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


A Mathematician Reads the Newspaper

Read about reporters who blew it when trying to write about stats. A mathematician leads you through the tricks, citing contemporary examples.
More information
-->

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

How to assess?
Message posted by Ruby Lo (via 203.193.106.164) on November 17, 2001 at 12:54 AM (ET)

Hello editor,
My name is Ruby from Hong Kong, I have a problem that I have got a set of data - e.g. the unemployment figues of each month from 1/2000 to 10/2001. My boss told me to divide them in a plot graph with three lines. (1) The mean, (2) A line above the mean and (3) A line below the mean. The purpose is to categorize the figues in three section:(1) The figures that over the upperest line is red that means it is high risk - the Government should be alerted and aware, (2) the middle section is amble - normal situation, (3) The section below the lowerest line is green - the least risk level. I tried to calcalate the mean, the (+)(-)standard deviation to divide them into three sections, however, it seems not right since the upperest line always higher than the biggest number of the set data. Thus, it never go to the red zone. I tried other sets of data but the result is the same. My boss says why all the figures always at the amble zone only?
Can you help me how to assess the data and how to set up the range? Thank you very much.
Best wishes,
Ruby


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

Re: How to assess?
Message posted by Tomi (via 154.32.143.180) on November 18, 2001 at 1:16 AM (ET)

Hi Ruby,

You may want to look at http://www.statsoftinc.com/textbook/stquacon.html for more detailed information on control charts, but here is a quick fix for you:

Use 1.96 * standard deviation / sqrt(n) where n is the number of observations.


Re: How to assess?
Message posted by Darius (via 200.23.217.10) on November 19, 2001 at 10:22 AM (ET)

What you were telling is just control charts. The standard deviation that you may be using (according to the effect) is taking in account total variation, not within subgroup variation (that is the one used for this type of chart).

Total Variation = Variation within Subgroup + Variation between subgroup

Its wrong to use std/sqrt(n) because of the estimate of variation you where using.

It looks like individual control charts so http://www.qualitymag.com/articles/jan98/0198wh.html could help you.

There are other interesting articles and sites:

http://www.qualitymag.com/articles/1998/feb98/0298wh.html

http://www.qualitymag.com/articles/feb99/0299wh.html

http://www.qualitymag.com/articles/1999/aug99/0899wh.html

http://www.qualitymag.com/articles/1999/sep99/0999wh.html

http://www.qualitymag.com/articles/1999/oct99/1099wh.html

http://www.qualitymag.com/articles/oct98/1098wh.html

look for "individual control charts" in

http://www.itl.nist.gov/div898/handbook/

I hope, this may help you.



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!