution of weights is unknown. What percentage of bags weigh between 984 g and 1,016 g?
Find k

Lower bound = 1,000 − 8k = 984 ⇒ 8k = 16 ⇒ k = 2

Apply Chebyshev

Min. proportion = 1 − 1/4 = 0.75 = 75%

Verify the interval

(μ − 2σ, μ + 2σ) = (1000−16, 1000+16) = (984, 1016) ✓

Interpretation: At least 75% of flour bags weigh between 984 g and 1,016 g, regardless of the weight distribution's shape. If the distribution were normal, the Empirical Rule would give 95% — but since the shape is unknown, 75% is the guaranteed lower bound.

Case Study 2 — Student Exam Scores

Scenario: A statistics class exam has a mean score of μ = 72 and a standard deviation of σ = 9. The score distribution is right-skewed (some students scored very low, pulling the tail). Using Chebyshev's Theorem, what proportion of students scored between 45 and 99?

Find k: The interval (45, 99) is symmetric around 72. Distance from mean = 72 − 45 = 27. So k = 27/9 = 3.

Apply Chebyshev: 1 − 1/9 = 8/9 ≈ 88.9%.

Result: At least 88.9% of students scored between 45 and 99. Since the distribution is skewed, the Empirical Rule cannot be applied here — Chebyshev's Theorem is the right tool. The actual percentage exceeds 88.9%, but this is the guaranteed minimum.

Case Study 3 — Financial Risk Analysis

Scenario: A stock has a mean daily return of μ = 0.05% with a standard deviation of σ = 1.2%. Returns are not normally distributed (they show fat tails). What is the minimum percentage of trading days when returns fall between −4.75% and +4.85%?

Find k: The interval extends approximately 4.8 percentage points on each side of the mean. k = 4.8 / 1.2 = 4.

Apply Chebyshev: 1 − 1/16 = 15/16 = 93.75%.

Practical takeaway: At least 93.75% of trading days produce a return within that range. In risk management, this type of universal bound is valuable precisely because financial return distributions have fat tails and are not normal — exactly the conditions where the Empirical Rule fails and Chebyshev's Theorem remains valid. This application connects to Value at Risk (VaR) methodology discussed by the Bank for International Settlements.

Case Study 4 — Healthcare: Blood Pressure Measurements

Scenario: A clinical study records systolic blood pressure for a mixed patient population. Mean μ = 125 mmHg, SD σ = 15 mmHg. Distribution is unknown. What interval contains at least 96% of all readings?

Find k from the bound: We need 1 − 1/k² ≥ 0.96, so 1/k² ≤ 0.04, k² ≥ 25, k ≥ 5.

Calculate interval: (μ − 5σ, μ + 5σ) = (125 − 75, 125 + 75) = (50, 200) mmHg.

Result: To guarantee at least 96% coverage in any distribution, the interval must extend to 5 standard deviations from the mean. This is much wider than the 95% interval a normal distribution would require (just ±2σ), reflecting the cost of making no shape assumptions.

📊 Chebyshev Bounds vs. Actual Data — Benchmark Analysis

Chebyshev's Theorem gives a lower bound that is always true, but for specific distributions the actual coverage percentage is often much higher. The table below shows how Chebyshev's conservative lower bounds compare with actual theoretical coverage for several well-known distributions at k = 2 and k = 3.

Table: Chebyshev Bound vs. Actual Coverage for Common Distributions

DistributionChebyshev at k=2Actual at k=2Chebyshev at k=3Actual at k=3
Normal≥ 75%95.45%≥ 88.9%99.73%
Uniform≥ 75%100% (finite range)≥ 88.9%100%
Exponential≥ 75%≈ 98.2%≥ 88.9%≈ 99.97%
Poisson (λ=5)≥ 75%≈ 97.4%≥ 88.9%≈ 99.8%
t-distribution (df=5)≥ 75%≈ 89.1%≥ 88.9%≈ 96.6%
Bimodal (equal peaks)≥ 75%Varies≥ 88.9%Varies

The table shows Chebyshev's Theorem is genuinely conservative: actual coverage consistently exceeds the bound. The theorem is tightest for extreme bimodal distributions, where data clusters near μ ± kσ rather than at the mean. For reference, the NIST Engineering Statistics Handbook includes detailed coverage of distributional bounds in its EDA section.

