Normality Test Calculator

Test whether data is normally distributed using skewness, kurtosis, Jarque-Bera, and Shapiro-Wilk statistics.

Share this calculator

Normality Assessment

Likely Normal

The Jarque-Bera p-value of 0.6960 exceeds the 0.05 significance level, suggesting the data is consistent with a normal distribution. Skewness (0.000) and excess kurtosis (-0.834) are within expected ranges.

Skewness
0
Excess Kurtosis
-0.83424
Jarque-Bera
0.724955
JB p-value
0.69595
Shapiro-Wilk W
0.982098
Count
25
Mean
4
Std Dev
0.204124

Interpreting normality tests

The Jarque-Bera test checks whether sample skewness and kurtosis match a normal distribution. A p-value above 0.05 suggests the data is consistent with normality; below 0.05 indicates significant departure. Skewness measures asymmetry (0 for symmetric data), while excess kurtosis measures tail weight relative to normal (0 for normal). The Shapiro-Wilk W statistic (shown for datasets of 50 or fewer values) measures how closely ordered data matches expected normal order statistics, with values near 1 indicating normality.

Also in Statistics

Descriptive Statistics

Normality test calculator: check if data follows a normal distribution

A normality test calculator evaluates whether a dataset is approximately normally distributed. It computes skewness, excess kurtosis, the Jarque-Bera test statistic with its p-value, and the Shapiro-Wilk W statistic for smaller datasets. Enter numbers separated by commas, spaces, or newlines.

Normality test statistics explained

Skewness measures the asymmetry of the distribution. A perfectly symmetric distribution has skewness of 0. Positive skewness indicates a longer right tail; negative skewness indicates a longer left tail.

Excess kurtosis measures the heaviness of the tails relative to a normal distribution. Normal data has excess kurtosis of 0. Positive kurtosis (leptokurtic) means heavier tails; negative kurtosis (platykurtic) means lighter tails.

The Jarque-Bera test combines skewness and kurtosis into a single test statistic that follows a chi-square distribution with 2 degrees of freedom under the null hypothesis of normality.

JB = (n/6)(S² + K²/4)

Jarque-Bera statistic where S is skewness and K is excess kurtosis.

Frequently asked questions

How many data points do I need for a reliable normality test?

The Jarque-Bera test is asymptotic and works best with larger samples (n > 30). The Shapiro-Wilk test is more reliable for small samples (n < 50). With very small samples (n < 8), normality tests have low statistical power.

What should I do if my data fails the normality test?

Consider non-parametric alternatives to tests that assume normality (e.g., Mann-Whitney instead of t-test), apply a transformation (log, square root), or use robust statistical methods that are less sensitive to non-normality.

Related

More from nearby categories

These related calculators come from the same leaf category, nearby sibling categories, or the same top-level topic.