02chapters1 6
02chapters1 6
02chapters1 6
INTRODUCTION TO PARAPSYCHOLOGY
In this first chapter, a socio-empirical1 approach is taken to explore the current state
of parapsychology2 and its subject matter: alleged paranormal3 phenomena, sometimes
referred to as psi.4 Skepticism from the point of view of polemicists who have “already
decided the issue” (Hansen, 1991, p. 202) about these phenomena is the main focus of this
chapter, and three forms of inquiry will be made that ultimately challenge this extremely
skeptical position on psi. Of course, only a limited number of issues can be covered in a
single chapter. First, the singular yet paradoxical nature of psi is scrutinized in terms of its
specifiability as an ‘acceptable’ scientific phenomenon by comparison with other allegedly
specified scientific phenomena. Second, the impact on parapsychology of the philosophical
and sociological factors of ‘paradigm incommensurability’ and ‘socio-cognitive
discontinuity’ is examined. Third, three socio-empirical effects on
1
The term ‘socio-empirical’ is used here to describe a fusion of, or inter-relatedness between, sociological
and scientific ideas and principles both of which underlie the so-called ‘objective’ findings and statements of
scientists.
2
Parapsychology is the “scientific study of [alleged] paranormal phenomena” (Thalbourne, 1982, p. 51). See
n3 for a definition of paranormal.
3
Both paranormal and psi phenomena “[exceed] the limits of what is deemed physically possible on current
scientific assumptions” (Thalbourne, 1982, p. 50). The general distinction in parapsychological circles is that
psi “simply denotes a communications anomaly” (Hyman & Honorton, 1986, p. 353), whereas paranormal
phenomena “must be adequately explained scientifically . . . by a theory or model the core assumption(s) of
which violate one or more of C.D. Broad’s so-called ‘Basic Limiting Principles’ ” (Palmer, 1988, p 154; see
also, Palmer, 1987).
4
Thouless and Wiesner, 1947, p. 179, proposed a unitary process underlying ESP and PK. They called it
‘psi’ from the Greek letter y.
1
parapsychology¾experience, belief, and the scientific method¾are considered. These
three factors are dynamic forces that drive the field of parapsychology onwards. On the
basis of the above considerations, it is argued that scepticism, particularly of a form that
denies rather than doubts the possibility of psi, has not produced cogent arguments that
undermine the psi hypothesis.
Having thus set a context for this thesis, it will be seen that subsequent chapters,
although concerned with more experimental and theoretical issues, will nevertheless give
us cause to refer again and again to the issues raised in this first chapter.
Chapters 2 and 3 are reviews of the meta-analytic literature in parapsychology,
which represent a period of more than 70 years of theoretical and experimental research in
the field. The main purpose of these two chapters is to establish justifiable reasons to
pursue serious investigations of psi, but the meta-analytic findings presented in these two
chapters will also be drawn upon throughout the thesis for theoretical and experimental
purposes (see especially Chapter 4 and Chapter 10).
In Chapter 4, a new theory¾Thalbourne’s (2000a) theory of psychopraxia¾will be
introduced as a paradigmatic ‘shift’ in parapsychological thinking that has emerged as a
result of the socio-empirical findings of more than a century of research into the
paranormal. It will be seen that the meta-analytic findings give some support to the claims
made for the theory of psychopraxia.
Seven subsequent chapters (Chapters 5-11) are reports on experimental
investigations into the theory of psychopraxia, and other research which has bearing on the
theory. Finally, Chapter 12 will be a discussion of the findings of these experiments and
research, with some general conclusions about the theory of psychopraxia.
1.2 Prologue
In any scientific field of research that involves experimental work and theoretical
speculation there is a widely held assumption that such activities are conducted in a
dispassionate way by investigators and theorists, as if the socio-cultural and behavioural
influences on that field had been isolated and removed as confounding variables. Far from
being removed, they are actively present. But these influences are more than just variables
since they not only underpin the whole enterprise of science, but also constitute its very
core. A number of critics of science have already made these observations. For example,
Latour and Woolgar (1979) state the case:
2
Whereas we now have fairly detailed knowledge of the myths and circumcision
rituals of exotic tribes, we remain relatively ignorant of the details of equivalent
activity among tribes of scientists whose work is commonly heralded as having
startling or at least extremely significant effects on our civilization. (p. 11)
Collins and Pinch (1982) also recognize this state of affairs as problematic for science and
society, and the following statement implies a solution:
Collins and Pinch recommend an ongoing sociological investigation into the scientific
enterprise. This approach may lead to an understanding of how science works, but without
realization, or even a tacit acknowledgement from scientists that transparent influences
(i.e., the social forces that act upon science) exist, they remain invisible and therefore
uncontrollable. More to the point, even if these social forces are identified, scientists may
still fail to see the relativity and the subjectivity of their claims and findings if they do not
adopt a reflexive (i.e., introspective) attitude to those forces. Consequently, they will
pursue the same time-honoured course they have always followed. This state of affairs is
analogous to a blinkered horse that can have no concept at all of the broader nature of its
environment¾not unless it bothers to turn its head and look.
Thus we may lack a keen insight into, or at least remain ignorant of, how these
influences have had considerable effects on (a) the way experimental and theoretical
problems are resolved, (b) the criteria for interpreting results, (c) the future path and
attitude of the investigator, and (d) the direction of a discipline. Metaphorically speaking,
these influences work in the same way as the invisible ‘forces’ that shape an environment.
To extend the environmental metaphor, if we want fully to comprehend and
appreciate a new territory, then not only must we circumnavigate its borders and coastline,
scale its peaks, and enter into its heartland, we must also consider the forces and actions
that have shaped and are still shaping that territory. In like manner, parapsychologists must
not only seek to discover empirically verifiable phenomena through direct observation,
whereupon they posit theories, but they must also consider the field’s current status
(specifically how it came to be that way) in terms of those crucial but sometimes invisible
influences that help give shape and form to those theories. Thus, a careful consideration of
3
the socio-empirical forces that have shaped parapsychology as a discipline, and continue to
shape that discipline, is a necessary undertaking in terms of the importance to the field of
maintaining its consistent identity and purpose in a constantly changing social and
intellectual environment.
5
ESP is “paranormal cognition” and PK is “paranormal action” (Thalbourne, 1982, pp. 27, 61). Life after
death refers to the survival of human consciousness (or conservation of identity) in one form or another after
the demise of the corporeal body.
6
From the frontispiece of the Journal of the American Society for Psychical Research.
4
intimidated by the specious claim that its subject is unspecifiable (for example, see Boring,
1955, 1966).
To clarify this point about the specification of the “parapsychological or
paranormal” from a socio-empirical perspective, it is important to recognize that the related
term ‘psi’ is often used interchangeably as both a ‘construct’ and a ‘paranormal
phenomenon’ (see Palmer, 1988, pp. 155-156). Palmer warns that the “traditional”
approach of positing paranormal theories to explain psi is sometimes confused with the
goal of confirming the psi construct. For example, the oft-used existentialist question
“Does psi exist?” suggests a need for “construct-validation” of psi in its descriptive
capacity as an ‘anomaly’ (Palmer, 1987, p. 156), yet psi was validated in this way as early
as 1886¾i.e., by the time of the founding of the American Society for Psychical Research
(Palmer, 1987, p. 111). Palmer argues that the investigator might be more gainfully
employed in seeking an “explanation of (ostensible) psi phenomena” as possibly being a
paranormal process¾clearly a different goal to that of the verification of the psi construct.
Palmer’s focus, then, is on reaching a greater clarity of understanding about psi than is
currently circulating inside and outside the parapsychological community.
Before discussing paranormal phenomena any further, we must ask how it is that
any phenomenon comes to be specified. Traditionally, specification has been regarded as a
process of recognizing how different one phenomenon is from another phenomenon by
means of a categorization procedure¾thus a phenomenon can be regarded as specified if it
has a unique identity. But, here we fall victim to taken-for-granted discontinuities because
some form of general agreement on what constitutes a system of difference must first be
established (such systems are clearly vulnerable to arbitrary decision rules, exclusion
criteria, and other “incommensurables”; Collins & Pinch, 1982, pp. 9-10, introduce these
ideas).
Specification has also been understood as a process that reveals the nature of a
phenomenon by establishing a certain level of predictability of that phenomenon. Edge
(1985a) points out the flaw in this kind of thinking. It does not follow that a suitable theory
set up to specify a phenomenon should emerge exclusively from the fact that the
phenomenon is predictable because prediction does not yield understanding, whereas a
theory makes the phenomenon intelligible. (It will be seen shortly that even a good theory
does not necessarily guarantee specification in any absolute way.)
Given these definitions, and the fact that attempts to specify an allegedly
paranormal phenomenon by conventional means may be faulted, we need to see how a so-
called normal phenomenon would bear up against such criteria. We can show this by
5
borrowing a construct from the field of physics (i.e., ‘gravity’) and probing into the socio-
empirical derivation of that construct. The effect we know as gravity is a well-accepted
(because repeatable, and therefore a predictable) phenomenon, so it came to be ‘talked
about’ as if it had been specified at some point in scientific history. This misappropriation
(pace Newton) emerged, first, because gravity, whatever it is, was deemed a genuine
phenomenon by consensus omnium7, and second, because Newton had the scientific
insight to propose a good theory to explain the phenomenon. But only the effects, not the
nature of gravity, were recognized and ‘explained’ by Newton’s theory.8 Still, in the
twentieth century, the search for gravity waves as opposed to the equally plausible
hypothesized entity, the ‘graviton’¾both given as possible ‘causes’ of gravity¾only
proves that, by the very standards of empirical science, unspecifiable phenomena lurk in all
corners, not just in parapsychology.
For gravity, repeatability of the phenomenon was a deciding factor in its acceptance
as a real phenomenon (notwithstanding the relatively powerful predictive model proposed
by Newton). Yet the community of scientists then, as now, failed to realize that there is no
such animal as a “common criterion” or “preconceived notion” (to use Schlitz’s, 1985, p.
79, words) in science that was not, or should not be construed as anything more than a
construction in the sense that it was subjectively derived by way of consensus. On this
basis, scientists accepted the ‘gravity’ construct on the grounds of very real (i.e., repeatable
and demonstrable), mathematically describable effects, and these ‘facts’ were set up as the
criteria underlying the general acceptance of gravity as an ‘entity’ about which one could
propose a theory. Only amongst a very limited circle of scientists in the twentieth century
were these criteria and subsequent assumptions questioned.9
7
Radin (1997, p. 44) refers to this social phenomenon as the “herd effect,” which describes the way scientists
herd together and come to a mutual agreement about which ideas or techniques will or will not be taken up as
acceptable. This ancient principle effectively describes the formative process of the Kuhnian paradigm (see
1.2.2).
8
Newton, in fact, had insight into the ‘problem’ of gravity: “That Gravity should be innate, inherent and
essential so that one body may act upon another at a distance through a vacuum without the mediation of
anything else, by and through which their action and force may be conveyed, is to me so great an absurdity
that I believe no Man [sic] who has in philosophical matters a competent a faculty of thinking can ever fall
into it. Gravity must be caused by an agent acting constantly according to certain laws: but whether this agent
be material or immaterial, I have left to the consideration of my readers” (Motte & Cajori, 1962, p. 364).
9
Note how Newton’s model is complemented by Einstein’s theory that gravity should be perceived as
geodesics in (possibly curved) spacetime. To complicate matters, masses do not ‘attract’ each other, they
simply follow a geodesic (a path of least resistance) in warped spacetime brought about by the effects of the
6
To specify the phenomenon of psi in a scientific way also requires a “common
criterion” by which to assess that phenomenon¾a “preconceived notion.” And it seems
that in most cases, the ‘common criterion’ for assessing the existence of a
phenomenon¾the ‘preconceived notion’ that assists in pronouncing with great certainty
the overwhelming evidence of a phenomenon¾is the prerequisite of repeatability, because
predictability ensues, and as a consequence, theory can be gainfully tested. But, how
concrete is this repeatability construct? Schlitz (1985) protests that “there is no common
criterion for assessing replication in science” (p. 79). In fact, to go further, it is not always
possible to specify what the term ‘replication’ itself is supposed to mean because, as
implied above, replication would be dependent on acceptability and the ‘preconceived
notions’ that we use to interpret experimental results (Schlitz, 1985, p. 79; see also Storm
& Thalbourne, 2000b, p. 348, who discuss the distinction between statistical and complete
replication).
Collins (1976) confirms the above point¾when replication is claimed as occurring,
it tends mostly to be the end product of negotiation amongst investigators and theorists,
and has very little to do with objective facts. For example, Edge (1985a) noted that
replication in science does not in fact take place as often as we are led to believe. A
‘replication’ experiment is often the same experiment with improvements, and/or
refinements, and these differences may be of such importance that they disqualify one
experiment as being a replication of the other. To use Diaconis’s (1991) words: “One
cannot judge what ‘really goes on’ in most areas [of research], and it is impossible to
demand wide replicability in others” (p. 386). For example, one “area,” the discipline of
physics (again!), delivers no more a consistency of results from experimentation than do
the social sciences. And Radin (1997) gives a specific example from that discipline by
asking: “How many times does an effect need to be repeatedly shown before it is accepted
as a ‘real’ phenomenon and not an artifact?” He was referring to the ‘omega-minus’
particle, which has been found only twice in 200,000 observations. This particle is now
considered ‘real’ by a majority of physicists by dint of a “sufficient” albeit “poor
replication rate” of production/observation of the said particle (Radin, 1997, p. 49).
Clearly, specification and replication of a phenomenon are ‘social actions’
undertaken in accordance with subjective decision rules and criteria, thus throwing into
question the validity of any so-called ‘objective’ investigation of a phenomenon intended
masses on the spacetime continuum. Thus, the specification of gravity (and by induction, any scientific
entity) can be seen as discourse dependent.
7
to establish the express or definitive nature of that phenomenon. Outside the demands of
conventional science and its often loosely defined criteria (like the expectation of
replication), psi is no different from gravity, yet it is ‘acted upon’ as if it were¾a purely
socio-empirical response to an allegedly unspecifiable phenomenon claimed to be such by
standards that are unspecifiable. Thus, the tacit assumption in science that repeatability
(and its concomitants, predictability and unique identity) may lead to specifiability does
not follow. In fact, the nature of any entity or construct is arguably unspecifiable.10
All the above statements may give rise to the assumption that the belief that
specification has been achieved, because replication and its concomitants are somehow
entity-confirming processes, is actually a defense mechanism of conventional science set
up as a legitimate scientific principle for arriving at some kind of consensus (cf.
McClenon, 1984, pp. 89-91). In fact, if parapsychology can be criticized as having not
demonstrated a repeatable phenomenon, then perhaps all the disciplines may often warrant
the same criticism. As an aside, and as an apparent contradiction to the socio-empirical
approach outlined above, parapsychologists do not have to defend themselves so forcefully
against the accusation that psi is not repeatable¾in fact, new evidence suggests psi is
repeatable. Admittedly, parapsychology has had to wait a long time for its day, but it did,
in the closing decades of the twentieth century, yield statistical evidence of repeatable
paranormal phenomena (see Chapters 2 and 3). Thus the argument that parapsychology has
failed to show replicability can be countered by the cumulative evidence of the meta-
analyses. Such findings and statements challenge, in quite compelling ways, the relevance
and even the validity of the claim that unrepeatability is parapsychology’s only finding
(Blackmore, 1985a, p. 183).11
The earlier relativisation of the demand for replication, in the sense that
repeatability is not necessarily the sine qua non of scientific principles, does nothing to
debase the validity of the repeatability principle as demonstrated in the meta-analyses. It is
accepted that repeatability is a generally accepted hard-line principle for confirming the
10
Laudan (1983, cited in Rockwell & Rockwell, 1986) has demonstrated that the Aristotelean “know-why”
of a phenomenon (the possibility of its being specifiable) was dropped by Galileo as a scientific criterion
because he admitted no understanding of the “causes or essence of gravity” (Rockwell & Rockwell, 1986, p.
107). However, he maintained the Aristotelean “know-how” because he could demonstrate how his results
were ‘true’. “Scientists no longer [claim] that science offer[s] apodictic [church established] certainty” (p.
107).
11
Other contributions to the repeatability issue have been made by Honorton (1985b) and Rao (1985). See
also Shapin and Coly (1985) and Murphy (1987) for further arguments which question the assumption that
repeatability is even essential to the purposes of parapsychology.
8
existence of phenomena. But the more important issue under consideration here is the way
in which the demonstrated repeatability of a phenomenon has falsely led many researchers
to (a) the illusory belief that said phenomenon is not likely to be an artifact of
experimentation or demonstration simply because it is replicable, and (b) the nature of such
a phenomenon has thus been specified, or indeed will be discovered, sooner or later.
Realistically, one needs to recognise the implications for science of this dichotomy
where replication is accepted while at the same time can be regarded as unnecessary, and
one cannot make a sound judgment about these implications by the standards of the
empirical scientist or the critical sociologist alone. On the one hand, repeatability is a
means of identifying (i.e., confirming the existence of) a phenomenon by the frequency of
its occurrence, but many, if not most, inferences about the nature (specification) of that
phenomenon might best be classified as speculative, or even philosophical. On the other
hand, repeatability is not always a possibility, but that circumstance should not give rise to
any suspicion about the existence of the phenomenon. Most researchers would surely
acknowledge these issues, but the literature does not always suggest it. For example, see
Honorton (1993) for a similar argument. Honorton identifies the ‘speculative’ nature of the
terms ESP and PK, preferring the terms “anomalous communication” and “energetic
processes,” respectively. He even recommends that the ‘anachronistic’ term “paranormal”
be “abandoned” (pp. 210-211).
In summary, when parapsychologists are expected to specify and replicate psi, yet
fail to satisfy the demands of skeptics, they are criticized on the basis of technicalities that
just happen to lend themselves very well to an ideology of ‘skeptical scientism’12 (to coin a
term), which blazons itself as good science. Thus do the same skeptics (often the
spokespersons of other disciplines) make unreasonable challenges against
parapsychologists by issuing socio-empirically derived demands they would not require
themselves to fulfil. (See also Irwin’s, 1999, p. 314, observation that the demand for
repeatability is largely a rhetorical exercise underscored by ulterior motives.)
To formulate a question based on the general statements posed in this section (viz.,
“Has the singular nature of psi been undermined by scientific criticism?”), it has been
argued here that specification of a phenomenon is an idealistic notion falsely associated
with repeatability and its concomitants. Some kind of scientific community must accept the
phenomenon, identify it, name it, and make inferences about its nature. Like gravity, psi
phenomena have been identified, and constructions and re-constructions have been
12
I shall tentatively define ‘skeptical scientism’ as an ‘anti-anomaly’, pro-scientific position dependent upon
doctrinaire scientific convention.
9
formulated around them within the paradigmatic frameworks of scientific discourse.
Again, like gravity, psi phenomena maintain anomalous dimensions. This pluralistic aspect
may forever be unrepresentable in science. To borrow from Rhine (1948/1954, p. 7) it
often seems more accurate to specify what we collectively believe about a phenomenon
rather than what we know of it. (In a later section, the influence of belief will be scrutinized
further¾see 1.3.4.2.)
13
See Stokes, 1987, pp. 77-189.
10
of the aesthetic of the objectively verifiable ‘trajectory’ of modernism, yet it maintains the
conflicting two-edged ‘postmodern’ premise that (a) paradigm comparison is not a
plausible undertaking, which leads to the inevitable conclusion that (b) progress cannot be
determined in ways other than subjective, and may not even take place at all by the
standards of a relatively broader, more objective (i.e., less subjective) criterion.
The argument seems clear: If paradigms are incommensurable with each other then
there is no standard by which they can be compared. Thus, to state that things can
somehow be ‘better’, or that progress (in a discipline, for example) can be objectively
substantiated, undermines the relativist argument that Kuhn puts forward. However, this
ostensibly logical conclusion can be given a radical twist, as Collins and Pinch (1982) have
shown. Their argument is that whatever other arbitrary system of decision-making
scientists might resort to in assessing new ideas, “scientists reserve their final commitment
to [these] ideas until they are forced to accept them by logic, or experiment” (Collins &
Pinch, 1982, p. 16). To go further, rather than being at a loss in deciding which knowledge
to keep and which to reject, scientific epistemology is evidently and productively a
composite of the ‘old’ and the ‘new’. Its diverse contents can be drawn upon like tools in a
toolbox, with each tool doing a specific job. Thus, we can live with the dichotomous
situations that science sometimes sets up without necessarily being frozen in our tracks.
For parapsychology, then, it is almost business as usual. There seems to be no
reasonable argument that would undermine the recommendation that things proceed as
normal, providing that one major concession be made. Parapsychologists may accept or
reject both old and new ideas according to their findings, but with reservations. It is now
not just a matter of assessing which theories do a better or worse job of explaining
experimental outcomes. Judgments may need to be made about outcomes in terms of
multiple paradigmatic viewpoints, since different paradigms recognize different
phenomenologies according to the discursive constructs of those paradigms
The investigator will still be free to design research, and assess the results of that
research, from within or without a specifically preferred, and/or generally approved,
paradigm or theory. In the past, this state of affairs suggested a discipline that was
fragmented, but it can now be regarded as a major philosophical strength suitable for any
discipline. In recent decades this pluralist approach has been adopted in parapsychology
(see also Irwin, 1999, pp. 318-319). Thus this approach accommodates, and is sympathetic
to, the subject of the current thesis¾the theory of psychopraxia¾and the ways in which
other theories might bear similarities and differences to the theory of psychopraxia. These
11
issues resurface throughout the experimental chapters, but will be brought into much finer
focus in Chapter 4.
12
problems to the philosopher.14 And justifiably so since, as science stands, the main point of
the above argument is that Pratt’s position is tenable.
The skeptic, who may also be oblivious to these questions, may enter the debate at
this point and start his/her argument, in agreement with Pratt, by also stating that psi
phenomena are incompatible with scientific positivism. Recall Pratt’s (1960) idea that the
very real psi event “def[ies] description in terms of any physical theory now available” (p.
25). But the skeptic (assumed here to be one who avows a total allegiance to ‘scientism’
alone) will be led to state that psi cannot exist because science cannot explain it. Collins
and Pinch (1982) make reference to this paradox: “These two arguments have at their core
the same claims regarding the incompatibility of psi phenomena and science though they
are deployed respectively by sceptics and believers . . . ” (p. 49).
Believer and skeptic both agree on what psi is according to what science cannot
explain. But for socio-cognitive reasons they split at this point¾the one convinced that
science needs to be more flexible, the other certain that there is no room for psi according
to the standards of scientific truth. These arguments create the socio-cognitive
discontinuity of which Collins and Pinch speak, but, to go further, both these arguments
are actually epistemologically discontinuous, and therefore, unsound¾both are derived
from situational presumptions. Thus, the believer introduces the condition that the
scientific viewpoint must change (not necessarily true, but it does square with the evidence
of science as an accumulative enterprise qua Kuhn, therefore fully endorsing the research
ethic), whereas the skeptic introduces the condition that the scientific viewpoint need not
change (not constructive for science because it undermines historical process, therefore
implying that all breakthroughs are explainable, and even that proof can apply without
experimentation!). (There are psycho-sociological reasons underlying the social category
within which one chooses to be; viz., believer or skeptic. This issue is explored in 1.3.4.)
