Content uploaded by Ken Drinkwater
Author content
All content in this area was uploaded by Ken Drinkwater on Nov 07, 2018
Content may be subject to copyright.
Journal of Parapsychology, 79(1), 87–104
SUGGESTION, BELIEF IN THE PARANORMAL,
PRONENESS TO REALITY TESTING DEFICITS AND
PERCEPTION OF AN ALLEGEDLY HAUNTED BUILDING
By Neil Dagnall, Kenneth Drinkwater, Andrew Denovan, and Andrew Parker
ABSTRACT: The present study investigated whether suggestion, level of belief in the paranormal, and
proneness to reality testing decits inuenced participants’ expectation of haunt-related phenomena.
Participants watched a short slideshow outlining the history of a ctitious, abandoned hospital. Sugges-
tion occurred in the nal sentence of the presentation narration and stated that the hospital adminis-
trative building had either a history of ghostly activity or structural problems. Following the slideshow,
to ensure participants attended to the suggestion, they read a transcript of the presentation narration.
The experimenter then informed participants that they would see the internal features of the admin-
istrative building via a soundless, black and white video tour. On conclusion of the lmed sequence,
participants completed measures assessing environmental perceptions and phenomena, haunt-related
opinions and feelings, belief in the paranormal (Revised Paranormal Belief Scale), and proneness to re-
ality testing decits (Inventory of Personality Organization). Within the experimental phase, only level
of paranormal belief and proneness to reality testing decits affected haunt-related ratings; suggestion
had no effect. Second phase inquiry, using path analysis, revealed that haunting history (the extent to
which participants believed the administrative building had a history of being haunted) mediated the
relationship between paranormal belief and expectation of haunt-related phenomena.
Keywords: suggestion, paranormal belief, reality testing, haunting
Denitions of ghosts vary over time and across cultures (Houran & Lange, 2001). The term “ghost”
refers traditionally to the notion that spirits of the dead (human and animal) persist after corporeal death and
exert an inuence on the physical world. More precisely, as dened by Laythe and Owen (2012), haunting
experiences denote internally perceived phenomena (e.g., sensations of a presence) or externally witnessed
phenomena (e.g., objects moving), ascribed to spirit activity.
Belief in and experience of ghosts persists within modern society. Indeed, opinion polls report con-
sistently that a substantial proportion of the general population believe in the existence of ghosts (Williams,
Ventola, & Wilson, 2010). Illustratively, a 2005 Gallup survey, incorporating telephone interviews with
1,002 American adults, found that 32% of interviewees believed that ghosts (spirits of dead people) could
return to certain places/situations; 37% considered houses could be haunted (Moore, 2005). These gures
are commensurate with an earlier 2001 Gallup survey (Newport & Strausberg, 2001). MORI polls evidence
similar levels of endorsement in Britain. The 2007 Survey on Beliefs, comprising telephone interviews
with a representative quota sample of 1,005 adults, noted that 38% of interviewees believed in ghosts and
36% claimed to have seen a ghost (MORI, 2007). These gures concur with the 1998 MORI Paranormal
Survey, which found that 40% of respondents believed in ghosts and 37% reported personal experience of
ghosts (MORI, 1998). The prevalence of ghost-related beliefs and relatively frequent reporting of haunting
experiences indicates the socially important nature of haunting phenomena and designates ghosts/hauntings
as an important research area worthy of academic consideration.
Empirical attempts to explain ghost and haunt-related perceptions centre frequently on psycho-
logical factors. One signicant variable is suggestibility. Generally, research has found associations be-
tween suggestibility and belief in the paranormal (Danoiu, 1995) and that experimental manipulation of
The Journal of Parapsychology
88
verbal suggestion can inuence perception and recall of paranormal phenomena. For example, Wiseman,
Greening, and Smith’s (2003) study of séance phenomena, using self-selected delegates attending a Fortean
Times convention, observed that participants who afrmatively answered the question “Do you believe
that paranormal phenomena sometimes occur during séances?” were more susceptible to verbal sugges-
tion about a séance-consistent phenomenon, movement of a hand-bell, than nonbelievers. Participants also
reported experiencing unusual phenomena often associated with “genuine” séances; about a fth believed
the staged séance contained authentic paranormal phenomena, and a signicantly greater percentage of
believers considered this to be the case.
Similarly, Wiseman and Greening (2005) found that verbal suggestion affected perception of al-
leged paranormal key bending. Participants viewed footage of a performer (professional close-up magi-
cian) and an interviewer sitting at a table containing several objects (keys, pack of cards, cutlery, etc.).
The camera showed a close-up of the performer’s hands as he selected the key. They then used specious
psychokinetic ability to produce a bend in the stem of the key (sleight of hand produced the distortion).
The performer placed the key back on the table and the videotape footage concluded with a 60-s close-up
of the distorted key. In the suggestion condition, a verbal comment on the soundtrack implied that the key
continued to bend. The no-suggestion condition did not include this “bending” comment. Participants in
the suggestion condition were more likely to report the key continuing to bend. In addition, participants
reporting bending were highly condent that their testimony was reliable and were less likely to report the
fake psychic’s suggestion.
Wilson and French (2008) examined whether suggestion inuenced recall of a psychic reading.
Participants watched a scripted video of a reading followed by a “manipulated” interview in which the sitter
commented upon the accuracy of the reading. In one version, the sitter declared correctly that the psychic
mentioned the name Sheila and that Sheila was their mother’s name. In the other version, the sitter asserted
incorrectly that the psychic stated that their mother’s name was Sheila. Wilson and French (2008) found
believers showed a strong tendency to misremember the manipulated section of the reading, regardless of
whether they received misinformation or not. Nonbelievers tended to remember the reading more accu-
rately if no misinformation was present; however, in the misinformation condition their memories were as
distorted as those of the believers.
Collectively, research shows that verbal suggestion can inuence the perception and recall of sup-
posed paranormal phenomena, especially when the suggestion is consistent with existing paranormal be-
liefs. Pertinently, research indicates that suggestion may play a critical role in the reporting of haunting
phenomena (O’Keefe & Parsons, 2009). For example, Lange and Houran (1997) investigated whether
the suggestion that a location was haunted would be sufcient to induce poltergeist-like perceptions (e.g.,
sensed presence). Participants attending an indoor performance theatre took part via random allocation to
either the suggestion (paranormal activity) or control (renovation) condition. Each group went on a tour
in which they visited ve main theatre areas and completed an experiential questionnaire assessing their
psychological and physiological perceptions. More intense experiences across measures demonstrated that
mere suggestion could stimulate paranormal-type experiences.
Similarly, Terhune and Smith (2006), using a psychomanteum (mirror gazing task), established
that suggestion could induce apparitional experiences. In the suggestion condition, instructions specied
that mirror gazing could result in anomalous sensations (including seeing an apparition); the nonsuggestion
condition advised only about the possibility of experiencing unusual bodily sensations/perceptual distor-
tions. Suggestion inuenced participants’ perceptions; the suggestion condition produced more reports of
visual and auditory apparitions.
Research has shown suggestibility (hypnotic and imaginative) is associated with the induction of
anomalous/unusual experiences (Kumar & Pekala, 2001; O’Connor, Barnier, & Cox, 2008) and the stimu-
lation of hallucinatory experiences (McConkey & Barnier, 2004). A notable example is the White Christmas
paradigm (Barber & Calverey, 1964). The White Christmas test instructs participants to close their eyes and
imagine hearing the famous Bing Crosby White Christmas song. After 30 s, participants rate the intensity of
their imagery. Typically, a signicant percentage of participants report hearing the song clearly. Mintz and
Inuences on Perception of an Allegedly Haunted Building 89
Alpert (1972) reported that the majority of schizophrenic patients (85%) and a nontrivial minority (40%)
of controls reported a clear auditory image during the test (Merckelbach & van de Ven 2001). This nding
illustrates that “normal” people will readily report suggested auditory events (Barber & Calverey, 1964).
