Gervase--

The test that you linked to above, at least at first review, seems not to have been designed very well.
The scientist used the astrologers feedback to structure the research, which seems fair enough at the face of it, but the fact is that astrologers, for the most part, have little to know idea how to structure a valid scientific test for anything, even astrology--they are not scientists--

A variety of practical problems skewed the selection of subjects, so that all had certain similar, and therefore confusing, elements in their charts. Another serious problem was that there was no overarching qualification for the astrologers, they were self-selected, and there was no attempt to discern the competency of the astrologers --or even to determine what the criteria for a competent vs an incompetent astrologer would be.

Mind you, I only read over the description briefly--but it didn't seem like a very good study.