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

Student's t-Test
Message posted by Jane on July 7, 2000 at 12:00 AM (ET)

I am illiterate when it comes to stats. However, I have a thesis to finish and need help!!
I have two groups of mice that are being fed different diets and I am using an unpaired Student's t-Test to compare their weights. An online calculator figured out all the details and returned these results to me:
"t=0.528 sdev=3.76
degrees of freedom=51
The probablility of this result, assuming the null hypothesis, is 0.600"

What does that tell me? Can anyone explain?


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

Re: Student's t-Test
Message posted by Phil on July 7, 2000 at 12:00 AM (ET)

Your null hypothesis was that the two feeding regimens produce no weight differencte (the two diets produced equal weights). The alternative (if you reject the null) is the two diets produce different weights.

Your results (small t-statistic) shows that you "cannot reject the null hypothesis" and you conclude that the diets had no significant effect on the weight.

The 0.60 is a "p-value". That is what the level of "Type I error" (alpha) would have to be before you would have rejected the null hypothesis in favor of the alternative.

So, if you were hoping the two diets would produce different weights, your data did not show that at all.


Re: Student's t-Test
Message posted by Phil on July 7, 2000 at 12:00 AM (ET)

This is a follow-up to my first response.

If you have the beginning weights for each subject, you could also consider doing an "analysis of covariance". This analysis would remove any influence of higher or lower starting weights. It could be that by correcting for starting weight you could get a significant difference after all.



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!