From the archives: Parapsychology – when is evidence sufficient?

Author

Chris Roe
Chris Roe is Senior Lecturer and Research Leader for the Psychology Division at the University of Northampton. He has been a Board member of the Parapsychological Association, Council member of the Society for Psychical Research, and is currently editor of the SPR’s Journal. His research has focused on the psychology of anomalous experience and belief and the effects of psychological factors on psi performance in the laboratory.

More from this author

- Advertisement -spot_img

This article originally appeared in The Skeptic, Volume 22, Issue 2, from 2012.

Ray Hyman would have us believe that “parapsychology is dead”. In this response I shall consider the reasons he provides for this claim and show – to paraphrase Mark Twain, an early member of the Society for Psychical Research and author of articles on ‘mental telegraphy’ (Bratcher, 2008) – that the report of parapsychology’s death is an exaggeration.

Hyman claims that the goal of parapsychology is (and implicitly always has been) to gain the recognition of the scientific community and to meet the most exacting of scientific standards of evidence, and in this respect parapsychology has been unsuccessful. I am happy to concede that this may have been one of a number of motivations behind the approach adopted by J.B. and Louisa Rhine in founding and developing their laboratory at Duke University from the 1930s to the 1960s (a period that Beloff, 1993, refers to as the “Rhine Revolution”), but of course it is far removed from the primary aim of parapsychology. The founders of the SPR did not set out on a public relations exercise but rather were committed to extend the reach of the scientific method to include more contentious claims, to “examine without prejudice or prepossession and in a scientific spirit those faculties of man, real or supposed, which appear to be inexplicable on any generally recognized hypothesis”.

Contemporary definitions of the discipline similarly emphasize the application of available methodological and conceptual tools to make best sense of an array of phenomena that superficially seem difficult to accommodate in our current worldview, so that parapsychology is the “study of experiences which, if they are as they seem to be, are in principle outside the realm of human capabilities as presently conceived by conventional scientists” (Irwin & Watt, 2007, p. 1, emphasis added). As Caroline Watt has described in this exchange, this means that parapsychologists are increasingly interested in “what seems psychic but isn’t” (Morris, 1986) in their efforts to understand phenomena rather than seeking support for some prior esoteric worldview (which at times they are unfairly caricatured as doing). However, parapsychologists recognise that this work towards conventional explanations must be complemented with more direct tests of the putative phenomena under controlled conditions if we are to do justice to the experiences that people report, particularly given that the results of such experiments are claimed to provide evidence for anomalies in information exchange that might require revisions in elements of our worldview (but see Dobyns, in press).

In the next section I shall briefly consider whether parapsychologists have grounds to claim persuasive evidence for the occurrence of effects under conditions that preclude normal explanations, but first would like to respond to Hyman’s claim that parapsychology’s attempts at scientific respectability are “unrealistic and unachievable”. Of course, acceptance is a matter of degree and much work remains to be done, but there has been some positive movement in the normalisation of academic parapsychology in recent years, in the UK at least. Delanoy (in press) has noted that at least 16 UK universities have fulltime academic staff whose doctoral training was solely or primarily in parapsychology, and a similar number are currently engaged in graduate training that is accredited by these institutions.

Parapsychology has featured regularly at conferences organised by the British Psychological Society (I personally have had papers accepted for the BPS’s annual conference as well as those organised by the sections concerned with Social Psychology, Transpersonal Psychology and Consciousness and Experiential Psychology), which tends to argue against an institutional antipathy. ‘Anomalistic psychology’ has been included in the specification for A2 Psychology (the standard pre-university qualification) by the Assessment and Qualifications Alliance (AQA), the assessment body that has much the largest market share of the 50,000 students who take Psychology each year. This specification includes elements on experimental testing of ESP and PK, so that future undergraduates will come to university with a grounding in parapsychology and an expectation that the subject will be represented on any comprehensive undergraduate syllabus. This is not really the profile of a subject that is “consigned to the fringes of academia”, as Richard Wiseman claims in his contribution here.

Hyman makes a number of rather vague allusions to a lack of quality in parapsychological research, which undermines any claim to have demonstrated the occurrence of an anomalous effect. He asserts, for example, that the evidence “fails to meet established scientific criteria”, but apart from some general concerns about the limitations of meta-analysis (which are clearly not specific to parapsychology) it is not clear in what way(s) parapsychological research is deficient; without specifying what “established criteria” are required for a study to qualify as of high quality this claim remains insubstantial and unfalsifiable – it is not a scientific claim.

