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


Statistics for the Utterly Confused

This book offers a super-accessible approach to the much-misunderstood subject of statistics.
More information
-->

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

Bimodal distribution
Message posted by Lena Kupriyanova (via 129.96.253.100) on November 6, 2001 at 1:18 AM (ET)

How can I test whether my distribution is bimodal or unimodal? It looks rather right skewed if plotted


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

Re: Bimodal distribution
Message posted by Tomi (via 213.133.214.2) on November 6, 2001 at 7:38 AM (ET)

There will be two distinct peaks if it is bimodal (having two modes).


Re: Bimodal distribution
Message posted by Lena Kupriyanova (via 129.96.253.100) on November 7, 2001 at 1:22 AM (ET)

Sorry, my question was not clear enough. I know what bimodal is (two peaks). I wonder if there is any statistical test to determine whether the distribution is bimodal or not even though it LOOKS very much unimodal. The problem is that the distribution of log-TRANSFORMED data has two peaks . Can I claim that the distribution of the original variable is also bimodal? Thank you!!!


Re: Bimodal distribution
Message posted by Tomi (via 154.32.143.198) on November 10, 2001 at 10:33 AM (ET)

I see. If the log-transformed data is bimodal, then the original data is usually bimodal, too.

However, the act of taking logarithms squeezes together parts of the number line and stretches out other parts. This can lead to the creation of modes that do not exist in the original data. For example, the log-transformed distribution of a uniform distribution that takes values between 1 and 10 will have a mode at 10. If there is a long right-hand tail to the original distribution, then it is possible that the squeezing process could create a second mode in the log-transformed data.

So what do you do? I suggest you undo the log-transformation by using the exponential function.


Re: Bimodal distribution
Message posted by Tomi (via 154.32.143.198) on November 10, 2001 at 11:02 AM (ET)

Oops!!!

... mode at log(10) = 2.3



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!