Cumulatively, research demonstrates that the presentation of haunt-related suggestions can induce
and heighten haunting-related perceptions (O’Keefe & Parsons, 2009). The relationship between sugges-
tion, paranormal belief, and the reporting of unusual (ghost-related) experiences, however, is not a simple
one and results across studies have been inconsistent. A classic illustration of this is the large-scale study
conducted by Wiseman, Watt, Greening, Stevens, and O’Keefe (2002). They found that the frequency with
which participants reported experiencing unusual phenomena in the past and attributed their experiences
to ghosts varied as a function of level of belief. Believers reported experiencing seven of the eight unusual
phenomena (feelings, sense of presence, sounds, changes in temperature, smells, sights, and tastes). The
only phenomenon not reported more frequently by believers was an unusual sense of dizziness. Believers
were also signicantly more likely to attribute their experiences to ghosts. Similarly, when walking around
Hampton Court Palace, an allegedly haunted location, believers noted more haunt-related experiences and
demonstrated a greater tendency to attribute these experiences to a ghost.
Wiseman et al. (2002) also manipulated suggestion. In the positive suggestion condition, research-
ers told participants that an area was associated with increased reports of unusual phenomena. Contrasting-
ly, in the negative suggestion condition, the researchers stated that the area was not associated with unusual
phenomena. Suggestion had no effect on participants’ expectations of experiencing unusual phenomena
during the experiment or their tendency to attribute unusual phenomena to ghosts. In addition, the belief by
suggestion interaction produced mixed results. When participants visited the allegedly haunted area, more
believers reported unusual experiences in the positive suggestion condition. Generally, however, ndings
proved nonsignicant.
Overall, the results of haunt-related suggestion research are inconclusive and difcult to interpret.
Generally, ndings suggest that believers in the paranormal are particularly susceptible to suggestion and
more easily deceived. However, this assumption is inconsistent with previous critical thinking research,
which has failed to nd consistent differences between believers and nonbelievers (Hergovich & Arendasy,
2005; Moore, Thalbourne, & Storm, 2010). In this context, philosophical bias may affect susceptibility to
suggestion. Believers are prone to wrongly endorsing paranormal phenomena, whilst sceptics are inclined
to deny the existence of paranormal effects. Hence, sceptics would be unlikely to detect genuine effects (if
they exist).
The present study built on the work of Wiseman et al. (2002) in a number of important ways.
First, to avoid problems linked to prior inuence, the location used was ctitious. Real, historical locations
(Hampton Court, Edinburgh Vaults, etc.) are often associated with hauntings and knowledge of this may in-
uence participants’ perceptions (Houran, Wiseman, & Thalbourne, 2002). Secondly, the manipulated sug-
gestion appeared consistently throughout the presentation phase. It occurred on three occasions: the end of
the narrative, within the narrative text, and as part of instructions prior to producing ratings. Contrastingly,
Wiseman et al.’s (2002) use of suggestion was subtle (embedded within a talk about their experiment) and
stated only once. Thirdly, only the haunting condition referred to the notion of haunting. This avoided any
potential confound arising from stating that the location was not haunted. Mention of haunting, regardless
of valence, may inuence expectations, attention, and report rates. Previous work has indicated that atten-
tional mechanisms play an important role in haunting experiences (Houran & Lange, 1996, 2001; Terhune,
Ventola, & Houran, 2007). For this reason, the present study employed a haunted versus control (structure)
manipulation.
Finally, to avoid the possible inuence of external variables (magnetic elds, drafts, etc.) the cur-
rent study took place in a controlled experimental setting. This ensured that participants’ experiences were
similar and prevented the introduction of physical conates (temperature uctuation, lighting variations,
magnetic, electric elds, etc.). Using a controlled, nonhaunted location avoided issues arising from individ-
ual differences in perceptual acuity/sensitivity. There is evidence to suggest that believers in the paranormal
may be more reactive to perceptual stimuli generally, and specically more sensitive to possible paranormal
The Journal of Parapsychology
90
effects (Houran, Hughes, Thalbourne, & Delin, 2006; Thalbourne, 2010).
A further development was the inclusion of a reality testing measure, the Inventory of Personal-
ity Organization, IPO-RT (Lenzenweger, Clarkin, Kernberg, & Foelsch, 2001). Several studies report an
association between proneness to reality testing decits and unconventional beliefs, particularly belief in
the paranormal and endorsement of urban legends and conspiracies (Drinkwater, Dagnall, & Parker, 2012;
Irwin, 2004), and there have been no reported failures. Reality testing refers to the preference to test crit-
ically the coherent credibility of beliefs (Irwin, 2004). Reality-testing decits bias individuals away from
analytical-rational processing towards intuitive-experiential interpretations of anomalous events. Believers
in this context are dependent upon, or favour, intuitive-experiential processing and consequently appraise
perceptions and experiences less critically (Lindeman, 1998). Accordingly, they are more inclined to report
unusual perceptual sensations (seeing things that do not exist, hearing things when there is no apparent
reason, etc.). Similarly, emotion-based reasoning predicts level of paranormal belief (Irwin, Dagnall, &
Drinkwater, 2012). Thus, believers tend to endorse paranormal occurrences because of their emotional
rather than rational appeal (Sappington, 1990). In this context, the authors anticipated that participants scor-
ing high on proneness to reality testing decits would score higher on haunt-related expectancy measures.
Similar to paranormal belief, the relationship between proneness to reality testing decits and susceptibility
to suggestion was less certain and, hence, worthy of investigation. This was tested in Phase 1 alongside the
paranormal hypotheses.
The study comprised two distinct but related phases. Phase 1, experimental manipulation, examined
the effects of verbal suggestion on paranormal believers (vs. nonbelievers) and perception of the ctitious hos-
pital building. It was hypothesised, consistent with previous research, that believers (vs. nonbelievers) would
expect the building to contain more haunt-related phenomena, and the differences would be greater with sug-
gestion. Given the inconsistent nature of suggestion effects, the latter prediction was tentatively stated.
As outlined above, Phase 1 also addressed the relationship between suggestion and proneness to
reality testing decits. Previous research has found a positive association between belief in the paranormal
and proneness to reality testing decits (Drinkwater et al., 2012; Irwin, 2003, 2004). This relationship
suggests, consistent with belief-in-the-paranormal research, that participants scoring high on proneness to
reality testing decits would be more susceptible to manipulation of suggestion. In the context of the pres-
ent study, the suggestion that the building was haunted was likely to coincide with personal subjective, in-
tuitive-experiential evaluations and existing paranormal-related schemas of haunted locations (antiquated,
deserted, isolated, former hospital, etc.). Thus, we hypothesized that participants scoring high for proneness
to reality testing decits would correspondingly rate the expectation of haunt-related phenomena higher,
and that this effect would be heightened when it was suggested that the location was haunted.
Phase 2 explored relationships between paranormal belief, proneness to reality testing decits, the
degree to which participants believed the building was haunted, and anticipation of haunt-related phenomena.
Method
Participants
One hundred eighty participants, recruited via opportunity sampling from undergraduate students
and staff at Manchester Metropolitan University (faculties of Humanities, Languages & Social Sciences
and the Department of Psychology), participated in the study. Mean age 20.89 years (SD = 4.99), range
18–48. The sample was 75% female (M = 20.96, SD = 5.42, 18–48 years) and 25% male (M = 20.71, SD
= 3.48, 18–38 years). Participants responded to general e-mails, noticeboard advertisements, or personal
requests to take part in a psychological study.
Hospital Photographs
A slideshow containing 24 photographs of hospitals was constructed. Pictures were obtained via
an Internet search using the term “disused hospitals.” Selected images were public domain, colour, non-dis-
torted, taken during daytime, and representing a range of internal and external architectural features. Col-
Inuences on Perception of an Allegedly Haunted Building 91
lectively, images appeared to depict a single derelict hospital. Each image was mounted and tted to a
single PowerPoint 2010 slide (4:3 on screen; 25.40 cm x 19.05 cm). Slide presentation was set to 5 s and
the slideshow ran for 2 min in total. A standard narration, read by a male member of staff, accompanied the
slideshow and outlined the building’s (ctitious) history. Recording of the narration was via a tablet and the
initial recording was edited using Audacity software.
Suggestion Manipulation
The suggestion stated that the building had either a history of ghostly activity (haunted manipu-
lation) or structural problems (structure manipulation): “In the years prior to its closure patients, staff and
visitors reported ghostly sightings/structural problems and the building developed a reputation for being
haunted/outdated. The centre of spectral activity/architectural concerns was the Administrative Building.”
To ensure equivalence between conditions, changes were restricted to three statements of the key manipu-
lation phrase. These occurred at the end of the narrative, within the narrative text, and as part of instructions
prior to producing ratings.