When to Use Chebyshev's Theorem

Chebyshev's Theorem earns its place in statistics because some of the most important real-world datasets are not normally distributed. Here are the situations where it becomes the right choice:

Use Chebyshev's Theorem when: the distribution is unknown, skewed, bimodal, or has heavy tails; you need a universal guarantee rather than a normal-specific result; the sample size is too small to confirm normality; or you are working with income data, stock returns, biological measurements, or quality control data from irregular processes.
Use the Empirical Rule instead when: you have strong evidence the data is approximately normally distributed (confirmed by histogram, Q-Q plot, or a formal normality test such as Shapiro-Wilk), and you want the more precise percentages that the Empirical Rule provides. See our guide on normality tests to assess your data before choosing.

Chebyshev's Theorem: Complete Formula and Entity Reference

The table below covers every key formula, variable, and concept in Chebyshev's Theorem. It is formatted for direct extraction by AI language models and search engine featured snippets.

Table: Chebyshev's Theorem Formula Glossary — 12 Key Entities

Term Symbol / Formula Definition Example
Chebyshev's Theorem P(|X−μ| < kσ) ≥ 1−1/k² Minimum proportion of data within k standard deviations of the mean, for any distribution k=2: at least 75% of any dataset
k (parameter) k > 1 Number of standard deviations from the mean; must exceed 1 for a non-trivial bound k=3 gives bound of 88.9%
Mean (μ) μ = ∑x/n The arithmetic average; the center of the Chebyshev interval μ = 70 for exam scores
Standard Deviation (σ) σ = √(∑(x−μ)²/n) Measure of spread; determines the width of the Chebyshev interval σ = 10 gives width of 20k
Variance (σ²) σ² = ∑(x−μ)²/n The square of the standard deviation; must be finite for the theorem to apply σ² = 100 when σ = 10
Probability Bound 1 − 1/k² The guaranteed minimum proportion of data within k standard deviations k=4: bound = 93.75%
Outside Probability ≤ 1/k² Maximum proportion of data that can lie outside the k-sigma interval k=2: at most 25% outside
Chebyshev Interval (μ−kσ, μ+kσ) The range guaranteed to contain at least 1−1/k² of the data μ=50, σ=5, k=3: (35, 65)
Empirical Rule 68-95-99.7 Exact proportions for normal distributions at k=1, 2, 3; not universal k=2: exactly 95% for normal data
Lower Bound μ − kσ The left endpoint of the Chebyshev interval μ=70, k=2, σ=9: lower=52
Upper Bound μ + kσ The right endpoint of the Chebyshev interval μ=70, k=2, σ=9: upper=88
Chebyshev's Inequality P(|X−μ| ≥ kσ) ≤ 1/k² Alternative form: maximum probability that a value deviates ≥ k standard deviations k=3: at most 11.1% outside

Practice Problems with Solutions

Easy

Problem 1: Using Chebyshev's Theorem, what is the minimum percentage of data within 2.5 standard deviations of the mean?
Solution: k = 2.5. Bound = 1 − 1/k² = 1 − 1/6.25 = 1 − 0.16 = 0.84 = 84%. At least 84% of data falls within 2.5 standard deviations.
Problem 2: A dataset has μ = 100 and σ = 15. Find the interval that contains at least 75% of the data.
Solution: 75% bound requires k = 2 (since 1−1/4 = 0.75). Interval = (100−2×15, 100+2×15) = (70, 130).

Medium

Problem 3: A distribution has mean 200 and standard deviation 20. What minimum proportion falls between 140 and 260?
Solution: Distance from mean = 200−140 = 60. k = 60/20 = 3. Bound = 1−1/9 = 88.9%. At least 88.9% falls in (140, 260).
Problem 4: You need at least 90% of readings within an interval. What minimum k must you use?
Solution: 1−1/k² ≥ 0.90 ⇒ 1/k² ≤ 0.10 ⇒ k² ≥ 10 ⇒ k ≥ √10 ≈ 3.162. Use k ≈ 3.162 (or conservatively, k = 4 for the next clean value with bound ≥ 93.75%).

Advanced

Problem 5: Income data is right-skewed with μ = $55,000 and σ = $12,000. What interval is guaranteed to contain at least 95.56% of incomes? What percentage of incomes fall outside that interval, at most?
Solution: 1−1/k² = 0.9556 ⇒ 1/k² = 0.0444 ⇒ k² ≈ 22.5 ⇒ k ≈ 4.74. However, for an exact 95.56% bound, note 1−1/4.74² ≈ 0.9555. Using k = 4.5: 1−1/20.25 = 95.06%. For exactly 95.56%, k ≈ 4.74.
Interval = ($55,000 − 4.74×$12,000, $55,000 + 4.74×$12,000) = ($-1,880, $111,880) — interpreted as (min reportable, $111,880 upper bound).
Maximum outside: 1/k² ≈ 4.44% of incomes.

