Isn't this just calculating the difference and then using -ttest diff = #-
Example:
. sysuse auto
(1978 Automobile Data)
. gen diff = disp - length
. ttest diff = 10
One-sample t test
------------------------------------------------------------------------------
Variable | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]
---------+--------------------------------------------------------------------
diff | 74 9.364865 8.632391 74.25863 -7.839465 26.56919
------------------------------------------------------------------------------
Degrees of freedom: 73
Ho: mean(diff) = 10
Ha: mean < 10 Ha: mean != 10 Ha: mean > 10
t = -0.0736 t = -0.0736 t = -0.0736
P < t = 0.4708 P > |t| = 0.9415 P > t = 0.5292
If you want to set it equal to the standard deviation you could define a local
macro as the standard deviation, or use it immediately following -summarize-
. sum diff
Variable | Obs Mean Std. Dev. Min Max
-------------+--------------------------------------------------------
diff | 74 9.364865 74.25863 -80 204
. ttest diff = `r(sd)'
One-sample t test
------------------------------------------------------------------------------
Variable | Obs Mean Std. Err. Std. Dev. [95% Conf. Interval]
---------+--------------------------------------------------------------------
diff | 74 9.364865 8.632391 74.25863 -7.839465 26.56919
------------------------------------------------------------------------------
Degrees of freedom: 73
Ho: mean(diff) = 74.25863
Ha: mean < 74.25863 Ha: mean != 74.25863 Ha: mean > 74.25863
t = -7.5175 t = -7.5175 t = -7.5175
P < t = 0.0000 P > |t| = 0.0000 P > t = 1.0000
Hope this helps,
Scott
----- Original Message -----
From: "Wandland, Kristen" <[email protected]>
To: <[email protected]>
Sent: Wednesday, January 14, 2004 11:01 AM
Subject: st: specifying difference in Ho of ttest
> Does anyone know a way to specify the difference in a ttest? The default
> result is:
>
> Ho = diff = 0
>
> I hope to be able to do:
>
> Ho = diff = #
>
> where # is the standard deviation or other pre-determined value.
>
> Thanks all.
>
> Kristen Durocher Wandland
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
