Cognition and perception

Extra-sensory perception - a controversial debate

Eric Robinson, winner in the postgraduate category of our student writer competition, weighs up the evidence

06 July 2009

Extra-sensory perception (ESP) is a term often scoffed at in psychology and wider science: an alleged ‘paranormal’ or supernatural phenomenon that many believe is best suited to science fiction films. Yet, to the surprise of many academics, a significant body of scientific evidence exists which may suggest otherwise.

ESP has been defined as ‘anomalous processes of information or energy transfer, processes such as telepathy… that are currently unexplained in terms of known physical or biological mechanisms’ (Bem & Honorton, 1994, p.4). Although unexplained by current scientific thought, surveys continually indicate that belief in ESP is extremely common (Moore, 2005) and many individuals are adamant that they have experienced ESP at first hand (Greeley, 1987). Additionally, psychologists and sociologists have collected thousands of retrospective case reports of alleged ESP phenomena (Irwin & Watt, 2007).

However, our knowledge of the human mind raises considerable doubts over the accuracy of such anecdotal evidence. Extensive research has shown us that memory can be unreliable, and that cognitive biases result in events regularly and easily being misinterpreted (Henkel & Mather, 2007; Kahneman et al., 1993). Furthermore, the laws of probability mean that weird coincidences that may well appear ‘ESP-like’ can’t help but happen. For example, thinking of a friend you have not spoken to for sometime and then receiving a telephone call from that very person may be seem too much of a coincidence, best explained by some form of ESP. Alternatively it may just be pure coincidence; how many times is a distant friend thought of and they don’t call? If we also consider fraud and mistakenness, anecdotes are reduced to something only vaguely resembling scientific evidence.

Yet the assumption that belief in ESP is related to lower IQ and poorer reasoning skills has been shown to be inaccurate (Roe, 1999). Indeed, education level has even been shown to be positively correlated with belief in ESP (Rice, 2003). Those that believe in the possibility of ESP are also in good company; William James, Carl Jung and Nobel Prize winner Charles Riche to name but a few great minds.

Although many feel antipathy towards proposed paranormal and occultist suggestions such as ESP, Freud wrote that ‘This disinclination must ultimately be overcome. What we are dealing with is a question of fact’ (1940/2003, p.29). Parapsychology scientifically investigates the possibility of ESP, and has collected a large body of evidence which some suggest support such a remarkable claim (Bem & Honorton, 1994; Sherwood & Roe, 2003). The most common experimental design that has been used to examine the ‘ESP hypothesis’ is the ganzfeld procedure.

The ganzfeld experiment

A typical ganzfeld experiment involves two participants. Participant 1, known as ‘the receiver’ is seated in a comfortable chair in an acoustically isolated room. Translucent ping-pong ball halves are taped over their eyes, and a red floodlight shone over them while white noise is played through headphones. These measures are taken to reduce external noise and place participants in a comfortable ‘dreamy state of awareness’. The reasoning behind such procedures is that anecdotal reports of ESP often occur during altered states of consciousness. Participant 2, otherwise known as ‘the sender’, is located in a different room. A computer randomly selects a stimulus or ‘target’ (typically  a photograph or video) from a large pool. It is the sender’s job to concentrate on the target and attempt to mentally send it to the receiver. During this time the receiver provides a continuous verbal report of any imagery or thoughts. After the session the receiver is presented with four stimuli (one is the target and the other three serve as decoys) and asked to select which of the four was most similar to their mentation.

By chance we would expect participants to select the target stimuli (otherwise known as a ‘hit’) on 25 per cent of trials.

Early ganzfeld experiments produced highly significant results, well above what chance would predict for performance. Meta-analyses by parapsychologist Charles Honorton and sceptic Ray Hyman, examining 28 studies taking place between 1974 and 1981, reported a hit rate of 35 per cent (Honorton, 1985). Although the 10 per cent deviation may seem small, over so many trials this is a robust finding that is extremely unlikely to be explained by chance deviation. The studies also yielded an impressive effect size (0.5 is normally considered a medium-sized effect in the social sciences) of 0.63 (Bem & Honorton, 1994).

Furthermore, the effect was replicated by numerous researchers (Honorton, 1985). However, concern was raised over a number of possible methodological flaws surrounding the experiments, including cues through sensory leakage and poor randomisation of target stimuli (Hyman, 1985). Interestingly, a separate US National Research Council report by esteemed social psychologist Robert Rosenthal also accepted these flaws, but suggested that they were highly unlikely to explain the remarkably consistent effect (Harris & Rosenthal, 1988).

Nevertheless, accepting such a controversial hypothesis based on evidence coming from methodologically flawed experiments is poor science. Based on the previous problems of early experiments, a new form of ganzfeld protocol was developed; ‘the auto-ganzfeld’, whereby randomisation and selection of stimuli were completely computerised and procedural rigour tightened.