Socio-cognitive discontinuity, then, arises out of a failure to discern (or
acknowledge) the possibility of alternative meanings that are necessarily implied in any
number of scientific statements that have been posited as certainties. Such statements may
be exposed for their inconsistencies. In support of the believer’s argument, the only truly
responsible scientific premise that would fairly represent the psi hypothesis (or any
hypothesis) is that researchers proceed in their quest to accumulate an ever-increasing
14
Thalbourne’s (2000a) discussion of what may essentially be called ‘the ontological question’ is taken up in
Chapter 4, where we shall see that psi (as both ESP and PK) can be characterized from within three mutually
exclusive philosophical systems: Metaphysical Dualism, Metaphysical Materialism, and Metaphysical
Idealism. The distinction between ‘x’ and ‘not-x’ will depend on one’s ontological perspective.
13
epistemology. In recognising the implication of the skeptic’s argument, researchers on both
sides of the fence must not delude themselves into thinking that they have not undermined
the fundamental Cartesian tenet of creating certainty from doubt. This certainty may be
based on knowledge claims already established and accepted when, on the one hand, the
reality is that these claims may not have been fully applied. On the other hand, they may
have been carelessly applied to (say) the psi hypothesis (for examples of knowledge claims
not fully applied, or carelessly applied to psi, see Collins and Pinch,15 1982, pp. 47-65;
Meehl and Scriven, 1987). To do the latter responsibly, the practitioners of science
(opposing groups of scientists) must avoid the all-too-easy task of constructing socio-
cognitive and epistemological discontinuities that can often arise out of the presumptive
thinking that attempts to eliminate doubt. Such constructs ultimately propagate doubt itself,
not certainty. The introduction of insubstantial and/or unsubstantiated conditions and
provisos (for example, rhetorical devices, over-opinionatedness, emotional appeals,
possible futures, etc.) merely to peddle an argument is not, and never has been, good
science, no matter how good it looks ‘on the ground’ (i.e., at face value.16). Relevant to the
theme of this chapter is the fact that this type of science, inside and outside
parapsychology, seems more like the social actions of the practitioners of scientism, rather
than like science.
Thus, we have to be cautious about how we construct so-called ‘good’ science
(including ‘good’ parapsychology). Socio-cognitive discontinuity, as well as
epistemological discontinuity, should but seldom do create cognitive dissonance amongst
those prone to its effects. Antithetical statements can be drawn from a single observation,
yet often neither the poverty of these statements, nor the ensuing irony of their
contradictory natures are acknowledged, if they are ever noticed at all.
The next section will focus on other ironies, with particular focus on the influences
of experience, belief, and the scientific method as socio-empirical forces that also give
form to the field of parapsychology. These factors necessarily emerge from the issues just
discussed. It will be seen how it is often very difficult to discuss experience, belief, and the
15
It is taken, for example, that psi does not obey the inverse-square law, but “the inverse-square law does not
necessarily apply, even to ordinary electro-magnetic radiations” (Margenau, 1966, p. 222, cited in Collins &
Pinch, 1982, p. 58). And “the inverse-square propagation of energy is seldom realized in practice” (Rush,
1943, p. 48, cited in Collins & Pinch, 1982, p. 58). Thus the “radiation analogy” can be used to show that
“psi radiation” need not conform to the inverse-square law (Collins & Pinch, 1982, p. 58).
16
The above statements do not invalidate scientific ‘conjecture’, or ‘speculation’, etc., which usually invite
further research for the purposes of verification. Absolute declarations, however, are criticized here as
typifying the poorest quality science.
14
scientific method independently of each other since there is so much conceptual overlap
between them. But it will also be seen how all three factors have shaped the field of
parapsychology, and how they continue to contribute to its evolution and our
understanding of the anomaly called psi.
17
John Locke advocated experience above all things¾even rationalism¾as the raw material of all
knowledge and reason (Drever, 1968).
18
Quine and Ullian (1970, p. 92) hold that “in a person’s web of beliefs there is no strand that does not help
support some value judgement.” Judgement is a subjective, non-intellectual process, like feeling-toned
evaluation. The process, like thinking, can function in abstract mode, and is thus capable of arriving at
sophisticated concepts and ideas about phenomena (Jung, 1987, para. 724).
19
Rosnow and Rosenthal (1993, pp. 17-19) note that the orienting attitude of the scientist, which includes
enthusiasm, open-mindedness, inventiveness, confidence, etc., has an influence on scientific practice. In fact,
some of these ‘variables’, known as “psi-conducive states” (Braud, 1975) are claimed to have real psi
influences on outcomes in parapsychological experiments. Some of these variables are tested in later
chapters.
15
particularly unique and pivotal influence on the field, as will become clear in the following
sections. These three factors also constitute the formative elements of the socio-empirical
approach that an investigator may adopt. For these reasons, they are discussed next.
1.3.4.1 Experience: Ordinarily, one would hold that there is no teacher like experience.
Psi phenomena usually seem the most real when they are experienced first hand, and they
have a numinous quality about them that ‘informs’ and enhances that experience. Such an
experience may derive from a spontaneous case. However, for the experimental
parapsychologist, the ‘experience’ (or better, ‘evidence’) of psi is often represented only as
a cumulative statistic, where individual trials in laboratory situations provide units of data
that appear rather ordinary or chance-like on their own. Often, therefore, due to a failure to
demonstrate psi directly (such as in the spontaneous case) or inferentially (through a
statistic), many parapsychologists have probably entertained Beloff’s doubt: the fleeting
thought that “perhaps, the whole field had been misconceived from the start” (1990, p. 11).
Beloff’s Doubt (to coin a term) ultimately implies two possible assumptions. On
the one hand, psi cannot be convincingly and consistently demonstrated (or better, induced
at call) in the laboratory because it does not exist. Of necessity, therefore, the
parapsychologist’s techniques (and even his or her own influence, otherwise referred to as
the ‘experimenter effect’) must be ineffectual.20 On the other hand, if psi does exist, the
way we go about looking for it may be inappropriate given its phenomenology. There are
many who agree with the former assumption (Hyman, 1985a; Hansel, 1966, 1980), while
others have expressed sympathy or agreement with the latter (Rhine,21 1948/1954; von
Franz, 1980).
For most researchers the idea of questioning the techniques of laboratory
investigations of psi, and the existence of the psi effect itself, would be considered
unreasonable, whether or not these researchers have the ‘evidence’ of first-hand
20
Akers (1984) has brought attention to ‘explanations’ that skeptics have used to justify their doubt regarding
the existence of psi, such as randomization failure, sensory leakage, cheating, and procedural errors. Added
to the skeptic’s list of ‘explanations’ of the psi event are “delusion, . . . coincidence, [and] unconscious
inference” (Stokes, 1987, p. 84). The notion that psi might be an artifact of the methodologies used in
parapsychology is discussed below (see 1.3.4.3).
21
Rhine acknowledged the general problem of the possible inappropriateness of testing in parapsychology. In
speaking about the decline effect, for example, he said: “We destroy the phenomena in the very act of trying
to demonstrate them. Evidently the tests themselves get in the way of the abilities they are designed to
measure” (Rhine, 1948/1954, p. 161).
16
experience. Such loyalty is praiseworthy, but a working lifetime devoted to a single ideal
can create an insurmountable prejudice. Beloff (1990) has stated a profound truth in this
regard: “By the laws of cognitive dissonance the longer you commit yourself to some
cause and the more effort you devote to it the harder it becomes to renounce it” (p. 11).
True words, though not true for all scientific researchers. Many investigators in the
past, armed with certain prejudices, have entered a field in order to disprove that field’s
major claims, only to be convinced otherwise by their own experience. For example,
French psychologist Michel Gauquelin (1983) sought to prove a life-long belief that
astrological relationships as proposed by astrologers did not exist between the inner planets
of our solar system and the professional lives of certain eminent individuals. In
correlational studies spanning more than 40 years of Gauquelin’s working life, statistical
analyses consistently gave significant results that challenged his prior beliefs (see also
West, 1992, p. 233, pp. 284-289, 302-308, who provides evidence of successful
replications of Gauquelin’s work).
Again, starting off with his own brand of skepticism, Walter Mischel (1968, 1973)
was most emphatic in his criticism of the idea of a consistent personality (temporally and
situationally) as proposed and explained by psychodynamic theories (Freudian, etc.). Not
until he conducted his own longitudinal studies (showing that character traits, which
developed in childhood, could persist into adulthood) did he classify the findings of this
work as being akin to that of the psychodynamic theorists (Mischel & Shoda, 1995). Thus
experience and the scientific method when taken together may temper our prejudices.
Experience, then, is an important component of scientific research, as it can
determine the future path of an investigator. The history of parapsychology, too, has been
dotted with examples of the effects that experience has had on its researchers. Susan
Blackmore (1985b, 1986, 2001a), for example, joined ranks with the skeptics when she
repeatedly failed to demonstrate the presence of psi in her experimental studies (probably a
case of Beloff’s Doubt, perhaps indicating too that cognitive dissonance needs time to take
effect). On the other hand, some researchers, like the late Charles Honorton, have run
consistently successful experiments (Beloff, 1990, p. 38), encouraging others to join or
continue in the field.
The socio-empirical value of experience should not be ignored or conceptualized
out of existence, on the grounds that its chance-like (i.e., coincidental) nature renders it
unworthy as a contributing element towards scientific knowledge and the scientific
enterprise. Utts (1991) has stated that “experience is a poor substitute for the scientific
method” (p. 363), and she has also claimed that “there is little more [in the way of
17
suggestive evidence of psi] to be offered to anyone who does not accept the current
collection of data” (Utts, 1995, p. 290). Perhaps she is correct for the most part, but as she
herself has discovered, a large cross-section of the scientific community is not prepared to
accept that which they would deem to be unexplained anomalies (Diaconis, 1978; Hyman,
1991).
Such denial, however, is typical of the outsider (specifically, the skeptic) bereft of a
personal experience in the area of parapsychology (especially a direct experience of
ostensible psi). So perhaps there is more to be offered to anyone who doubts the evidence.
Of course, one cannot reject the fact that a novel encounter of the psi kind can wear off
quickly¾“every miracle, no matter how incontrovertible it may once have appeared, will
lose its lustre” (Beloff, 1990, p. 33)¾especially if the “miracle” is not reinforced by
replication. Beloff (1990) gives examples illustrating how difficult it is to quell doubt and
“sustain,” or even instill a belief in psi, even when the evidence seems beyond doubt (pp.
30-32). Little wonder that the lack of an experience gives good reason for some critics to
follow their doubts and criticize mostly the methods used by parapsychologists (especially
the statistics, but more on this shortly), there being little recourse to do otherwise.
While the efforts of many critics do help to fine-tune scientific methods in
parapsychology, the psi experience still seems to count once again as a key influence
towards believing in the paranormal (especially if recent techniques such as meta-analysis
are questioned). In other words, skepticism may be instrumental in strengthening the
methods of parapsychology, but nothing seems to inform an individual’s belief about the
paranormal quite like the direct psi experience, even though many may believe without it.
1.3.4.2 Belief: Quine and Ullian (1970) describe the complex processes of belief
formation and modification that many of us employ: Self-evidence, observation, testimony
(i.e., vicarious observation), and hypothesis testing. Most researchers are familiar with
these processes, since they may also be applied in the procedures of the scientific method.
(In a formal scientific investigation, however, these processes may not necessarily result in
any particular belief about the nature of the phenomenon under investigation.)
The discipline of philosophy has provided additional means of belief construction
through deduction, induction, analogy, and intuition. Each of these approaches can be used
independently or collectively as a means of arriving at a belief, and they may extend
outside the formal disciplines to the layperson, even leading him or her to skepticism of, or
belief in, paranormal phenomena.
18
As already noted, however, no amount of evidence may be enough to change the
“closed-minded thought processes” or beliefs of the skeptic, to use Stokes’s (1987, p. 84)
words. To explain skepticism, Tyrrell (1945) argued decades ago that “the constitution of
our minds, as a result of biological evolution, causes us to reject whatever is entirely
foreign to the world of common experience.” For many, such is the basis for the denial of
psi phenomena too.
On the other hand, many millions the world over in modern (westernized) and pre-
industrial cultures do believe in anomalous acquisition of knowledge, communication with
discarnate entities, and so on. One might almost surmise, therefore, that belief formation
for all peoples cannot stem from the application of any one or more of the above-
mentioned processes alone. For if the opposite were true, unanimity of belief about most
phenomena, including psi, would have been reached by now in all populations. However,
such is not the case. Thus, there appears to be more to belief formation than we might
realize.
Having reviewed much of the literature on paranormal belief, Irwin (1993, 1999)
discussed four hypotheses that may serve as more general explanations for belief in the
paranormal¾the social marginality hypothesis, the worldview hypothesis, the cognitive
deficits hypothesis, and the psychodynamic functions hypothesis:
(i) The social marginality hypothesis: Believers are seen as ‘isolated’ in some
way¾by age, gender, socioeconomic status, ethnicity, or culture, etc.,¾so
that belief acts as a buffer to the vicissitudes of a life so often associated
with marginal status. (After an extensive literature review, Irwin, 1999, p.
285, reports that this hypothesis has not been supported.)
(ii) The worldview hypothesis: Believers hold paranormal belief as only one
subjective and esoteric aspect of a much broader picture of life, humanity,
and the world in general, where mental and metaphysical religious beliefs
and practices are incorporated into that worldview. (There is partial support
for this hypothesis, according to Irwin, 1999, pp. 285-287.)
(iii) The cognitive deficits hypothesis: Believers are said to have uncritical,
naïve, and irrational thought processes based on deficits in education, lower
average levels of intelligence and poor reasoning skills. (Evidence for this
aspect of the hypothesis is inconclusive at this stage, according to Irwin,
1999, pp. 287-289.) Believers may also have overly creative imaginations.
19
(There is some support for this aspect of the hypothesis, according to Irwin,
1999, p. 289.)
Overall, there is some evidence supporting the last three hypotheses, with
particularly strong support from the correlational studies for the psychodynamic functions
hypothesis. In respect of this hypothesis, and as an extension to it, Irwin (1993, pp. 21-30;
1999, pp. 291-292) notes that paranormal belief may be the result of physical abuse and
childhood trauma, a mechanism whereby a psychodynamic need is fulfilled by an “illusion
of control” which the belief serves (1993, p. 28). There is no evidence, however, that such
an aetiology excludes paranormal ability. That is, while it might be found that some
believers in the paranormal base their belief on illusory states of mind or fantasy
proneness, it might also be based on genuine paranormal ability. (In fact, although
controversial, it may not be all that unreasonable to propose that trauma may be a sufficient
psi-conducive condition!)
In parapsychology, too, it must surely be more than a state of mind that motivates
many or most investigators in their research. Irwin (1999, p. 291) actually includes
“anomalous experiences” as one of two possible threats to an individual’s intellectual and
emotional autonomy (the other, as mentioned, being “traumatic events”), so it is feasible
that experiences of genuine paranormal phenomena may be amongst those anomalous
experiences. Thus may any number of parapsychologists be of a type actively engaged in
the scientific pursuit of an understanding of the paranormal based on personal experience
of such anomalies. Their intellectual and emotional autonomy may not be ‘threatened’ at
all, but may be modified to include an openness to, and acceptance of, the reality of psi,
based on an informed understanding, and/or experience, of the paranormal. In concurrence
with this view, Irwin (1999, p. 290) proposes that belief in psi might logically stem from
“the data of parapsychological research,” but he does not actually say that direct
experiences of psi might be sufficient causes in themselves for belief in the very same
phenomena.
20
To continue this argument, one “specific version” of the psychodynamic functions
hypothesis, by which at least one underlying psychodynamic cause of belief in psi might be
explained, lies in the mental/neuro-psychological processes described in Thalbourne’s
(2000d) theory of transliminality. Transliminality is defined as “a hypothesised tendency
for psychological material to cross (trans) thresholds (limines) into or out of
consciousness” (Thalbourne & Houran, 2000, p. 853). (Transliminality is similar to
Hartmann’s Boundary concept and indeed the two correlate highly. Houran, Hartmann,
Thalbourne, & Hughes, 2002.) Thalbourne (2000d) has found significant relationships
between transliminality and a number of factors including belief in psi (see 5.3 for a review
of the transliminality literature), but more to the point, transliminality has been shown to be
a predictor of paranormal ability, as found and stated in Storm and Thalbourne (1998-
1999):
The highly transliminal individual was more likely to be warm and out-going (16PF
Factor A), idea-oriented and imaginative (16PF Factor M), and to believe in his or
her own ability to achieve a successful outcome in a psi-task, a belief which was
vindicated in part by the fact that [a significant] 40% of the highly transliminal
participants obtained a hit, suggesting that they . . . might be ‘psi-able’. (p. 115)
In short, belief in psi need not necessarily be associated with pathology, but may follow
from psi ability.
The argument does not end here: It may have been shown that paranormal belief
can engender a feeling of control (Irwin, 1999, p. 292), but again, such control may not
necessarily be an illusory one. Even if we were to disregard Storm and Thalbourne’s
(1998-1999, p. 115) results, there is still stronger evidence that belief in ESP is a psi-
conducive factor, as has been borne out in the many “sheep-goat”22 studies since 1947
(Lawrence, 1993). Lawrence found that belief in ESP correlated positively with ESP scores.
(See also Haraldsson, 1993, where various forms of belief other than belief in ESP also
correlated with ESP scores.) Certainly, the socio-empirical effects of belief can determine
the direction of scientific research, but belief also appears to have real motive force as well
(psychologically and parapsychologically) so that it is indeed a variable of quite
considerable impact worthy of continued investigation as a psi-conducive factor.
Belief, then, appears to have manifest potency, and this fact may give reason for the
reverence in which we hold our beliefs. It is not surprising, therefore, that our beliefs are
22
Sheep are believers in ESP, and they tend to score significantly better than goats, who are disbelievers in
ESP.
21
the very foundations of our truths. But however ‘true’ our truths may seem, the often
personal nature of such truths must not be overlooked. We run the risk of deceiving
ourselves if the so-called truths, which arise from our beliefs, can be argued as misguided
due simply to their being nothing more than something we firmly believe to be true (i.e.,
are found not to be substantiated). In such cases Irwin’s (1993) cognitive deficits
hypothesis holds certain validity.
A few years after Tyrrell wrote the words quoted above (see 1.3.4.2), Jung (1960,
para. 821) drew a distinction between the unique, “just so” quality of the chance-like
(coincidental), personal truth and the more scientifically acceptable “statistical truth.” Jung
claimed:
1.3.4.3 The scientific method: In the nineteenth century, science became defined by its
methodology¾the scientific method¾and in many disciplines, statistical analysis, became
one of the fundamental components of the scientific method. Having successfully made the
transition to the behavioural sciences, the scientific method naturally became a crucial part
of experimental parapsychological research, and has served the field remarkably well for
over a century.23 But in the process, the effects of which statistics is not designed to
23
Richet (1884) and Edgeworth (1885, 1886) were the earliest to propose the use of statistical approaches to
the analysis of data generated in psi experiments. Fisher (1924, 1929) followed with improved methods that
credited “close” guesses in card-guessing experiments as evidence supporting the psi hypothesis. See also
22
explain or give account (particularly psi effects) have been too easily labeled as artifacts,
based on the assumption that statistical and methodological procedures may be
fundamentally flawed.24 For example, for some time there was the assumption that true
randomness cannot be achieved in any finite system (Spencer-Brown, 1957), but Scott
(1958) refuted this claim (Spencer-Brown overestimated the importance to the argument of
true randomness). Gilmore (1989) later warned that researchers in psi “should take account
of the possible inappropriateness of classical inferential statistics” (p. 338).
Once again, the average parapsychologist would have no difficulty in rejecting as
untenable the consequent idea that these effects, which they usually take as suggesting the
presence of psi phenomena, are to be taken as the perennial artifact of fundamental flaws.
Such a rejection seems warranted, given the vast improvements in methodologies over the
last 150 years or so (see Pratt, 1973; Braud, 1991). Braud, in particular, objected to the
possibility that a “singularly powerful flaw” or “more subtler confounds” could be
responsible for the “positive outcomes” in psi studies (1991, p. 58). To argue the case, he
cites a “vast number of successful experiments” by many different investigators, all of
whom have used “very different methodologies” (for example, ganzfeld, autoganzfeld,
forced-choice, etc. These and other domains are introduced in Chapters 2 and 3).
To expand the range of techniques and statistical tools for assessing psi, and to help
avoid hackneyed experimental and statistical approaches, Rosenthal (1986) and others
(e.g., Schmeidler, 1968; Morris, 1991) have advised the use of additional methodologies,
and even an alternative to significance testing of hypotheses. Rosenthal, for example,
recommends “sub-division of studies as a function of different experimental procedures or
individual difference variables such as sex, age, degree of belief in psi effects, and the like”
in order to arrive at “moderator variables” (p. 322). He also recommends the use of
multiple dependent variables, and replicability considerations based on effect size (p. 322).
These advances would be (or have been) constructive for parapsychology, but a
larger problem may lurk in the wings. If statistical procedures are flawed because classical
statistics does not offer a valid model of reality, then all the claims of science, not just
those of parapsychology, are in jeopardy in so far as they employ just such procedures. To
use Mosteller’s (1991, p. 396) words: “ if there is no ESP, then we want to be able to carry
out null experiments and get no effect, otherwise we cannot put much belief in work on
small effects in non-ESP situations.”
Burdick and Kelly (1977) for outlines of the statistical analyses of forced-choice and free-response data
(defined in Chapter 3).
24
In defense of the statistical procedures used in parapsychology, see Mauskopf and McVaugh (1980).
23
Like researchers in other fields, the parapsychologist must continue as before in the
belief that the “point null hypothesis makes some sense” (Utts, 1991 p. 401). But, to be
realistic, is such a criterion enough to satisfy even the most hardened skeptic? The likely
answer is generally a resounding “No!” because often the skeptic believes only when
everybody else believes. In order to justify this seemingly facetious remark, one has to
consider the social ramifications of statistical findings. In some cases particular
methodologies are not even given due recognition by reviewers of scientific journals, or
the scientific community in general, regardless of the claimed validity of these
methodologies (see Greenhouse, 1991).
The issue of prejudice surfaces once again when “prevailing substantive beliefs and
theories held by scientists at any given time” work against the parapsychologist
(Greenhouse, 1991, p. 387). When Jung (1960, para. 821) implied that “the possibility of
[rare] events” was a “rationalistic assumption” decided by the consensus omnium, he failed
to mention that the process could sometimes be a biased procedure. Our age’s ‘rationalistic
assumptions’ are based on the epistemological foundations of our sciences, such as
Newtonian physics, causality, etc. They form the commonsense basis by which we think
and form our ideas, and we must remember, as Beloff (1990) claims, that a commonsense
approach means doing “justice to the evidence while, at the same time, seeking to do the
least violence to our reason and our general knowledge” (p. 13). (As it happens, relativity
theory and discoveries in quantum physics are slowly modifying our commonsense
approach, as well as our reason and general knowledge.)
Psi phenomena, therefore, are irrational according to generally held scientific
precepts, and that adds to their anomalous nature. But it is to the highest degree ironic how
the irrationality of emotions and other prejudices can lend support to arguments that
criticize psi on the basis that psi is an irrational phenomenon (see Stokes, 1987, p. 84-90).
Thus do some critics, perhaps many, try to have their cake and eat it too. It is even more
ironic when we realize that the reasoned approach of statistical law can suggest the psi
anomaly in the first place. Taking all the foregoing statements into consideration, we can
see how experience, belief, and the scientific method may all intertwine and give us much
food for thought.
