Concept Architecture
Concept
Theoretically, Kurtosis is a statistical measure that quantifies the shape of a probability distribution by describing the relative weight of its tails compared with a reference distribution, typically the normal distribution. It is the fourth standardised central moment and reflects the propensity of a distribution to produce extreme observations. Kurtosis exists to characterise distributional shape, assess deviations from normality and support the selection of appropriate statistical methods.
Mathematically, kurtosis is represented by the fourth central moment divided by the fourth power of the standard deviation. Excess kurtosis is commonly reported by subtracting three from the kurtosis of a distribution so that a normal distribution has an excess kurtosis of zero. Positive excess kurtosis indicates heavier tails than a normal distribution, whereas negative excess kurtosis indicates lighter tails.
In practice, kurtosis is estimated from sample data using standard estimators of the fourth central moment. In health economics it is used during exploratory data analysis to evaluate the distribution of costs, resource utilisation, quality-of-life scores and other variables. The measure informs decisions regarding statistical modelling, transformation of variables and the choice between parametric and non-parametric analytical methods.
Purpose
Used to assess the tail behaviour and shape of statistical distributions, evaluate departures from normality, identify the presence of extreme observations and support the selection of appropriate statistical methods in health economic analyses.
Mathematical Formulae
Primary Formula
Population kurtosis:
? = E[(X ? ?)?] / �?
Sample kurtosis:
g? = m? / m?�
where:
- m? = second sample central moment
- m? = fourth sample central moment
Excess kurtosis:
Excess Kurtosis = ? ? 3
Supporting Formulae
Second central moment:
m? = (1/n) ? ?(x? ? x?)�
Fourth central moment:
m? = (1/n) ? ?(x? ? x?)?
Related Mathematical Methods
- Skewness
- Central Moments
- Normality Assessment
- Jarque?Bera Test
- Shapiro?Wilk Test
- Anderson?Darling Test
Example
A health economist analyses annual healthcare costs for 500 patients before fitting a regression model.
The calculated sample kurtosis is:
g? = 5.8
Excess kurtosis is therefore:
5.8 ? 3 = 2.8
The positive excess kurtosis indicates heavier tails than a normal distribution, suggesting that unusually high-cost patients occur more frequently than expected under normality. The analyst therefore considers robust estimation methods or transformation of the cost variable before modelling.
Excel Implementation
| Function | Example Formula | Health Economics Application |
|---|---|---|
| KURT | =KURT(A2:A501) | Calculate sample excess kurtosis for healthcare costs or outcomes. |
| AVERAGE | =AVERAGE(A2:A501) | Calculate the sample mean before distributional assessment. |
| STDEV.S | =STDEV.S(A2:A501) | Calculate the sample standard deviation used in distribution diagnostics. |
| SKEW | =SKEW(A2:A501) | Assess distributional asymmetry alongside kurtosis. |
| IF | =IF(KURT(A2:A501)>0,"Heavy tails","Light or normal tails") | Classify the distribution during exploratory analysis. |
VBA (Optional)
A VBA routine can automatically calculate kurtosis for multiple variables and flag distributions that substantially deviate from normality during model diagnostics.
Sources
- Joanes DN, Gill CA. Comparing Measures of Sample Skewness and Kurtosis. Journal of the Royal Statistical Society: Series D (The Statistician). 1998;47(1):183?189.
- Westfall PH. Kurtosis as Peakedness, 1905?2014. The American Statistician. 2014;68(3):191?195.
- Kendall MG, Stuart A. The Advanced Theory of Statistics. Charles Griffin.
- Agresti A. Statistical Methods for the Social Sciences. Pearson.
- Briggs A, Claxton K, Sculpher M. Decision Modelling for Health Economic Evaluation. Oxford University Press.
- Drummond MF, et al. Methods for the Economic Evaluation of Health Care Programmes. Oxford University Press.
Related Concepts (2)
Library
Publications
1
Bayesian Methods in Health Economics — Gianluca Baio, 1st Edition ed., 2012 (Chapman & Hall / CRC Press)
An overview of Bayesian statistical methods for the analysis of health economic data, covering economic evaluation concepts, statistical cost-effectiveness analysis, Bayesian computation and MCMC, and applied health economic evaluation.
BookView source →
Frequently Asked Questions (6)
What is kurtosis?
A measure describing a distribution's tail shape relative to a normal distribution, higher values indicating heavier tails and more extreme outliers.
Source: Pearson 1895
What does kurtosis describe about a distribution's tails?
Kurtosis describes the shape of a distribution's tails compared with the normal bell curve, indicating how prone it is to extreme values. High kurtosis means heavy tails, where outliers far from the centre occur more often than a normal distribution would predict, while low kurtosis means light tails. Unlike skewness, which concerns lopsidedness, kurtosis is about the weight of the tails and the sharpness of the peak. Measuring the propensity for extreme values is its purpose. Kirkwood and Sterne (2003) describe this.
Source: Kirkwood & Sterne 2003
How is kurtosis interpreted?
Kurtosis is interpreted relative to the normal distribution: excess kurtosis above zero, sometimes called leptokurtic, indicates heavier tails and more outliers than a normal distribution, while excess kurtosis below zero, platykurtic, indicates lighter tails and fewer outliers. The normal distribution has zero excess kurtosis. So kurtosis is interpreted as how heavy or light a distribution's tails are compared with a normal one, with high kurtosis warning of a greater chance of extreme values, which matters because heavy tails can affect analyses that assume normality and increase the influence of outliers, making kurtosis useful for assessing distributional shape.
Source: Pearson 1895
How does kurtosis differ from skewness?
Kurtosis describes the tailedness of a distribution, how heavy or light its tails are and its propensity for extreme values, while skewness describes its asymmetry, whether it leans to one side with a longer tail in one direction. Kurtosis concerns the tails and peak; skewness concerns the lopsidedness. So kurtosis and skewness are distinct aspects of distributional shape, with kurtosis based on the fourth moment and skewness on the third, and both complement measures of centre and spread, since a full description of a distribution considers its central tendency, dispersion, asymmetry through skewness, and tailedness through kurtosis.
Source: Pearson 1895
Why does kurtosis matter?
Kurtosis matters because heavy-tailed distributions produce extreme values more often than a normal distribution, which can affect statistical analyses that assume normality, increase the impact of outliers, and have practical consequences such as underestimating the risk of extreme outcomes. So kurtosis matters for understanding a distribution's behaviour in the tails, which is important for checking the assumptions of methods, for anticipating outliers, and for assessing the likelihood of extreme events, since ignoring high kurtosis can lead to underestimating the frequency of extreme values, which is why examining kurtosis is part of describing and diagnosing data.
Source: Pearson 1895
How is kurtosis measured?
Kurtosis is measured from the fourth standardised moment of the distribution, the average of the fourth powers of the deviations from the mean divided by the standard deviation raised to the fourth power, often reported as excess kurtosis by subtracting the value for a normal distribution so that a normal distribution has zero. So kurtosis is measured through the fourth moment, giving a number that captures the weight of the tails, with excess kurtosis expressing this relative to normality, and it is estimated from sample data, though such estimates can be sensitive to outliers and sample size, which is considered when interpreting them.
Source: Casella & Berger 2002
Trust Record
Verified by Dr Darrin Baines
British health economist
Professional identity: darrinbaines.org
Verification date: 17 Dec 2025
Content version: 1.0.0
Canonical Identity
- Persistent URI
- https://healtheconomics.wiki/concept/kurtosis
- Term code
- HE-ES-SA-094
Stable URI · Machine-readable · Resolvable · CC BY 4.0