Natal astrology under controlled matching

The most direct controlled design asks astrologers to match charts to independently measured profiles, or asks participants to identify their own interpretation among decoys. Blinding prevents names, conversational feedback, and obvious demographic cues from doing work attributed to the chart.

Carlson’s widely discussed 1985 double-blind test used astrologer-approved procedures and found performance consistent with chance. The study has been criticised over analysis and implementation, which is why it should not stand alone as the single trial that decides the subject.

Reviews have repeatedly found that apparent successes shrink when controls remove flexible scoring, multiple testing, and cues available after birth. Large-scale work comparing birth date with measured personality and intelligence has likewise failed to recover the ordinary zodiac predictions.

This pattern matters more than whether one old study can be re-analysed into a marginal result. A durable effect should survive new samples, explicit outcomes, and analysis choices written before the data are inspected.

A public register can contribute something a retrospective argument cannot: a protocol fixed in advance, every completed entry retained, an exact chance baseline, and data available for hostile as well as friendly reanalysis. It also tests the material readers actually meet now rather than a reconstruction of a decades-old consultation.

The expected result under the current evidence is chance performance. Publishing that expectation before collection makes a contrary result more informative, not less, while a null result becomes one transparent addition to a long and uneven record.