Video Tour of Administrative Building
After the slideshow, participants undertook a virtual (video) tour of the Administrative Building.
The video comprised handheld camera footage of a building walkthrough, was in colour, and contained no
sound. Filming was from the rst person perspective and no people appeared in the shoot. The building
was actually a disused location on one of the University campuses. Prior to the video shoot, the internal
features of the building (i.e., doorways, staircase, furniture, room size, and décor) and the photographs of
disused hospitals were carefully matched. This control ensured that participants were unlikely to perceive
a mismatch between the slideshow and the virtual tour. Editing reduced the original recording to 10 min;
coverage featured a range of internal features. Filming started on the ground oor, moved down to the cel-
lar, and then progressed up through the second and third oors. Shooting concluded as the camera started
to move back down the stairs. Recording used a Panasonic HC-V130EB-R Camcorder. Editing was via
CyberLink PowerDirector.
Test Booklet
After watching the slideshow and nishing the virtual tour, participants completed a test booklet
containing the following measures.
Environmental Perceptions and Phenomena Scale (EPPS). The EPPS is an 18-item measure
adapted from previous research on hauntings and contains questions on a range of perceptions and sen-
sations typically associated with hauntings (Laythe & Owen, 2012; Wiseman et al., 2002). Instructions
asked participants to consider how they would feel when visiting and exploring the hospital Administrative
Building late at night: “Within the Administrative Building would you expect to encounter the following
phenomena?” Participants rated the extent to which they expected to encounter the following unusual,
haunt-related phenomena (sensations and perceptions): unexplained voices; feeling of a sensed presence;
eeting visual shadows; heard footsteps; bouts of ringing in the ears; headache/migraine; perception of
being touched; bouts of dizziness; periodic feelings of foreboding; heard bangs/raps; heard music from an
unexplained source; unexplained odours; mist, lights, shadows, or other unexplained visual phenomena;
notice of animals reacting to something that isn’t there; sudden extreme cold or heat; heard or seen unex-
plained movement; and feelings of being watched. In this context, the EPPS provided a measure of the de-
gree to which participants expected to encounter haunt-related phenomena within the Administrative Build-
ing. Participants rated each phenomenon on a Likert scale (0% = “certainly not” to 100% = “certainly”.
History of Haunting (HH). A further question assessed whether participants believed the location
was haunted: “To what extent do you believe that there is a history of paranormal activity (ghostly activity)
at the location observed in the video (the Administrative Building)?” Participants responded on a 7-point
The Journal of Parapsychology
92
Likert scale (1 = “denitely not” to 7 = “denitely”).
Haunting-Associated Opinions and Feelings (HAOF). A single item assessed general attitudes to
the existence of ghosts: “To what extent do you believe that ghosts exist?” Responses were again assessed
using a 7-point Likert scale (1 = “denitely not” to 7 = “denitely”).
Two items measured the degree of anxiety (“How anxious do you feel when you think about
ghosts?”) and fear evoked by the notion of ghosts (“How fearful of ghosts are you?”). Responses were
recorded on a 7-point Likert scale (1 = “not at all” to 7 = “extremely”).
Revised Paranormal Belief Scale (R-PBS). The R-PBS is a self-report measure containing 26
questions measuring belief in seven facets of paranormal belief: Traditional Religious Belief, Psi Belief,
Witchcraft, Spiritualism, Superstition, Extraordinary Life Forms, and Precognition (Tobacyk, 2004). The
scale is a modied version of Tobacyk and Milford’s (1983) Paranormal Belief Scale. Items are presented
as statements (e.g., “I believe in God” and “Black magic really exists”) measured on a 7-point Likert scale
ranging from 1 (“strongly disagree”) to 7 (“strongly agree”). Prior to analysis, scores were recoded 0–6
to facilitate Rasch scaling (Irwin, 2009). Final potential scores ranged from 0–156, with higher scores re-
ecting greater belief in the paranormal. Rasch scaling produced a 2-factor solution: New Age Philosophy,
NAP, and Traditional Paranormal Belief, TPB (Lange, Irwin, & Houran, 2000). NAP (11 items) assesses
belief in psi, reincarnation, altered states, and astrology, whilst TPB (5 items) measures traditional Western
religious concepts such as the devil and witchcraft (Irwin, 2004). The Rasch scaling procedure (Andrich,
1988) produces scores ranging from 6.85 to 47.72 on NAP, and 11.16 to 43.24 on TPB. Previous research
has established that the R-PBS is psychometrically and conceptually satisfactory (Tobacyk, 2004).
The Inventory of Personality Organization Reality Testing Subscale (IPO-RT). The IPO-RT
(Lenzenweger et al., 2001) assesses proneness to reality testing decits, and the capacities to differentiate
self from non-self and intrapsychic from external stimuli, and to maintain empathy consistent with ordinary
social criteria of reality (Kernberg, 1996). The IPO-RT also measures information-processing style (e.g.,
“I have heard or seen things when there is no apparent reason for it’) without reference to psychotic symp-
tomology. The scale contains 20 items assessed via a 5-point Likert scale (1 = “never true” to 5 = “always
true”). Scores range from 20 to 100 (low scores indicate high reality-testing ability). The IPO-RT is tem-
porally stable with nonclinical populations and is psychometrically established, demonstrating good retest
reliability and construct validity (Lenzenweger et al., 2001).
Subjective Paranormal Experiences Scale (SPES). In addition to the previously outlined mea-
sures, participants completed the SPES, which measures participants’ general paranormal history and as-
sesses incidence of subjective paranormal experiences. The measure was not analysed in the present study
because the number of respondents reporting hauntings was low.
Procedure
Ethical approval was granted as part of a wider research project examining the relationship between
paranormal beliefs and anomalous experiences. Prior to testing, all participants provided informed consent
and were aware of their right to withdraw from the study at any time.
Participant testing occurred in groups or individually and haunt condition allocation was random
(suggestion manipulation: haunted vs. structure).
Prior to testing, instructions informed participants that the session would involve attending to a
brief slideshow and accompanying audio narrative about an old disused hospital. Participants rst watched
the slideshow containing the hospital photographs and simultaneously listened to the supporting audio nar-
rative. The rst presentation of the suggestion manipulation occurred within the concluding sentence of the
narrative. On conclusion of the presentation, participants read a text version of the narrated commentary
in preparation for the virtual (video) tour of the old Administrative Building. This served to reinforce the
previously narrated suggestion (second presentation of the suggestion manipulation). After allowing suf-
cient reading time, the recording (virtual tour) was played. Following the virtual tour, instructions guided
participants to complete the test booklet containing the study measures. The instructions preceding the test
Inuences on Perception of an Allegedly Haunted Building 93
measures stated the suggestion manipulation a third and nal time.
Participants completed the environmental measures (EPPS and HH) and the HOAF rst to avoid
priming paranormal ideation. To avoid potential order effects the sequence of the remaining measures
(R-PBS and IPO-RT) was counterbalanced. Guidelines instructed participants to complete all questions
and that there was no time limit. At the end of the testing session, testers debriefed the participants, thanked
them for engaging with the research, and provided follow-up details.
Results
Primary Analysis: Reliability and Scale Descriptives
Prior to the main analysis, Cronbach’s alpha (α) assessed the internal reliability of the measures. All
scales proved psychometrically acceptable: The Environmental Perceptions Phenomena scale (EPPS; α =
.93), proneness to reality testing decits measure (IPO-RT; α = .90) and Revised Paranormal Belief Scale,
(R-PBS; α = .89) demonstrated excellent internal reliability. The two R-PBS factors (New Age Philosophy,
NAP; and Traditional Paranormal Belief, TPB) produced alpha coefcients in the good (NAP, α = .83) to
acceptable TPB (α = .70) range (see also George & Mallery, 2003). Scale descriptive statistics appear in
Table 1.