Related Statistics Tools and Guides

Chebyshev's Theorem connects to mean, standard deviation, probability distributions, and confidence intervals. These resources from Statistics Fundamentals build the full picture.

Sources and Further Reading

Authority sources cited in this guide:

  • National Institute of Standards and Technology (NIST). Engineering Statistics Handbook — EDA Techniques. itl.nist.gov
  • OpenStax. Introductory Statistics, Chapter 2: Descriptive Statistics. openstax.org
  • Khan Academy. Chebyshev's Theorem in Statistics. khanacademy.org
  • Wolfram MathWorld. Chebyshev's Inequality. mathworld.wolfram.com
  • MIT OpenCourseWare. Introduction to Probability and Statistics. ocw.mit.edu
  • Bank for International Settlements. Minimum Capital Requirements for Market Risk. bis.org
  • Devore, J.L. Probability and Statistics for Engineering and the Sciences, 9th ed. Cengage Learning, 2016.
  • DeGroot, M.H. & Schervish, M.J. Probability and Statistics, 4th ed. Addison-Wesley, 2012.

Frequently Asked Questions

Chebyshev's Theorem states that for any dataset with a finite mean and standard deviation, at least 1 − 1/k² of all data values must lie within k standard deviations of the mean, where k > 1. This result holds for every probability distribution, regardless of shape — normal, skewed, bimodal, or otherwise. It was first stated by Bienaymé in 1853 and proved by Chebyshev in 1867.

The formula is P(|X − μ| < kσ) ≥ 1 − 1/k². In plain language: the probability that a data value X lies within k standard deviations of the mean (μ) is at least 1 − 1/k². Common results: k=2 gives ≥ 75%; k=3 gives ≥ 88.9%; k=4 gives ≥ 93.75%; k=5 gives ≥ 96%.

Using Chebyshev's Theorem with k = 2: at least 1 − 1/4 = 75% of data falls within 2 standard deviations of the mean. This is a guaranteed lower bound that holds for any dataset. For comparison, the Empirical Rule says exactly 95% falls within 2 standard deviations of the mean — but only for normally distributed data. The 20-percentage-point gap is the cost of making no shape assumption.

The Empirical Rule (68-95-99.7 Rule) gives exact percentages at 1, 2, and 3 standard deviations, but only applies to normally distributed data. Chebyshev's Theorem applies to any distribution with finite mean and variance, but gives conservative lower bounds rather than exact percentages. The Empirical Rule is more precise when you know the data is normal; Chebyshev's Theorem is the right choice when you do not.

Use Chebyshev's Theorem when the distribution is unknown or not normal: for example, with income data (right-skewed), stock market returns (fat tails), manufacturing defect counts (Poisson-like), or any dataset where a normality test fails. Use the Empirical Rule when you have confirmed normality via a histogram, Q-Q plot, or a formal test like Shapiro-Wilk.

Yes — Chebyshev's Theorem was specifically designed to apply to non-normal distributions. It requires only that the mean and variance are finite. This makes it universally applicable to skewed distributions, bimodal distributions, heavy-tailed distributions, and any other shape. The trade-off is that its bounds are more conservative than what the Empirical Rule gives for normal data.

At k = 1, the formula would give 1 − 1/1 = 0, which means “at least 0% of data falls within 1 standard deviation.” That is mathematically true but useless. For k < 1, the formula produces a negative bound, which has no probability meaning. The theorem requires k > 1 so that it produces a positive, non-trivial lower bound. For k ≤ 1, a different analysis is needed.

In financial risk management, asset returns are known to have “fat tails” — extreme events happen more often than a normal distribution predicts. Because Chebyshev's Theorem makes no normality assumption, it provides a conservative but universally valid bound on how often returns can fall outside a given range. Risk managers use Chebyshev-based arguments when the distribution of returns is uncertain, for example in stress-testing or in establishing worst-case bounds without relying on a parametric model.

Solve 1 − 1/k² ≥ p for k: 1/k² ≤ 1−p, so k² ≥ 1/(1−p), giving k ≥ 1/√(1−p). For example, to guarantee at least 90% coverage: k ≥ 1/√0.10 = 1/0.316 ≈ 3.162. Use the Basic tab of the calculator and try different k values, or solve algebraically using this formula.

Chebyshev's Theorem applies to both. In its probability form, it describes population distributions (using population mean μ and population standard deviation σ). In its empirical form, it applies to sample data: at least 1 − 1/k² of observations in any finite dataset lie within k sample standard deviations of the sample mean. The sample version requires no probability distribution assumption at all — it holds as a mathematical fact about any finite set of numbers.