[R] Cross-over Data with Kenward-Roger correction

[peter dalgaard]

On 08 Jun 2015, at 22:59 , Ben Bolker <bbolker at gmail.com> wrote:

> As a statistical point, I don't understand why you can't just
> do a paired t-test on these data??

It's a cross-over trial, so you need to account for the period effect. However, a standard way of treating it is by to calculate period differences within subjects and use a t-test (welch or classical) to see if the differ according to sequence (trt1-trt2 or v.v.).

-- 
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Office: A 4.23
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com