Abstract
Given a random variable X with finite mean, for each 0 < p < 1, a new sharp bound is found on the distance between a p‐quantile of X and its mean in terms of the central absolute first moment of X. The new bounds strengthen the fact that the mean of X is within one standard deviation of any of its medians, as well as a recent quantile‐generalization of this fact by O'Cinneide.
Original language | English |
---|---|
Pages (from-to) | 279-283 |
Number of pages | 5 |
Journal | Statistica Neerlandica |
Volume | 47 |
Issue number | 4 |
DOIs | |
State | Published - Dec 1993 |
Keywords
- absolute central first moment
- convex function
- mean
- median
- quantiles