Merits of Quartile Deviation:
(1) It is easy to calculate and simple to follow.
(2) It is not affected by the extreme values and is, therefore, useful in skewed distributions.
(3) It is the only method of dispersion applicable in case of ‘open-end classes’.
(1) Since Quartile Deviation is based on Quartiles, sometimes it is not rigidly defined.
(2) It does not take into account all the observations in the series. Hence it is not representative.
(3) It is not capable of further algebraic treatment.
(4) It is not a stable measure of dispersion as it is affected very much by fluctuations of sampling.
(5) In a symmetrical distribution, the value of the median which is the second Quartile of the distribution lies midway between Q1 and Q3. If however, the series in non-symmetrical, Md does not lie midway between Q1 and Q3. If the series is symmetrical md + Q. D should equal Q3 and Q1.
If it is non-symmetrical md + Q.D will not be equal to Q3 and Q1.
To that extent Quartile Deviation gives a poor picture of dispersion.