scipystatsnormaltest

scipystatsnormaltest refers to the normaltest function in the SciPy library’s statistics module (scipy.stats.normaltest). It implements D’Agostino and Pearson’s test for normality, used to evaluate whether a dataset may be drawn from a normal distribution.

The function tests the null hypothesis that the data are normally distributed against the alternative that they are not. It combines measures of skewness and kurtosis into a single statistic, known as D’Agostino’s K^2 statistic, which under the null follows a chi-square distribution with two degrees of freedom. The resulting p-value indicates how unlikely the observed skewness and kurtosis would be if the data were normal.

Input to the function is an array-like sequence of observations, with optional parameters such as axis to specify the computation direction and nan_policy to handle missing values. The function returns two values: the test statistic and the p-value.

Interpretation of the results follows standard statistical practice. A small p-value (commonly below 0.05) leads to rejection of the normality assumption. However, the test can be sensitive to sample size: with very large samples, even small deviations from normality may yield small p-values, while with small samples it may lack power. Therefore, results are typically considered alongside visual assessments (e.g., Q-Q plots) and other normality tests.

Scipystatsnormaltest is part of SciPy’s statistical toolbox and is frequently used in data analysis workflows to assess distributional assumptions before applying parametric tests. It is one option among several normality tests, each with its own strengths and assumptions.