Table 1
Scale Descriptive Statistics (N = 180)
α M SD Range
EPPS .93 43.73 18.09 2.35–82.94
R-PBS .89 51.97 24.01 4.00–116.00
NAP .83 21.60 3.88 6.85–31.41
TPB .70 22.59 4.56 11.16–39.23
IPO-RT .90 42.81 11.91 20.00–83.00
HH NA 4.12 1.84 1.00–7.00
OG NA 4.18 2.04 1.00–7.00
FA NA 3.61 1.86 1.00–7.00
FF NA 3.41 1.92 1.00–7.00
Note. EPPS = Environmental Perceptions Phenomena Scale, R-PBS = Revised Paranormal Belief Scale,
NAP = New Age Philosophy, TPB = Traditional Paranormal Belief, IPO-RT = Proneness to Reality Testing
Decits, HH = History of Haunting, OG = Opinion Ghosts Exist, FA = Feeling Anxiety, FF = Feeling Fear.
EPPS correlated positively with level of paranormal belief and IPO-RT, indicating that expectancy
of haunt-related phenomena increased as a function of level of paranormal belief and reality testing decit.
Paranormal belief (R-PBS, NAP, and TPB) correlated positively with IPO-RT. All observed interscale cor-
relations were in the moderate range (see Table 2).
Experimental Manipulation: Paranormal Belief, Proneness to Reality Testing Decits,
Suggestion, and Haunting-Related Ratings
Consistent with previous research, median splits classied participants as either low or high (be-
low vs. above median) on R-PBS and IPO-RT. This procedure is an established methodological conven-
tion within parapsychological and individual differences research (Wilson & French, 2014; Wiseman &
The Journal of Parapsychology
94
Greening, 2005) that has recently received disapproval. Ubiquitously cited criticisms of median split often
reference MacCallum, Zhang, Preacher, and Rucker (2002), who outline potential negative analytical con-
sequences arising from dichotomization of continuous variables (i.e., loss of information about individual
variability, ensuing loss of power and effect size, and the undermining of measurement reliability). Accord-
ingly, MacCallum et al. (2002) concluded that dichotomization is rarely defensible because it produces
misleading results.
Table 2
Matrix of Pearson Product Moment Correlations Between Study Variables
1 2 3 4 5 6 7 8 9
1. EPPS
2. R-PBS .57*
3. NAP .52 .85
4. TPB .44 .81 .65
5. IPO-RT .40 .35 .37 .21
6. HH .66 .59 .52 .49 .20
7. OG .52 .61 .54 .52 .21 .74
8. FA .41 .32 .23 .31 .22 .39 .43
9. FF .33 .29 .17 .29 .19 .33 .39 .84
Note. EPPS = Environmental Perceptions Phenomena Scale, R-PBS = Revised Paranormal Belief Scale,
NAP = New Age Philosophy, TPB = Traditional Paranormal Belief, IPO-RT = Proneness to Reality Testing
Decits, HH = History of Haunting, OG = Ghosts Exist, FA = Feeling Anxiety, FF = Feeling Fear.
*All p < .01.
A recent paper by Iacobucci, Posavac, Kardes, Schneider, and Popovich (2014) challenges and
re-evaluates this view. Iacobucci et al. (2014) conducted an examination of median splits, which revealed
that spurious effects were negligible and limited (usually) to instances where predictor variables correlated
signicantly among themselves, that is, cases of acute multicollinearity. Indeed, even when variables are
nonorthogonal, the robust nature of a 2 by 2 factorial design typically negates extreme distortions (Iaco-
bucci et al., 2014). Finally, it is worth noting that multicollinearity generally does not promote type I errors.
Instead, it reduces effect sizes and increases the likelihood of type II error (Iacobucci et al., 2014).
We assessed the validity of our ndings in two ways. Firstly, following Iacobucci et al.’s (2014)
recommendation, we demonstrated orthogonality by computing correlation coefcients between the me-
dian split variables (belief in the paranormal and proneness to reality testing decits) and the experimental
factor (level of suggestion). The correlations were nonsignicant: R-PBS, r(178) =. 02, p = .37; IPO-RT,
r(178) = .00, p = .50.
As a second validity check, consistent with Wilson and French (2014), we conducted a series of
multiple regressions (not reported), which replicated the pattern of results observed from the ANOVAs re-
ported below (see Tables 3 and 4).
ANOVA analysis was preferred to regression analysis because suggestion was an experimental
manipulation and we were concerned primarily with exploring potential interactions. Separate ANOVAs for
level of paranormal belief (R-PBS) and proneness to reality testing decits (IPO-RT) were justied because
the correlation between these two factors was in the low (see Hinkle, Wiersma, & Jurs, 2003) to moderate
(see Cohen, 1992) range. Thus, effects of level and suggestion on environmental perceptions (EPPS) and
history of haunting (HH) were analyzed via separate 2 (level of paranormal belief/level of reality testing:
high vs. low) x 2 (suggestion: structure vs. haunted) independent ANOVAs.
Inuences on Perception of an Allegedly Haunted Building 95
Paranormal Belief
Means and standard deviations appear in Table 3.
Table 3
Level of Paranormal Belief (R-PBS) as a Function of Haunt-Related Attitudes
(Environmental Perceptions Phenomena Scale, EPPS, and History of Haunting, HH)
Level of Paranormal Belief
Below Median Above Median Overall
M SD M SD M SD
EPPS
Suggestion Type
Structure 35.58 18.21 51.15 16.02 43.28 18.78
Haunted 36.62 16.67 52.63 14.33 44.19 17.48
Overall 36.12 17.35 51.88 15.14
HH
Suggestion Type
Structure 3.53 2.02 5.07 1.32 4.29 1.87
Haunted 3.10 1.73 4.88 1.40 3.95 1.81
Overall 3.31 1.88 4.98 1.36
Environmental Perceptions Phenomena Scale (EPPS). No signicant main effect on EPPS was
found for suggestion (structure, M = 43.28, SD = 18.78 vs. haunted, M = 44.19, SD = 17.48), F(1, 176) =
0.27, p > .05, ηp2 = .002. A signicant main effect was found for level of paranormal belief, F(1, 176) =
41.68, p < .001, ηp2 = .19. Participants scoring above the median on paranormal belief scored higher on the
EPPS (M = 51.88, SD = 15.14) than participants below the median (M = 36.12, SD = 17.35). Higher EPPS
scores indicate a higher expectancy of haunt-related phenomena within the Administrative Building. No
signicant interaction was found between suggestion and level of paranormal belief on EPPS, F(1, 176) =
.01, p > .05, ηp2 = .00.
History of Haunting (HH). No signicant main effect on HH was found for suggestion (structure,
M = 4.29, SD = 1.87 vs. haunted, M = 3.95, SD = 1.81), F(1, 176) = 1.56, p > .05, ηp2 = .009. A signicant
main effect was found for level of paranormal belief, F(1, 176) = 45.53, p < .001, ηp2 = .21. Participants
scoring above the median on paranormal belief scored higher on HH (M = 4.98, SD = 1.36) than participants
below the median (M = 3.31, SD = 1.88). Higher HH scores indicate a higher expectancy of haunt-related
phenomena within the Administrative Building. No signicant interaction was found between suggestion
and level of paranormal belief on HH, F(1, 176) = 0.25, p > .05, ηp2 = .001.
Proneness to Reality Testing Decits
Means and standard deviations appear in Table 4.
Environmental Perceptions Phenomena Scale (EPPS). No signicant main effect on EPPS was
found for suggestion (structure, M = 43.28, SD = 18.78 vs. haunted, M = 44.19, SD = 17.48), F(1, 176) =
0.12, p > .05, ηp2 = .001. A signicant main effect was found for IPO-RT, F(1, 176) = 17.92, p < .001, ηp2
= .092. Participants scoring above the median on IPO-RT scored higher on the EPPS (M = 49.28, SD =
16.98) than participants below the median (M = 38.32, SD = 17.57). Higher EPPS scores indicate a higher
The Journal of Parapsychology
96
expectancy of haunt-related phenomena within the Administrative Building. No signicant interaction was
found between suggestion and IPO-RT on EPPS, F(1,176) = .05, p > .05, ηp2 = .00.
