Talk:Quartile coefficient of dispersion

agreement on definition?
How accepted is this definition? In a quick google I found 2 others, one using the mean in the divisor and one using the median. Melcombe (talk) 10:54, 1 April 2008 (UTC)


 * Well, I'm finding an item published in 1942 whose definition agrees with what's here. Another journal article gives
 * $$ {Q_3 - Q_1 \over 2\times\text{median}}. $$
 * Either of these can make a certain amount of sense for positive-valued variables. The one in 1942 was on distribution of incomes.  I'll come back to this...... Michael Hardy (talk) 14:19, 1 April 2008 (UTC)
 * Either of these can make a certain amount of sense for positive-valued variables. The one in 1942 was on distribution of incomes.  I'll come back to this...... Michael Hardy (talk) 14:19, 1 April 2008 (UTC)
 * Either of these can make a certain amount of sense for positive-valued variables. The one in 1942 was on distribution of incomes.  I'll come back to this...... Michael Hardy (talk) 14:19, 1 April 2008 (UTC)

For the data set provide, the two definitions give the same result. —Preceding unsigned comment added by Seeker Seeking (talk • contribs) 07:09, 3 May 2008 (UTC)

no agreement on definition!
This is probably the wrong definition! Correct is to divide only through one median an not through twice median.
 * $$ {Q_3 - Q_1 \over \Q_2}. $$

To divide the standard deviation through the mean means to be a coefficient of variation, the parametric version of the quartile coefficient of dispersion. The article regarding coefficient of variation or relative standard deviation is not referenced to this nonparametric version. Vohumano (talk) 06:57, 12 May 2013 (UTC)


 * It is correct: https://www.itl.nist.gov/div898/software/dataplot/refman2/auxillar/qcd.htm Cerberus (talk) 03:37, 17 October 2018 (UTC)