Displaying 1 result from an estimated 1 matches for "kleinbaum".
2009 Feb 12
0
Comparing slopes in two linear models
...t2)$coefficients
db <- (s2[2,1]-s1[2,1])
#### TEST DEGREES OF FREEDOM
df <- (fit1$df.residual+fit2$df.residual)
### Pooled Variance ####
sd <- sqrt(s2[2,2]^2+s1[2,2]^2)
td <- db/sd
2*pt(-abs(td), df)
}
Spencer(dP, dHT)
[1] 0.0001717730
3- I wrote one based on Zar (1999) and Kleinbaum (1987) procedure:
Zar <- function(df1, df2){
fit1 <- lm( log(df1$varCDI)~log(df1$CDI) )
fit2 <- lm( log(df2$varCDI)~log(df2$CDI) )
s1 <- summary(fit1)$coefficients
s2 <- summary(fit2)$coefficients
db <- (s1[2,1]-s2[2,1])
#### TEST DEGREES OF FREEDOM
n1 <- fit1$df.residu...