Table 4
Level of Proneness to Reality Testing Decits (IPO-RT) as a Function of Haunt-Related Attitudes
(Environmental Perceptions Phenomena Scale, EPPS, and History of Haunting, HH)
Level of RT
Below Median Above Median Overall
M SD M SD M SD
EPPS
Suggestion Type
Structural 37.57 18.54 49.12 17.34 43.28 18.78
Haunted 39.05 16.73 49.43 16.82 44.19 7.48
Overall 38.32 17.57 49.28 16.98
HH
Suggestion Type
Structural 3.80 2.04 4.80 1.53 4.29 1.87
Haunted 3.74 1.88 4.16 1.73 3.95 1.81
Overall 3.77 1.95 4.47 1.66
History of Haunting (HH). No signicant main effect on HH was found for suggestion (structure,
M = 4.29, SD = 1.87 vs. haunted, M = 3.95, SD = 1.81), F(1, 176) = 1.69, p > .05, ηp2 = .010. A signicant
main effect was found for IPO-RT, F(1, 176) = 6.86, p = .01, ηp2 = .038. Participants scoring above the
median on IPO-RT scored higher on the HH (M = 4.47, SD = 1.66) than participants below the median (M
= 3.77, SD = 1.95). Higher HH scores indicate a higher expectancy of haunt-related phenomena within the
Administrative Building. No signicant interaction was found between suggestion and IPO-RT on HH, F(1,
176) = 1.15, p > .05, ηp2 = .007.
Haunt Associated Opinions and Feelings (HAOF)
In light of the lack of a suggestion effect and for brevity’s sake, subsequent analysis focused on only
belief in the paranormal and proneness to reality testing decits. A series of independent t tests assessed
differences on the opinion and feeling measures (HAOF), as shown in Table 5.
Participants high (vs. low) on R-PBS scored higher on the opinion (ghosts exist) and feeling mea-
sures (anxious and fearful). Participants high (vs. low) on IPO-RT scored higher on opinion (ghosts exist).
Following application of a Bonferroni correction for multiple comparisons, the feeling measures (anxious
and fearful) were not signicant. The adjusted alpha level was based upon the notion of family: the smallest
set of meaningful inferences within a set of analyses. There were three planned comparisons within each
independent variable assessing levels of paranormal belief and proneness to reality testing decits; hence,
the corrected alpha was p = .017.
Path Analysis: Paranormal Belief, Reality Testing, and Haunt-Related Phenomena
Within these data, there were no issues with multicollinearity; all correlations were below .9 (see
Table 2 for zero-order correlations). All correlations were signicant and in the expected direction. A path
model examining direct effects (R-PBS on EPPS) and indirect effects (R-PBS on EPPS, mediated by IPO-
RT and HH was constructed.
Inuences on Perception of an Allegedly Haunted Building 97
Table 5
Paranormal Belief and Reality Testing Comparisons (Low vs. High)
on Haunt-Associated Opinions and Feelings (HAOF)
Level
Below Median Above Median
M SD M SD t df p d
Paranormal Belief
Opinion
Ghosts Exist 3.32 2.00 5.09 1.65 -6.49 175.50 <.001 .97
Feelings
Anxious 3.17 1.96 4.08 1.64 -3.38 175.80 <.001 .50
Fearful 2.96 1.94 3.90 1.79 -3.39 177.96 <.001 .51
Proneness to Reality Testing Decits
Opinion
Ghosts Exist 3.81 2.05 4.55 1.96 -2.46 178 .008 .37
Feelings
Anxious 3.34 1.90 3.89 1.79 -1.99 178 .025 .30
Fearful 3.13 1.98 3.70 1.82 -1.99 178 .024 .30
p <. 05*, p < .01**, p < .001***
Figure 1. The outcome of AMOS 19 path analysis, putative relations between paranormal beliefs subscales
(NAP and TPB), level of paranormal belief (R-PBS), proneness to reality testing decits (IPO-RT), belief
the building was haunted (HH) and haunt-related expectations (EPPS). The effects of extraneous variables
are depicted by the use of “e” to denote error term.
Standardized estimates, covariance, and regression coefcients (R2) appear in Figure 1. Kline
(2005) recommends several criteria for assessing goodness-of-t: chi-square, the comparative t index
The Journal of Parapsychology
98
(CFI), standardized root mean square residual (SRMR), and root mean square error of approximation (RM-
SEA). These indices indicated very good overall model t; chi-square is nonsignicant, χ2(7, N = 180) =
7.79, p = .35; CFI = .99; SRMR = .02; and RMSEA = .02. Hu and Bentler (1999) examined indices cutoffs
and suggested that type I and type II errors were best minimised via a combination of relative t indexes
(CFI ≥ 0.95 is indicative of good t) and absolute t indices (SRMR, good models < .08; or RMSEA, good
models < .06). Path coefcients were signicant at the p < .05 level. Paranormal belief (R-PBS) had a sig-
nicant effect on proneness to reality testing decits (IPO-RT), the belief the building was haunted (HH),
and haunt-related expectations (EPPS). IPO-RT and HH also signicantly affected EPPS.
Direct and Indirect Effects
To assess whether direct and indirect effects were statistically signicant, a mediation analysis us-
ing the bias-corrected bootstrap 95% condence intervals (CI) procedure (Hayes, 2013) was applied with
5,000 bootstrap samples (ndings are reported in Table 6). This analysis further examined the specic inu-
ence of each proposed mediator; AMOS is unable to examine the unique inuence of two or more mediators
when they are simultaneously included in a path diagram.
Table 6
Mediation Analysis Direct and Indirect Effects
Standardized Bias-Corrected Standardized Bias-Corrected
Direct Effect 95% CI Indirect Effect 95% CI
(DE) ( DE) (IE) (IE)
Haunt-Related Expectations (EPPS)
Paranormal Belief (R-PBS) .18 .48–.66* .37* .28–.47*
Belief Building Haunted (HH) .50 .19–.46* None None
Reality Test Decits (IPO-RT) .24 .07–.31* None None
* p < .001
Signicant direct effects were observed on EPPS, R-PBS, HH, and IPO-RT. Paranormal belief
(R-PBS) had a signicant indirect effect on haunt-related expectations (EPPS). To discern the inuence of
each proposed mediator (HH and RT) on the relationship between R-PBS and EPPS, Preacher and Hayes’
(2008) INDIRECT bootstrapping macro was run. The direct relationship between R-PBS and EPPS was
nonsignicant (b = .12), indicating that belief the building was haunted (HH) and reality testing decit
(IPO-RT) mediated the relationship between R-PBS and EPPS. Further inspection revealed that IPO-RT
was not a signicant mediator between R-PBS and EPPS at the 99% condence level across bias-correct-
ed (99% CI = -.01 to .13) point estimates. HH, however, emerged as signicant, and the indirect effect of
R-PBS and EPPS through HH was signicant at the 99% condence level across bias corrected (99% CI =
.05 to .34) point estimates. The overall model accounts for 53% of the total variance on EPPS.
Discussion
Prior to replication, signicant effects within the present study require cautious interpretation. The
present paper produced a number of key ndings. Within the experimental phase, expectation of haunt-re-
lated phenomena as measured by the Environmental Perceptions Phenomena Scale and belief the building
was haunted increased as a function of level of paranormal belief and proneness to reality testing decits.
Suggestion, however, had no effect on expectation of haunt-related phenomena. Phase 2 conrmed and
expanded upon these ndings. Path analysis revealed signicant relationships between belief in the paranor-
Inuences on Perception of an Allegedly Haunted Building 99
mal, proneness to reality testing decits, belief the building was haunted, and expectation of haunt-related
phenomena. Further analysis found belief the building was haunted signicantly mediated the relationship
between paranormal belief and haunt-related expectations. Finally, whilst correlating positively with each
other, both belief in the paranormal and proneness to reality testing decits (information processing style)
explained unique variance within haunt-related expectations. The contribution of reality testing decits,
however, in comparison to belief was relatively minor.
The observed association between level of paranormal belief and haunt-related expectations con-
curs with Wiseman et al. (2002), who found a relationship between belief in ghosts, reporting paranormal
experiences, and attributing unusual experiences to a ghost. In the present study, believing in the paranormal
predisposed participants to anticipate haunt-related phenomena within the ctitious Administrative Build-
ing, especially when participants considered the location haunted. This nding is consistent with previous
work on context mediation and psychological haunting research generally (Harte, 2000; Houran & Lange,
1996, 2001; Lange & Houran, 1997). Expectation in the context of real-world locations (disused structures,
historical buildings, battleelds, etc.) may arise from physical characteristics (appearance, age, structure,
etc.) and social characteristics (e.g., history and reputation). Similarly, proneness to reality testing decits,
and an associated preference for intuitive-experiential information processing, predisposed participants to
expect more haunt-related phenomena within the ctitious location. This nding adds to the emergent body
of research revealing associations between proneness to reality testing decits and unconventional beliefs
(Drinkwater et al., 2012; Irwin, 2004).
