Statistical Proportions – A Guide To the Quick and Dirty Estimator
I use statistical proportions as a very fast way to make predictions when you have three numbers and need to calculate the fourth one. My algebra teacher in Junior High School taught me how to apply them to many situations. Statistical proportions – often referred to as just proportions – can be a lifesaver if you need a number quickly.
Below is an example of how it works:
Current Workers Comp Policy = 136,000 in premiums
The Experience Modification Factor is .92. What would be the premium if the E-Mod moves to .99?
Proportions have been referred to as cross-multiplication and here is why –
.92/136,000 = .99/x
.92x = .99 * 136,000
x = 146,347
Now, many variables go into an X-Mod or E-Mod. At least you have a fast way to estimate a number for budgeting purposes, etc.
That was a very raw example for Workers Comp but it does work to a degree. Many actuaries will argue with me that I cannot shortcut the full E-Mod calculation. However, the numbers do come out as I have compared the prediction to the outcome that actually happened with the premium.
The further you go away from .92, for instance 1.35, the more inaccuracy that may be built in, but not necessarily.
No, I do not use this calculation to make predictions or comparisons all the time – however – it is not a bad way to figure any type of what if situaion when you know three numbers and need to predict the fourth.
A great example is the dog’s head drawing as to how statistical proportions work. The webpage has many different ways to work with proportions.
This may seem too simple, but when you only have a minute to calculate an unknown number for comparison, this is a powerhouse. I have used them since the 8th grade – thank you Mr. Epperson.
©J&L Risk Management Inc Copyright Notice
2 Responses
Thanks іn support of shaｒing such a goⲟd thinking, piece of writing is good,
thats wһy i һave rеad it fully
Thanks for the compliment. There are many articles on this blog concerning statistics. Check out the search here.