I'm trying to understand the confidence intervals for the effect sizes (omega^2) of a regression, but the Stata output does not include CI's for the omega^2.
Here is the output without the omega-squared specification:
regress STARx_Total FullTimeWork CollegeDegree
Source | SS df MS Number of obs = 71
-------------+---------------------------------- F(2, 68) = 0.92
Model | 104.306805 2 52.1534026 Prob > F = 0.4037
Residual | 3858.31291 68 56.7398958 R-squared = 0.0263
-------------+---------------------------------- Adj R-squared = -0.0023
Total | 3962.61972 70 56.6088531 Root MSE = 7.5326
-------------------------------------------------------------------------------
STARx_Total | Coefficient Std. err. t P>|t| [95% conf. interval]
--------------+----------------------------------------------------------------
FullTimeWork | 2.532173 1.86773 1.36 0.180 -1.194825 6.25917
CollegeDegree | -.9444689 2.784248 -0.34 0.735 -6.500349 4.611411
_cons | 49.21507 2.544936 19.34 0.000 44.13673 54.29341
-------------------------------------------------------------------------------
. estat esize
Effect sizes for linear models
-----------------------------------------------------------------
Source | Eta-squared df [95% conf. interval]
--------------------+--------------------------------------------
Model | .0263227 2 . .1175344
FullTimeWork | .0263188 1 . .1363192
CollegeDegree | .0016893 1 . .0654512
-----------------------------------------------------------------
Note: Eta-squared values for individual model terms are partial.
Here is the output with the omega-squared specification:
estat esize, omega
Effect sizes for linear models
-------------------------------------------
Source | Omega-squared df
--------------------+----------------------
Model | -.0022822 2
FullTimeWork | .0118306 1
CollegeDegree | -.0128037 1
-------------------------------------------
Note: Omega-squared values for individual
model terms are partial.
*In the above effect size output, it doesn't provide any column with CI's*
Is there code that I could include that could allow me to gain the CI's for the omega-squared?
Here is the output without the omega-squared specification:
regress STARx_Total FullTimeWork CollegeDegree
Source | SS df MS Number of obs = 71
-------------+---------------------------------- F(2, 68) = 0.92
Model | 104.306805 2 52.1534026 Prob > F = 0.4037
Residual | 3858.31291 68 56.7398958 R-squared = 0.0263
-------------+---------------------------------- Adj R-squared = -0.0023
Total | 3962.61972 70 56.6088531 Root MSE = 7.5326
-------------------------------------------------------------------------------
STARx_Total | Coefficient Std. err. t P>|t| [95% conf. interval]
--------------+----------------------------------------------------------------
FullTimeWork | 2.532173 1.86773 1.36 0.180 -1.194825 6.25917
CollegeDegree | -.9444689 2.784248 -0.34 0.735 -6.500349 4.611411
_cons | 49.21507 2.544936 19.34 0.000 44.13673 54.29341
-------------------------------------------------------------------------------
. estat esize
Effect sizes for linear models
-----------------------------------------------------------------
Source | Eta-squared df [95% conf. interval]
--------------------+--------------------------------------------
Model | .0263227 2 . .1175344
FullTimeWork | .0263188 1 . .1363192
CollegeDegree | .0016893 1 . .0654512
-----------------------------------------------------------------
Note: Eta-squared values for individual model terms are partial.
Here is the output with the omega-squared specification:
estat esize, omega
Effect sizes for linear models
-------------------------------------------
Source | Omega-squared df
--------------------+----------------------
Model | -.0022822 2
FullTimeWork | .0118306 1
CollegeDegree | -.0128037 1
-------------------------------------------
Note: Omega-squared values for individual
model terms are partial.
*In the above effect size output, it doesn't provide any column with CI's*
Is there code that I could include that could allow me to gain the CI's for the omega-squared?