At the end of this necessary first chapter, any responsible investigator who
considers these issues and takes them seriously must be left with the resounding feeling
that social factors form the undercurrent of a seemingly unimpeded flow of scientific
progress. These factors combined with the business of science as usual, give a socio-
empirical structure to parapsychology as much as they do to science in general. The whole
24
field of science is a human endeavour that requires of us that we give heed to that fact
consistently and thoughtfully.
1.4 Conclusion
1.5 Epilogue
Partly by way of a criticism of hard-line skepticism, this first chapter has been an
attempt to lay the foundations for the main theme of this thesis: the investigation of
psychopraxia. I have withheld a full-scale launch into the theory of psychopraxia until
Chapter 4 because a failure to pinpoint the particular issues that have faced parapsychology
in the past, and still do in the present, would be to present the theory of psychopraxia
without a context. That would be an injustice to the theory for one very important reason:
the theory of psychopraxia (a theory that describes the nature of the ‘psyche at work’
manifesting its ‘goals’ by way of a ‘pro attitude’ and concomitant necessary conditions)
wrestles with the same problems that parapsychology has long wrestled with¾Chapter 1 is
about some of those problems. It would not provide the theory of psychopraxia with the
opportunity of fair representation, when put through the rigors of an investigation, if the
investigator failed to make sense of the major problems facing parapsychology in general,
and the theory of psychopraxia in particular.
26
These problems challenge the parapsychologist, and through that challenge
emerges one of the active ingredients of science¾change. Science, as an evolutionary and
dynamic process, depends on, and is represented by, constant change, and change
characteristically comes about because new goals are constantly being set. The theory of
psychopraxia emerged out of the same processes¾like any theory it exemplifies and
crystallizes the idea of science as a process of necessary change brought about by the
challenges set by problems such as those addressed above.
The next two chapters continue the case for the psi hypothesis, and give
justification for the investigation of the theory as psychopraxia. Those two chapters will be
a review of the meta-analytic literature in parapsychology. The broader, conceptual and
practical issues of meta-analysis¾a relatively new and distinct form of statistical
procedure¾will also be reviewed. Like statistics in general, meta-analysis has maintained
the evidence for an anomalous effect that we may call psi, though we must also
acknowledge, as already stated, that such evidence should not do “violence to our reason
and our general knowledge” (Beloff, 1990, p. 13).
27
CHAPTER 2
META-ANALYSIS IN PARAPSYCHOLOGY: I.
THE GANZFELD DOMAIN
28
(Naturally, these assumptions have been criticized, and these assumptions will be
addressed in the next section.)
The importance and usefulness of the ‘meta-analytic’ approach was demonstrated
decades ago by prominent statisticians, psychologists, and medical researchers (Cochran,
1954; Edgington, 1972a, 1972b; Fisher, 1932; Mantel & Haenszel, 1959; Mosteller &
Bush, 1954; Tippett, 1931). The flexibility and diversity of applications of meta-analysis
have also been well reviewed and described.25
Bullock and Svyantek (1985) point out that meta-analysis is “objectively verifiable,
using measured concepts, quantitative data, and statistical analysis” (p. 112). As opposed
to the traditional literature review, where the reviewer is prone to his or her own
“subjective interpretation of results across studies” (p. 112), the reviewer’s selection bias in
a meta-analytic review is more explicit, since the inclusion and exclusion of particular
domains (i.e., ‘experimental types’ in the context of the present study) is so readily
apparent. The advantage for the reader is that the philosophical persuasion and/or general
intention of the author can be determined with ease so that the overall merits of the work
can be more easily assessed.
As mentioned, meta-analysis produces results that are not attainable by other means
(at least, not without difficulty), but even these results may not end the debate over a
specific area of controversy. Meta-analysis can never be conclusive. It provides a critical
examination of the current status of research in a given area. However, by finding flaws in
current meta-analytic techniques, and by identifying the more successful experimental
domains through up-to-date meta-analytic procedures, newer, more focussed directions for
parapsychology can be established, which, in the case of the parapsychologist, may help
“settle the question” of whether or not psi exists (Broughton, 1991, p. 284).
As will be seen in the next section and Chapter 3, meta-analysis has been a godsend
for parapsychologists. Specifically, the direct benefits of meta-analysis for parapsychology
become clear once it is understood that meta-analysis is the means by which the “signal” of
25
For examples, see Glass et al.’s (1981, pp. 25-26) meta-analyses of studies relating to the efficacy of
psychotherapy in treatment of asthma and alcoholism. See also the meta-analyses by Cook et al. (1992) of
juvenile delinquency studies, and studies on psychoeducational care with adult surgical patients, among
others.
29
the psi effect can be ‘distilled’ from the “noise” of chance with greater sensitivity than in
any single study (Broughton, 1991, p. 281-282).26
Closely related to meta-analysis is the concept of replication (introduced in Chapter
1, see 1.2.1). Rosenthal (1986) lists three main reasons for the so-called failure of many
single studies to elicit significant psi effects, after previous successes with the same
experimental design:
(ii) Pseudo-successful replications, where p values are less than .05, but the
effect sizes are significantly different, and therefore have not been
replicated.
(iii) Successful replication of Type II error (due to small sample size, or very
weak effects).
26
In fact, the presence (‘signal’) of a paranormal effect can never be specifically demonstrated as and when it
occurs, even if 100% of trials are hits. For example, an experiment by Rhine (1937/1950, pp. 74-76) showed
25 hits in 25 trials, but in such an experiment there is no means of distinguishing the successful calls that
suggest psi effects from those calls attributable to chance. The statistical inference of a psi effect is always
made in the knowledge that the effect is produced within a stochastic framework.
30
that data from methodologically flawed experiments are thrown into the ‘soup’ with the
data from better-designed experiments, therefore corrupting the result. Different
procedures in ostensibly the same kind of experiment may also yield a tainted result.
Of course, the general implication derived by critics (that meta-analysis is a flawed
procedure) also applies to meta-analysis in parapsychology. But these criticisms have not
been ignored, and improvements have been made to control for the confounding factors.
Rosenthal (1984) advocates differential weighting as an effective way of dealing with
“variation in the quality of research” (p. 127). Hence the ‘blocking’ procedure is used to
code experiments according to their ‘quality’ and type (i.e., methodology, hypothesis, etc.).
Sample size and the population from which the sample is drawn are also critical
considerations. Credit is given to studies if sample size is specified in advance, as well as
the nature of the analyses¾pre-planned or post hoc. Acceptable randomization methods
are also credited, and even the date of the experiment and the identity of the investigator
are now important criteria in meta-analytic studies (Broughton, 1991, p. 283). Although
these procedures may be seen as subjective, some degree of qualitative assessment can be
made about studies, and these assessments are converted to numerical values to arrive at a
more objective, albeit pseudo-precise, numerical result that is still seen as a gain over
previous methods which did not consider study quality.
One important criticism made against meta-analysis in parapsychological research
(applicable to meta-analysis in general) is that significant results are inevitable, since the
majority of studies used in the analysis have significant results already. On the other hand,
studies with nonsignificant results are never or are rarely published and, therefore, cannot
be included in the meta-analysis (Hyman, 1985a). There are two facts that belie this
criticism.
First, parapsychology journals go to great efforts to publish studies with
nonsignificant results¾they tend not to end up in the ‘file-drawer’. In 1975 the
Parapsychological Association Council adopted a policy of opposing the exclusive
publication of studies with positive outcomes. Thereby, “negative findings have been
routinely reported at the association’s meetings and in its affiliated publications” since that
date (Bem & Honorton, 1994, p. 6; see also Honorton, 1985a, p. 66.)
Second, estimates can be made which account for unpublished studies. Usually, the
number of nonsignificant studies that would be needed to reduce a significant meta-
analytic result to a chance outcome is shown to be far in excess of that which would be
possible for the few researchers in the field of parapsychology (Broughton, 1991, p. 286;
Utts, 1991, pp. 370, 372, 375-376).
31
Rosenthal (1984) has also addressed many criticisms leveled at meta-analysis,
including exaggeration of significance levels, which can be a problem since increasing the
number of studies in a meta-analysis increases the probability of rejecting the Null
hypothesis. Rosenthal argues that when the Null is false it ought to be rejected, but notes
that if such a characteristic of meta-analytic procedure increases its accuracy and decreases
the likelihood of Type II errors, it must be an advantage. Alternatively, should it be
possible that meta-analysis increases the likelihood of Type I errors (i.e., when the Null is
really true), an increase in the number of studies does not increase the probability of
rejecting the Null, or the size of the estimated effect.
Finally, Oakes (1986) questions the validity of a procedure that doesn’t really test
for a ‘directional hypothesis’ (e.g., “do men perform better than women?” Oakes, 1986, p.
162). For parapsychology, at least, domain-specific meta-analyses clearly refer to
directional hypotheses. Generally, what are sought in the meta-analyses of these specific
domains are significant effect sizes, significant differences in psi performance (above or
below mean chance expectation), or significant differences in scoring between comparison
groups.
The dubious value of “meta-meta-analysis,” criticized by Oakes (1986, p. 162), but
endorsed by Glass (1981, cited in Oakes, 1986, p. 162) as a reasonable undertaking, is also
acknowledged (Glass, McGaw, & Smith, 1981, p. 218, believe it is a valid exercise to mix
studies on “apples and oranges” if your hypotheses are about fruit!). Such a ‘method’ may
be superfluous and even of no validity, furnishing (it would seem) a rather vague,
nebulous, and therefore, ambiguous finding, whether significant or not.
As it happens, parapsychology is bereft of such adventurous undertakings. The
idea, for example, that ganzfeld studies could be meta-analyzed with dice-throwing studies
is untenable in the extreme. For example, a significant result in one single meta-analysis of
both domains combined would not draw out the possible effect size difference between
these two domains. Should there be a significant effect size difference, it would only draw
out the distinction that either two types of psi were in operation, or the paradigms are more
conducive to psi in one domain, and less conducive in the other. Thus there are limits to
how far we can apply Glass et al.’s fruit analogy.
Having described the current status of meta-analysis, and argued that meta-analysis
has a relative degree of validity, and therefore, that there is scientific value in its processes,
the following section is a presentation of the findings of the meta-analyses of ganzfeld
studies dating back to the mid-1970s.
32
2.3 Review of the Ganzfeld Meta-Analyses
The meta-analyses now reviewed used as their sources bibliographic databases for
parapsychology, and all the parapsychological journals, including publications of technical
reports, conference proceedings, and manuscripts, etc., and, in some cases, physics journals
(e.g., Radin and Nelson, 1989).
33
1985b, p. 5) was inflated due to the fact that many of the studies were not independent
(they were more like subsets of ongoing experiments).
Hyman also cited evidence that suggested bias in how the studies were reported.
For example, some studies were not planned as such, but were “given this status
retrospectively just because they yielded significant results” (Hyman, 1985b, p. 16).
Hyman reduced the success rate to 31% (he actually argued for less than 30%, given that
there must be unsuccessful but unknown ganzfeld studies yet to be considered¾the ‘file-
drawer’ problem mentioned above).
Hyman (1985b) further criticized many of the studies for their multiple analyses
(e.g., use of a number of measures of ESP), which gave increased opportunity for a good
result, especially since investigators were not adjusting their criterion significance levels
according to the number of statistical tests they performed. He also claimed that
independence had been violated in some meta-analytic studies because “agents were
friends of the percipient . . . [or were even] members of laboratory staff” (1985b, p. 26).
Of interest is Child’s (1986, pp. 337-343) comment on procedural flaw where
pooling results based on groups or conditions can actually conceal an effect rather than
erroneously identify one. Optimal randomization could not be assumed for such studies.
Child indicated that hit-rates can vary systematically from individual to individual, or
group to group, so that “genuinely high performances of some may well be buried by the
chance performance of many others” (1986, p. 339).
Honorton (1985a) accepted the criticism of multiple analysis, and he applied a
Bonferroni correction across all studies. He found that only 45% of the 42 studies were
significant¾not 55% which he originally claimed (but 45% was still higher than Hyman’s
lower estimate of 31%). Honorton then used the proportion of direct-hits as a common
index, since it was the most common measure in the studies (also the most conservative).
A total of 28 studies using direct hits alone were thus employed in the meta-analysis, 12
(43%) of which were significant at p £ 05 (see Table 2.1).
Honorton (1985a) noted that of the 28 studies, 23 (82%) had positive z scores. (The
probability of this outcome is shown in Table 2.1, along with the number of studies, also
expressed as percentages, and their significance levels.) Honorton reported a composite
Stouffer27 Z score of 6.60 across the 28 studies.
27
“Stouffer’s Z is found by dividing the sum of the z scores for the individual studies by the square root of
the number of studies” (Rosenthal, 1978, p. 6).
34
Table 2.1
Meta-Analysis28 of the 28 Direct-Hit Ganzfeld Studies (and their Subgroups)
Number of Independent Proportion Effect Size Stouffer Z Probability
studies p value of Hits (p) (p)
7 (25%) .01 0.47 0.73 8.63 9.80 x 10-9
12 (43%) .05 0.46 0.72 10.46 3.50 x 10-9
23 (82%) n/aa 0.40 0.67 8.42 4.60 x 10-4
28 (100%) n/ab 0.38c 0.65 6.60 2.10 x 10-11
a
Notes: Studies with positive z scores (Exact binomial test with p = q = .5)
b
Includes five studies with negative z scores
c
Rosenthal’s (1986, p.333) more conservative estimate of the proportion of hits is “about 1/3.” Bem
and Honorton (1994, p. 8) give a value of .35, and thus calculate p as .62
Table 2.1 also includes effect size measures as p values because of their ease of
interpretation¾p “depends simply on k, the number of alternative choices available, and P,
the raw proportion of hits”29 (Rosenthal & Rubin, 1989, p. 333). Using the mean effect size
(ES) formula, S[z/Ön]/k, the ES for the 28 studies was .26.
Honorton (1985a, p. 59) also calculated a more conservative estimate of
significance by including 10 additional blind-judging studies that did not report direct-hit
information. Assuming a mean z score for these 10 studies of zero, the Stouffer Z becomes
5.67, p = 7.30 x 10-9. Such a probability still indicates how extremely unlikely it would be
that these successful ganzfeld studies were all the result of chance.
Using the blocking technique, six of the ten independent investigators who
produced these studies achieved significant results, so that neither a specific investigator,
nor a specific laboratory was single-handedly responsible for the significant results. The
suggestion of a file-drawer problem was also rendered less plausible by the fact that 15
nonsignificant and unknown studies would have to exist for every one of the 28 direct hit
studies to reduce the result to a chance outcome.
But other problems had to be addressed. Hyman (1985b, pp. 30-35) found that
flaws correlated positively with significant results. He identified 12 major flaws, such as
inadequate randomization of targets, and failure to use a duplicate set of targets for judges.
Hyman (1985b, pp. 35-36) used cluster and factor analysis on these 12 flaws, combining
them into 3 new variables: General Security, Statistics, and Controls¾upon which were
28
Most of these data come from Honorton (1985a), or are calculated from the data provided in that study.
29
p = P(k – 1)/[1 + P(k – 2)]. Bem and Honorton (1994, p. 8) point out the advantage this measure has in
providing a “straightforward intuitive interpretation” of the effect size, because p is the “proportion correct,
transformed to a two-choice standard situation” so that PMCE = Ptest = .50 (Rosenthal & Rubin, 1989, p. 333).
35
conducted several analyses. The most detailed (factor) analysis was one consisting of
supposedly 17 variables, from which emerged four factors (actually, there were only 16
variables, according to Saunders, 1985, p. 97). Utts (1991) paraphrased the findings of this
analysis:
From these [four factors], Hyman concluded that security had increased over the
years, that the significance level tended to be inflated the most for the most
complex studies and that both effect size and level of significance were correlated
with the existence of flaws. (p. 371)
Hyman’s adjusted figure for the number of successful studies which would not possess
these flaws was 27% of all the studies considered¾“well within the statistical
neighborhood of the 25% chance rate” (1985, p. 37). Honorton acknowledged the
problems with the studies, but Saunders (1985), on behalf of Honorton, repudiated
Hyman’s “meaningless” analysis and its “logical problems” (Saunders, 1985, p. 87).
Saunders found a violation in statistical procedure in Hyman’s factor analysis: “the size of
the available database marginally suffices to support [only] one factor” (p. 87).
Hyman had performed a multiple analysis that included the three flaws just
mentioned, but out of nine potential flaws (giving 84 sets of three) Hyman (conveniently)
selected the set that correlated highly with effect size. Thus the impression was given that,
as Hyman implied (1985b, p. 37), effect size was a function of procedural flaws (the more
flaws in an experiment, the higher the effect size). Saunders noted that Hyman’s multiple
correlations that resulted from selective testing should be regarded as nonsignificant,
rendering Hyman’s adjusted figure of 27% meaningless.
Rather than continue the debate, Hyman and Honorton produced a “Joint
Communiqué” (Hyman & Honorton, 1986) addressing fundamental issues in
parapsychological experimentation. The Communiqué recommended that “more stringent
standards” be implemented in experiments, which should also be conducted by a “broader
range of investigators” (p. 351). Utts (1991 listed these standards as including:
controls against any kind of sensory leakage, thorough testing and documentation
of randomization methods used, better reporting of judging and feedback protocols,
control for multiple analyses [and statistics] and advance specification of number of
trials and type of experiment. (p. 371)
Hyman and Honorton (1986) also believed that meta-analysis had a growing role in the
evaluation of “research quality and the assessment of moderating variables” (p. 361).
36
A number of researchers commented on the “Joint Communiqué” and most were in
general agreement with its recommendations, though all had unique points to make about
the state of affairs of parapsychology. Hövelmann (1986, p. 366) felt that the participant
should be left alone at the judging stage (no presence of the experimenter), since even the
non-verbal behavior of a non-blind experimenter may have an influence on the judging
outcome. Usually, the use of ‘blind’ experimenters avoids this problem.
Palmer (1986, p. 379) argued that the presence of Hyman’s identified flaws in a
ganzfeld experiment does not mean replication of positive results will continue in the
future. He added that the absence of flaws would not necessarily guarantee positive results
either. Nor should it be assumed that failure to replicate when the flaws are removed means
that past successes were due to the presence of the flaws.
Stanford (1986, p. 384) expressed his unease about both Hyman’s and Honorton’s
readiness to make a cause célèbre out of the ganzfeld by yielding it up to the National
Science Foundation for extensive replicability studies. The worst-case scenario of dismal
failure could damage the field of parapsychology, not to mention the careers and
professional lives of parapsychologists. It is too early for parapsychologists to be so
confident when ganzfeld-ESP success looks more like “art” than “science” (p. 386).
Utts (1986), as a statistician, felt that power considerations must be undertaken
more often in experimental design since the “replicability” problem in so many
experiments can be due to poor consideration of the sample sizes needed in certain
experiments. Real effects can be lost if N is too small, whereas a larger N (as in meta-
analytic studies) only increases the chance of getting a significant result.
Rosenthal’s (1986) commentary on the meta-analysis also focused on replication,
and so included a consideration of effect size. He used Cohen’s h, which is the transformed
proportion of direct hits.30 Rosenthal calculated that 23 (82%) of the direct hit studies had
effect sizes greater than zero. The mean effect size h was 0.28, corresponding to the
significant direct-hit-rate of 0.38 reported in Table 2.1 (where PMCE = .25). Rosenthal
recommended that effect size be considered a more favorable measure of replication
success over and above that of significance testing which has “nothing to do with success
of replication” (1986, p. 334).
30
Cohen’s h = 2(arcsin Öp’ - arcsin Öp), where p’ is the proportion of observed direct hits, and p is the
proportion of expected number of direct hits.
37
Milton (1997) conducted one recent meta-analysis relevant to the ganzfeld domain.
She meta-analyzed 46 free-response31 studies (including 42 ganzfeld studies) to determine
which measure¾direct hits or sums of ranks¾was the more sensitive of the two. The
cumulative result of the 46 studies was significant, suggesting a psi effect. In considering
only effect sizes and p values, sums of ranks “outperformed” direct hits (Milton, 1997, p.
227). However, there was no statistically significant degree of difference between the two
measures. Milton called for caution until further research might be more conclusive about
which of the two techniques should be considered ‘superior’, since the nonsignificant
difference may have been a chance result.
31
The free-response method describes any test of ESP using a relatively unlimited range of possible targets,
thus permitting the participant to “respond freely with whatever impressions come to mind” (Thalbourne,
38
as suggestive evidence that, generally, dynamic targets may be “more effective than static
targets.” There was also a distinct difference in success rate between experienced
participants and novices, suggesting that experienced participants (those previously tested)
yield better results than novices (experienced participants’ hit-rate: 37%, p = 0.64; novices’
hit-rate: 32.5%, p = 0.59).
When Rosenthal (1986, p. 333) adjusted for the flaws in the earlier ganzfeld
studies, he arrived at a conservative estimate of “about 1/3,” thus reducing the original
38% hit-rate to a hit-rate roughly equivalent to the 10 autoganzfeld studies of 32.2%. The
ordinary ganzfeld and the autoganzfeld appeared to be equally effective, since they
produced effect sizes in roughly the same vicinity (p = 0.62, p = 0.59, respectively). After
two nonsignificant performance comparisons (on effect sizes and z scores) between
Honorton’s (1985a) database of 28 studies and the new Honorton et al. (1990) database of
11 autoganzfeld studies, Honorton et al. (1990, p. 99) combined the two databases into a
39-study database, Stouffer Z = 7.53, p = 9.00 x 10-14 (Cohen’s h = 0.28).
One criticism leveled at the autoganzfeld meta-analysis was that the eleven
experiments were conducted by only eight experimenters, all of whom were at the same
laboratory. Consequently, Milton and Wiseman (1999) conducted a meta-analysis of new
ganzfeld studies dating from 1987 to 1997. Studies prior to 1987 were not used because it
was assumed that investigators needed time to familiarize themselves with Hyman and
Honorton’s (1986) guidelines so that earlier studies would be too flawed for serious
consideration in a meta-analysis.
Milton and Wiseman (1999) deemed suitable for analysis thirty studies by “10
different principal authors from 7 laboratories” (p. 388). They calculated a Stouffer Z of
0.70, p = .24, one-tailed (ES = 0.013), and concluded that a significant psi effect for the
ganzfeld had not been replicated by a “broader range of researchers” (p. 391).
Storm and Ertel (2001) singled out Milton and Wiseman’s (1999) main finding of a
nonsignificant ES of 0.013 and disputed its derivation. Storm and Ertel argued that Milton
and Wiseman did not adopt a ‘responsible’ attitude in their meta-analysis. A thorough
meta-analysis requires a comprehensive literature search and an accumulative approach to
the available databases (this approach was not adequately demonstrated in Milton &
Wiseman’s paper). Arbitrary exclusion rules and unjustifiable, a posteriori periods of
analysis should be considered unacceptable in any meta-analysis.
1982, p. 28). The participant may, for example, respond by drawing a pictorial representation of the target.
39
Storm and Ertel (2001) found 11 pre-Communiqué studies not previously meta-
analyzed, and after step-by-step performance comparisons, combined them with the three
ganzfeld databases currently extant: Honorton’s (1985a) database of 28 studies, Bem and
Honorton’s (1994) databases of 10 studies, and Milton and Wiseman’s (1999) database of
30 studies. The resulting 79-study database had a significant mean ES of 0.14 (Z = 5.66, p
= 7.78 x 10-9). Milton and Wiseman’s negative conclusion about the failure of the ganzfeld
to replicate is rather misleading and premature as it is pertinent to a limited pool of only 30
studies.