So what might those criteria include? Watt and Nagtegaal (2004) have described the use of double-blind methods – in which the experimenter and others who may have some involvement in the data generation and recording phases are at that time unaware of the nature of the target and so cannot  unwittingly cue the participant – as a possible indicator of study quality (while acknowledging that otherwise poor-quality designs could still include blinds). It is well established that the outcome of a study can be affected by the experimenter’s prior knowledge and expectations, leading to artifacts if this is not adequately addressed (Rosenthal, 1976). But when Sheldrake (1999) surveyed various journal publications he found that the use of double-blinds was very rare in the physical and biological sciences (0% and 0.8% respectively), somewhat more common in medical sciences (24.2%) but most common in parapsychology (85.2%); surprisingly, other research in psychology and animal behaviour used blinded methods even less than in medicine (4.9%), despite the emphasis on the problems of demand characteristics in social science methods training. Watt and Nagtegaal (2004) conducted a second survey to see if Sheldrake’s findings could be replicated and found a similar profile (parapsychology: 79.1%, Psychology: 14.5%). By this criterion, then, it is difficult to justify the general claim that parapsychological research is methodologically poor in comparison with other social science research.

Hyman bemoans parapsychology’s poor replication record and, taking the Ganzfeld method as his example case, rather selectively uses one study by Broughton and Alexander (1997) to illustrate his point – one could just as easily have chosen Parker’s (2000) automated Ganzfeld database, consisting of 150 trials that gave a hit rate of 36% (z = 3.02), or Dalton’s (1997) series of 128 trials that gave a hit rate of 47% (h = .46). Which of these should we prefer? None. While it is tempting to focus on the outcomes of individual studies, particularly when so few people are professionally engaged in parapsychological research and it takes so long to build up anything like a reasonable database, we must accept that individual studies are susceptible to giving outcomes that reflect sampling error and also are affected by idiosyncratic features of the experimental environment – not obscure or mystical features, but ordinary factors such as experimenter-participant rapport or differences in recruitment strategy. For this reason we should prefer summary reviews (while remaining cognizant of their shortcomings) when making judgments about the robustness of effects. That is not to suggest that the outcomes from ganzfeld studies are robust and heterogeneous (they are not), but a good case can be made to argue that at least some of this variation is due to different researchers posing different research questions as they move from simply demonstrating a phenomenon to characterising and explaining it. This typically leads to variations in the procedure followed so that more or less conducive conditions can be compared or more radical departures from established procedure can be tested, with the expectation that not all variations are likely to be psi-conducive. In this context simply combining outcomes across studies to give a bottom-line effect size would be rather naïve (see Roe, in press a, for a fuller consideration of this point). Where ganzfeld studies have been coded according to the degree to which they adhere to Honorton’s ‘standard’ protocol (Bem, Palmer & Broughton, 2001), there is a very strong correlation between ‘standardness’ and effect size, with the subset of studies that closely replicate Honorton’s approach giving a hit rate of 31.2% that is highly significant and falls comfortably within Bem and Honorton’s 95% confidence interval for their autoganzfeld effect size estimate. That looks rather like replication to me.

A number of other strands of research in parapsychology offer some promise of also reliably capturing (admittedly small) effects, including work that takes more direct physiological measures (e.g., Radin, 2004) and studies that look for incidental psi effects in overtly cognitive tasks (e.g., Bem, 2008; Luke, Roe & Davison, 2008). Space does not permit me to consider these in any detail, but work to date suggests that these effects are reasonably replicable – certainly in comparison to the rather weak and inconsistent effects that are regarded as support for the more sceptical claim that paranormal beliefs are associated with cognitive deficits or social marginality (see Roe, in press b). I would be interested to know what levels of replication would be regarded as appropriate or reasonable for parapsychological work with human participants.

Finally, I should like to respond to the tendency of sceptical commentators to refer to the achievements of parapsychology in relation to the time since it became recognisable as a separate discipline. In this exchange, for example, Wiseman complains that “after over a century of work mainstream science remains sceptical of psi” and Hyman begins by noting that parapsychology has lasted “approximately 160 years”. This rhetorical device encourages the reader to assume that extension over time also translates into extensive and intensive activity, so raising their expectations about the degree of progress that might be reasonable. Against this, any actual progress pales, and the reader is moved to conclude that the phenomena are non-existent rather than elusive. But Caroline Watt has already drawn attention to the fact that very few persons are professionally involved in parapsychology, and even these typically have to combine their research activity with other academic duties. This situation is not new: Thouless (1953, p. 23) warned that “experimental workers in our subject are so few that we cannot afford wasted effort”, and Schouten (1993) calculated that the person-hours invested across the lifetime of parapsychology from its beginnings with the establishment of the Society for Psychical Research in 1882 equates to only two months’ of research in conventional psychology in the United States. In that context the body of evidence may seem more impressive.

