meanmedian

Meanmedian is not a standard statistical term with a universally agreed definition. In practice, it may refer to either the pair of central-tendency measures—the arithmetic mean and the median—or to a single blended statistic that combines the mean and the median to summarize a data set.

Common implementations treat meanmedian as the average of the mean and the median: M = (mean + median)/2. A more general form is M_w = w*mean + (1-w)*median for a weight w in [0,1].

Example: data 1, 2, 3, 100 yield mean 26.5 and median 2.5, so the blended statistic is M = (26.5 + 2.5)/2 = 14.5.

Interpretation and use: It can serve as a compromise between the sensitivity of the mean and the robustness of the median, useful for quick descriptive summaries of skewed data or when outliers are present but not extreme. Since it is not standard, it is best to report both components or specify the exact blend.

Relation to other concepts: It is not as robust as the median and remains influenced by the mean component. It is a simple linear combination, one of several ad hoc composite estimators; more robust alternatives include trimmed means, Winsorized means, and M-estimators.