In reply to Storm and Ertel (2001), Milton and Wiseman (2001) argued that the 11
pre-Communiqué studies used in their meta-analysis should not have been used at all
because they were poor in quality due to their ostensible “methodological problems” (p.
434). Thus, Milton and Wiseman clearly ignored Storm and Ertel’s (2001) performance
comparisons of (a) pre-Communiqué authors with post-Communiqué authors, and (b) pre-
Communiqué studies with post-Communiqué studies, both of which yielded no statistical
evidence that the guidelines in the Communiqué had any “influence on effect size
outcomes” (p. 430). Logically, there was no indication that the mean effect size of the pre-
Communiqué database was ‘inflated’ (i.e., an artifact of flaws) because it compared
favourably with the allegedly ‘flawless’ post-Communiqué studies. It follows that there
was no evidence that the mean effect size of the post-Communiqué database was ‘deflated’
due to the removal of these flaws. Apropos to these findings, Palmer (1986) warned that
false conclusions can be drawn on account of, and by appeal to, the Communiqué’s
guidelines¾it should not be assumed that “past successes were due to the presence of the
flaws” (p. 379).
Milton and Wiseman’s (2001, p. 436) only other major criticism concerned the lack
of conservative calculations of some z scores for studies in the 11-study database. In fact,
only 3 of the 11 studies needed adjustment, thus reducing the quality-weighted mean z
score from 0.32 (ES = 0.14; Stouffer Z = 1.06, p = .144) to 0.26 (ES = 0.13; Stouffer Z =
0.87, p = .192). The 11-study database is still not significantly different from Honorton’s
(1985) 28-study database, t(37) = 0.61, p = .543, two-tailed. The Old Ganzfeld Database
can still be formed. It has a mean z of 0.97 (ES = .225; Stouffer Z = 6.05, p = 7.24 x 10-10;
cf. Storm & Ertel’s, 2001, p. 429, original data for the Old Ganzfeld Database: mean z of
0.99 (ES = .227; Stouffer Z = 6.15, p = 3.93 x 10-10).
The ‘Old’ (Pre-Communiqué) and the ‘New’ (Post-Communiqué) ganzfeld
databases are significantly different, t(77) = 3.04, p = .003, w2 = 0.09, but the omega
squared value (9%) is now exactly that of the critical value stipulated in Storm and Ertel’s
40
(2001) paper. The difference might be considered important, but Cohen’s (1988) test, as
originally applied by Storm and Ertel, was again not significant. When the two databases
are combined the 79-study database has a mean z score only slightly reduced from 0.64 to
0.63 (ES = 0.14; Stouffer Z = 5.59, p = 1.14 x 10-8).
This larger database has not only unified the ganzfeld database, but also indicates
that over two decades of ganzfeld/autoganzfeld work has not been in vain. The ganzfeld
might even prove to be the ideal paradigm that Honorton hoped it might be for finding
“strong evidence for psychic functioning” (Milton & Wiseman, 1999, p. 391).
2.4 Conclusion
This chapter reviewed the ganzfeld meta-analytic literature and found that significant
results were obtained in all but one ganzfeld meta-analysis¾that of Milton and Wiseman
(1999). However, with combinatorial re-construction of the available databases and the
uncovering of 11 studies overlooked by Milton and Wiseman, Storm and Ertel (2001)
reconfirmed that the ganzfeld was still the paradigm that delivered one of the highest effect
sizes of all the experimental domains in parapsychology.
Thus, excluding the one disputable meta-analysis, there has not been one meta-
analysis of the ganzfeld domain to date that has not supported Honorton’s (1985a, p. 81)
claim that the ganzfeld demonstrates a “significant psi effect.” Even more importantly,
results from the meta-analyses have indicated, as Honorton also believed, that the ganzfeld
represents an ‘encouraging’ step toward replicability of psi effects. In the next chapter, a
review of the meta-analyses continues into non-ganzfeld domains.
41
CHAPTER 3
META-ANALYSIS IN PARAPSYCHOLOGY: II.
NON-GANZFELD DOMAINS
42
Thirty-second “influence epochs” in a session were reduced to a single score (the
unit of analysis). A percent influence score was then calculated, which was a percent
measure of the “total activity that occurred in the prescribed direction during the entire set
of influence (decremental or incremental aim) periods” (Braud & Schlitz, 1991, p. 5). A
score of 50% (MCE) set the baseline for influence outcomes (no effect) and the t test was
used to compare actual percent influence scores against this baseline.
Influence on remote biological systems was generally found to be significantly
above chance on all target systems except muscular tremor, although a total of 19 sessions
only were run, whereas the next lowest was 40 sessions, and the average number of
sessions was over 65 (see Table 3.1 for results of individual studies). Meta-analysis of the
eight studies showed a significant mean ES of 0.42 (Z = 7.72, p = 2.58 x 10-14). Braud and
Schlitz (1991, pp. 31-34) discussed rival hypotheses that might explain these successful
results, such as external stimuli, common internal rhythms, recording errors and biased
misreading of records, participants’ prior knowledge of when influence was to take place
followed by appropriate responses, and even fraud. All these ‘explanations’ were
disconfirmed.
Table 3.1
Summary of Results of Direct Mental Influence Experiments
Living target system No. of Mean z Mean Stouffer p
sessions score ES Z
Electrodermal activity (influence) 323 1.05 0.058 4.08 2.25 x 10-5
Electrodermal activity (attention) 78 0.84 0.095 1.68 4.65 x 10-2
Ideomotor reactions 40 1.72 0.272 2.98 1.44 x 10-3
Muscular tremor 19 -0.42 -0.096 -0.59 7.22 x 10-1
Blood pressure 41 1.35 0.210 1.91 2.81 x 10-2
Fish orientation 40 1.88 0.297 3.78 7.84 x 10-5
Mammal locomotion 40 1.90 0.300 3.81 6.92 x 10-5
Rate of hemolysis 74 2.43 0.282 4.20 1.33 x 10-5
All systems combined 655 1.34 0.420 7.72 2.58 x 10-14
Note: Table reproduced from Braud and Schlitz (1991, p. 30, Table IX)
43
the desired outcomes. Braud and Schlitz (1991) argued that participants’ “intentionality”
was the “key factor” that brought about the significant results”:
Maintaining a strong intention of a desired goal event, focusing attention upon the
relevant aspect of the target system, and filling oneself with strong imagery of the
desired biological activity are, under certain conditions, accompanied by a shift in
the target system’s activity in the intended direction (p. 41)
32
Precognition is a form of ESP “in which the target is some future event that cannot be deduced from
normally known data in the present” (Thalbourne, 1982, p. 55).
44
Table 3.2
Meta-Analyses of Nine Psi Domains: DMILS, Forced-Choice, Free-Response, Dice-
Throwing, RNG, Clairvoyance, Precognition, Ganzfeld, and Autoganzfeld.
Domain meta-analyzed and No. studies (k) Mean z Mean ES Stouffer p
author(s) and period of score Z
(S[z/Ön]/k)
analysis (Sz/k)
(Sz/Ök)
DMILS 8 1.34 0.420 7.72 2.58 x 10-14
(Braud & Schlitz, 1991) (1979-1991)
Forced-Choice 248 0.38 0.012a 6.02 1.10 x 10-9
(Honorton & Ferrari, 1989) (1935-1987)
Free-Response (GESP/Remote 75 0.68 0.170a 5.85 2.46 x 10-9
Viewing) (Milton, 1998)
(1964-1993)
Dice-Throwing 59 0.003b
0.42 3.19 7.16 x 10-4
(Radin & Ferrari, 1991) (1935-1987)
RNG 490 0.0003 b
0.42 9.74 » 10-23
(Radin & Nelson, 1989) (1959-1987)
Clairvoyance 22 0.60 0.009 2.81 2.50 x 10-3
(Steinkamp et al., 1998) (1935-1997)
Precognition 22 1.02 0.010 4.78 8.80 x 10-7
(Steinkamp et al., 1998) (1935-1997)
Ganzfeld 71 0.71 0.154 5.98 1.12 x 10-9
(Storm & Thalbourne, 2000a) (1974-1999)
Autoganzfeld 17 0.58 0.117 2.35 9.39 x 10-3
(Storm & Ertel, 2000) (1983-1997)
Note: The most conservative values were used in the table. DMILS = direct mental influence on living
systems; ES = effect size; GESP = general extra-sensory perception
a
Calculated from a homogeneous data set
b
Quality-weighted value calculated from a homogeneous data set
45
A later study by Steinkamp, Milton, and Morris (1998) meta-analyzed forced-
choice studies for the period 1935-1997, while at the same time comparing clairvoyance33
with precognition in order to ascertain statistical evidence of a phenomenological
difference between the two. They hypothesized that clairvoyance studies would have a
significantly higher effect size because precognition had an extra “calculational step,”
involving “real-time ESP” (clairvoyance) and then extrapolation from that information “to
make an informed prediction about future events” (p. 193).
Steinkamp et al. (1998) used a total of 22 comparable study-pairs in their meta-
analysis, where procedures were effectively the same in both types of studies. Effect sizes
for precognition and clairvoyance were almost identical (see Table 3.2). Being such a
small sample (N = 22 study-pairs) N-weighted effect sizes were calculated, again with
essentially no difference in outcome (precognition: 0.034; clairvoyance: 0.030). Steinkamp
et al. felt that their coding method may have been responsible for this nonsignificant result,
and that a different method for coding study comparability may yield different results.
They concluded that the burden of proof rested with those “who argue for a difference
between effect sizes under real-time and future ESP” (p. 209).
3.1.3 Free-Response34
The free-response experiment tests participants in a “normal, waking state of
consciousness,” thus differentiating it from the ganzfeld experiment (Milton, 1998, p. 31).
Milton (1998) meta-analyzed all available free-response studies published during the
period 1964-1993 (these studies included remote viewing studies, in which the percipient
“attempts to describe the surroundings of a geographically distant agent,” Thalbourne,
1982, p. 67).
Milton found a mean effect size for 78 studies of .16 (Z = 5.72, p = 5.33 x 10-9)
(see Table 3.2 for other results). A file-drawer of 866 studies would be necessary to reduce
this significant result to a chance outcome. An homogenized database of 75 studies had an
ES of .17 (Z = 5.85, p = 2.46 x 10-9). She also wondered if the large effects sizes in the
ganzfeld domain were due to the ganzfeld technique or to the free-response methodology
itself as used in ganzfeld research. When she compared mean effect sizes between her
database and Honorton’s (1985a) database, there was no significant difference, t(104) =
1.49, p > .05. This result suggests that there is no basis as yet for assuming that the
33
Clairvoyance is “paranormal acquisition of information concerning an object or contemporary physical
event” (Thalbourne, 1982, p. 11).
34
See p. 38, n31, for a definition of the free-response test.
46
ganzfeld condition provided a more superior method of eliciting psi than the standard free-
response protocol.
3.1.4 Dice-Throwing
The dice-throwing experiment is one of a number of experiments designed to test
whether consciousness can influence physical systems at the ‘macro’ (‘greater than
molecular’) level. Radin and Ferrari (1991) examined dice-throwing studies spanning more
than 50 years (1935 to 1987). There were 148 experimental studies and 31 control studies
considered. A total of 2,500 participants attempted to influence over 2.5 million dice
throws.
Forty-four percent of the 148 experimental studies gave results significantly above
the 5% level. The weighted mean ES for the experimental studies was 0.012, which was
“19 standard errors from chance” (Radin & Ferrari, 1991, p. 79). The control studies’
weighted mean, however, was a low 0.00093, which was within one standard error from
chance (p. 79). The combined Stouffer Z for the experimental studies was 18.20, but the
control studies gave a low 0.18. The fail-safe N was 17,974 (121 nonsignificant studies to
every one significant study). (See Table 3.2 for other results.)
Given that die faces are rarely equal in mass due to scooping out of the die face to
mark the numbers, biases would have existed in many of the 148 studies. Radin and Ferrari
(1991, pp. 74-76) took into consideration the fact that only 69 studies used protocols where
targets were evenly balanced among all six die faces. A conservative quality-weighted ES
of 0.007 was calculated (Z = 7.62, p = 1.30 x 10-14). Eliminating the outlier studies that
contributed to the heterogeneity of the database resulted in a database of 59 studies with an
even more conservative, but still significant, quality-weighted ES of .003 (Z = 3.19, p =
7.16 x 10-4).
Radin and Ferrari (1991, p. 68) found no evidence that the overall effect size was
due to a “few exceptional investigators.” Of note was their finding that methodological
quality improved over time, but they also found, in a first analysis, that quality correlated
negatively and significantly with effect size, suggesting that design flaws present in low
quality studies were contributing to the success of earlier experiments. However, analysis
of a homogeneous subset of the original database (from which outliers were removed)
found no suggestive evidence for a possible ‘regression to the mean effect’ in the “perfect”
dice-experiment. The general conclusion, based on the “homogeneous subset of balanced
protocol studies,” was that, if not strong, the mean effect size for the dice-throwing
47
experiments was still significant and consistent over time, indicating a “genuine mental . . .
intention effect on dice” (Radin & Ferrari, 1991, pp. 79-80).
It is important to note that, after RNG experiments (reviewed next), dice-throwing
experiments produce the weakest effect sizes on average of all the domains. This fact must
be seriously considered by investigators wishing to conduct research in the areas of micro-
PK and macro-PK since the probability of a ‘pseudo-failure to replicate’ (i.e., Type II error)
would be high.
48
have an overall significant Z of 2.86, p = .002 (chief author being Helmut Schmidt in each
case).35
Although replicability and robustness were demonstrated in this meta-analysis,
statistical power must be considered in RNG experiments since their effect sizes are
typically very low. In fact, RNG experiments to date represent the domain that delivers the
lowest mean effect size of all domains meta-analyzed (see Table 3.2). Broughton (1991, p.
290) stresses the point, however, that these results are still dramatic evidence that
consciousness could have an effect on micro-systems. The extreme odds against the RNG
effect sizes being the result of chance represent the most convincing evidence so far, after
dice-throwing, for an anomalous effect of the psychokinetic kind.
35
Particular attention is drawn to these four studies because each experiment was supervised by “independent
observers” to avoid accusations of fraud, error, and experimental negligence. This type of improvement in
methodology speaks volumes about the current direction of parapsychological research in the RNG domain,
particularly when significant psi results can still be produced under the critical eye of possibly skeptical
supervision. This same ‘witness’ protocol will be employed in the ‘I Ching Experiment’ (see 5.6.4).
49
of belief be given before the ESP task, since outcome feedback from the ESP task could
have a biasing effect on belief responses.36
Table 3.3
Possible Predictors of ESP Performance (Haraldsson, 1993)
Predictor variable z ES p
Religiosity (n = 383). 2.48 0.13 .007
Belief in an Afterlife (n = 382). 1.92 0.10 .028
Belief in Psychic Phenomena (n = 448). 0.73 0.04 .232
Note: p values are one-tailed
Finally, a regression analysis revealed that religiosity was the only significant
predictor of ESP performance. Haraldsson (1993) concluded: “religiosity might be a more
efficient predictor of ESP performance than the traditional sheep-goat variable” (p. 270,
since sheep-goat scores did not correlate significantly with ESP performance.
However, Haraldsson added that inferences from these findings may only be
applicable to Iceland’s population. Religiosity has “differences in content” in Iceland
compared to the USA, “which may be the reason for the less consistent relationship in the
US between religiosity and belief and reporting of psychic experiences than in Iceland”
(Haraldsson, 1993, p. 270).
36
Lawrence’s (1993) finding has been applied to all the experiments in this thesis that involve measures of
belief.
50
3.1.8 Hypnosis/Comparison Condition ESP
Hypnosis has long been associated with paranormal events, and Dingwall’s (1967;
cited in Stanford & Stein, 1994, p. 235) review of the literature showed an association
between hypnosis and paranormal events, including ESP performance. Twenty-five studies
which tested hypnosis and comparative conditions (controls) for their effects on ESP
performance, were meta-analyzed by Stanford and Stein (1994). An unweighted
cumulative Z score of 8.77 (highly significant) was found for the 25 studies deemed
suitable for analysis, whereas for comparison conditions, Z = 0.34 (not significant). Effect
size was rather small for hypnosis (p = 0.52), but was essentially at chance for comparative
conditions (p = 0.51).
While the hypnotic state appears conducive to psi performance (judging from the
cumulative Z values) further statistical analysis showed a tendency only toward psi-hitting
among both hypnotic participants and the comparative condition participants when
consideration was given to the chief investigator. Some investigators were better than
others at inducing an effective hypnotic state. Nevertheless, Stanford and Stein (1994, pp.
260-261) reached the conclusion that hypnosis generally speaking may still enhance psi
performance, as long as the expectations of the investigator, and the skill and personal
attributes necessary in the participants and the investigator, are present or can be
implemented in the experimental situation.
51
Generally, the ESP task consisted of 40 clairvoyance trials, followed by 40 precognition
trials (except one experiment where the order was reversed). These ten experiments were
planned in advance to test the replicability of the DMT-ESP relationship. For all ten
experiments the DMT-ESP correlation was confirmed, Z = 2.60, p = 4.66 x 10-3, one-tailed
(ES = 0.12).
There was no decline effect in the correlation over the series of ten experiments.
However, a “meta-analysis demolition effect” was found, said to result from the
experimenter’s focussed attention on the combined result of a meta-analysis¾a form of
negative experimenter effect (Haraldsson & Houtkooper, 1995, p. 267). Haraldsson and
Houtkooper explored the possibility of a “DMT-correlated scoring pattern” within the psi
task (e.g., position effect) but no pattern was found. They concluded that overall Z score
appeared to be the best indicator of ESP performance to date.
3.1.10 Extraversion-ESP
A number of studies done in the past give firm evidence that psi hitting and
extraversion are positively correlated (Eysenck, 1967; Palmer, 1977; Sargent, 1981).
Honorton, Ferrari, and Bem (1998) conducted a meta-analysis using a total of 60
independent studies involving 2,963 participants and 17 independent investigators. Rather
than look at psi performance per se, Honorton et al. focused on the relationship between
psi-performance and extraversion as a predictor variable. Thus the effect size index was a
correlation coefficient calculated on these two variables.
Overall, the mean weighted r for the 60 studies indicated an extremely weak
relationship, but nonetheless a significant one, r = 0.09, Z = 4.63, p = 4.00 x 10-5. The
extraversion-ESP correlation was even smaller for the 45 forced-choice studies, r = 0.06, Z
= 2.86, p = 4.20 x 10-3, but for the 14 free-response studies the correlation was higher, r =
0.20, Z = 4.32, p = 1.5 x 10-6, though conventionally signifying a weak relationship. (For
all three extraversion-ESP meta-analyses c2 tests were significant at p < .05.)
One problem raised in Honorton et al.’s paper was the fact that the extraversion
measure was administered after the ESP task in some of the forced-choice studies, meaning
that there was a conceivable inflation in extraversion scores for those participants
successful in the ESP task:
52
(ii) ESP task after extraversion assessment, r = -0.02¾no extraversion-ESP
relationship.
53
(1999), for example, note that the types of extraversion tests used in the forced-choice
studies should have been reported, as was the case for the free-response studies (Honorton
et al., 1998, p. 267). Thalbourne and Jungkuntz (1983) reported two significant negative
extraversion-ESP correlations when they used the Minnesota Multiphasic Personality
Inventory (MMPI) Social Introversion Scale. If Honorton et al. are proved incorrect in their
negative assumption about forced-choice experiments then the forced-choice paradigm
might still serve in ESP-extraversion research.
Utts (1991) has already given a comprehensive review of most of the foregoing
meta-analytic studies (pre-1991). She also stressed the importance of replicability, power
analysis, and effect size. In discussing the general state of parapsychology she concluded
that:
(ii) “It may be that the [significant] nonzero effects observed in the meta-
analyses can be explained by something other than ESP, such as
shortcomings in our understanding of randomness and independence.
(iv) “If ESP does not exist, there is little to be lost by erring in the direction of
further research, which may in fact uncover other anomalies.
(v) “If ESP does exist, there is much to be lost by not doing process-oriented
research, and much to be gained by discovering how to enhance and apply
these abilities to important world problems.” (p. 377)
All these issues are familiar to parapsychologists, and some implications of those issues, to
varying degrees, have been addressed above. Skeptics will continue to criticize the validity
of the significant nonzero effect-sizes, while proponents may base their conviction that psi
is real, based exclusively on the statistical evidence, although the very low effect sizes may
raise serious doubts from both skeptics and believers about the ultimate utility of psi. As it
54
happens, the problem of low effect sizes is not unique to parapsychology. This problem is
quite common in other fields (see the examples from the field of medicine in Greenhouse
& Greenhouse, 1988, and Rosenthal, 1990).
Studies in extraversion and other personality variables address (iii) above (“a
promising direction for psi research . . .”), but it is difficult to imagine how we might
entertain the idea that the ‘wild goose-chase’ hypothesis implied in (iv) above (“ESP does
not exist . . .”) would win a legion of motivated investigators. The safest bet for practicing
parapsychologists is to maintain (v) as a goal and a certainty (“ESP does exists . . .”). This
way the futurity of parapsychology is assured by (1) the conviction of its investigators, (2)
the promise of a plausible paradigm that appeals to many, if not all, in its ‘explanation’ of
paranormal phenomena, and (3) the belief that psi has some practical value.
Some of the comments made for and against the inferences of Utts’s article are
worthy of note. Diaconis (1991, p. 386) feels that parapsychology is “worth serious study,”
but efforts at producing valid and reliable results that allegedly indicate psi can be taken
seriously if and only if “magicians” and/or “knowledgeable” psychologists “skilled at
running experiments with human participants” are used. This unwarranted claim flies in
the face of the rigorous methodologies that form the basis of studies in parapsychology.
Experimenters in psi are often “knowledgeable” psychologists, many of whom consult
magicians, or use them to scrutinize experimental proceedings (Thalbourne, 1995a, pp.
362-363; Utts, 1991, pp. 396-397).
Greenhouse (1991, p. 386) praises Utts’s presentation, but does not believe that
“anyone’s views regarding the existence of paranormal phenomena” will change. He
makes it clear that only until certain prejudices are dissolved can new paradigms be
adopted. Greenhouse (1991, p. 367) cites the aspirin/heart-attack study used by Utts (1991,
p. 367) to clarify this point. In spite of the very low effect size reported in this study, the
relationship was easily accepted because the biological mechanism that explained this
relationship is a causal process that is easily understood, even by the layperson. The point
to be noted here is that parapsychology must focus on process studies that aim to explain
the mechanism of psi, rather than just demonstrate its presence.
Hyman (1991, p. 392) yields to Utts’s conclusion that there is an “anomaly in the
parapsychological findings.” His only reservation in this regard is that the effect sizes vary
so much, both within experimental domains (“the effect sizes obtained by [Robert G.] Jahn
are much smaller than those obtained by [Helmut] Schmidt with similar experiments on
random number generators,” Hyman, 1991, p. 392), and between domains (for example,
meta-analysis shows that the RNG domain yields the lowest mean effect size, while the
55
biological systems domain yields the highest). Hyman asks what it is we are supposed to
account for when there are “nonchance departures from a statistical model” (p. 392)
concomitant with such varied effect sizes. He uses an anecdotal and misleading
astronomical example to illustrate his point (Uranus’s anomalous orbit is explained as
being due to another planet, which led to the discovery of Neptune), which implies that
non-parapsychological science can always give account of its anomalies. However,
unaccounted anomalies abound in orthodox science, where in many cases, it is “empirical
departures from expectation [that have] led to important findings or theoretical models”
(Utts, 1991, p. 400). Parapsychologists have merely followed the example of science.
