Skip to content
Prev 223 / 585 Next

[epitools] Risk Difference

Incidentally a colleague of mine pointed out that you might be after a RISK difference, i.e. a difference of proportions.

Same trick:

D <- c(15,28) # Numerator
N <- c(22,61) # Denominator

# Hand calculation:
D/N
diff(D/N)

# Glm to give the sd easily:
summary( glm( cbind(D,N-D)~factor(0:1),family=binomial(link="identity") ) )

Bendix
Message-ID: <2F73FA3AF524144C863B7C2C903DBC3302B5D3D0EB@exdkmbx002.corp.novocorp.net>
In-Reply-To: <2F73FA3AF524144C863B7C2C903DBC3302B5D3D077@exdkmbx002.corp.novocorp.net>