Non-parametric tests are based on which central tendency measure?

Non-parametric tests are based on which central tendency measure?

• A. Means
• B. Medians
• C. Mode
• D. Range

Answer: (B)

Non-parametric tests center on the median because they rely on the order of the data rather than on the exact data values. They use ranks or ordinal information and do not assume that the data come from a normal distribution, which makes them robust to outliers and skewed data. In practice, many non-parametric tests (like those that compare two groups or multiple groups) assess whether there is a difference in location that is best interpreted as a difference in medians across groups, rather than a difference in means. This is in contrast to parametric tests, which rely on the mean and assume a normal distribution with certain variance conditions. The mean is more sensitive to extreme values, and the range describes spread rather than central tendency; the mode reflects the most frequent value but is not the focus of these tests. That’s why medians—and the associated ranking approach—are the foundation of non-parametric procedures.

Non-parametric tests center on the median because they rely on the order of the data rather than on the exact data values. They use ranks or ordinal information and do not assume that the data come from a normal distribution, which makes them robust to outliers and skewed data. In practice, many non-parametric tests (like those that compare two groups or multiple groups) assess whether there is a difference in location that is best interpreted as a difference in medians across groups, rather than a difference in means. This is in contrast to parametric tests, which rely on the mean and assume a normal distribution with certain variance conditions. The mean is more sensitive to extreme values, and the range describes spread rather than central tendency; the mode reflects the most frequent value but is not the focus of these tests. That’s why medians—and the associated ranking approach—are the foundation of non-parametric procedures.