< Back to Volume 11, Number 2
Volume 11: Number 2: Article 3
The Astrology of Time Twins: A Re-Analysis
Christopher C. French and Antony Leadbetter, Department of Psychology,
Goldsmiths' College, University of London, New Cross, London SE14 6NW
United Kingdom
Geoffrey Dean, Analogic, Box 466, Subiaco 6008, Western Australia
Roberts and Greengrass (1994) tested the astrological claim that persons
born close together in time are more alike than those born far apart.
They collected a total of 128 subjects born on six dates spanning a
period of thirty years. Each subject provided their time of birth and
completed the short form EPQ. After comparing the interval between births
with the difference in EPQ scores the authors concluded that their data
showed no strong support for astrology. However, they claimed that the
proportion of close personalities increased as the birth interval decreased,
which would provide some support for astrology. We have re-analyzed
their data and find that their results can be explained as an artifact
of data division, sampling errors and age trends. A careful examination
of correlations and serial correlations (which are more appropriate
tests than the ones used by Roberts and Greengrass) confirmed the absence
of astrological effects in their data. If anything, the results were
in the wrong direction. A new finding not reported by Roberts and Greengrass
was a very small and non-significant difference between younger and
older subjects, consistent with the effect of prior knowledge, which
suggests that their sample may be slightly contaminated. We conclude
that an astrological interpretation of their results is unwarranted.
To purchase back issues contact Allen Marketing & Management: 1-800-627-0629
