When an astrophysicist starts testing mediums scientifically, the academic world at least listens. That is exactly what Prof. Archie Roy (1924–2012, Chair of Astronomy at the University of Glasgow) and his colleague Tricia Robertson did between the mid-1990s and the early 2000s. Their experimental series, developed over five years – today often referred to as the Glasgow Study or the Robertson-Roy Protocol – is regarded as one of the methodologically cleanest investigations of mediumistic work ever published. The result is striking: the probability that the mediums' hits arose by pure chance was around one in a million.
Who Were Roy and Robertson?
Archie Roy was no esoteric figure. He had been Professor of Astronomy at the University of Glasgow since 1977, a specialist in astrodynamics, celestial mechanics and archaeoastronomy, and later professor emeritus. For decades he was also active in the Society for Psychical Research (SPR) – a scientific society founded in London in 1882, whose stated aim is the methodologically rigorous investigation of paranormal phenomena. Tricia Robertson, also a long-standing SPR member and today President of the Scottish Society for Psychical Research, jointly developed with him an experimental design that addressed the classical skeptic's objection head-on.
PRISM: The Starting Point
The background was the PRISM initiative (Psychical Research Involving Selective Mediums), with which the SPR began in the mid-1990s to investigate selected mediums under controlled conditions. Roy and Robertson took the methodological lead and refined the experimental protocol step by step over five years – from single-blind through double-blind to triple-blind designs. Their results were published in 2001 and 2004 in the Journal of the Society for Psychical Research (JSPR).
The Hypothesis: The Forer Effect
The study tested a central skeptic hypothesis in a precise form:
"All mediumistic statements are so general that they could apply to anyone."
This is the so-called Forer or Barnum effect, familiar from horoscopes: vague statements such as "you are sometimes unsure" or "there has been an important loss in your family" can be applied to almost any life. If the hypothesis holds, people who actually received a reading should accept the statements as relevant to their lives no more often than people for whom the reading was not intended. That is exactly what was tested.
The Experimental Setup
The protocol roughly worked as follows:
- A medium gave a reading in the presence of a recipient – in the blinded variants without being able to see or directly speak to the recipient (sensory isolation, separate rooms).
- All of the medium's statements were recorded in full.
- The transcript was then given to the actual recipient and to several non-recipients – people for whom the reading had not been intended.
- Both groups had to assess independently how many statements applied to their own lives.
In the later experimental series, neither the medium nor the raters knew who the actual recipient was (double-blind); in some configurations the experimenters were also blinded during analysis (triple-blind). This consistently ruled out classic cold-reading clues – facial expression, gesture, vocal reaction, prior knowledge.
The Numbers
Across the experimental series, the central publications cover:
- 10 mediums
- 44 recipients in the first study, 407 non-recipients as a comparison group
- In the final series, around 300 participants and 73 readings over roughly 2.5 years
The Result
The actual recipients identified a significantly higher proportion of mediumistic statements as relevant to their own lives than the non-recipients – even though all sensory clues were excluded by the blinded setup. The statistical probability that this result arose by chance alone was, according to the authors, around one in a million. Under the conditions of the study, the Forer hypothesis was therefore not tenable.
The Publications
The main results appeared in three publications in the Journal of the Society for Psychical Research:
- JSPR 65/2 (April 2001), pp. 91–106: "A preliminary study of the acceptance by non-recipients of mediums' statements to recipients."
- JSPR 65/3 (2001), pp. 161–174: "A double-blind procedure for assessing the relevance of a medium's statements to recipients."
- JSPR 2004: "Results of the application of the Robertson-Roy protocol to a series of experiments with mediums and participants."
Putting It in Context
Like Jeffrey Long's NDERF work or the EREAMS study by Oliver Lazar, the Glasgow Study is not a laboratory experiment in the strict natural-science sense; it cannot prove how mediums obtain their information. What it does show, however – and with high statistical robustness – is this: the simple skeptic explanation that "it's all just generalities" does not hold. Anyone who wants to dismiss the phenomenon must develop alternative hypotheses that account for both the blinded conditions and the level of accuracy achieved – the simple Forer argument is not enough, given the Roy-Robertson data.
Roy himself – whose principal profession had trained him in scientific discipline – emphasised throughout his life that he did not favour any hypothesis on ideological grounds. What counted, he said, was the data. It is precisely this stance that makes the study, more than two decades after its publication, a key reference in serious mediumship research.
Sources:
• A. E. Roy & T. J. Robertson, A preliminary study of the acceptance by non-recipients of mediums' statements to recipients, Journal of the Society for Psychical Research 65/2 (April 2001), pp. 91–106.
• A. E. Roy & T. J. Robertson, A double-blind procedure for assessing the relevance of a medium's statements to recipients, JSPR 65/3 (2001), pp. 161–174.
• T. J. Robertson & A. E. Roy, Results of the application of the Robertson-Roy protocol to a series of experiments with mediums and participants, JSPR 2004.
• Psi Encyclopedia (SPR): Archie Roy.
• Tricia J. Robertson, Things you Can do When You're Dead / More Things you Can do When You're Dead (White Crow Books).
For more on scientific mediumship research, see our articles on Gary Schwartz and the VERITAS program, Julie Beischel's quintuple-blind work at the Windbridge Institute, Oliver Lazar's EREAMS study, Jeffrey Long's NDERF research, and Patrizio Tressoldi's meta-analysis.