In a transformation model h(Y)=X'beta+epsilon for some smooth and usua
lly monotone function h, we are often interested in the direction of b
eta without knowing the exact form of h. We consider a projection of h
onto a linear space of B-spline functions which has the highest corre
lation with the design variable X. As with the Box-Cox transformation,
the transformed response may then be analysed by standard linear regr
ession software. The direction estimate from canonical correlation cal
culations agrees with the least squares estimate for the approximating
model subject to an identifiability constraint. This approach is also
closely related to the slicing regression of Duan & Li (1991). The di
mensionality of the space of spline transformations can be determined
by a model selection principle. Typically, a very small number of B-sp
line knots is needed. A number of real and simulated data examples is
presented to demonstrate the usefulness of this approach.