[R] PREDICT NEW VALUES FROM REGRESSION MODEL, EST. ST.ERROR, AND CI

Ricardo Gomez ricardogomeznz at yahoo.co.nz
Wed Dec 17 16:46:41 CET 2008


Greetings, 

I'd be grateful if a good Samaritan  helps me to approach this problem.... 

with my data, I've created the following model 

lm(formula = OUTCOME ~ VAR1 + VAR2) 
 summary(model) 

Call: 
lm(formula = OUTCOME ~ VAR1 + VAR2) 

Residuals: 
Min 1Q Median 3Q Max 
-1.4341 -0.3621 0.1879 0.4994 0.7696 

Coefficients: 
Estimate Std. Error t value Pr(>|t|) 
(Intercept) 1.89020 0.26826 7.046 5.92e-07 *** 
VAR1 0.04725 0.06001 0.787 0.440 
VAR2 0.04139 0.05655 0.732 0.472 

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1 

Residual standard error: 0.6618 on 21 degrees of freedom 
Multiple R-squared: 0.9474, Adjusted R-squared: 0.9424 
F-statistic: 189.2 on 2 and 21 DF, p-value: 3.696e-14 

but now, I need to predict OUTCOME (Y) when VAR1=8 and VAR2 =64; 
estimate the standard error of the predicted value, and construct a 95% CI 

Your help is much appreciated 

RG 
*****************************************
Ricardo L Gomez
Center for International Education
University of Massachusetts-Amherst
Telephone: (413)545-0465 | Fax: (413)545-1263
Web Address http://www.umass.edu/cie
E-mail: cie at educ.umass.edu



      Get the world's best email - http://nz.mail.yahoo.com/



More information about the R-help mailing list