RobertNiles.com
About Robert
Reporting Help
Finding Data on the Internet
Questions for Candidates
Stats Lessons
Mean
Median
Percent
Per capita
Standard Deviation
Margin of Error
Data Analysis
Sample Sizes
Stats Tests
Help Board
Bookstore


The Cartoon Guide to Statistics

This isn't some dumb-downed whitewash. It's a must-read for a beginner taking his or her first steps toward understanding stats.
More information
-->

Home > Statistics Every Writer Should Know > The Stats Board > Discusssion

percent and weighted results
Message posted by Xavier V-P (via 12.8.181.4) on September 7, 2001 at 9:18 AM (ET)

I collect data on a tool that is divided into 3 sections each having a weight.
each section is based on numper of mets / number of applicables then * their section weight.
Sec-a = 20%
Sec-b = 50%
Sec c = 30%
I know the first calculation is a percent.
My question is
when I take the section percentage results multiplied by their weight then add all three section to obtain a final result is this final result also clasified as a percentage ?

Ex:
a= 8/10=.80*100=80%*.2=16%
B= 29/50=.580*100=58%*.5=29%
c= 15/20=.750*100=75%*.3=22.5%

These three results 16%+29%+22.5%=67.5%

is this 67.5 considered a percentage or a composite score.
why is it not a percentage

Thank you
Xavier


READERS RESPOND:
(In chronological order. Most recent at the bottom.)

Re: percent and weighted results
Message posted by Tomi (via 154.32.142.241) on September 8, 2001 at 3:32 AM (ET)

General principle: if it looks like a percentage and behaves like a percentage then it probably is a percentage.

The value you end up with certainly looks like a percentage. It takes values between 0 and 100.

Of course, percentages are only a convenient way of expressing numbers. People are generally happier with the concept of a 75% score than a 0.75 score. So even if what you have is a composite score, then using the notation of percentages is fine.

Big question: will using percentages confuse or clarify? After all, statistics is all about making sense of number. If you give a result of 80%, the impression given will be that the total number of mets/total number of applicables = 0.8

Now because you are weighting, this will not be true. In the example you gave, your weighted score comes out at 67.5%

In fact, the "true" percentage is (8+29+15)/(10+50+20)*100%
=52/80*100%=65%

Clearly there isn't much of a difference, as your weighting (20:50:30) is not very different from the effect of not weighting at all (10:50:20).

Presumably you have good reasons for wanting to emphasise the "a" and "c" measurements, so the 67.5% is "better" because it is being influenced more strongly by the high percentages in "a" and "c".

For clarity, it might be wise in reporting your results to explain your reasons for weighting and give both the unweighted and weighted percentages.



Your $5 contribution helps cover part the $500 annual cost of keeping this site online.

Niles Online Sites:RobertNiles.comTheme Park InsiderViolinist.com

RobertNiles.com™, the site, content and services 咀opyright 1996-2002, Robert Niles.
All rights reserved. Questions? Comments? Read my Privacy Policy, or E-mail me!