Utts’s (1991) response to Hyman is to propose “intensive investigation” as the means by
which we may arrive at explanations for psi effects (p. 400).
Finally, Morris (1991) notes that meta-analysis allows the parapsychologist the
opportunity of identifying the moderator variables that need to be incorporated into
experiments in order to yield the strongest effects. For example, it appears to be better to
test individuals rather than groups, better to use dynamic rather than static targets, and
better to use experienced participants rather than novices. Currently, though, there are
doubts about the last two of these three recommendations (Milton & Wiseman, 1999).
3.3 Conclusion
This chapter continued the main theme of Chapter 2¾a review of the meta-analytic
literature in parapsychology. Though meta-analysis is a generally accepted methodology, it
does have its critics. Those critics of meta-analysis, and statistics in general, will assure us
that statistical procedures can “mislead and be misused” (Diaconis, 1991, p. 386). There
seems little doubt that such an assurance is based on fact, but to avoid such pitfalls a basic
tenet of science has always been to proceed with caution.
Nevertheless, the evidence to date from the meta-analyses¾that there are
significant effect sizes for the nine major domains featured (see Table 3.2)¾suggests a
very real (albeit statistical) anomaly worthy of continued investigation. Thus the field of
parapsychology justifies itself as a discipline, and in the spirit of Rosenthal’s (1986, p.
333) words¾“at any point in time some judgment can be made”¾it is hoped that the
chapters to follow, including the findings from those experiments described therein, will
foster sound judgements that might help strengthen the integrity of the field by reaching a
deeper and more critical understanding of the nature of psi. The theory of psychopraxia
56
will be the means by which this task will be operationalised. The next chapter (Chapter 4)
introduces this theory.
57
CHAPTER 4
THE THEORY OF PSYCHOPRAXIA37
In the previous two chapters nine major paranormal domains were reviewed:
Biological systems, forced-choice, free-response (including remote viewing), dice-
throwing, micro-PK (RNG), clairvoyance, precognition, ganzfeld and autoganzfeld. These
and other domains are used to test various forms of ESP and PK, both of which can be
represented in tabular form (see Tables 4.1 & 4.2, adapted from Thalbourne, 1981). Table
4.1 shows that ESP target-events are traditionally classified as either mental in nature, or
physical in nature.
Table 4.1
The Traditional 3 x 2 Taxonomy of ESP according to Temporal Location and Nature of
the Presumed Target-Event (SOURCE: Thalbourne, 1981, p. 35)
Nature of Temporal Location of Presumed Target-Event
Presumed Target-
Past Present Future
Event
Mental Retrocognitive Contemporaneous Precognitive
Telepathy Telepathy Telepathy
Physical Retrocognitive Contemporaneous Precognitive
Clairvoyance Clairvoyance Clairvoyance
37
Parts of this chapter were published in a refereed article (see Appendix AD).
58
Table 4.2 shows that PK target-events are restricted to physical events.
Table 4.2
The Traditional Taxonomy of PK according to Temporal Location
Nature of Temporal Location of Presumed Target-Event
Presumed Target-
Past Present Future
Event
Physical Retroactive Contemporaneous Preactive
Psychokinesis Psychokinesis Psychokinesis
J. B. Rhine (1934) coined the terms extra-sensory perception (ESP) and psychokinesis (PK)
as ostensibly useful categories for describing two seemingly different paranormal
phenomena. Even though, by the 1940s, Rhine (1948/1954) regarded ESP and PK as “one
single fundamental two-way process” (p. 112), both terms are still used as two basic
categories of paranormal phenomena. Parapsychologists have grown accustomed to the
terms ESP and PK, and generally believe that there is a clear distinction between the two
phenomena. This distinction or ‘dichotomy’ is discussed next.
and PK] may have blinded us to their similar core essence, and that ESP and PK may be the
same process” (p. 70).
On the basis that there might be some degree of overlap between ESP and PK it
becomes possible to propose an ESP-PK continuum, as Schmeidler (1987, p. 38) did. She
suggested that at one extreme there may be pure PK (no measurable ESP). Movement
along the continuum would see an increase in ESP fused with PK, until only ESP with no
measurable PK would be evinced at the other end of the continuum. Schmeidler (1994)
59
later claimed that supporting evidence for a difference between ESP and PK would come
from experimentation that might reveal a distinction in scoring between the two, but if
“procedural conditions for success are found to be similar in PK and ESP, and if their
physiological and psychological correlates are also found to be similar, then the two are
functionally indistinguishable” (p. 199).
One problem with testing Schmeidler’s proposal would be in determining just how
the procedural conditions of the ESP task and the PK task are to be considered the same
when the tasks would have to be procedurally different from each other just by design. A
further problem, as Schmeidler (1994, pp. 198-199) recognized, is whether the tasks can
be exclusively described as ESP or PK in the first place, and she admits there are “gray
areas” in this regard. (See Knowles, 1967, 1968; Osis, 1953; Schmidt, 1969, 1973; and
Storm & Thalbourne, 1998-1999, for examples of these gray areas, where it is possible to
interpret the paranormal effects as either ESP, or PK, or both in combination.) In fact, some
theorists suggest that ESP and PK are so similar to each other that they might very well be
the same thing, or that one may be a form of the other (S. E. Braude, 1979; Nash, 1983;
Rhine, 1974; Rhine, cited in Rao, 1983; Schmidt & Pantas, 1972; Stanford, 1977a,
1977b).
Although it seems that the psychophysiological state-trait of the participant has an
influence on paranormal performance, there is no consensus yet as to whether ESP and PK
function best under mutually exclusive conditions. For example, some studies seem to
suggest that PK works best when the participant is tense or aroused (Braud, 1985;
Steilberg, 1975), whereas other studies report a negative correlation between PK
performance and anxiety (which is related to tension and arousal;38 Broughton &
Perlstrom, 1986, 1992). Still other studies suggest that ESP can function similarly under
conditions of arousal or lack of arousal (see Stanford, 1977a; Van der Sijde & Snel,
1992). Therefore, Schmeidler’s (1987) schema and similar schemata only raise the
question: “What grounds are there for dichotomizing what may be a single paranormal
process?” For some theorists, it appears there are no grounds, and a few of them have
attempted a more complete unification of ESP and PK than has been offered to date. Their
theories are now reviewed.
38
Clark, Beck, and Beck (1994) found, through discriminant-function analysis, that anxiety was
discriminated by subjective anxiety and tension as measured on standard anxiety disorder scales. Hoehn,
Rudolf, Pourmotabbed, and McLeod (1997) found that increased muscle tension and heightened arousal are
consistent symptoms among anxiety-prone individuals.
60
4.2.2 The Thouless-Wiesner Hypothesis
The first major consideration of the ESP-PK dichotomy predates Schmeidler’s
attempts at unification and comes from Thouless and Wiesner (1947). They proposed a
unitary process underlying ESP and PK, but paradoxically, they went on to use the Greek
symbols Yg (psi-gamma) and Yk (psi-kappa) as labels for ESP and PK, respectively, in order
to indicate that ESP and PK were “different aspects of one process” (Thouless & Wiesner,
1947, p. 179). Rhine accepted the term ‘psi’, and it has been used ever since in
parapsychology to designate “paranormal causation” or “paranormal process” (Thalbourne,
1982, p.56).
Thouless and Wiesner (1947), therefore, never quite escaped the dualistic argument
that ESP and PK were qualitatively different phenomena. Nevertheless, they then proposed
that “normal thinking and perceiving,” which are endosomatic processes (i.e., they take
place within the body) are established by the same means as an ESP event (e.g.,
clairvoyance), which has an exosomatic target (a target outside the body). They also
proposed that motor activity of the body is conducted by the same means as psychokinetic
activity. So under the Thouless and Wiesner schema there are fundamentally two
phenomena: Yg and Yk, manifesting endosomatically and exosomatically, thus re-
formulated as a bi-modal ‘diasomatic’39 process (see Figure 4.1).
SPHERE OF OPERATION
Figure 4.1. The Thouless-Wiesner reduction of four phenomena to two. (SOURCE: Thalbourne,
2000a, p. 52)
39
“From the Greek dia ‘through’ (as in ‘diaphanous’) or ‘throughout’, + soma, meaning ‘body’. The term is
intended to mean something like ‘capable of occurring either endosomatically or exosomatically, the body
being like a ‘permeable membrane’.” Thalbourne, 2000a, p. 35).
61
Any kind of mental event (thinking a thought, or a telepathic mind-to-mind
communication) is Yg, and any kind of motor action (moving an arm, or paranormally
influencing the fall of coins) is Yk. Again, note that Thouless and Wiesner maintained the
ESP-PK distinction, even though they seemingly opted for a unitary psi process. It is noted
too that psi has invariably come to be associated more with paranormal phenomena and
less, if at all, with normal phenomena (Beloff, 1985, p. 226). Few parapsychologists today
seem to be familiar with the Thouless and Wiesner theory.
62
fundamental similarity between ESP and PK that was overlooked by Thouless and Wiesner.
However, the only way of determining that ESP and PK might be a unitary, synchronistic
process, as Jung’s theory suggests, is to show how they are consistently contingent upon
coincidence and meaningfulness. Thus far, evidence is limited, but certainly the
experimental approach to synchronicity (e.g., see Braud, 1983) does not require that ESP
and PK terminology be evoked.
The organism nonintentionally uses psi to scan its environment for need-relevant
objects or events or for information crucially related to such events, and when
obtained, the organism tends to act in ways which are instrumental in satisfying its
needs in relation to the particular object or event in question. (p. 35)
(Note that the “scanning component” was later discarded from the PMIR model¾see
Stanford, 1990, pp. 57-59.) Stanford (1974b) extended the above definition to include PK
phenomena. He then developed his conformance behavior model, which “subsumes both
ESP and PK” events under one model (Stanford, 1978, p. 198). Any paranormal process
was thereby typified as a “conformance” (i.e., a disposition-serving outcome) respondent
to the “need or other disposition” of an individual (p. 207). Conformance behavior theory
regards both ESP and PK as “goal-oriented rather than information-based” (pp. 203, 208).
At a later date, Stanford (1990) was “less than fully convinced about the goal-
oriented character of psi processes” (pp. 58–59), but maintained it for the most part in his
resurrected PMIR model. Stanford (1990) also abandoned the conformance behavior
model, but even though the PMIR model now maintains that the same underlying elements
are necessary for ESP and PK effects (e.g., a “disposed” system) he still discusses them as
discernibly different processes (cf. Stanford, 1990, on ESP, with Stanford, 1974b, on
40
PK). For example, Stanford (1974b, pp. 326–328) holds that “implicit extrasensory
guidance” is “part and parcel” of many PK events and that “active-agent telepathy” may
really be a special form of PK (p. 344). He thus appears to identify qualitative differences
40
Stanford (1990, p. 57) regards Stanford (1974b) as the definitive reference for PMIR in terms of
psychokinetic effects (see especially Stanford, 1974b, pp. 348–351).
63
between ESP and PK, thereby failing to merge acceptably the two into a truly unitary
process.
The major theories just reviewed have dealt with the ESP-PK dichotomy either
directly or indirectly. In some respects, most theorists have regarded the dichotomy as a
real problem in need of a solution, with the exception of Jung, who never criticized the
dichotomy, yet ironically, identified a common phenomenology between ESP and PK.
According to Thalbourne (2000a, pp. 15-25), theorists (such as those identified above)
41
There are traditional-minded researchers who would regard the RNG effect as a form of PK. The RNG effect
has been referred to as micro-PK since 1969, although Schmidt (1969) preferred precognition (i.e., ESP).
64
have failed to identify the fact that they were making specific philosophical assumptions
about the phenomena. In this next section, the ontological question, and its bearing on the
ESP-PK dichotomy, will be addressed.
Scientists for the most part depend heavily on terminology. The kind of
terminology they use depends on the philosophical world-view they hold. The investigator,
therefore, is discourse-dependent in regard to interpreting, describing, defining, and
otherwise explaining the phenomena they ‘discover’ or ‘invent’. Likewise, a careful
consideration of the ESP-PK problem reveals that ultimately any interpretation of
paranormal phenomena will depend absolutely on one’s metaphysical outlook, which may
logically include normal phenomena in this consideration, as will be seen.
Traditionally, there are three fundamental metaphysical perspectives on the
ontology of normal phenomena: Metaphysical Dualism, Metaphysical Materialism, and
Metaphysical Idealism (see Table 4.3). Thalbourne (2000a, pp. 15-25) applies these same
perspectives to paranormal phenomena. The terminological problem (or discursive
problem), just raised, has to do with the way in which one prefers to characterize or
interpret phenomena. The difficulty for science that this problem raises is readily apparent,
since scientific consensus about any phenomenon also requires some implicit agreement
about the philosophical structure underlying the terms to be used in describing that
phenomenon. Often there is a tacit understanding amongst scientists regarding the use of
the various ontologies. For example, in the case of dualism, there may be a mixed use of
terms that entails a switching to and from materialistically derived, ‘entity’-like words, to
idealistically derived, conceptual words.
Thalbourne (2000a) argues that it is rare that the above actions, and the
implications of the various ontological foundations, are brought forward for analysis. As a
case in point, the ongoing debate in parapsychology over ESP and PK is essentially driven
by the three metaphysical perspectives (see Table 4.3). The idealist may prefer ESP-type
definitions, while the materialist may prefer PK-type definitions. Of course, the dualist has
the dubious advantage of wrestling with the category they find the most parsimonious!
(i.e., ESP or PK; see Edge, 1985b). This conundrum forms the basis of the theoretical ideas
postulated earlier by Thouless and Wiesner (1947), and revived by Thalbourne (2000a),
and these ideas are discussed next.
65
Table 4.3
Interpretation of Normal and Paranormal Phenomena in Terms of the Three Major Metaphysical Ontologies: Dualism, Materialism, and Idealism
Ontology Perspective on Normal Phenomena Perspective on Paranormal Phenomena
Dualisma a) All events are considered possible so long as they do not a) The processes underlying paranormal events would be
defy the known physical laws, or known mental processes. described in terms of ESP if mind-to-mind or mind-to-object
There may be other physical laws yet to be discovered. The information transfer is involved (telepathy, clairvoyance,
absolute limits of the mind are yet to be determined. and precognition), or PK when mind influences matter.
b) Normal human events are classified as physical events b) ESP and PK are distinct categories.
(e.g., motor function, etc., originate in the brain) or mental
events (e.g., cognition, volition, etc., originate in the mind).
Materialismb a) All events are considered possible so long as they do not a) The processes underlying paranormal events would be
defy the known physical laws. There may be other physical described in terms of PK, where ‘ESP’ is actually a brain
laws yet to be discovered. state otherwise called a mental state (telepathy may work
like a brain-to-brain field transmission; clairvoyance may
work like radar or sonar).
b) Normal human events (i.e., cognitive and motor functions) b) ESP reduces to PK.
are mediated by the brain.
Idealism a) All events are considered possible so long as they can be a) The processes underlying paranormal events would be
experienced as sense-data. The absolute limits of the mind described in terms of ESP, where ‘PK’ is actually a
are yet to be determined. paranormal transformation of sense-data otherwise called
‘the perceived physical object’ (telepathy is mind-to-mind
sense-data; clairvoyance works via mind-to-external sense-
data or information).
b) The source of all normal human events (i.e., cognitive and b) PK reduces to ESP.
motor functions) is attributed to mind.
a
Notes: Assumes Interactionism
b
Assumes Identity Theory
4.4 The Theory of Psychopraxia
67
4.4.3 ESP and PK as ‘Actions’
The third and final commonality is that ESP and PK are both instances of action,
whether cognitive or motor. “Mind and matter may, in the final analysis, be ontologically
different substances (as the Dualists believe), but the more important fact is that from the
point of view of the active agent, mind and matter are manipulated in fundamentally the
same way” (Thalbourne, 2000a, p. 75). Psychopraxia is thus defined as:
A . . . principle underlying all interactions between the self, or ego, and the realm
consisting of mental and physical events, whereby under certain conditions (as yet
unspecified, but probably psychophysiological) the adoption of a desire
automatically results in its fulfillment in reality; psi is thus seen as a special
instance of psychopraxia being those manifestations of goal-achievement which are
exosomatic rather than endosomatic, that is, which are not mediated by the normal
sensory-motor apparatus. (Thalbourne, 2000a, p. 76)
Thalbourne (2000a) later drops the term “desire,” replaced it with “pro attitude” (p. 96),
and referred to other “certain conditions” as being “necessary” (p. 96) in the sense that they
were sufficient in bringing about a psychopractic effect. This important stipulation
eliminated the “wish-fulfillment” allegation that might be made against the theory of
psychopraxia (“someone might be tempted to describe ‘psychopraxia’ as simply a more
formal name for ‘magic’,” p. 95). In the present thesis, a ‘sufficient condition’ shall refer to
a necessary condition that was sufficient in bringing about a psi effect at the level of the
sample or a group formed from that sample, but certainly not at the level of each and every
individual. Were that the case, we would expect perfect correlations, and 100% hit-rates.
By noting that exosomatic and endosomatic processes have the same underlying
cause, the theory of psychopraxia is thus an attempt to unify normal and paranormal
phenomena, cognitive and motor (Thalbourne, 2000a, pp. 87-88).
68
1. Honorton et al. (1990) identified three psi-conducive ‘features’ (i.e., conditions):
All experiments in this thesis involve individual testing, so that (i) above has been
optimized as a psychopraxia-conducive condition.
3. More recently, belief (or disbelief) in the paranormal process, and high (or low)
scores on the Transliminality Scale (Form B; Thalbourne, 1998; see 5.3) have been
correlated with psi performance (in the case of belief, see Lawrence, 1993, and in
the case of transliminality, see Sanders, Thalbourne & Delin, in press, and Storm &
Thalbourne, 1998-1999). Belief in paranormal processes (usually given as ESP) and
transliminality may therefore be taken as conditions that are conducive to
psychopraxia. Belief is tested as a psychopraxia-conducive condition in the
experiments featured in Chapters 5, 7, and 9. Transliminality is also tested in
Chapter 5.
69
Loomis Type Deployment Inventory (Singer & Loomis, 1996), were each tested as
possible correlates of paranormal performance (see Chapter 8).
Some of the last six of these conditions can be corollaries of the first (i.e., physical
relaxation), particularly (v) and (vi) above. It is hypothesized that relaxation is conducive
to psychopraxia (see Chapter 9).
6. Thalbourne (2000a, p. 96) identified six ‘state of mind/body’ factors that are
conducive to normal (i.e., psychopsychological) functioning, and given that
psychopraxia is a principle hypothesized as applying to normal as well as
paranormal functioning, these same ‘state of mind/body’ factors may be conducive
to exosomatic psychopraxia:
70
7. Finally, based on the low effect size found in the RNG meta-analysis (Radin &
Nelson, 1989; see Chapter 2), it is hypothesized that perceived simplicity of the
apparatus/mechanical set-up and perceived ease of the paranormal task (as opposed
to perceived complexity of the apparatus/mechanical set-up and difficulty of the
task) would be conducive to psychopraxia (see Chapter 11). This hypothesis was
based on the assumption that the low effect size in the RNG meta-analysis came as a
result of participants’ being intimidated by the experimental set-up, and thus
performances were worse than might ordinarily have been expected in comparison
with other paranormal domains (see Chapter 10 for details).
There would be many other conditions conducive to psychopractic effects, but not
all of them are necessary. For example, Vernon Neppe (V. Neppe, personal
communication, 31st December, 2000) gave examples of altered states of consciousness
that are hypothesized as being conducive to “subjective paranormal (psi) experiences”:
Note that the ganzfeld state, (f) above, includes relaxation, which is tested in Chapter 9.
Stanford (1990, pp. 138-148) lists conditions (‘factors’) that may be aversive to psi
(i.e., may limit or prevent PMIR), including ‘rigidity’, ‘inhibition’, and ‘stereotopy’. The
polar opposites of these factors (namely, ‘flexibility’, ‘openness’, and ‘spontaneity’),
which would not put “constraints on behavior” (p. 139), and thus would not diminish the
possibility of PMIR, may also be seen as conditions that bring about psychopraxia. These
factors are not specifically tested in this thesis, but the dimensions, ‘openness’ vs.
‘inhibition’, and ‘spontaneity’ vs. ‘stereotopy’, recall Factor H on the 16PF (i.e., ‘Social
Boldness’, which has been polarized as “adventurous vs. withdrawn” by Kanthamani &
Rao, 1971, p. 194). This factor has correlated with psi performance (Storm & Thalbourne,
1998-1999), and has been retested in the replication study with the I Ching (see Chapter 5).
71
SET OF NECESSARY CONDITIONS
PSYCHOPRACTIC
PRO ATTITUDE EFFECT
73
4.5 Experimental Psychopraxia
1. The diasomatic hypothesis: Psychopractic effects that are endosomatic (i.e., normal
cognition and motor control) are givens (i.e., proved by common experience,
psychological verification, etc.), so testing the diasomatic hypothesis within a
parapsychological paradigm implies testing the extent to which psychopraxia may
function exosomatically. Exo-psychopraxia may operate in two ways:
74
correlations that are found to be significant in the planned analyses, these
conditions will be re-tested (post hoc through median-split analyses, etc.) to
determine whether they are sufficient as well. If they are sufficient, then they will
be classed as circumstantially necessary.
4.6 Conclusion
The traditional view of paranormal processes is that they can usually be placed in
either one of J. B. Rhine’s two categories, ESP or PK. Thouless and Wiesner (1947)
attempted a unification of these two categories, subsuming them under the super-ordinate
category of psi, and the psi process has since been used to refer to anomalous influences on
mental and physical systems. Since then, a number of theorists (Schmeidler, 1987, 1994;
Jung, 1960; Stanford, 1974a, 1974b, 1978, 1990; and May, Utts, & Spottiswoode, 1995a,
1995b) have tried to subsume ESP and PK more fully into one or other category, but
attempts have been impeded, either by theoretical and philosophical biases, or by
insufficient experimental evidence.
The theory of psychopraxia is an attempt by Thalbourne (2000a) to resolve the ESP-
PK dichotomy, first, by recognizing that the distinction exists only on the basis of
philosophical assumptions, and second, by unifying the two formerly categorical effects as
one singular psychopractic effect. To paraphrase Thalbourne’s (1982, pp. 76, 96)
definition: Psychopraxia underlies all interactions between the individual and
mental/physical events, whereby under ‘certain conditions’ the adoption by the self of a
pro attitude towards a certain goal results in the achievement of that goal.
Endo- and exo-psychopraxia only take effect if certain conditions are met, and
these mandatory conditions are regarded as necessary and sufficient. There would be many
other psychopraxia-conducive conditions similar to those listed above (see 4.4.4), and
some set of them would be hypothesized as sufficient for a given psychopractic effect. A
series of five experiments have been conducted to illustrate the theory of psychopraxia.
These experiments are presented chapter by chapter, starting with the next, Chapter 5.
75
CHAPTER 5
AN EXPERIMENT WITH THE I CHING
(A REPLICATION)42
The method of the I Ching does indeed take into account the
hidden individual quality in things and men [sic].
C. G. Jung (1989, p. xxviii)
2. Find experimental evidence that might substantiate Thalbourne’s (2000a, pp. 1-15)
claim that the traditional ESP-PK division can be unworkable in practice.
42
A shorter version of this chapter was published as a refereed article (see Appendix AE).
43
Recall from Chapter 1 (see 1.3.4.2) that Transliminality is defined as “a hypothesised tendency for
psychological material to cross thresholds into or out of consciousness” (Thalbourne & Houran, 2000, p.
