Content Type:

Glossary

Share Print

Kurtosis Definition

Kurtosis is a statistical measure that describes the heaviness of the tails of a distribution relative to a normal distribution.

Kurtosis is a statistical measure that describes the heaviness of the tails of a distribution relative to a normal distribution. It reflects the extent to which extreme deviations contribute to a dataset’s overall variability. When comparing to the normal distribution, analysts typically use excess kurtosis, where the normal distribution has a value of zero.

Distributions are often classified into three general categories:

• Mesokurtic distributions have excess kurtosis near zero, indicating tail behavior similar to a normal distribution.
• Leptokurtic distributions have positive excess kurtosis and heavier tails than a normal distribution, indicating a greater influence of extreme values.
• Platykurtic distributions have negative excess kurtosis and lighter tails than a normal distribution, indicating a lower influence of extreme values.

(Note: Kurtosis primarily reflects tail heaviness, not the “peakedness” of a distribution.)

In marketing research, kurtosis helps analysts better understand the distribution of survey responses, consumer behaviors and performance metrics beyond basic measures such as averages or standard deviation. 

Who relies on kurtosis?

Marketing researchers, data analysts, statisticians and insight professionals use kurtosis when evaluating survey data, customer behavior, sales performance and other market-related metrics. It is particularly useful for identifying outliers or assessing whether rare observations may disproportionately influence analysis, forecasting or business decisions.

Why should I care about kurtosis?

Kurtosis helps researchers interpret how likely extreme deviations are within a dataset. 

High kurtosis may indicate heavier tails, greater volatility, unusual market reactions or concentrated behavior in a small number of cases, while low kurtosis suggests a more stable distribution with fewer extreme deviations. 

Understanding kurtosis can improve data interpretation, support more accurate modeling and help researchers determine whether unusual results reflect meaningful patterns or statistical anomalies.

When should researchers pay attention to kurtosis?

Researchers should pay attention to kurtosis when datasets may contain infrequent but influential observations that could affect interpretation or predictive models. This is especially relevant when analyzing customer spending, campaign performance, satisfaction scores or behavioral metrics where a small number of extreme responses may significantly influence results.

Reviewing kurtosis alongside other distribution measures can help researchers better understand data structure and select appropriate analytical methods.