Hi guyz,

I might seem a bit stupid here but i need some assistance. I have data set for average call durations(ACD) in minutes. Sometimes the samples are close enough and doesn't vary much and sometimes very spread.The data set is

ACD = {6,6,6,4,6,6,7,6,7,6,2,24,2,2,5,5,17,51,4,1,1,9,3, 15,1,2,2,7,4,6,6,7,
7,7,8,5,7,11,17,25,20,3,16,10,7,17,16,52,22,389,52 6,43,151,359,
44,753,9,6,81,22,182,64,66,245,606,190,2238,480,32 7,536,1,249,
731,592,62,35,373,146,611,94,1393,456,1388,13,801, 12,1172,209,
193,44,1097,1345,889,2,87,20,420,113,23,505,58,16, 859,17,66,45,
184,20,514,35,10,19,269,263,10,10,12,11,129,14,9,4 89,8,13}

Now I have a very large variation in samples, how do i calculate mean??should it be mean or media or mode etc?????

Thanx

Fahdnoor

Though you have a very large variation, the ordinary mean formula is still sum(xi)/n, i = 1,2,...,n.
You may use the "trimmed mean", which aims to exclude the extreme values at both "ends": very small and very large values are excluded. You may designate to exclude the upper and lower 10% of the values.
However, the most "robust" method is the median. The extreme values will not influence median very much.