853).
76
3. Show that exo-psychopraxia is the result of a pro attitude (Thalbourne, 2000a, pp.
93, 95-96). In the present study, the ‘necessity’ of the pro attitude was tested.
4. Show that psi is not so much “goal-oriented” as “pro attitude serving” (Thalbourne,
2000a, p. 66). Irwin (1999, p. 160) describes goal-orientation as non-mediational
(i.e., not requiring a flow of information by various means such as is posited in the
“cybernetic” theories¾see also Stanford’s, 1978, comments in 4.2.4). In the
present study, exo-psychopraxia was tested as a possible mediational process.
5. Test Thalbourne’s (2000a, p. 96) claim that various ‘conditions’, such as level of
transliminality, belief in paranormal processes, and specific personality traits are
‘necessary’ (see 4.5) in bringing about effects, which Thalbourne describes as
psychopractic.
77
described as psi, and theorized as paranormal). Thus, the theory of psychopraxia draws
attention to the problems that are inevitable if parapsychological experimentation proceeds
with the ESP and/or PK constructions in mind. This chapter, in part, considers the
paranormal process from a ‘psychopractic’ (i.e., non-dichotomous) point of view in an
attempt to demonstrate the hypothesized redundancy of the terms ESP and PK.
78
processes. Kennedy (1978, 1995) also found that ESP and PK do not necessarily involve
information processing. There appears, then, to be some justification for models of
paranormal ‘process’ that can be described as non-mediational. That is, psi may not
necessarily take the form of information flow between ‘object’ and ‘subject’¾environment
and individual¾which may, for example, depend on wave-forms, particles, or so-far
undiscovered forces (see Stokes, 1987, pp. 111-135, for a review of some of these
theories). In fact, using theories or models that suggest information processing in one form
or another to explain paranormal knowledge acquisition or action may simply be a
prejudice of conventional parapsychological thinking.
While Thalbourne (2000a, p. 55) states that psi is a goal-oriented process, he later
prefers that psychopraxia be termed a ‘pro attitude serving’ process (p. 66). The theory of
psychopraxia allows for information processing to the extent that it fulfils the obligation of
being a sufficient condition in a given situation. Some paranormal phenomena merely
present in such a way as to suggest the goal-orientation hypothesis, leaving open the
possibility that subtle information processes may well occur in psychopractic action, but
may not be detected in any measurable way.
It is possible to test the hypothesis that psychopraxia may be a mediational (i.e.,
informational, or cybernetic) process in the I Ching experiment if so-called ‘hexagram-
hitting’ in Sample 1999 is significantly lower than it was in Sample 1998. This hypothesis
is plausible because the new Hexagram Descriptor Form (Appendix D) used in the follow-
up study does not contain the 64 hexagram symbols, as does the old Hexagram Descriptor
Form (Appendix E). If paranormal mediation is based on information processing then the
new descriptor form, being harder to use (since it has no symbols), would mean an
additional, anomalous step must be taken by the participants to reach their goals of
successfully achieving hits. If there is a significant difference in hitting between the two
samples, then exo-psychopraxia might be a process involving information processing that
is pro attitude serving rather than goal-oriented.
It was hypothesized, therefore, that the hexagram hit-rate for Sample 1999 is lower
than the hexagram hit-rate for Sample 1998. This hypothesis was tested using Rosenthal
and Rubin’s (1989, p. 334) procedure for testing the statistical significance of the
“heterogeneity” of the obtained p values by means of a chi-square test on the two
independent studies.
79
5.1.4 The Concept of the ‘Necessary Condition’
The necessary condition in the theory of psychopraxia can be rendered by analogy
as being like one domino in a series of dominoes, parenthetically embraced by the pro
attitude on one side and the psychopractic effect on the other (see Figure 4.2). Each
‘domino’ must be strategically ‘placed’ in that series so that, prior to the pro attitude, it
fulfils its obligation of maintaining a scientifically causal44 chain of events, where each
domino initiates the mandatory next step, and is itself a step towards psychopraxia (see
4.4.4 for further clarification of this model).
Chapter 4 more fully details the nature of the necessary condition, and includes a
partial inventory of psi-conducive conditions already identified in the literature (see 4.4.4).
Added to that inventory were some more recently identified psychopraxia-conducive
conditions that are hypothesized as being necessary and sufficient in a given experimental
situation for bringing about the psychopractic goal. (Note that this switch in terminology,
from psi-conducive to psychopraxia-conducive, serves the primary purpose of avoiding the
narrowly defined term ‘psi-conducive’ that has come to connote only paranormal
process.45)
Theoretically, psychopraxia-conducive conditions bring about effects that can be
classed as either normal or ostensibly paranormal (see again, 4.4.4, for a specific
clarification on the issue of ‘task-dependence’¾conditions identified as necessary because
they are sufficient in one task, normal or paranormal, may not be necessary in another).
Storm and Thalbourne (1998-1999) found that transliminality and certain 16PF
factors correlated with success at a psi task. An attitudinal variable¾belief in one’s “psi
ability”¾was reported as an indirect predictor of psi success so that belief (possibly in a
variety of forms) may also qualify as a possible psychopraxia-conducive condition (see
5.3.3 for details). Assuming the successful replication of the above correlates, the same
conditions may again be regarded as psychopraxia-conducive. These conditions will also
be tested to determine whether they are necessary and sufficient in the I Ching experiment.
44
See Storm (1999, pp. 260-261) for an explication of scientific causality. Essentially, in scientific circles a
causal connection between cause and effect is recognized as long as the requirement ‘if A, then B’ is
fulfilled.
45
Psychopraxia and psi are not interchangeable terms, but since the present thesis is fundamentally a
parapsychological work, psychopraxia-conducive conditions that are instrumental in normal cognition and
motor function will not be scrutinized here. Therefore, in this thesis both terms¾psi and psychopraxia¾will
describe, or otherwise imply, the same ostensible paranormal processes as suggested by the significant
findings.
80
Given that the hypotheses in these four sections are falsifiable and can be tested
experimentally, it was therefore deemed a suitable undertaking to attempt a replication
study of the I Ching experiment. The following three components form the basis of the
experiment: (1) the I Ching divinatory system, (2) the concept of transliminality, and (3)
Cattell’s 16PF Questionnaire (Cattell & Kline, 1937; Cattell, Eber, & Tatsuoka, 1970;
Russell & Karol, 1994). These three components are now described.
46
The Wilhelm-Baynes version and Hazel’s (1990) ‘New Age’ interpretations of the 64 hexagrams were
used to derive the I Ching Hexagram Descriptor Form in the present study (see 5.6.2).
81
underlies the outcome, but his claims are driven by certain assumptions made under the
rubric of synchronicity theory (see Storm, 1999).
It is equally valid to argue the case that the rules of chance can actually be
subverted by intention (conscious or unconscious) so that the effects of chance are reduced.
The outcome may then be less coincidental and more ‘veridical’ in nature. That is, the pre-
disposed participant introduces his or her own influence into the system, so that human
involvement¾the crucial part of the I Ching process¾may anomalously generate the
appropriate reading. Thus, the system may be dependent on other than normal processes
(i.e., a paranormal process).
Yin Yang
82
Four ‘duograms’ are then generated through various combinations of the two lines:
yin/yin, yin/yang, yang/yin, and yang/yang (see Figure 5.2).
The duograms represent a four-step continuum, moving from old yin to old
yang¾a gradual shift from one polar opposite to the other. By the addition of a yin or a
yang line to each of the duograms, eight trigrams (the pa kua) are formed (see Figure 5.3).
1 2 3 4 5 6 7 8
The trigrams also represent a continuum, but in eight steps, from Heaven (Chi’en)
through to Earth (K’un). The individual trigrams are magnified in meaning above and
beyond the simple degrees of yin and yang, or the duograms. For example, Heaven (º)
symbolizes creativity, masculinity, activity, and the Father.
The trigrams were also given a basic characteristic of the world of nature and social
experience. From the social world the eight trigrams were also associated with Mother,
Father, three sons, and three daughters (the ‘sons’ represent the principle of movement in
its various stages¾‘beginning of movement’, ‘danger in movement’, and ‘completion of
movement’, while the daughters represent devotion in its various stages: ‘gentle
penetration’, ‘clarity and adaptability’, and ‘joyous tranquility’ (Wilhelm, 1989; see Table
83
5.1). Therefore, natural phenomena, such as water, thunder, and earth, etc., appear in the
symbolism of the trigrams.
Table 5.1
Meanings and Associations of the Eight Trigrams.
Ch’ien heaven Creative, male, active Father
[SOURCE: Barrett, D. V. (1992). Destiny and your dreams. London: Treasure Press.]
Finally, pairing the trigrams produces the 64 hexagrams (see Appendix E). For
example, the combination of two trigrams, K’un plus Ch’ien, gives Hexagram #11, which
is called T’ai and means ‘Prosperity’ (see Figure 5.4). In most cases the hexagrams have
new names that describe their unique qualities.
+ =
Figure 5.4. Constructing Hexagram #11: ‘Earth (K’un) + Heaven (Ch’ien)’ = ‘Prosperity’ (T’ai).
The hexagrams (the paired trigrams) also range as a continuum from the most yang
(Hexagram 1: ‘Heaven above Heaven’) to the most yin (Hexagram 64: ‘Earth above
Earth’; see Figure 5.5). The 64 hexagrams each have their own unique reading, and it is the
readings that form the basic text of the I Ching.
84
Chi’en/The Creative K’un/The Receptive
Figure 5.5. Hexagram 1: Heaven above Heaven (Chi’en) and Hexagram 64: Earth above Earth
(K’un).
Covello (1977) found that there was a rationally derived, systematic substructure to
the I Ching, and he claimed:
One of the more fascinating aspects of the I Ching is that it can be viewed as a
mathematically ordered cosmology. Its surprisingly systematic structure renders
many of its assertions amenable to controlled investigation. (p. 115)
On this basis, Covello investigated the various symbolic systems in the I Ching for their
meaningfulness. He found, by analyzing the line structures making up the hexagrams, and
classifying the words used in the text, that “concreteness” (associated with “earthy,
pictorial, and regressive”) was characteristic of the “Yin principle,” and “abstractedness”
(associated with “heavenly” and “potency”) was characteristic of the “Yang principle.”
Within the abstract-concrete dimension exists a continuum or spectrum of meanings as
determined by the number of yin or yang lines in the hexagrams.
It is important to recognize that yin and yang are not either good or bad, as Von
Franz (1980) makes clear:
The Chinese were detached and philosophical enough to say that even if it is bad
for me it might be good as a whole. From the beginning they had a wiser or more
objective view of what we call good and bad, and saw it more as something in the
ensemble of existence. (p. 47)
Earlier, Jung (1949, cited in Jung, 1989) claimed that “the method of the I Ching does
indeed take into account the hidden individual quality in things and men [sic], and in one’s
own unconscious as well” (p. xxviii). This claim was partly prompted by Jung’s use of the I
Ching, but he was also responding in terms of an ancient Chinese philosophical way of
thinking where “natural laws are merely statistical truths and must necessarily allow for
exceptions” (Jung, 1989, p. xxii).
85
5.2.4 Use of the I Ching
To consult the ‘oracle’ of the I Ching, one must first generate a hexagram, which
involves, first, posing a meaningful question to the I Ching (yes/no questions are
excluded), followed by the repeated casting of 64 yarrow sticks to generate all six lines of
the hexagram. (The modern method involves throwing three coins six times. The ‘coin
method’ was used in the present I Ching experiment.)
In the coin throwing method, each throw of the coins produces one of the six yin or
yang lines, so that, throw by throw, the hexagram is built from the bottom up. The coins
may show three-of-a-kind on some occasions, which means the yin or yang line is a
‘changing’ line (with its own unique reading: the so-called yao text) that changes into its
opposite¾either yin to yang, or yang to yin. In other words, a second hexagram is
produced. One reads the first hexagram reading (the ‘judgement’) in the context of past
and present events, and the second hexagram reading is said to be a forecast or
prognostication of the scenario discussed in the first reading to be read in the context of the
question.
86
(i) Although the mean difference-score was above chance, it was not
significantly so.
(iii) Those who believed in the efficacy of the I Ching scored significantly
higher on difference scores than those who disbelieved.
Regarding finding (ii), Lawrence (1994) used Rosenthal’s (1991, p. 63) Formula 4.2 for
testing the significance of the difference between effect sizes to test the nonsignificant
finding of Thalbourne et al. (1992-1993) against Rubin and Honorton’s (1971, 1972)
significant finding. When the original mean difference scores were converted to effect size
estimates, Lawrence found no significant difference between the two. The conclusion was
that Thalbourne et al. had effectively replicated Rubin and Honorton’s result.
Thalbourne (1994) conducted another experiment using the I Ching, but this time
focussed on changing lines exclusively. Using his own ecologically sound database,
Thalbourne found that the number of changing lines in mid-1992 was significantly higher
than mean chance expectation (1.66 changing lines where MCE = 1.5 changing lines).
Later, in early 1994, the number of changing lines dropped to 0.70 lines, which was
significantly lower than MCE. Thalbourne (1994) concluded that “familiarity” with the I
Ching system had over time reduced his “information-hunger.” Thus the I Ching process
seemed to be influenced by an experimenter effect, so that more “static” hexagrams (i.e.,
no changing lines) were generated in the later period compared to the earlier period (p.
133). This “motivational hypothesis” (p. 133) may explain the earlier result.
While these studies have, to varying degrees, supported the hypothesis that the I
Ching process may involve anomalies, the study by Storm and Thalbourne (1998-1999)
went further by looking for relationships between paranormal ability in the I Ching
‘setting’ and two dimensions not previously tested in any I Ching
experiment¾transliminality (see 5.3.4) and other aspects of personality (see 5.4.2).
In relation to paranormal performance in the I Ching experiment, Storm and
Thalbourne’s (1998-1999) sample of 93 participants (mainly university students) produced
an overall hit-rate of 32% on hexagram-hitting (PMCE = 25%), which was described as
“marginally significant” (p. 109; p = .067). In a follow-up research note (Storm [&
87
Thalbourne], 2001a), a direct significance test of the effect size was found to be significant
(p = .59, p = .048). Number of changing lines was at chance.
88
‘subliminal’ mind (the ‘subconscious’ or ‘unconscious’), the ‘supraliminal’ mind (for
example, states of consciousness and conscious activity), and the external environment.
The origins of transliminality trace back to 1991, when it was first conceptualized
as “an openness or receptiveness to impulses and experiences whose sources are in
preconscious (or unconscious) processes” (Thalbourne, 1991, pp. 181-182). This
conceptualization very closely mirrors the above definition, although it does not account
for the loss of this material to the subconscious or unconscious mind.
A similar two-way process of “leakage” of information between the subliminal
mind and supraliminal mind was described as “transliminal” by Usher and Burt (1909,
cited in Thalbourne, 2000d, p. 3). Rugg (1963, cited in Thalbourne, 2000d, p. 4) used the
same word to describe the attraction of non-conscious creative material into consciousness.
MacKinnon (1971; cited in Thalbourne, 2000d) referred to “transliminal experiences.”
These were the experiences in consciousness of material that came “out of the non-
conscious into the vestibule of the conscious mind,” but which could also go back again
(Thalbourne, 2000d, p. 4).
These various descriptions of the ‘transliminal’ tend to converge in meaning¾they
all refer to material either coming into, or going out of consciousness, or both. Having
established the origins and nature of transliminality, we now turn to the mainly
correlational studies that bear on the transliminality concept.
89
Another constituent of transliminality¾‘attitude to dream interpretation’, from the
MMPI (Dahlstrom, Welsh, & Dahlstrom, 1972)¾was later identified by Thalbourne and
Delin (1999).
Finally, Thalbourne, Bartemucci, Delin, Fox, and Nofi (1997) identified a further
four potential constituents: schizotypal personality (Claridge & Brooks, 1984), ‘fantasy-
proneness’ using the Inventory of Childhood Memories and Imaginings (Form C; Myers,
1983), ‘absorption’, which is “the ability to become totally absorbed in the object of
attention” (Tellegen & Atkinson, 1974), and ‘hyperæsthesia’ (hypersensitivity to
environmental stimulation; see Thalbourne, 1996b).
In the same study, Thalbourne devised the Transliminality Scale (TLS)¾a 29-item scale of
true/false questions derived from nine constituent variables (see Appendix B).
Given that the Launay-Slade Hallucination Scale (Launay & Slade, 1981) already
correlated with transliminality, Thalbourne (1999) hypothesized that vivid visual imagery
90
would also correlate with transliminality. The hypothesis was not confirmed. However, in
a study on creativity, Thalbourne’s (2000b) hypothesis that transliminality would correlate
with the Creative Personality Scale (see 5.3.3) was replicated, although transliminality
failed to correlate with the Revised Barron-Welsh Art Scale (Welsh & Barron, 1963).
Looking further into the nature and correlates of transliminality, Thalbourne and
Houran (2000) used two measures: the Mental Experience Inventory (Kumar & Pekala,
1992) and the AT-20 Scale (MacDonald, 1970) to measure tolerance of ambiguity. They
found that participants scoring high on transliminality also tended to:
Table 5.2
Correlations between Transliminality and Five 16PF Factors
Variable r p
91
Highly transliminal participants were more likely to be ‘warm’ and ‘outgoing’
(Factor A), ‘expedient’ and ‘nonconforming’ (Factor G), ‘imaginative’ and ‘idea-oriented’
(Factor M), ‘receptive’, ‘open-minded’, and ‘intuitive’ (Factor TM), and ‘unrestrained’
(Factor SC). Transliminality also correlated with a belief measure (“Do you believe in your
own abilities to predict the outcome hexagram, or influence the fall of coins so that the
outcome hexagram matches one of your sixteen choices?”).
Transliminality went through three changes in definition during these six years of
experimental research. At first it was identified as an exclusively inwardly generated
process, in the sense that the material seemed to originate in the subliminal mind only. As
other sources were identified, particularly from the work of Thalbourne et al. (1997), the
definition changed in order to include supraliminal sources deemed responsible for
experiences such as absorption and active fantasy-proneness, and also sensitivity to
external stimuli, such as hyperæsthesia.
Finally, the current definition of transliminality (see 5.3.1) was arrived at through
recognition of the fact that the transliminal process was analogous to a two-way street
giving passage to material coming and going in both directions, across a number of
thresholds, into and out of consciousness.
92
ability and an earlier version of the TLS (Form A)¾the transliminality “scores of receivers
making direct hits on the film clips were not discriminated from those making misses”
(Parker, 2000a, p. 10).
In the next section the third and final component of the I Ching experiment will be
explored: Cattell’s 16PF.
93
guessing task) correlated significantly with scores on the paranormal task: Factor C
(Emotional Stability), which was a positive correlation, and Factor O (Apprehension) and
Factor Q4 (Tension), which were both negative correlations. These factors contribute to
Global Factor AX (Anxiety). In fact, high ESP scores were likely to be held by low anxiety
participants.
Nicol and Humphrey (1955) attempted to replicate these results, but failed. There
were no significant correlations with 16PF factors and no directional trends. When they
combined both samples, however, there was a return to the earlier significant results.
Kanthamani and Rao (1971, 1972, 1973) and Rao (1974) used the HSPQ (Cattell’s
High School Personality Questionnaire), which incorporates factors common to the 16PF.
The following results were obtained:
(i) Kanthamani and Rao (1971) used concealed drawings as targets in their ESP
task. Factor A (Warmth), Factor E (Dominance), Factor F (Liveliness), and
Factor I (Sensitivity) correlated significantly and positively with ESP scores.
(ii) Kanthamani and Rao (1972) used extraversion (E) scores only in an ESP
card test. Extraverts’ scores were significantly above MCE, introverts were
significantly below MCE, and there was a significant difference between the
two types.
(iv) Rao (1974) noted that (a) belief was related to performance on the ESP task;
(b) extraversion and neuroticism were related to psi, and (c) combined
personality factors were better predictors of psi performance than individual
factors.
Sargent, Barlett, and Moss (1982) used the ganzfeld setting to test participants’
performance against Factor EX (Extraversion) on the 16PF. The correlation found was
positive, but nonsignificant.
94
Sudhakar and P. Rao (1986) also tested participants in a ganzfeld-GESP setting.
Factor A (Warmth) correlated positively and significantly with ESP scores. This correlation
may have been the result of multiple analyses, given that there are 21 factors in the 16PF.
Some of the results of Storm and Thalbourne’s (1998-1999) experiment with 16PF
factors have already been reported (see 5.3.3). The following results are also relevant to the
16PF:
(i) Significant correlations of hexagram-hitting with six 16PF factors (see Table
5.3). Hitters were therefore more likely to be ‘lively’ and ‘animated’ (Factor
F), ‘socially bold’ and ‘venturesome’ (Factor H), ‘group-oriented’ and
‘affiliative’ (Factor Q2), ‘relaxed’ and ‘patient’ (Factor Q4), ‘extraverted
(Factor EX), and ‘independent’ and ‘persuasive’ (Factor IN).
Table 5.3
Correlations between Hitting and Six 16PF Factors
Variable r p
(ii) Significant correlations of number of changing lines with five 16PF factors
(see Table 5.4). Participants scoring high on changing lines tended to be
‘reserved’ and ‘impersonal’ (Factor A), ‘reactive’ (Factor C), ‘imaginative’
(Factor M), ‘self-reliant’ and ‘solitary’ (Factor Q2), and ‘introverted’ (Factor
EX).
95
Table 5.4
Correlations between Number of Changing Lines and 16PF Factors
Variable r p
47
Eysenck’s E and Cattell’s ‘exvia’ probably refer to the same personality variable, since they correlated
very highly with a single factor in a study by McKenzie, Tindell, and French (1996, p. 272): Eysenck’s E, r =
.91, and Cattell’s Factor EX, r = .85. Gentry, Wakefield, and Friedman (1985) constructed a new MMPI
Extraversion scale in which MMPI items had to correlate significantly with E on the EPQ and found the new
factor overall correlated significantly with Eysenck’s E (r = .79, p < .01).
96
Nash (1966) found that scores on the Social Introversion Scale of the MMPI
correlated significantly and negatively with psi scores in one of eight experiments. In five
experiments, the correlations were consistently negative also (though not significant).
Perhaps Nash’s (1966) eight experiments should have been tested for differences in
performance amongst themselves rather than individually tested for significance. These
studies preceded Rosenthal and Rubin’s (1989) recommendation that an effect size
comparison between studies can be more fruitful for the psi hypothesis.
Interestingly, Eysenck (1967, p. 68) considered the introvert likely to be a poor “psi
participant,” but Palmer’s (1977) and Irwin’s (1986) reviews gave evidence that the
introvert was actually a more likely to be a ‘psi misser’ (a participant who tends to score
significantly below chance).
Palmer (1977, p. 187, Table 1) presented a table of 33 extraversion-ESP
experiments, of which 23 (70%) showed positive relationships between extraversion and
ESP scoring (see also Sargent, 1981, and the extraversion-ESP meta-analyses in 3.1.8 in
Chapter 3). However, Irwin (1999, p. 102) warns that extraversion-ESP correlations may be
more frequently significant and positive because high scorers on extraversion scales have
the ability to relax and feel comfortable in the test situation. Irwin argues that introverts too
might produce significant positive scores if they were able to relax as well in the test
situation as they would do in other more familiar settings.
97
The analytical component of this study involved testing for overall hitting, and
determining correlates, if any, between paranormal ability, transliminality, and 16PF
personality factors, in accordance with the hypotheses given below (see 5.5.2 and 5.5.3).
