Non-parametric tests are used for data that are not normally distributed and/or nominal or ordinal, and are based on what measure?

Non-parametric tests are used for data that are not normally distributed and/or nominal or ordinal, and are based on what measure?

• A. Means
• B. Medians
• C. Modes
• D. Ranges

Answer: (B)

Non-parametric methods are used when data don’t meet normal-distribution assumptions or are on an ordinal scale. In these cases, the central quantity they rely on is the median (often via ranks), not the mean. The median remains meaningful for skewed data and for ordinal data, and many non-parametric procedures test whether medians differ from a value or whether two groups share the same median. That robustness to outliers and distribution shape is why medians (and the related ranking of data) form the basis of these tests. In contrast, means can be distorted by skew and outliers, and modes or ranges don’t provide the same reliable basis for inference in this context.

Non-parametric methods are used when data don’t meet normal-distribution assumptions or are on an ordinal scale. In these cases, the central quantity they rely on is the median (often via ranks), not the mean. The median remains meaningful for skewed data and for ordinal data, and many non-parametric procedures test whether medians differ from a value or whether two groups share the same median. That robustness to outliers and distribution shape is why medians (and the related ranking of data) form the basis of these tests. In contrast, means can be distorted by skew and outliers, and modes or ranges don’t provide the same reliable basis for inference in this context.