| |
 |
|
|
Science Forum Index » Statistics - Math Forum » significance testing on indexed data
Page 1 of 1
|
| Author |
Message |
| u9946675 |
 Posted: Tue Mar 13, 2007 9:30 am |
|
|
|
Guest
|
Hi all
I have a data set which is split into several groups. The values of
various items (all integer values) are indexed on the average (which
is equal to 100). Can anyone tell me a simple formula or method to
calculate whether the average score for a given group is significantly
different to the indexed (=100) average score, using the indexed
values, not the absolutes.
Thanks! |
|
|
| Back to top |
|
| Richard Ulrich |
| Posted: Tue Mar 13, 2007 11:01 pm |
|
|
|
Guest
|
On 13 Mar 2007 07:30:00 -0700, "u9946675"
<chris@chrislooney.wanadoo.co.uk> wrote:
Quote: Hi all
I have a data set which is split into several groups. The values of
various items (all integer values) are indexed on the average (which
is equal to 100). Can anyone tell me a simple formula or method to
calculate whether the average score for a given group is significantly
different to the indexed (=100) average score, using the indexed
values, not the absolutes.
What you want is not at all clear.
If you want to test a group against a normative sample,
you can plug in the means and standard deviations,
for both groups (one is mean=100?, s.d.= ??) using the
formula for a t-test for two samples.
If the normative sample is large enough, or if you just
want to test against a mean of 100, you can use a one-sample
t-test.
--
Rich Ulrich, wpilib@pitt.ed
http://www.pitt.edu/~wpilib/index.html |
|
|
| Back to top |
|
| Ray Koopman |
| Posted: Wed Mar 14, 2007 1:11 am |
|
|
|
Guest
|
u9946675 wrote:
Quote: Hi all
I have a data set which is split into several groups. The values of
various items (all integer values) are indexed on the average (which
is equal to 100). Can anyone tell me a simple formula or method to
calculate whether the average score for a given group is significantly
different to the indexed (=100) average score, using the indexed
values, not the absolutes.
Thanks!
What do you mean by "group" and "item"? I suspect you're using those
terms somewhat differently than they are usually used in statistical
discussions. Please be specific about how the data were generated and
what the questions of interest are. |
|
|
| Back to top |
|
| |
|
Page 1 of 1
All times are GMT - 5 Hours
The time now is Mon Dec 01, 2008 9:11 am
|
|