With the completion of this replication study, there were two samples: Storm and
Thalbourne’s (1988-1999) ‘Sample 1998’, and the sample from this replication study,
‘Sample 1999’. These two samples will be analyzed further and compared in the next
chapter (see Chapter 6).
2. Hitting for second-hexagram ‘hitters’ who were also first hexagram ‘hitters’ is
above MCE (binomial test).
3. Hitting for second-hexagram ‘hitters’ who were first hexagram ‘missers’ is above
MCE (binomial test).
4. Hexagram-hitting for Sample 1999 is lower than hexagram-hitting for Sample 1998
(chi-square test¾see Appendix N).
48
The Pearson r tests were one-tailed for all hypotheses that used this test since they were all directional
hypotheses.
98
6. There is a positive correlation between transliminality and number of changing
lines (Pearson r).
The following psychological hypotheses were proposed. (The tests used are given
in parentheses with each hypothesis.):
(In this study, the only 16PF factors that were tested were those that yielded significant
results in Storm and Thalbourne’s, 1998–1999, first study.)
49
The relationships hypothesized in Hypotheses 9, 10, 11, and 12 were all based on significant post hoc
discoveries of Storm and Thalbourne (1998-1999, pp. 110-111; 114-115).
99
5.6 Method
5.6.1 Participants
A total of 107 participants volunteered for this experiment.50 Thirty-one (29%) were
Adelaide University psychology students from all levels (undergraduate, honours, or
postgraduate), and the remaining 76 (71%) were Adelaide University students not enrolled
in any psychology course, at any level. These 76 participants were either students from
other departments or were found through friends and colleagues by word of mouth. Fifty-
four percent of the total sample were women. The mean age was 26 years (SD = 8.90).
5.6.2 Measures
(i) The ‘states of mind’ scales (see Appendix A). This scale was used in
another experiment to test for relationships between perceived simplicity of
mechanical set-up(s) of paranormal task(s) and paranormal success, and
perceived ease of paranormal task(s) and paranormal success¾see Chapter
11.
(ii) Thalbourne’s (1998) TLS (Form B), which contains 29 items taken from
various scales, 14% of which refer to paranormal phenomena (see Appendix
B). The participant answers “true” or “false” to each item, and the total
number of “true” answers out of 29 is his or her transliminality score.
(iii) The I Ching Hexagram Descriptor Form, which includes a question about
previous use of the I Ching, and 2 questions about belief in the I Ching
process (see Appendix C). The form also contains 64 two-word descriptors
representing each of the 64 hexagrams (Appendix D), but does not contain
the 64 hexagram symbols as used in Storm and Thalbourne (1998-1999, p.
118; see Appendix E).
50
Note that the odd number of participants (viz., N =107) should not raise the suspicion that optional
stopping had taken place in the experiment. It was decided before the experiment began to test 107
participants with a view to combining them wherever possible with Sample 1998 (N = 93), thus yielding a
round total of 200 participants (see Chapter 6 for the analyses of these two samples as a pooled data set).
100
(iv) Cattell’s 16PF, designed to measure and “identify the primary components of
personality,” including five global factors (Russell & Karol, 1994, p. 7).
5.6.3 Apparatus
Ten sets of material were used in the experiment: (1) invitation to volunteers
(Appendix F); (2) information sheet (Appendix G); (3) a consent form (Appendix H); (4)
an I Ching hexagram file, containing an introductory page, a how-to-score page, and the 64
hexagram readings (one reading per page, totaling 64 pages, Wing, 1982, with the
changing line readings on the back of each page, Wing, 1979); (5) three coins (Australian
10-cent pieces¾75% copper, 25% nickel), a coin cup (for shaking the coins), and a felt-
lined box (as a receptacle for the falling coins); (6) a score record sheet for recording coin
throws (Appendix I); (7) a “How to generate an I Ching hexagram” sheet which are
instructions to the experimenter on how to convert the outcomes of the coin tosses to “yin”
and “yang” lines, and whether they were so-called changing lines (Appendix J); (8) an
“eight-by-eight (8 x 8) trigram matrix” for calculating hexagrams (Appendix K); (9) a
debriefing sheet for “hitters” (see Appendix L); and (10) a debriefing sheet for “missers”
(see Appendix M).
5.6.4 Procedure
Once ethics approval was granted from the relevant departmental ethics committee,
psychology students were approached to participate in the experiment by way of a written
invitation lodged in their pigeonholes in the psychology department. Non-psychology
students placed response slips in a ‘ballot’ box placed in the university library.
At the experimental sessions, participants first read the information sheet and then
signed the consent form. The information sheet outlined the general nature of the
experiment, describing it in three stages. Participants were instructed to take their time and
start when they felt ready because there was no time limit. The three stages were as
follows.
Part 1: Participants are required to completie the I Ching Hexagram Descriptor
Form, which also asks three questions:
2. Do you think it is possible for at least some people to exhibit paranormal effects in
this experiment, by predicting the outcome hexagram, or influencing the fall of
coins so that the outcome hexagram matches 1 of their 16 choices?
101
3. Do you believe in your own abilities to exhibit paranormal effects in this
experiment, by predicting the outcome hexagram, or influencing the fall of coins so
that the outcome hexagram matches 1 of your 16 choices?
Question 1 separates the naïve from the sophisticated participants, and Questions 2 and 3
measure the participants’ beliefs about the ostensible paranormal effects involved in the I
Ching process.
Participants were required to choose 16 descriptor-pairs that they felt to be relevant
to their feelings “Lately, or right now . . .” 51 These choices were not ranked. Under the
watchful eye of the experimenter and a witness,52 the participant then threw three coins six
times, recording the number of heads and tails of each throw on the score record sheet,
from the bottom up, according to the conventions of the I Ching.
Each of the six ‘heads-and-tails’ counts was converted to its respective hexagram
line, as shown on the how-to-score page of the hexagram file (see “How to generate an I
Ching hexagram”¾Appendix J). A second hexagram was also generated if changing lines
were produced from throws of three-of-a-kind. Hexagrams were decoded using the 8 x 8
trigram matrix (Appendix K). The bottom three lines and the top three lines each form
trigrams, which are collated with each other with the aid of the trigram matrix to form the
hexagram.
When the participant and the witness were satisfied that the hexagram(s) had been
calculated correctly, they signed and dated the bottom of the score sheet. (N.B.: the
experimenter was also witness to this whole process from the coin-throwing stage to the
signing stage.) A ‘hit’ was a match of the participant’s outcome hexagram with 1 of his or
her 16 selections, as marked on the Hexagram Descriptor Form, whereas in the case of a
“miss” there was no such match.
Part 2: When the I Ching component of the experiment was over, participants
completed the TLS (Form B).
Part 3: The 16PF component (Russell & Karol, 1994) was the third and last stage of
the experiment. Participants completed this component, and their tasks were thus finished.
51
Traditionally, the I Ching process requires a “general question” (Hazel, 1990, p. 7), or a question
“preferably of great personal relevance” (Thalbourne et al., 1992-1993, p. 13). Therefore, the use of the I
Ching in this unorthodox study is itself somewhat unorthodox.
52
The presence of a witness ensured that accurate recordings of coin throws were made. The witness also
certified that no ‘unsuccessful’ coin throws were neglected (and re-thrown).
102
During the three stages, no feedback was given to participants as to whether or not
they were successful at the paranormal task (i.e., whether or not they generated a hexagram
that matched 1 of the 16 chosen on the descriptor form). Neither were participants given
performance feedback or results on the TLS or the 16PF. These precautions were taken in
order to eliminate the rival hypothesis that some (or all) of the significant personality
correlates with hitting were caused by artifactual responses to 16PF questions as a result of
knowledge of success or failure at the paranormal task (e.g., a participant’s mood may shift
from habitual introversion to temporary extraversion if he or she got a hit, or he or she
might suddenly feel that there is some truth to the paranormal process after all). The same
rival hypothesis would apply to significant correlates of transliminality with hitting and
changing lines. (It is noted that lack of performance feedback on the paranormal task might
also produce disgruntlement effects on the personality measures.)
Alternatively, it can be argued that normal feedback may be redundant given the
possibility that anomalous knowledge of a hit or miss could still have a biasing effect on
responses to the test items. It is also conceivable that the choices made on the Hexagram
Descriptor Form could affect item responses. There may be no way of avoiding all these
problems. As was stated in Storm and Thalbourne’s (1998-1999, p. 108) initial study, there
are methodological problems regardless of order of administration of materials and
paranormal task, and in this study the author decided to avoid just some of the more
pertinent problems, for example, the possible motivational obstacle participants might face
when spending one hour, on average, on personality testing before the paranormal task.
Some time after testing, once scores were calculated on both the TLS and the 16PF,
debriefing sheets (stating “you got a hit” or “you did not get a hit,” transliminality scores,
and instructions on how to interpret the 16PF results) were issued to all participants.
Included with each debriefing sheet was a copy of the participant’s consent form and
copies of hexagram readings and changing line readings (if any). Participants were thanked
for their participation in the experiment.
5.7 Results
103
Eighty participants (a surprising 75% of the sample) believed that it was “possible”
for other people in the sample to achieve a hexagram outcome matching 1 of the 16
designated on the descriptor form (i.e., to get a hit) by paranormal means, but only 30
participants (28%) believed in their “own [paranormal] abilities” to get a hit.
40
35
30 32
29
28
25
Frequency
20
15
13
10
5
5
0
0 1 2 3 4 5 6
The average number of changing lines was 1.42 (SD = 1.14), where MCE = 1.5.
This performance was not significantly below average. There is a nonsignificant positive
104
skew53 of 0.393 (SE = 0.234), which only suggests that the task of generating more than
four changing lines might be difficult.
5.7.3 Transliminality
As was the case in Storm and Thalbourne’s (1998-1999) initial study, the TLS was
shown to have a satisfactorily high reliability coefficient, expressed as a Cronbach’s alpha
of .82. This result was only slightly smaller than the previous alpha of .86 reported by
Storm and Thalbourne (p. 108).
The theoretical range of the TLS is 0 to 29, but the observed range was 4 to 28 (N =
107). The mean transliminality score was 16.96 (SD = 5.43), which was not significantly
larger than Storm and Thalbourne’s (1998-1999, p. 108) result of 16.38 (SD = 6.08). The
distribution of transliminality scores is shown in Figure 5.7.
12
11
10
10
9
9
8
8 8
Frequency
7
7 7
6
6 6 6
5
5
4
4 4
3
3 3 3 3 3 3
2
2 2 2
1
1 1 1
0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Transliminality Score
53
Dividing the skew value (.393) by its SE (.234) yields 1.68 which is less than 1.96, and therefore not
significant.
105
Figure 5.7 shows a slight negative skew of -0.173 (SE = 0.234; Kurtosis = -0.467).
Storm (1998, p. 29) reported a slight positive skew (0.094, SE = 0.250), but neither of these
two skews departs significantly from chance. Consequently, any apparent variability of
transliminality scores between the two samples can be taken as chance fluctuation.
106
on the Hexagram Descriptor Form. Note, however, that the typical volunteer for
experiments may likely be the ‘experimenting’ type open to new experiences (i.e., one who
is ‘open to ‘change’).
107
Hypothesis 4: Hexagram-hitting for Sample 1999 is lower than hexagram-hitting for
Sample 1998. Sample 1999 actually scored higher on average (35% hit-rate) than Sample
1998 (32% hit-rate), which is not in the direction hypothesized. The difference was not
significant, c2(1, N = 200) = 0.120, p = .729; see Appendix N for calculations). Thus,
evidence was not found that psi is a mediational process that depends on information
processing (see 5.1.3).
108
possibility that paranormal phenomena might be involved in the I Ching process tended to
generate significantly more changing lines than those who did not believe. (Note that this
measure of paranormal belief does not conform to Schmeidler’s, 1945, original paranormal
belief question, which strictly refers to a generic belief in psi.)
Hypothesis 11: Answers to Question 2 (the sheep question) and Question 3 (the super-
sheep question) correlate positively with each other. Answers to Question 2 (the sheep
question) and Question 3 (the super-sheep question) correlated significantly and positively
with each other, r(105) = 0.32, p < .001. Hypothesis 11 was therefore supported, and Storm
and Thalbourne’s (1998-1999, pp. 114-115) post hoc finding was replicated. Participants
who believed in their own psi ability believed it was possible for other participants in the
sample to have psi ability.
Hypothesis 12: Transliminality scores correlate positively with Factor A (Warmth) and
Factor M (Abstractedness) and negatively with Factor G (Rule-Consciousness), Factor TM
(Tough-Mindedness), and Factor SC (Self-Control) of the 16PF. Table 5.5 shows the results
for these correlations. Transliminality did not correlate significantly or positively with
Factor A, but did correlate significantly and positively with Factor M, suggesting that
highly transliminal participants tended to be idea-oriented. Transliminality also correlated
significantly and negatively with Factor G and Factor TM. Highly transliminal participants,
therefore, tended not to be rule-conscious (Factor G) or tough-minded (Factor TM).
As can also be seen from Table 5.5, this last correlation of transliminality with
Factor SC just missed significance. Nevertheless, it can be regarded as “tell[ing] about the
109
same story” (Rosenthal & Rubin, 1979, p. 1165) as Storm and Thalbourne’s (1998-1999, p.
111) statistically significant result (see Table 5.2), since Rosenthal and Rubin (1979)
would say: “Both results are in the same direction and the studies are of similar size” (p.
1165). Thus, highly transliminal participants tended to be lacking in self-control (Factor
SC).
Table 5.5
Correlations between Transliminality and 16PF Factors
Variable r p
Factor M (Abstractedness) 0.23 .008
Factor G (Rule-Consciousness) -0.24 .006
Factor TM (Tough-Mindedness) -0.17 .042
Factor SC (Self-Control) -0.15 .062
Notes: N = 107; df = 105; p values are one-tailed
110
5.7.7 Belief in the Possibility of Psi as a Necessary Condition
Pursuant to the protocol for testing for necessary conditions, the significant result
found in testing Hypothesis 9 warrants a further (i.e., post hoc) test on the belief variable
‘Possibility’. Those who answered ‘Yes’ to this question (n = 80), were tested separately
on number of changing lines, but the mean number of changing lines, although above MCE
(1.53, where MCE = 1.50), was not significantly above chance. Therefore answering ‘Yes’
to this ‘sheep’ question was not sufficient or necessary in bringing about paranormal
performance on changing lines generation.
5.8 Discussion
The I Ching is an ancient Chinese form of divination based on the principle of
duality, the yin and the yang. From a simple yin/yang polarity are derived the 64
hexagrams. Hexagram symbols carry with them certain meanings, which come in the form
of readings. The traditional view is that in the uncertain flux of world events is the
possibility of knowing that in the present are the seeds for solutions in the future, and that
these solutions can be derived through divination. Replicable evidence was found in this
study that suggested that an anomalous process underlies the I Ching process. Given a
significant 35% hit-rate (where MCE = 25%), the readings do not appear to be generated
purely by chance. Thus there might be some validity in the traditional view.
111
questionnaires, including the time-consuming 16PF. Nevertheless, most aspects of each of
the five aims (see 5.1) were supported. The aims of this study will be given next in terms
of supported hypotheses.
(b) Hypothesis 9: Number of changing lines correlated positively and significantly with
Question 2 (the ‘sheep’ question).
(c) Hypothesis 10 (Part I): Transliminality correlated positively and significantly with
Question 2.
(d) Hypothesis 10 (Part II): Transliminality correlated positively and significantly with
Question 3.
(e) Hypothesis 11: The sheep question (Question 2) and the super-sheep question
(Question 3) correlated positively and significantly.
(f) Hypothesis 12: Transliminality correlated positively and significantly with 4 out of
5 16PF factors.
(a) Hypothesis 5: Transliminality did not correlate positively and significantly with
hexagram-hitting.
(b) Hypothesis 6: Transliminality did not correlate positively and significantly with
changing lines.
(c) Hypothesis 7: Hexagram-hitting did not correlate positively and significantly with 6
factors on the 16PF.
(d) Hypothesis 8: Number of changing lines did not correlate positively and
significantly with 5 16PF factors.
Six of ten hypotheses were confirmed. Generally, Aim One was achieved, where it
was proposed that Storm and Thalbourne’s (1998-1999) findings would be replicated.
112
However, four hypotheses were not confirmed so that replication was demonstrated in the
1999 study only to a limited degree. Furthermore, the four failed hypotheses happen to be
parapsychological, so that these failures confirm, yet again, the well-recognised fact that
psi effects are difficult to replicate, although the successfully replicated psi effects (see
Hypotheses 1 and 9) help to quell some of this doubt.
5.8.1.2 The ESP-PK dichotomy: That the traditional ESP-PK division can be unworkable in
practical parapsychology has been discussed in Chapter 4 (see also 5.1.1). Given that a psi
process took place in the I Ching experiment, the paranormal task would be conventionally
described as involving a physical process (coin-throwing) and/or a mental process
(predicting or precognising the hexagram outcome). Thus, the traditional view is that ESP
or PK may have taken place, or indeed, both ESP and PK may have taken place
simultaneously.
Theorists might like to argue the point over whether the coins are influenced by the
user. If that is the case, the theorist might decide in favour of PK, or blind-PK. The blind-PK
explanation is proposed because naïve participants in Sample 1999 performed better than
non-naïve participants: Naïve (n = 91), P = 35%, p = .017; non-naïve (n = 16), P = 31%, p
= .370). However, results so far do not indicate that blind-PK is necessarily involved in the
I Ching process (see Storm & Thalbourne, 1998-1999, p. 111, where non-naïve
participants performed significantly above chance, but naïve participants did not).
Thus, a middle ground position offers the theorist ‘real time’-ESP (clairvoyance) in
conjunction with PK (causing the coins to fall in the required way), and this process might
involve blind-PK. The blind-PK conclusion raises the possibility of psi mediation through
information acquisition, which implies extra steps for novices (i.e., the non-naïve
participant may already know how to score, whereas the naïve participant does not).
Alternatively, the theorist might decide in favour of ESP (viz., precognition). No
doubt, the debate will continue. Nevertheless, the argument that the ESP-PK division can be
unworkable in practice was supported by the experimental results of the present study.
5.8.1.3 The pro attitude: The results of three hypotheses relevant to this aim are
presented:
(a) Hypothesis 1: Significant hexagram-hitting was achieved (Pobs. = 0.35, where Ptest =
0.25).
113
(b) Hypothesis 2: Significant hexagram-hitting was not achieved for first hexagram
hitters who were also second-hexagram hitters.
(c) Hypothesis 3: Significant hexagram-hitting was not achieved for first hexagram
missers who were second-hexagram hitters.
114
the front page of the I Ching Hexagram Descriptor Form). (See Storm, 1999, who points to
similarities between the paranormal and synchronicity.)
5.8.1.5 Necessary conditions: Four of five hypotheses were not supported (Hypotheses
5, 6, 7 & 8). Generally speaking, Aim Four was not achieved. However, at this early stage,
it may be too soon to conclude that these failed hypotheses indicate that transliminality and
some 16PF factors were not necessary for, or at least conducive to, exo-psychopraxia. (The
failures of these hypotheses to be confirmed, and indeed the appropriateness of these
hypotheses to this study¾at least in their current form¾will be taken up in Chapter 6.)
Hypothesis 9 was supported. There is a relationship between number of changing
lines and the belief (‘sheep-goat’) measure used in the descriptor form (viz., ‘possibility’—
i.e., belief in the possibility of getting a hit through the psi process). This correlation is also
a replication. However, a post hoc analysis (see 5.7.7) revealed that this form of belief in
the paranormal process was not necessary or sufficient in bringing about exo-psychopraxia.
5.9 Conclusion
The overall results of this study send out a few mixed signals about exo-
psychopraxia. Definite support for the theory came in the way of highlighting the ESP-PK
dichotomy and the pro attitude (since psi effects were replicated). But this study failed in
115
its aim to replicate the 1998 results that transliminality54 correlates with hexagram-hitting,
and may therefore be a condition necessary for, or at least conducive to, exo-psychopraxia.
One replicated correlation—changing lines with ‘possibility’ (Hypothesis 9)—gave good
evidence that at least one form of belief might have been necessary for exo-psychopraxia,
but further testing failed to show it. However, generally, it must be conceded that although
psychopraxia may be a more parsimonious description of the events involved in the present
experiment, the study considered as a single research project throws little or no light on the
necessary mediating conditions that must theoretically be involved. Specifically,
knowledge of the level of transliminality and of the 16PF factors was of little use in
predicting hitting or number of changing lines.
The next chapter contains the results of various analyses that were performed on
the sample used in this study (N = 107), as well as the 1998 sample (N = 93). Post hoc
analyses were undertaken as necessary steps towards explaining the failures of Hypotheses
5 and 6 to be confirmed, and the appropriateness of Hypotheses 7 and 8 in their current
form to this study. Further analyses were performed in order to unify two disparate data
sets, thus allowing the calculation of better estimates of the relevant population parameters.
Multiple regression analyses and median-split analyses were also performed in the attempt
to replicate some additional post hoc findings in the 1998 study (see Storm & Thalbourne,
1998-1999, pp. 112-114).
54
Note that replication of the transliminality-belief and transliminality-16PF correlates confirmed the
relevance of transliminality for psychology (see Hypotheses 8, 9 and 10 in 5.7.6)
116
CHAPTER 6
THE I CHING EXPERIMENTS
(ADDITIONAL ANALYSES)55
1. To determine possible reasons for the failures of four hypotheses (i.e., Hypotheses
5, 6, 7, & 8¾see 5.7.5; bivariate correlation analyses were performed).
2. To pool the two samples, where feasible, thus yielding the combined sample (N =
200), from which more accurate estimates of the relevant population parameters
can be calculated.
4. To ascertain the success rates of hexagram hitting for ‘high’ scoring participants on
the transliminality scale in Sample 1999, and the combined samples, as originally
performed on Sample 1998 by Storm and Thalbourne (1998-1999, p. 114) using
55
This chapter was published as a refereed article (see Appendix AF).
117
median-split analyses (MSA). Transliminality was investigated as a possible
necessary condition for bringing about paranormal effects. The 16PF Factors F, H,
Q4, EX and IN were also investigated as possible necessary conditions.
In this section, the results of bivariate correlation analyses are presented, the aim
being to determine possible reasons why Hypotheses 5, 6, 7 and 8 failed to confirm. For
convenience, the results of three performance comparisons between Sample 1998 and
Sample 1999 on transliminality, hitting, and number of changing lines are given in advance
of the Performance Comparison section (see 6.3), so that the combined sample can also be
subjected to a similar analysis in the following section.
118
and Sample 1999 were made on the relevant variables. Transliminality scores did not differ
significantly between samples, t(198) = -0.71, p = .479, two-tailed (w2 = 0). Note that a
significant t value implies the existence of an association, but the estimate of effect size
‘omega-squared’ tells how strong that association is. Hays (1963) recommends that omega-
squared (w2) accompany the result of an independent-samples t test. When t £ 1, the
estimate of w2 = 0; when t > 1, estimated w2 = (t2 - 1)/(t2 + N1 + N2 –1), where Nj is the
size of each sample. Where omega-squared is less than 0.09, the association is regarded as
“functionally” unimportant because there is little predictive power in that association.
Consequently, the samples will be combined on the basis that they may be regarded as
coming from the same population.