The observation that suggestion had no inuence in the current study requires careful elucidation.
Previous research has produced mixed results. Generally, research indicates that suggestion affects the per-
ception and recall of paranormal phenomena (French & Wilson, 2006; Wiseman & Greening, 2005; Wise-
man et al., 2002), especially in believers, when the suggestion is consistent with a particular belief. Haunting
ndings, however, are less conclusive. For example, Wiseman et al. (2002) found mixed results. Suggestion
had no effect on perception of unusual phenomena in the current study, nor did it inuence the degree to
which participants attributed unusual events to a ghost. The only prediction to reach signicance was that
believers (vs. disbelievers) in the positive suggestion condition reported more experiences.
Although the present manipulation failed to produce signicant effects, suggestion may still play
a role in priming haunt-related experiences. Indeed, previous work has evidenced that suggestion can in-
uence haunt-related perceptions (cf. Houran & Lange, 1996). Clearly, further work is required to identify
factors effecting susceptibility to haunting experiences. One obvious potential manipulation is cue salience.
The present study used a virtual tour, whilst the signicant effect found by Houran and Lange (1996) oc-
curred within a “real” physical setting. Note that Wiseman et al. (2002) failed to nd compelling effects in a
real-life setting (Hampton Court).
Collectively, haunting research indicates that the effects of suggestion are subtle and less important
than belief in the paranormal. More generally, suggestion may inuence perception and recall of paranormal
events such as spoon bending, fake séances, and psychic readings. Thus, future studies need to explore the
conditions under which suggestion inuences haunt-type perceptions and recollections. The main factors to
consider are whether participants high in haunting belief are more inclined to be affected by haunt-related
suggestions (specic beliefs), environmental manipulation (allegedly haunted vs. manufactured setting),
and past experience (haunting vs. no haunting). This would produce a set of complex variable manipulations
sensitive to potential suggestion effects. Finally, believers gave higher ratings for believing the building was
haunted and ghost existence. In addition, believers produced higher anxiety and fear ratings.
This study has some potential limitations. For example, participants completed general question-
naire measures in a single test session. Whilst the order of measures was counterbalanced, previous stud-
ies found that personality measures administered in this way may be prone to context effects (Council,
1993). Context effects occur when correlations between measures vary as a function of whether scales
were completed in a single testing session (resulting in a high correlation) or in separate/unrelated sessions
(resulting in a lower or no correlation). Context effects arise within single session testing situations because
participants wrongly infer relationships between scales and become inadvertently motivated to provide
consistent responses (Council, Kirsch, & Hafner, 1986). For example, Council (1993) found the typically
The Journal of Parapsychology
100
cited positive correlation between hypnotisability and absorption only when participants completed both
measures in a single testing session. Whilst context effects are a consideration, likelihood of occurrence
differs according to the characteristics of specic personality measures and research situations. Pertinent
to the present study, Roig, Bridges, Renner, and Jackson (1998), using the Paranormal Belief Scale and the
Irrational Beliefs Inventory, found that whilst correlation coefcients between the measures became weaker
as a function of context controls, there were no statistically signicant effects of context. Similarly, Nadon,
Hoyt, Register, and Kihlstrom (1991) failed to nd evidence that the repeatedly observed correlations be-
tween absorption and hypnosis measures were artifacts of testing context.
Subsequent studies could extend the current study by including a suggestibility measure. Previous
research has found associations between the reporting of anomalous experiences and belief in paranormal
phenomena (Kumar & Pekala, 2001). Thus, individuals scoring high on suggestibility may be more likely to
respond to the verbal suggestion and report higher expectations, especially when the suggestion is consistent
with their beliefs. Suggestion may be less likely to affect participants with low suggestibility regardless of
whether the suggestion is consistent with their belief or not (cf. McGeown et al., 2012).
Further evaluation of haunt-related expectations could examine the distinction between proneness
to anomalous experiences and tendency to draw upon paranormal interpretations. The present study in-
cluded a range of phenomena associated with, but not exclusive to, haunt-related experiences. Participants
imagining entering an unfamiliar environment may anticipate strange sensations and perceptions but not
necessarily attribute them to ghosts. Further work could consider the relative contribution of attentional and
attributional processes and investigate whether believers and nonbelievers differ on both expectations for
unusual and haunt-related phenomena, using, for example, the Survey of Anomalous Experiences (Irwin,
Dagnall, & Drinkwater, 2013).
Analysis of direct and indirect effects demonstrated that the belief the building was haunted me-
diated the relationship between paranormal belief (R-PBS) and expectations of haunt-related phenomena.
The emergent haunting model contributes to the work of Houran and Lange (Houran & Lange, 1996; Lange
& Houran, 1997). They found that context cues and demand characteristics could induce poltergeist-like
perceptions and stimulate paranormal-type experiences. The present study explicates that this is most likely
to occur when belief in the paranormal is high and the location considered haunted, especially when the
individual believes hauntings are genuine. This nding is an important addition to the literature because, as
Laythe and Owen (2012) note, the generally utilised measure of paranormal belief, the R-PBS (Tobacyk,
2004), fails specically to assess specic haunt-related content and hence underestimates the relationship
between belief and perceived ghostly activity. In the present study we used the R-PBS for two reasons. First-
ly, the measure is conceptually and psychometrically satisfactory (Tobacyk, 2004). Secondly, the R-PBS is
the most widely used instrument of paranormal belief (Goulding & Parker, 2001). Our ndings, however,
suggest that specic haunt-related beliefs rather than general belief in the paranormal increases haunt-relat-
ed expectations. Thus, specic beliefs may provide a cognitive framework for structuring haunting-related
events and experiences (Houran et al., 2002). This is consistent with Wiseman et al.’s (2002) nding that
those who believed in ghosts (vs. nonbelievers) reported more unusual experiences as they walked around
Hampton Court Palace. Thus, believers in hauntings, via their beliefs, are inclined to expect haunt-related
phenomena in purportedly haunted locations.
As a brief postscript, we would like to contextualize our conclusions. Whilst our results indicate that
beliefs may give rise to haunting-like perceptions, they do not imply that all haunting/ghost experiences are
explainable via psychological mechanisms. As Broughton (1991) eloquently notes, conventional explana-
tions of hauntings and ghost-related phenomena have yet to provide comprehensive theories that account
for the breadth of encounters reported. The present study adds to the body of research demonstrating that
psychological processes can explain haunting-like phenomena.
References
Andrich, D. (1988b). Rasch models for measurement. Newbury Park, CA: Sage.
Barber, T. X., & Calverey, D. S. (1964). An experimental study of hypnotic (auditory and visual) hallucinations. Journal
Inuences on Perception of an Allegedly Haunted Building 101
of Abnormal and Social Psychology, 68, 13–20.
Broughton, R. S. (1991). Parapsychology: The controversial science. New York, NY: Ballantine.
Cohen, J. (1992). A power primer. Psychological Bulletin, 112, 155–159.
Council, J. R. (1993). Context effects in personality research. Current Directions in Psychological Science, 2, 31–34.
Council, J. R., Kirsch, I., & Hafner, L. P. (1986). Expectancy versus absorption in the prediction of hypnotic respond-
ing. Journal of Personality and Social Psychology, 50, 182–189.
Danoiu, I. (1995). Self-attribution, expectancy, suggestibility. Revue Roumaine de Psychologie, 39, 139–145.
Drinkwater, K., Dagnall, N., & Parker, A. (2012). Reality testing, conspiracy theories, and paranormal beliefs. Journal
of Parapsychology, 76, 57–78.
French, C. C., & Wilson, K. (2006). Incredible memories: How accurate are reports of anomalous events? European
Journal of Parapsychology, 21, 166–181.
George, D., & Mallery, P. (2003). SPSS for Windows step by step: A simple guide and reference. 11.0 update (4th ed.).
Boston, MA: Allyn & Bacon.
Goulding, A., & Parker, A. (2001). Finding psi in the paranormal: Psychometric measures used in research on para-
normal beliefs/experiences and in research on psi-ability. European Journal of Parapsychology, 16, 73–101.
Harte, T. M. (2000). Contextual mediation of perceptions during hauntings and poltergeist-like experiences: A repli-
cation and extension. Perceptual and Motor Skills, 91, 451–459.
Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-based
approach. New York, NY: Guilford Press.
Hergovich, A., & Arendasy, M. (2005). Critical thinking ability and belief in the paranormal. Personality and Individ-
ual Differences, 38, 1805–1812.
Hinkle, D. E., Wiersma, W., & Jurs. S. G. (2003). Applied statistics for the behavioral sciences (5th ed.). Boston, MA:
Houghton Mifin.
Houran, J., Hughes, L. F., Thalbourne, M. A., & Delin, P. S. (2006). Quasi-experimental study of transliminality, vi-
brotactile thresholds and performance speed. Australian Journal of Parapsychology, 6, 54–80.
Houran, J., & Lange, R. (1996). Diary of events in a thoroughly unhaunted house. Perceptual and Motor Skills, 83,
499–502.
Houran, J., & Lange, R. (Eds.). (2001). Hauntings and poltergeists: Multidisciplinary perspectives. Jefferson, NC:
McFarland.
Houran, J., Wiseman, R., & Thalbourne, M. (2002). Perceptual-personality characteristics associated with naturalistic
haunt experiences. European Journal of Parapsychology, 17, 17–44.
Hu, L.T., & Bentler, P. M. (1999). Cutoff criteria for t indexes in covariance structure analysis: Conventional criteria
versus new alternatives. Structural Equation Modeling, 6, 1–55.
Iacobucci, D., Posavac, S. S., Kardes, F. R., Schneider, M. J., & Popovich, D. L. (2014). Toward a more nuanced
understanding of the statistical properties of a median split. Journal of Consumer Psychology. Retrieved from
http://dx.doi.org/10.1016/j.jcps.2014.12.002
Irwin, H. J. (2003). Reality testing and the formation of paranormal beliefs. European Journal of Parapsychology,
18, 15–27.
Irwin, H. J. (2004). Reality testing and the formation of paranormal beliefs: A constructive replication. Journal of the
Society for Psychical Research, 68, 143–152.
Irwin, H. J. (2009). The psychology of paranormal belief: A researcher’s handbook. Hateld, England: University of
Hertfordshire Press. Amended.
Irwin, H. J., Dagnall, N., & Drinkwater, K. (2012). Paranormal belief and biases in reasoning underlying the formation
of delusions. Australian Journal of Parapsychology, 12, 7–21.
Irwin, H. J., Dagnall, N., & Drinkwater, K. (2013). Parapsychological experience as anomalous experience plus para-
normal attribution: A questionnaire based on a new approach to measurement. Journal of Parapsychology, 77,
39–53.
Kernberg, O. F. (1996). A psychoanalytic theory of personality disorders. In J. F. Clarkin & M. F. Lenzenweger (Eds.),
Major theories of personality disorder (pp. 106–140). New York, NY: Guilford Press.
Kline, R. B. (2005). Principles and practice of structural equation modeling (2nd ed.). New York, NY: Guilford Press.
Kumar, V. K., & Pekala, R. J. (2001). Relation of hypnosis-specic attitudes and behaviours to paranormal beliefs and
experiences: A technical review. In J. Houran & R. Lange (Eds.), Hauntings and poltergeists: Multidisciplinary
perspectives (pp. 260–279). Jefferson, NC: McFarland.
Lange, R., & Houran, J. (1997). Context-induced paranormal experiences: Support for Houran and Lange’s model of
haunting phenomena. Perceptual and Motor Skills, 84, 1455–1458.
The Journal of Parapsychology
102
Lange, R., Irwin, H. J., & Houran, J. (2000). Top-down purication of Tobacyk’s Revised Paranormal Belief Scale.
Personality and Individual Differences, 29, 131–156.
Laythe, B., & Owen, K. (2012). Paranormal belief and the strange case of haunt experiences: Evidence of a neglected
population. Journal of Parapsychology, 16, 79–108.
Lenzenweger, M. F., Clarkin, J. F., Kernberg, O. F., & Foelsch, P. A. (2001). The Inventory of Personality Organiza-
tion: Psychometric properties, factorial composition, and criterion relations with affect, aggressive dyscontrol,
psychosis proneness, and self-domains in a nonclinical sample. Psychological Assessment, 13, 577–591.
Lindeman, M. (1998). Motivation, cognition and pseudoscience. Scandinavian Journal of Psychology, 39, 257–265.
MacCallum, R. C., Zhang, S., Preacher, K. J., & Rucker, D. D. (2002). On the practice of dichotomization of quanti-
tative variables. Psychological Methods, 7, 19–40.
McConkey, K. M., & Barnier, A. J. (2004). The highly hypnotisable person: Unity and diversity in behaviour and ex-
perience. In M. Heap, R. Brown, & D. Oakley (Eds.), The highly hypnotisable person: Theoretical, experimental
and clinical issues (pp. 61–84). London, England: Brunner-Routledge.
McGeown, W. J., Venneri, A., Kirsch, I., Nocetti, L., Roberts, K., Foan, L., & Mazzoni, G. (2012). Suggested visual
hallucination without hypnosis enhances activity in visual areas of the brain. Consciousness and Cognition, 21,
100–116.
Merckelbach, H., & van de Ven, V. (2001). Another White Christmas: Fantasy proneness and reports of “hallucina-
tory experiences” in undergraduate students. Journal of Behavior Therapy and Experimental Psychiatry, 32,
137–144.
Mintz, S., & Alpert, M. (1972). Imagery vividness, reality testing and schizophrenic hallucinations. Journal of Abnor-
mal Psychology, 79, 310–316.
Moore, D. W. (2005, June 16). Three in four Americans believe in paranormal. Gallup News Service, p.161. Retrieved
from www.gallup.com
Moore, T., Thalbourne, M. A., & Storm, L. (2010). A study on coincidences. Australian Journal of Parapsychology,
10, 154–176.
MORI. (1998). Paranormal survey [Conducted for the Sun newspaper]. Retrieved from http://www.ipsosmori.com/
researchpublications/researcharchive/poll.aspx?oItemId=2082
MORI. (2007). Survey on beliefs [Conducted for Schott’s Almanac]. Retrieved from http://www.ipsos-mori.com/
researchpublications/researcharchive/246/Survey-on-Beliefs.aspx
Nadon, R., Hoyt, I. P., Register, P. A., & Kihlstrom, J. F. (1991). Absorption and hypnotizability: Context effects reex-
amined. Journal of Personality and Social Psychology, 60, 144–153.
Newport, F., & Strausberg, M. (2001, June 8). Americans’ belief in psychic and paranormal phenomena is up over last
decade. Gallup News Service. Retrieved from http://www.gallup.com/poll/releases.
O’Connor, A. R., Barnier, A. J., & Cox, R. E. (2008). Déjà vu in the laboratory: A behavioral and experiential com-
parison of posthypnotic amnesia and posthypnotic familiarity. International Journal of Clinical and Experimen-
tal Hypnosis, 56, 425–450.
O’Keefe, C., & Parsons, S. (2009). Haunting experiences: An investigative approach. In M. D. Smith (Ed.), Anoma-
lous experiences: Essays from parapsychological and psychological perspectives (pp.108–119). Jefferson, NC:
McFarland.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect
effects in multiple mediator models. Behavior Research Methods, 40, 879–891.
Roig, M., Bridges, K. R., Renner, C. H., & Jackson, C. R. (1998). Belief in the paranormal and its association with
irrational thinking controlled for context effects. Personality and Individual Differences, 24, 229–236.
Sappington, A. A. (1990). The independent manipulation of intellectually and emotionally based beliefs. Journal of
Research in Personality, 24, 487–509.
Terhune, D. B., & Smith, M. D. (2006). The induction of anomalous experiences in a mirror-gazing facility: Sugges-
tion, cognitive perceptual personality traits, and phenomenological state effects. Journal of Nervous and Mental
Disease, 194, 415–421.
Terhune, D. B., Ventola, A., & Houran, J. (2007). An analysis of contextual variables and the incidence of photograph-
ic anomalies at an alleged haunt and a control site. Journal of Scientic Exploration, 21, 99–120.
Thalbourne, M. A. (2010). Transliminality: A fundamental mechanism in psychology and parapsychology. Australian
Journal of Parapsychology, 10, 70–81.
