The Shapiro–Wilk test assesses whether numeric data is consistent with a normal distribution. Its null hypothesis states that the data is normally distributed; using a 5% threshold, a p-value below 0.05 rejects normality, while a larger p-value fails to reject it. section 1.4 section 4.2.1
Apply the test in Python with SciPy:
- Pass each numeric series to `stats.shapiro()` and inspect the returned statistic and p-value. section 4.2.1
- Interpret the p-value against 0.05 rather than treating the statistic alone as the decision. section 4.2.1
Testing two measurement series
For the 20-row `mds` dataset, `stats.shapiro(mds.stage1)` returns p = 0.2933 and `stats.shapiro(mds.stage2)` returns p = 0.1479. Both exceed 0.05, so both tests fail to reject the normality null hypothesis. section 4.2.1
