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

ordinal data - can you provide mean and sd?
Message posted by Shelly (via 64.228.61.150) on October 22, 2001 at 7:22 PM (ET)

I have a variable which has ordinal values (0=never; 1=once; 2=2-3times; 3=4-10 times; 4=more than 10 times).

Can I provide in my writeup of my study the MEAN and STANDARD DEVIATION of these variable? Does that mean anything?


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

Re: ordinal data - can you provide mean and sd?
Message posted by Tomi (via 154.32.142.7) on October 22, 2001 at 10:32 PM (ET)

You could calculate the mean and sd of the variable taking values {0, 1, 2, 3 and 4}, but they will be pretty meaningless.

However your variable is based on a real numerical variable: the number of times something occurs, by the sound of it. If you can identify the maximum possible (or likely) value for the "more than 10 times" response, you can calculate a meaningful mean and sd by using the techniques for grouped data.

Suppose that you replace the "more than 10 times" response with "11-19" times (obviously with no knowledge of the context I have no idea whether 19 is a sensible maximum value). The following example gives the calculations for mean and sd.

[variable]#[mid point]#[frequency]#[mid point * frequency]#[frequency*(mid point)^2]

[0]#[0]#[5]#[0]#[0]
[1]#[1]#[8]#[8]#[8]
[2-3]#[2.5]#[12]#[30]#[75]
[4-10]#[7]#[9]#[63]#[441]
[11-19]#[15]#[6]#[90]#[1350]

[Totals]#[n/a]#[40]#[1874]

Mean = sum(mid point*frequency)/sum(frequency) = 191/40=4.775

Variance = (sum(frequency*(mid point)^2)/sum(frequency) - (mean)^2 = 1874/40 - 4.4775^2 = 24.05 (2 d.p)

sd = sqrt(Variance) = 4.9 (1 d.p)

These measures will give good estimates of the sample mean and standard deviation. Note that they are only estimates because you are assuming that within each group values tend to be at the centre of the group (the mid point). This may not be true, especially with the "more than 10 times" group.

Hope this helps.


Re: ordinal data - can you provide mean and sd?
Message posted by Tomi (via 154.32.142.7) on October 22, 2001 at 11:05 PM (ET)

Mistake - Totals row should read:

[Totals]#[n/a]#[40]#[191]#[1874]



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!