Hexagram hit rates did not differ significantly between samples either, c2(1, N =
200) = 0.12, p = .729, two-sided (f = 0.03). A phi-squared value (i.e., the coefficient of
determination equivalent to r2) that falls below 0.09 will also normally be regarded as
unimportant, but this criterion does not apply to the categories of sex and academic
affiliation and, therefore, is not given for the relevant chi-square tests that follow later (see
6.2.3 & 6.2.4).
When the individual group correlations of the combined sample are plotted (see
Figure 6.1), a decline in the transliminality/hitting correlation was sought at the 8th group
before the 9th and 10th groups were formed, which turned out to be significant, rs(6) =
-0.88, p = .004, one-tailed. The decline was expected to continue, but the remaining two
groups when collected went against the linear trend, to produce in fact an overall
significant quadratic trend, rs(8) = 0.57, p = .044, one-tailed. A U-shaped trend might
appear if the experimenter was having an effect on the correlations and enthusiasm and
motivation picked up toward the end of the experiment after “burn out” or “heightened
anxiety to keep up the promising results” (see Broughton & Alexander, 1997, p. 223).
(Note that the two significant correlations in Figure 6.1 are probably nonsignificant when
adjusted for multiple analyses.)
In conclusion, neither Sample 1998, nor Sample 1999, nor the combined sample
produced significant declines. However, the overall transliminality/hitting correlation was
significant for the combined sample, r(198) = 0.12, p = .040, one-tailed, but this very weak
correlation might well have been stronger had it not been for the position effect.56
56
One commentator (S. Ertel, personal communication, September 9, 2000) advised the author to consider
the possibility that the ostensible change in Pearson r values across groups may have been attributable to
declines in the variance on the transliminality measures for those groups, but this rival hypothesis was not
119
.60
.40
.30
.20
.10
-.00
-.10
-.20
-.30
-.40
-.50 p = .012
-.60
1 2 3 4 5 6 7 8 9 10
Figure 6.1. The combined sample (N = 200): Pearson correlations between transliminality and
hitting for each of 10 consecutive groups of 20 participants.
confirmed. There were no significant declines in variances for transliminality across groups, nor were there
for hitting and number of changing lines.
57
Number of changing lines did not differ significantly between samples, t(198) = 0.56, p = .579, two-tailed
2
(w = 0). The samples were therefore combined for this analysis (see 6.3.5 for the result of the
transliminality/changing lines correlation for the combined sample).
120
6.2.3 The Hitting/16PF Correlations
As reported in Chapter 5, not one of the six hitting-16PF correlations that were
significant in Sample 1998 was significant in Sample 1999 (Hypothesis 7, see 5.7.5). It
was considered highly likely that the two samples were not homogeneous on the following
six relevant 16PF factors: Factor F (Liveliness), Factor H (Social Boldness), Factor Q2
(Self-Reliance), Factor Q4 (Tension), Factor EX (Extraversion), and Factor IN
(Independence). This assumption came from an earlier observation (see Storm &
Thalbourne, 1998-1999, p. 116) that there was a preponderance of female and psychology
students in Sample 1998. Conducting tests on one sample with a view to replicating the
results of an earlier sample is unwarranted if there happen to be fundamental differences
between the two samples due to biases of one form or another.58 Therefore, both samples
were investigated for biases in sex composition (because sampling procedures in the 1998
study were different from those of the 1999 study59), and for the same reasons, it was
decided that the number of participants studying psychology should also be investigated.
There was a drop in female participants from 72% to 54% across the two samples
(i.e., 9 fewer female participants in Sample 1999), which proved to be significant using the
Pearson chi-square test, c2(1, N = 200) = 6.75, p = .009 (two-sided). Another bias exists in
the form of a significant drop in the number of psychology students: 50.5% in 1998 to 29%
in 1999 (i.e., 16 fewer psychology students in Sample 1999), c2(1, N = 200) = 9.73, p =
.002 (two-sided). One final chi-square test was conducted, this time comparing four
nonoverlapping groups (female psychology students, female nonpsychology participants,
male psychology students, and male nonpsychology participants). The Fisher Exact Test
was significant (p < .001, two-sided). It was then hypothesized that the original hitting-
16PF correlations may have been specific to groups only and not the whole of Sample 1998.
Pearson r tests were conducted on each of the four nonoverlapping groups in Sample 1998
between hitting and the six relevant 16PF factors.
58
The literature on sex differences on paranormal belief is quite substantial, but for studies on sex differences
in ESP performance, see Palmer and Johnson (1991), Peretti (1971), and Rao and Kanthamani (1981, 1983).
See Parker, Frederiksen, and Johansson (1997) and Parker, Grams, and Pettersson (1998) for studies on ESP
performance differences between psychology students and participants recruited from New Age groups or
groups with paranormal experiences.
59
In the 1999 study, volunteers were sought from all disciplines of Adelaide University, whereas the 1998
study comprised mainly volunteers from the Departments of Psychology, Asian Studies, Architecture, and
Computer Science.
121
Prior to running the Pearson r tests, five one-way analyses of variance (ANOVAs)
were run to determine possible differences between groups on mean scores for the five
16PF factors. Only the mean scores on Factor EX scores were heterogeneous across groups,
but a post hoc Tukey’s test showed that there was only a marginally significant difference
on Factor EX scores between male nonpsychology participants and female psychology
students (p = .087). Consequently, all the relevant Pearson r tests were run for the four
groups. Only one of the four groups produced significant correlations, and that was the
female psychology students (n = 39; see Table 6.1). They produced five of the six
significant correlations originally found for the whole of Sample 1998; the hitting/Factor
IN (Independence) correlation was not significant.
Table 6.1
Correlations Between Hitting and 16PF Factors for Female Psychology Students 1998
Factor r p
F (Liveliness) 0.58 < .001
H (Social Boldness) 0.50 .001
Q2 (Self-Reliance) -0.40 .012
Q4 (Tension) -0.48 .002
EX (Extraversion) 0.48 .002
Notes: p values are one-tailed; n = 39; df = 37
122
remaining participants (n = 83), which excludes the female psychology students, the
hitting-Factor H correlation was not significant, r(81) = -0.08, p = .224, one-tailed.
Given these results, sex and academic affiliation alone do not explain the lack of
replication of four correlations that were found in 1998 (viz., hitting with Factors F, Q2, Q4,
and EX). Nevertheless, one correlation (hitting/Factor H) replicated for female psychology
students, so that, once again, those participants appeared to have disproportionately
contributed to that correlation.
Table 6.2
Correlations Between Changing Lines and 16PF Factors: Four Groups (Sample 1998)
Factor Male psych. Male nonpsych. Female psych. Female nonpsych.
(1998: n = 8) (1998: n = 18) (1998: n = 39) (1998: n = 28)
A (Warmth) -0.67* -0.19 -0.19 -0.22
C (Emotional Stability) 0.34 -0.20 -0.30* -0.34*
Q2 (Self-Reliance) 0.32 0.38 0.33* 0.22
EX (Extraversion) -0.48 -0.45* -0.12 -0.32
Notes: psych. = psychology students
* p < .05 (two-tailed)
On the basis of five significant correlations (see Table 6.2), five respective Pearson
r tests were run to see if these correlations might replicate in the Sample 1999 data. (Note
again that prior to the Pearson r tests, all four groups in Sample 1998 and Sample 1999
were compared using independent-samples t tests. There were no significant differences
between respective groups on the relevant factors.) Only one correlation was significant for
123
male nonpsychology participants on Factor EX, but it was not in the direction
hypothesized.
124
suggested that some unrecognised factor (e.g., relative task difficulty) may account for the
failure of participants to throw five or more changing lines, as shown in Figure 6.2.
Specifically, the task of generating at least four changing lines is up to 10 times more
difficult (PMCE = .037) than the relatively easier task of matching a hexagram (PMCE =
.250). It is possible that participants were intuitively aware of this increased difficulty and,
as a consequence, may have been intimidated by the difficulty of the task (see Chapter 10
for a discussion of the hypothesized effects of perceived difficulty).
70
65
64
60 62
55
50
45
Frequency
40 42
35
30
25
20 24
15
10
5 8
0
0 1 2 3 4 5 6
Added to the difficulty of throwing three-of-a-kind is the fact that participants were
required to perform two paranormal tasks simultaneously. Expectation of significant
performances in both tasks is perhaps analogous to expecting someone to demonstrate a
full comprehension of a text or television program while simultaneously holding a coherent
and uninterrupted conversation with a friend. Future I Ching experiments might yield
significant performances on changing line generation if hexagram targeting is irrelevant.
125
6.3.3 Transliminality
The t test on the two samples showed no significant difference on transliminality
scores (see 6.2.1). The two samples were therefore combined. A mean transliminality score
of 16.69 (SD = 5.73) was found, and with 95% confidence, the population from which
these two samples were drawn would produce a mean transliminality score between 15.89
and 17.49.
The distribution of transliminality scores for the combined samples is shown in
Figure 6.3, which shows an extremely small and nonsignificant negative skew of -0.044
(SE = 0.172). The distribution is comparatively normal, given the larger N, and certainly
more evenly distributed than the scores for Sample 1999 (1999 kurtosis = -0.649. Cf.
Figure 5.7, where the 1998 kurtosis = -0.467).
16
15
15
14
14 14
13
13
12
12 12
11
10
Frequency
10 10 10 10 10
9
8
8 8
7
7 7 7
6
6 6
5
5 5
4
4
3
2
2 2
1
0 1 1 1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
Transliminality Score
126
sizes for the hexagram hitting task for the two samples (again, see 6.2.1), the
transliminality/hitting correlation can be re-calculated for the combined samples.
A significant correlation was found, r(198) = 0.12, p = .040, one-tailed. Although
weak, the non-chance relationship between the two variables suggests that highly
transliminal participants tended to be more successful at hexagram hitting than other
participants. Therefore, transliminality was shown to be a condition conducive to an
expected paranormal effect, even though this relationship was not shown in Sample 1999.
Table 6.3
Correlations between Hitting and 16PF Factors
Variable r p
Factor F (Liveliness) 0.16 .011
Factor H (Social Boldness) 0.21 < .001
Factor Q2 (Self-Reliance) -0.08 .128
Factor Q4 (Tension) -0.13 .044
Factor EX (Extraversion) 0.16 .012
Factor IN (Independence) 0.12 .042
Notes: N = 200; df = 198; p values are one-tailed
127
These five significant correlations reproduce the earlier significant findings of
Storm and Thalbourne (1998-1999, p. 110), suggesting that participants who were
successful at hexagram hitting tended to be lively (Factor F), socially bold (Factor H), free
of tension (Factor Q4), extraverted (Factor EX), and independent (Factor IN).
Given the results of 6.2.3 above, the hitting/Factor H correlation was calculated for
female psychology students only in the combined samples (n = 63), and the result was
found to be significant, r(61) = 0.47, p < .001, one-tailed.
128
transliminality/Ability correlation was significant, r(198) = 0.30, p < .001, one-tailed. High
scorers on transliminality tended to believe that their own paranormal abilities would
directly contribute to a successful outcome on the paranormal task.
Table 6.4
Correlations between Transliminality and 16PF Factors
Variable r p
Factor A (Warmth) 0.13 .031
Factor G (Rule-Consciousness) -0.25 < .001
Factor M (Abstractedness) 0.33 < .001
Factor TM (Tough-Mindedness) -0.26 < .001
Factor SC (Self-Control) -0.23 .001
Notes: N = 200; df = 198. p values are one-tailed
Note that it may still be possible, as stated in Chapter 5 (see 5.7.6, Hypothesis 12),
that the transliminality/Factor A correlation is an artifact of multiple analyses since at least
one significant correlation, which is 5% of 21 correlations (there are 21 16PF factors, and
therefore 21 Pearson r tests were originally performed on Sample 1998 data) could be the
129
result of chance alone. Nevertheless, the combined results indicate that highly transliminal
participants in the combined sample tended to be warm and outgoing (Factor A) and idea-
oriented (Factor M). They also tended not to follow rules (Factor G), were not tough-
minded (Factor TM), and were lacking in self-control (Factor SC).
In this section, the results of an MRA on Sample 1999 using the ‘forward’ method
are presented in an attempt to replicate the regression results in Storm and Thalbourne’s
(1998-1999, p. 112) initial study. An MRA using the ‘forward’ method is also performed on
the combined samples (this method is most favoured at Adelaide University). The aim of
this section is to find predictors of transliminality and hitting.
130
two samples were regarded as homogeneous because the phi-squared value was less than
the critical 9% (see 6.2.1).
A t test was performed on Factor M scores for the two samples, and the result was
also significant, t(198) = -2.69, p = .008, two-tailed (w2 = 0.03). However, the omega-
squared value also did not reach the critical level. The differences between the two samples
on ability and Factor M scores were deemed not important, so they were combined. The
two variables (ability and Factor M) were both entered into an MRA for the combined
sample.
Ability and Factor M both entered the model, with Factor M entering first and
making a moderate contribution as a predictor, R = 0.33 (R2 = 0.11, adjusted R2 = 0.10).
Ability followed, raising the R value to a moderate 0.42 (R2 = 0.18, adjusted R2 = 0.17).
Thus 17% of the variance of the dependent variable (transliminality) was explained by the
two predictor variables. Once again, two of our three findings were replicated (Storm &
Thalbourne, 1998-1999, p. 112). The multiple R was significant, F(2, 197) = 21.20, p <
.001. The beta coefficients were as follows: ability, b = 0.27, and Factor M, b = 0.30.
Factor M made the greater contribution as a predictor, but only a little more than ability.
In this section, the results of MSAs on Sample 1999 and the combined samples are
presented. The first aim is to determine the success rates of hexagram hitting for ‘low’ and
‘high’ scoring participants on (i) the TLS (N = 98; N = 107 and N = 200), (ii) Factor
H¾female psychology students only (Sample 1998: n = 39; Sample 1999; n = 24; and the
combined samples: n = 63), and (iii) Factors F, H, Q4, EX, and IN (N = 200).
The second aim is to determine whether high scores on transliminality, low scores
on Factor Q4, and high scores on Factor H (for female psychology students only), and
Factors F, H, EX and IN were necessary conditions that brought about significant
131
hexagram hit-rates. The binomial test (two-tailed) was used to calculate hexagram hit-rates
as proportions of hits P, where PMCE = 0.25.
132
psychology students with ‘low’ scores (< 6), and the hit-rate was not significant, Pobs. =
0.07 (p = .580). Thus there was evidence that high scoring on Factor H was a necessary
condition, but in this case, only for female psychology students.
It is of interest to note that the hit-rate for all female psychology students in Sample
1999 (n = 24) was also significant, Pobs. = 0.42 (p = .05). However, once again, we are
looking at Factor H as a possible necessary condition. Female psychology students in
Sample 1999 have a median Factor H score of 6. There were only 8 female psychology
students with ‘high’ scores (> 6), but the hit-rate was very high, Pobs. = 0.63 (p = .027).
There were only 10 female psychology students with ‘low’ scores (< 6), and the hit-rate
was not significant, Pobs. = 0.30 (p = .474). Thus there was evidence that high scoring on
Factor H was a necessary condition so long as participants were female psychology
students.
For the combined samples, female psychology students (n = 63) again have a
median Factor H score of 6. There were 24 female psychology students with ‘high’ scores
(> 6), and the hit-rate was again high, Pobs. = 0.54 (p = .002). There were only 25 female
psychology students with ‘low’ scores (< 6), and the hit-rate was not significant, Pobs. =
0.16 (p = .214). Thus again, so long as participants were female psychology students, there
was evidence that high scoring on Factor H was a necessary condition.
133
Table 6.5
Hexagram Hit-Rates for ‘High’ Scorers on 16PF Factors F, H, Q4, EX, and IN
Variable n Median Score Pobs. p
Factor F (Liveliness) 86 6 0.41 .001
Factor H (Social Boldness) 94 5 0.40 < .001
Factor Q4 (Tension) 92a 6 0.37 .006
Factor EX (Extraversion) 65 6 0.42 .002
Factor IN (Independence) 88 6 0.40 .001
Note: N = 200;
a
Low scorers
Table 6.6
Hexagram Hit-Rates for ‘Low’ Scorers on 16PF Factors F, H, Q4, EX, and IN
Variable n Median Score Pobs. p
Factor F (Liveliness) 67 6 0.31 .145
Factor H (Social Boldness) 67 5 0.21 .263
a
Factor Q4 (Tension) 48 6 0.21 .309
Factor EX (Extraversion) 95 6 0.29 .187
Factor IN (Independence) 65 6 0.32 .112
Note: N = 200
a
High scorers
In Table 6.5, it can be seen that participants with high scores (or low scores in the
case of Factor Q4) on Factors F, H, EX and IN produced significant hit-rates. Table 6.6,
however, shows no hit-rates that were significant. Thus it is evident that high scoring on
any, or all, Factors F, H, EX and IN were necessary conditions, and low scoring on Factor
Q4 was also a necessary condition.
In the Bivariate Correlation Analyses section above (see 6.2), four nonoverlapping
groups were individually tested for significant correlations between 16PF factors and the
two paranormal tasks (hitting and changing lines), but only 1 test for replication was
successful out of a total of 10 relevant tests: That is, the hitting/Factor H (Social Boldness)
correlation for female psychology students in Sample 1999 was significant.
134
In the Performance Comparisons section above (see 6.3), 24 specific statistical tests
were performed to test 11 hypotheses. Of these tests, 16 produced significant results that
confirmed or partially confirmed 8 of those 11 hypotheses (i.e., 67% of all tests were
successful, resulting in the confirmation of 73% of all hypotheses tested).
More specifically, for the parapsychological hypotheses, there were 8 successful
tests out of 16 (50%), and for the psychological hypotheses, there were 8 successful tests
out of 8 (100%). All these percentages are greater than 5% and therefore are unlikely to be
explained by chance alone. (Note that these percentages may be inflated due to non-
independent correlations.)
Finally, in the Median-Split Analyses section (see 6.5), all but one of 14 binomial
tests (for Sample 1999 and the combined samples only) produced significant hit-rates on
hexagram hitting. These tests showed that high scoring on Factors F, H, EX and IN were
necessary conditions that brought about a specific paranormal effect (i.e., hexagram
hitting). Low scoring on Factor Q4 was also a necessary condition. It was necessary for
scoring on transliminality to be low in Sample 1999, but high in the combined samples,
leaving only one thing clear¾extreme scoring on transliminality was a necessary condition
that brought about hexagram hitting.
6.7 Discussion
In an initial study with the I Ching, Storm and Thalbourne (1998-1999) sought to
find evidence in the I Ching process of ostensible paranormal effects. In the tradition of
ESP-personality research, they also attempted to find relationships between (a) 16PF factors
and success at two I Ching tasks (i.e., hexagram hitting and number of changing lines) and
(b) transliminality and success at the two I Ching tasks. There were a number of significant
findings (see Storm & Thalbourne, 1998-1999, pp. 108-112).
In a follow-up I Ching study (see Chapter 5) it was shown that the ESP-PK
dichotomy was in this case unworkable in practice because the main paranormal effect
(i.e., hexagram hitting) could not be categorized exclusively as ESP or PK. This effect could
be explained as an example of exo-psychopraxia. However, in the same study, little light
was thrown on the conducive mediating conditions that must theoretically be involved in
the paranormal process which Thalbourne refers to as psychopraxia, although answering
yes to a paranormal belief question (“Do you think it is possible for at least some people to
exhibit paranormal effects?”) was conducive to a paranormal effect (specifically, the
generation of changing lines).
135
In Chapter 5, transliminality and nine factors on the 16PF (Factors A, C, F, H, M,
Q2, Q4, EX, and IN), each of which previously correlated with paranormal performance of
one form or another (see Storm & Thalbourne, 1998-1999), failed to replicate as conducive
conditions of the same paranormal effect.
The analyses in the present chapter were run to determine possible reasons why so
many previously significant findings in Storm and Thalbourne’s (1998-1999) initial study
were not found to be significant in the follow-up study featured in Chapter 5. In the present
chapter, performance comparisons were made in order to combine, where possible, the
relevant variables in the two samples.
The binomial test yielded significant hitting on designated hexagrams¾the official
pro attitude asked of participants (a disposition toward hitting one of these 16 pre-selected
hexagrams) was fulfilled to a significant extent for Sample 1998, Sample 1999, and the
combined sample.
For the combined samples, the fluctuations of the transliminality/hitting
correlations for 10 groups of 20 participants were first shown to conform at the 8th group
mark to a decline effect, which was hypothesized at that point, but later, when all the
relevant data were collected, the correlations of the 10 groups conformed to a U-shaped
trend. Thus, high scores on transliminality were associated with hitting at the beginning
and the end of the experiment, whereas low scores on transliminality were associated with
hitting at the middle of the experiment. The fact that the decline did not continue may be
related to the experimenters focusing on that effect or to extraneous influences coming into
play (see Broughton & Alexander, 1997, p. 223). Nevertheless, transliminality was
repeatedly confirmed (at least suggestively) as being a condition that might help bring
about exo-psychopraxia.
The alternative paranormal measure, number of changing lines, did not replicate in
Sample 1999 as a significant correlation with transliminality, and no evidence of a decline
effect or U-shape trend in 10 transliminality/changing lines correlations across 10 groups
was present that might help explain the failure to replicate. Nor did changing lines
correlate with five relevant 16PF factors in Sample 1999 or the combined sample.
For Sample 1999, six factors on the 16PF did not replicate as significant correlates
with hitting, contrary to previous findings (Storm & Thalbourne, 1998-1999, p. 109),
although a replicated hitting/Factor H (Social Boldness) correlation for female psychology
students was found (see Table 6.1 and 6.2.3). For the combined sample, however, five of
the six significant hitting-16PF correlations reappeared (see Table 6.3). Likewise for
Sample 1999, five factors on the 16PF did not replicate as significant correlates with
136
changing lines, contrary to previous findings (Storm & Thalbourne, 1998-1999, p. 111).
For the combined sample, there were no significant changing lines/16PF correlations.
All five 16PF correlates of transliminality (four of which replicated in 1999) were
significant in the combined sample (see Table 6.4).
Finally, some necessary and sufficient conditions were identified in Sample 1999
and the combined samples. These were high or low scoring on transliminality, high scoring
on Factors F, H, EX, and IN, and low scoring on Factor Q4.
Despite the poor level of replication of correlations in Sample 1999 (especially of
psi-personality correlates), reproduction of many significant correlations was obtained in
the combined sample for (a) parapsychological effects, particularly, the overall hexagram
hitting rates and hitting/16PF correlations, and (b) psychological effects (i.e., the 16PF
correlates of transliminality and various belief correlations). Thus, regarding (a), the effects
of hitting and its correlates are at least consistent with the notion that exo-psychopraxia
was operating and that necessary conditions may apply (i.e., some personality types tend to
be more able than others to produce paranormal effects). Further studies with the I Ching
may put the researcher on firmer ground in regard to establishing the form psychopraxia
and its correlates may take.
Returning to the hexagram-hitting task, it was noted that evidence for one other pro
attitude was not present (at least in sufficient strength)¾second-hexagram-hitting was at
chance (see again, 5.8.1.3). Had there been psi-hitting on second hexagrams, the
experimenter might be forced to conclude that there was good evidence that participants
‘ran amok’ by holding pro attitudes not intended by the experimenter. In accordance with
these results, it is clear that another issue needs to be raised, and this issue, the problem of
compliance, is introduced in the next chapter. In this chapter, a forced-choice experiment is
reported, which was conducted in the attempt to clarify some of the difficulties raised in
respect of the concept of the pro attitude and the problem of compliance. The so-called
‘Gambling’ experiment is also an attempt to discover other conditions that might be
conducive to, and possibly necessary in, bringing about paranormal effects.
137