Empirical Rule vs Chebyshev's Theorem
How much data lies near the mean, and what can we guarantee when the distribution shape is unknown?
How this lesson unfolds.
Begin with What is a distribution? The sequence then connects A recurring bell-shaped pattern and builds toward Choose the rule from the evidence.
What is a distribution?
A distribution records where observations fall and how densely they gather across possible values.
- Learning goalExplain distribution foundation and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionA recurring bell-shaped pattern
What did statisticians repeatedly observe?
Many measured phenomena looked roughly bell-shaped, motivating the empirical 68-95-99.7 pattern.
- Learning goalExplain historical observation and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionThe Empirical Rule needs a bell shape
When does the 68-95-99.7 rule apply?
For an approximately normal distribution, about 68%, 95% and 99.7% lie within one, two and three standard deviations of the mean.
- Learning goalExplain empirical rule and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionA classroom with mean 150 cm and SD 10 cm
What interval is three standard deviations from the mean?
With mean 150 cm and standard deviation 10 cm, the three-standard-deviation interval is 120 to 180 cm.
- Learning goalExplain classroom setup and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionUnknown shape calls for a guarantee
What if the distribution shape is unknown?
Chebyshev's theorem gives a minimum fraction near the mean without assuming normality.
- Learning goalExplain shape free motivation and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionVariance charges squared distance
How does one far-away observation use the variance budget?
A height 40 cm from the mean contributes 40 squared, or 1600, to the total squared deviation.
- Learning goalExplain squared distance and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionThe dataset has a fixed variance budget
What does SD 10 cm imply for 100 students?
Using population variance, total squared deviation is n times sigma squared: 100 times 100 equals 10000 square-centimetres.
Using population variance, total squared deviation is n times sigma squared: 100 times 100 equals 10000 square-centimetres.
- Learning goalExplain variance budget and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionStandard deviation is horizontal spread
What is the biggest standard-deviation misunderstanding?
Standard deviation measures distance along the value axis, not the vertical height of a density curve.
- Learning goalExplain sd misconception and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionThree SD defines a region, not a headcount
What does within three standard deviations actually mean?
Three standard deviations is 30 cm, so the region is 150 plus or minus 30, namely 120 to 180 cm; a theorem is still needed to guarantee a count.
- Learning goalExplain interval meaning and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionCan 80 students be outside the interval?
Could most students lie more than three SD from the mean?
If 80 students were at least 30 cm away, they would consume at least 80 times 900 equals 72000 units of squared deviation.
If 80 students were at least 30 cm away, they would consume at least 80 times 900 equals 72000 units of squared deviation.
- Learning goalExplain contradiction setup and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionThe variance budget limits outsiders
How many students can remain outside 120 to 180 cm?
If x students are outside, then 900x is at most 10000, so x is at most 11.11; with whole students, at most 11 are outside and at least 89 are inside.
- Learning goalExplain chebyshev intuition and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionGeneralize the variance-budget argument
What guarantee holds for any k greater than 1?
Chebyshev guarantees P(|X-mu| less than k sigma) at least 1 minus 1 over k squared for finite nonzero variance.
- Learning goalExplain chebyshev formula and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionOne spread measure, two ways to reason
Can two differently shaped distributions share mean and variance?
Yes. Mean and variance do not determine distribution shape, so shape-free and normal-specific rules answer differently.
- Learning goalExplain same moments different shapes and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionWhy bell curves often appear
Why does nature produce many bell-shaped patterns?
Many small, roughly independent additive effects can combine into an approximately normal pattern.
- Learning goalExplain bell shape intuition and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionNot every dataset is bell-shaped
Should every dataset use the Empirical Rule?
Skewed, heavy-tailed or multi-peaked data may violate the bell-shape assumption; use a shape-free bound when shape is unknown.
- Learning goalExplain non normal shapes and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionChoose the rule from the evidence
Which rule should you use?
Use the Empirical Rule for approximately normal data and Chebyshev's theorem when only finite mean and variance are justified.
- Learning goalExplain method selection and state the assumptions behind the bound.
- Helpful foundationsmean · variance · standard deviation
- Common trapThe Empirical Rule is approximately valid for bell-shaped data; Chebyshev is a conservative guarantee for any distribution with finite nonzero variance.
- Next connectionApply the correct spread rule to a new dataset