# Statistics with Varying Volumes

Printable View

• Oct 8th 2009, 09:01 AM
Pherion
Statistics with Varying Volumes
Hello everyone, I have a question with a large number of precentages being compared.

We are working on calculating precent increases of account volumes over time in each of the 50 states (and a few additional countries codes/etc).

The problem I am having is that when we hav a small starting volume (say between 1 and 100) precent changes can go throug the roof really fast (1 to 2 is a 100% increase).

What I want to do is find a way to compare all 50 states and determin which has the most signifigant increases in volume. A state where the volume moves from 500 to 1000 is signifigatly more important than one who goes from 1 to 2.

Any suggestions on how to calculate a number, say between 0 an 100, to classify each state on a scale of how signifigant the change is?

Cheers,
Pherion
• Oct 8th 2009, 10:46 AM
Ond
Which state is more important in your opinion in the two following examples:

State A: An increase from 250 to 750?
State B: An increase from 500 to 1000?

How about this:

State C: An increase from 250 to 999?
State D: An increase from 500 to 1000?
• Oct 8th 2009, 11:33 AM
Pherion
In the first situation it would be A, because the inital volume was small. With two states and the same increase, the smaller inital volume state would be more of a signifigant increase.

In the second situation I would say C, primarly because the increase is 749 opposed to 500.

I guess I'm really trying to identify if an issue has arrisen in any particular state, the actual total volumes are easy to coordinate and keep ontop of. I need a way to identify states where things are changing quickly, or getting out of hand.