1 Sample Sign

The 1 Sample Sign test is one of the most straightforward non‑parametric tools available in the Analyze phase. It is designed for situations where you want to compare a sample’s median to a known or target value, but the data does not meet the assumptions required for a 1‑sample t‑test. This often occurs when data is heavily skewed, contains outliers, or is measured on an ordinal scale. In these cases, the Sign test provides a clean, reliable way to evaluate whether the process has shifted meaningfully. 

The strength of the Sign test lies in its simplicity. Instead of relying on the magnitude of differences, it focuses solely on direction—whether each observation is above or below the target. This makes the test extremely robust to outliers and distributional irregularities. It also makes it ideal for processes where the median is a more meaningful measure of central tendency than the mean. 

The procedure is straightforward. For each observation, you determine whether it is greater than, less than, or equal to the target value. Observations equal to the target are discarded. You then count the number of “positive signs” (values above the target) and “negative signs” (values below the target). If the process median truly equals the target, you would expect roughly half of the observations to fall above and half below. A significant imbalance suggests that the median differs from the target. 

Because the test is based on counts rather than magnitudes, it is resistant to extreme values. A single outlier cannot distort the result. This makes the Sign test particularly useful in service environments, transactional processes, and human‑driven activities where variability can be unpredictable. 

However, the simplicity of the Sign test comes with trade‑offs. It is less powerful than alternatives like the Wilcoxon signed‑rank test because it ignores the size of the differences. This means it may fail to detect meaningful shifts when sample sizes are small. Still, when assumptions are violated or when data quality is imperfect, the Sign test provides a reliable, assumption‑light option. 

In the Analyze phase, the 1 Sample Sign test helps you determine whether a process has shifted relative to a target or historical baseline. It supports confident decision‑making in situations where traditional parametric tests are inappropriate, ensuring that your conclusions remain grounded in the realities of the data.