The auto-ganzfeld procedure

In 1994, Psychology Bulletin published an article by Cornell’s Daryl Bem and the late Charles Honorton analysing all auto-ganzfeld studies. Up to then, 354 auto-ganzfeld sessions had taken place during 11 studies. Again, results appeared to be in support of the ESP hypothesis. Reminiscent of the earlier ganzfeld studies, a significant hit rate of 32 per cent was observed. In search of better understanding of the effect taking place, the authors also examined internal effects and suggested that degree of extroversion and belief in the possibility of ESP predicted performance (see Bem & Honorton, 1994).

Such findings inevitably encouraged researchers to continue ganzfeld experiments. In 1999 Milton and Wiseman conducted their own meta-analysis of all auto-ganzfeld studies in the published literature. Surveying 30 studies, the authors’ analysis found no significant effect (Milton & Wiseman, 1999).

Nonetheless, there have been objections to the conclusions drawn from the analysis. The quality of the experiments analysed has been called into question. During this period many studies were ‘process-orientated’ and altered various aspects of the standard ganzfeld procedure, potentially removing conditions that may have facilitated ESP (Irwin & Watt, 2007). Nevertheless, over the 30 studies that explicitly investigated the existence of ESP, no evidence in support of the hypothesis was found. Additionally, the previous significant effects of individual differences in performance were not replicated.

Due to such criticisms another meta-analysis followed in 2001. Finding an additional 10 later studies, Bem and Palmer (2001) analysed 40 studies overall. The average hit rate was 30.1 per cent; a significant effect had returned. Furthermore, the proposal that studies tampering with the previously successful standard ganzfeld procedure may have been responsible for non-significant findings gained some support. Studies conforming to the conventional procedure yielded significant results, whereas those that altered the methodology tended to produce chance scoring (Bem & Palmer, 2001). Yet, the effect size was much smaller than those reported in earlier experiments, suggesting that if an effect was taking place, it was only faint.

A future for ESP?

Although it is not overwhelming evidence, experiments may suggest that  a small effect is taking place. However, the small number of papers providing evidence in reputable psychological journals do not appear to have changed opinion or attracted much more scientific attention towards ESP.

Shermer (2003) suggests the major reasons are that (a) the effect is extremely difficult to replicate and (b) parapsychology lacks a unified and valid theory to explain such an anomaly. But how important are these? Science is primarily based on observation followed by explanation through theory. Some parapsychologists suggest the size of the field may explain why a substantial ‘breakthrough’ has yet to be made. Schouten (1993) calculated that in the last 111 years, the total amount of human and financial resources dedicated to parapsychology is the equivalent to the resources available to sustain all psychological research for a mere two months in the US. An alternative view may be: if such an ability or phenomenon exists, then surely 111 years of academic study should have provided enough evidence for opinion to be swayed?
Nevertheless, parapsychologists have reported a number of findings that they suggest may explain the phenomena. Among a number of findings, analysis suggests that believers in ESP (possibly due to motivational effects) tend to outperform sceptics (Lawrence, 1993) and participant mood may also be related to experimental performance (Carpenter, 2001). Yet, as with most significant findings in this field, the sizes of these effects are very small.

When considering the possibility of ESP, Freud was correct in reminding us that ‘the easiest explanation is not always right one: the truth is often not terribly simple’ (1940/2003, p.34). It is of importance to remember that our thoughts on physics, biology and psychology have been way off the mark before and continue to evolve. Hypotheses (backed by scientific evidence) that make the mainstream academic feel a little uncomfortable, whether later accepted or rejected, are what push understanding forward – not scientific dogma.

BOX: Weird coincidences in the laboratory?

The following extract from a paper by Westerlund et al. (2004) reporting ganzfeld experiments at the University of Edinburgh, shows strong similarities between the video the ‘sender’ was watching and the receiver’s thoughts at the time.

‘On this tape, the target video clip is shown and at the same time the mentation of the receiver can be heard. One of the most remarkable excerpts shows a man who is running through a forest; it seems that he is being hunted (at the same time, the receiver says: “Trees. People running. Fleeing…”). Suddenly, the man falls down in a deep muddy pool (at the same time, the receiver says: “Falling. Muddy…”). The camera zooms in on the man’s face (at the same time, the receiver says: “Blond hair. 70´s hairstyle. Curly-ish. White face…”. All the utterances appear to describe exactly what is being shown on the film. The next thing that happens in the clip is that the man can no longer keep his head above the surface, so he disappears into the mud (at the same time, the receiver says: “Dead man in the water”)

- Eric Robinson is in the Ingestive Behaviour Group at the University of Birmingham, and the Psychology of Paranormal Phenomena Research Group at the University of Derby
[email protected]