| Author |
Message |
< ASReml ~ Fixed effects in animal model
|
| kabuck |
Posted: Tue Jul 03, 2012 4:45 pm |
|
|
|
Joined: 23 Feb 2012
Posts: 6
|
I have run a conditional test for fixed effect in my univariate animal model, and am trying to interpret the direction of the effect (e.g. females lower than males, older vs. younger individuals, etc.). How can I get ASReml to report the solution and standard error for each parameter (both male and female, etc.) of the fixed effect? The Wald-F table doesn’t seem to report all values.
Wald F statistics
Source of Variation NumDF DenDF_con F-inc F-con M P-con
11 mu 1 48.0 3826.70 1306.35 . <.001
5 DATE 1 680.8 16.56 19.42 A <.001
4 AGE 1 151.7 10.84 9.83 A 0.002
2 SEX 1 80.2 19.85 19.85 A <.001
Notice: The DenDF values are calculated ignoring fixed/boundary/singular
variance parameters using numerical derivatives.
Solution Standard Error T-value T-prev
2 SEX
Female -0.600090 0.134764 -4.45
4 AGE
1 -0.842220E-03 0.268754E-03 -3.13
5 DATE
1 -0.502139E-01 0.113967E-01 -4.41
11 mu
1 5.21659 0.151644 34.40 |
|
|
| Back to top |
|
| Arthur |
Posted: Thu Jul 05, 2012 5:49 am |
|
|
|
Joined: 05 Aug 2008
Posts: 277
Location: Orange, NSW
|
Dear Kluback,
From
Solution Standard Error T-value T-prev
2 SEX
Female -0.600090 0.134764 -4.45
4 AGE
1 -0.842220E-03 0.268754E-03 -3.13
5 DATE
1 -0.502139E-01 0.113967E-01 -4.41
11 mu
1 5.21659 0.151644 34.40
Assuming sex was a factor with Level 1 Male, Level 2 Female,
the effect for Male is 0 (see .sln file)
the effect for female is -0.6 so males
so females are 0.6 lower than males.
I can't tel whether AGE is a factor with 2 levels, or a covariate. Assuming it is a covariate (say in days), the Y variable descrease 0.00084 per day
Si,ilarly I can't tell if DATE is factor or covariate but assuming a covariate in days, Y decreases .05 per day with respect to DATE.
Does that help. |
_________________ Arthur Gilmour
Retired Principal Research Scientist (Biometrics) |
|
| Back to top |
|
|
|