But limitations in human resources and in funding affect more than gross productivity; they also shape the character of the research that is carried out. Elsewhere (Roe, in press c) I have described parapsychology as a “butterfly science” in which key researchers seem to shift from one protocol to another as they fall in and out of fashion much as a butterfly flits from flower to flower. Sceptical commentators have tended to regard this as suspect, much as Wiseman does in his commentary here. However, in my view this pattern occurs because the small number of researchers in parapsychology includes a disproportionately large number of innovators and early adopters and too few technicians. Innovators thrive on developing new protocols or adapting methods from other areas and in demonstrating ‘proof of principle’ by reporting significant psi effects using such methods; ‘early adopters’ are quick to seize on these new approaches and technologies and are responsible for the first wave of independent replications. However, relatively quickly the innovators lose interest in simple confirmations and move on to develop yet more methods and approaches, with the early adopters soon following suit. I am sure that this pattern also occurs in other disciplines, but with their greater numbers they also include many able technicians who are willing to conduct the kinds of modest replication extensions that Kuhn would have called ‘normal science’. (We have fewer technicians, but they do exist; for example, I found that there had been ten further ganzfeld studies conducted after the last meta-analytic reviews, and this database was independently significant – see Roe, in press a). The pattern is exacerbated by the intense competition for funding, which encourages novelty and innovation and which requires that results are forthcoming in the short to medium term; both of these select against the systematic and methodical programmes of the would-be technician that could provide the empirical foundation that sceptics are looking for.

At best this research cycle is frustrating in diverting resources away from a potentially fruitful avenue of research; at worst it looks suspicious to the outsider, who expects to see continuing and systematic work using a particular method for so long as it is productive, particularly where great claims were initially made for it – why are there so few micro-PK studies? So few ganzfeld studies? I agree with Wiseman and Watt that collectively parapsychologists need to address this issue by collaborating on a more systematic programme of research, one that goes beyond proof of principle and first wave independent replications. I believe that parapsychologists would be resolved to commit to such a programme, but in practice it would require a radical overhaul of the few funding mechanisms that are available to parapsychology for it to succeed; without a commitment of significant resources over the longer term it is difficult to see how parapsychologists could commit to Wiseman’s three basic rules.

References

  • Beloff, J. (1993). Parapsychology: A concise history. London: Athlone Press.
  • Bem, D.J. (2008). Feeling the future III: Additional experimental evidence for apparent retroactive influences on cognition and affect.International Conference of the Society for Psychical Research held Jointly with the Parapsychological Association, University of Winchester, UK, 13th to 17th August, 2008.
  • Bratcher, C. (2008). Mark Twain and ‘mental telegraphy’. International Conference of the Society for Psychical Research held Jointly with the Parapsychological Association, University of Winchester, UK, 13th to 17th August, 2008.
  • Dalton, K. (1997). Exploring the links: Creativity and psi in the ganzfeld. Proceedings of Presented Papers: The Parapsychological Association 40thAnnual Convention in Conjunction with the Society for Psychical Research, 119-134.
  • Delanoy, D. (in press). Parapsychology in a university setting. In C.A Roe, L. Coly & W. Kramer (Eds.) Utrecht II: Charting the future of parapsychology. NY: Parapsychology Foundation, Inc.
  • Dobyns, Y. (in press). Physics with an open mind. In C.A Roe, L. Coly & W. Kramer (Eds.) Utrecht II: Charting the future of parapsychology. NY: Parapsychology Foundation, Inc.
  • Luke, D.P., Roe, C.A., & Davison, J. (2008). Testing for forced-choice precognition using a hidden task: Two replications. Journal of Parapsychology, 72, 133-154.
  • Morris, R.L. (1986). What psi is not: The necessity for experiments. In H.L. Edge, R.L. Morris, J. Palmer, & J.H. Rush, Foundations of parapsychology: Exploring the boundaries of human capability (pp. 70-110). London: Routledge & Kegan Paul.
  • Radin, D.I. (2004). Electrodermal presentiments of future emotions. Journal of Scientific Exploration, 18, 253-273.
  • Roe, C.A. (in press a). The role of altered states of consciousness in extrasensory experiences. In M. Smith (Ed.), Developing perspectives on anomalous experience. Jefferson, NC: McFarland & Co.
  • Roe, C.A. (in press b). Anomalistic psychology. In N. Holt & R. Lewis (Eds) A2 Psychology 2009 AQA A Specification: The Student’s Textbook. London: Crown House Publishing.
  • Roe, C. A. (in press c). Personal reflections on Utrech II. In C.A. Roe, L. Coly, & W. Kramer (Eds) Utrecht II: Charting the future of Parapsychology. NY: Parapsychology Foundation, Inc.
  • Schouten, S. (1993). Are we making progress? In L. Coly & J. McMahon (Eds.) Psi research methodology: A re-examination. (pp. 295-322). NY: Parapsychology Foundation, Inc.
  • Thouless, R.H. (1953). Experimental design in parapsychology. Proceedings of the first international conference of parapsychological studies (pp. 21-23). New York, NY: Parapsychology Foundation, Inc.
  • Watt, C., & Nagtegaal, M. (2004). Reporting of blind methods: An interdisciplinary survey. Journal of the Society for Psychical Research, 68, 105-114.

The Skeptic is made possible thanks to support from our readers. If you enjoyed this article, please consider taking out a voluntary monthly subscription on Patreon.

- Advertisement -spot_img

Latest articles

- Advertisement -spot_img

More like this