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

Normal Distribution
Message posted by mitch on September 30, 1999 at 12:00 AM (ET)

my questions for a survey are usually categorical or nominal data.question: for variables like age, sex, educ attainment...do i need to know if they are distributed normally? what's the diff. bet norma l and uniform distribution?how can i apply normality in purely categorical data?in a sample of 400, do i have to know if it is normaly distributed? cant i make any statistical comp. if the data is not distributed normally?


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

Re: Normal Distribution
Message posted by Bill on October 1, 1999 at 12:00 AM (ET)

Data from ordinal or nominal (categorical) variables are not properly analyzed using the theory or tests based on the normal distribution. Assuming age is measured as an interval scale (as opposed to Age = 1 if age is > 20 and <= 49, Age = 2 if age > 49) then you would need to know if this variable is normally distributed if you plan to apply a statistical test that is based on the normal distribution, e.g., a t-test. However, it makes no sense to discuss "sex" (a categorical variable) as a normally distributed variable.

I assume you know the characteristics of a normal distribution. Now, suppose that we take an infinitely large population of random numbers that are EVENLY distributed between zero and 100. The population will have a uniform distribution, that is, every number between zero and 100 will be equally likely. I think that this is also called a rectangular distribution.

In any sample, regardless of size, you need to know if a variable is normally distributed if you plan to apply a statistical test that is based on the normal distribution. The concept of normality applies to variables, not the sample. Yes, you can make statistical comparisons if the data is not normally distributed if you use statistical tests which do not make assumptions about the distribution of the data, i.e., nonparametric or distribution-free tests.



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!