Tobacyk, J. J. (2004). A revised paranormal belief scale. International Journal of Transpersonal Studies, 23, 94–99.
Tobacyk, J. J., & Milford, G. (1983). Belief in paranormal phenomena: Assessment instrument development and im-
plications for personality functioning. Journal of Personality and Social Psychology, 44, 1029–1037.
Inuences on Perception of an Allegedly Haunted Building 103
Williams, B., Ventola, A., & Wilson, M. (2010). Apparitional experiences: A primer on parapsychological research
and perspectives. Retrieved from http://www.publicparapsychology.org/Public%20Parapsych/Apparitional%20
Experiences%20Primer%20Final.pdf.
Wilson, K., & French, C. C. (2008). Misinformation effects for psychic readings and belief in the paranormal. Imag-
ination, Cognition and Personality, 28, 155–171.
Wilson, K., & French, C. C. (2014). Magic and memory: Using conjuring to explore the effects of suggestion, social
inuence, and paranormal belief on eyewitness testimony for an ostensibly paranormal event. Frontiers in Psy-
chology, 5:1289, 1–9.
Wiseman, R., & Greening, E. (2005). “It’s still bending”: Verbal suggestion and alleged psychokinetic metal bending.
British Journal of Psychology, 96, 115–127.
Wiseman, R., Greening, E., & Smith, M. (2003). Belief in the paranormal and suggestion in the séance room. British
Journal of Psychology, 94, 285–297.
Wiseman, R., Watt, C., Greening, E., Stevens, P., & O’Keefe, C. (2002). An investigation into the alleged haunting of
Hampton Court Palace. Psychological variables and magnetic elds. Journal of Parapsychology, 66, 387–408.
Acknowledgements
The authors would like to thank the Bial Foundation (Fellowship Programme 39/10) for nancial
support and anonymous referees for providing helpful suggestions.
Manchester Metropolitan University
Department of Psychology
Faculty of Health, Psychology and Social Care
Birley Building, 53 Bonsall Street
Manchester, UK M15 6GX
n.dagnall@mmu.ac.uk
Abstracts in Other Languages
French
SUGGESTION, CROYANCE AU PARANORMAL, DEFICITS DANS L’INCLINATION
A TESTER LA REALITE ET PERCEPTIONS DANS UN IMMEUBLE REPUTÉ HANTÉ
RÉSUMÉ : La présente étude cherche à savoir si la suggestion, le niveau de croyance au paranormal et une
inclination à tester la réalité de manière défectueuse inuencent les attentes des participants dans le cas de
phénomènes relatifs à la hantise. On présente aux participants un court diaporama sur l’histoire ctive d’un
hôpital abandonné. La suggestion est faite lors de la dernière phrase de la présentation et consiste à afrmer
que l’immeuble administratif de l’hôpital a soit connu des activités de hantise, soit présenté des problèmes
structuraux. Suite au diaporama, pour s’assurer que les participants ont suivi la suggestion, ils lisent une
transcription de la présentation. L’expérimentateur informe ensuite les participants qu’ils vont regarder
l’intérieur du bâtiment administratif grâce à une vidéo touristique sans son et en noir et blanc. A la n de
la séquence vidéo, les participants remplissaient des échelles évaluant les perceptions et les phénomènes
dans l’environnement, les opinions et sensations relatives à la hantise, la croyance au paranormal (Echelle
révisée de croyance au paranormal), et l’inclination à tester la réalité de manière défectueuse (Inventaire
d’organisation de la personnalité). Lors de la phase expérimentale, seuls les niveaux de croyance au para-
normal et l’inclination à tester la réalité de manière défectueuse ont affecté les mesures relatives à la hantise
; la suggestion n’a pas eu d’effet. La seconde phase d’enquête, employant l’analyse des causes, a révélé que
l’histoire de hantise (le degré de conviction des participants dans le fait que l’immeuble administratif avait
véritablement été hanté) médiatisait la relation entre la croyance au paranormal et l’attente des phénomènes
relatifs à la hantise.
The Journal of Parapsychology
104
German
SUGGESTION, GLAUBE ANS PARANORMALE, NEIGUNG ZU DEFIZITEN IN DER REAL-
ITÄTSÜBERPRÜFUNG UND WAHRNEHMUNG EINES ANGEBLICHEN SPUKGEBÄUDES
Die vorliegende Studie untersuchte, ob eine gegebene Suggestion, die Stärke des Glaubens ans Paranor-
male und die Neigung zu Deziten in der Realitätsüberprüfung die Erwartung von Teilnehmern in Bezug
auf spukähnliche Phänomene beeinussten. Die Teilnehmer wurde eine kurze Diashow präsentiert, in der
die Geschichte eines ktiven, verlassenen Krankenhauses vorgestellt wurde. Die Suggestion wurde im letz-
ten Satz der Erzählung gegeben, in dem behauptet wurde, dass das Verwaltungsgebäude des Krankenhauses
entweder eine Vergangenheit mit Geistererscheinungen oder bauliche Mängel aufwiese. Um sicher zu ge-
hen, dass die Teilnehmer die Suggestion mitbekommen hatten, lasen sie nach der Diashow ein Transkript
der vorgestellten Geschichte. Anschliessend informierte der Versuchsleiter die Teilnehmer, dass sie das In-
nere des Verwaltungsgebäudes mittels einer tonlosen Videotour in Schwarzweiss zu sehen bekämen. Nach
Abschluss der Filmsequenz füllten die Teilnehmer Fragebögen zur Erfassung ihrer jeweiligen Umgebung-
swahrnehmungen und Phänomenen, ihrer Meinungen und Gefühle in Bezug auf Spukhaftes, ihren Glauben
ans Paranormale (Revised Paranormal Belief Scale) und ihrer Neigung zu Deziten in der Realitätsüber-
prüfung (Inventory of Personality Organization) aus. Innerhalb der Experimentalphase beeinussten nur
die Stärke des Glaubens ans Paranormale und die Neigung zu Deziten in der Realitätsüberprüfung die auf
Spuk bezogenen Einschätzungen; die Suggestion selbst blieb wirkungslos. Eine Auswertung in der zweit-
en Phase mittels einer Pfadanalyse ergab, dass die Spukgeschichte (das Ausmaß, zu dem die Teilnehmer
glaubten, dass das Verwaltungsgebäude im Ruf stünde, dass es in ihm spuke) die Beziehung zwischen dem
Glauben ans Paranormale und der Erwartung von auf Spuk bezogenen Phänomenen vermittelte.
Spanish
SUGESTIÓN, CREENCIA EN LO PARANORMAL, PROPENSIÓN A DÉFICITS EN LA PRUEBA
DE REALIDAD Y PERCEPCIÓN DE UN EDIFICIO SUPUESTAMENTE ENCANTADO
RESUMEN: Este estudio investigó si la sugestión, el nivel de creencia en lo paranormal, y la propensión a
décits en la prueba de realidad inuyeron en las expectativas de participantes en fenómenos relacionados
con casas encantadas. Los participantes vieron una breve presentación de diapositivas que resumía la his-
toria cticia de un hospital abandonado. La sugestión ocurrió en la última frase de la narración en donde se
mencionó que el edicio administrativo del hospital tenía una historia de actividad fantasmal o problemas
estructurales. Después de la presentación de diapositivas, para asegurar que los participantes habían presta-
do atención a la sugestión, leyeron una transcripción de la presentación. El experimentador le dijo entonces
a los participantes que iban a ver las características internas del edicio administrativo a través de una visita
guiada en video en blanco y negro y sin sonido. Al concluir el video los participantes contestaron cues-
tionarios para medir percepciones y fenómenos ambientales, opiniones y sentimientos relacionados con las
casas encantadas, creencia en lo paranormal (Revised Paranormal Belief Scale), y propensión a décits en
la prueba de realidad (Inventory of Personality Organization). Dentro de la fase experimental, sólo el nivel
de creencia paranormal y la propensión al décit de prueba de realidad afectaron las puntuaciones relacio-
nadas a la casa encantada; la sugestión no tuvo ningún efecto. Una segunda fase de evaluación, utilizando
análisis de rutas, reveló que la historia sobre la casa encantada (el grado en que los participantes creyeron
que el edicio administrativo tenía una historia de ser casa encantada) medió la relación entre la creencia
paranormal y la expectativa de los fenómenos relacionados con el encantamiento.
