Well the interview went well and I received an offer so thought I would come back and add to the topic in case it helps someone else one day!
It was for a Psychology PhD and the exam involved interpreting effect sizes in meta-analyses, a few simple stats questions (identifying the correct test to use), a question about longitudinal modelling and one on interpreting data by "eyeballing" it.
It is a heavily quantitative PhD though, so don't panic if you're invited to something similar, it may be easier or less technical.