Basic HTML Version
RELATIONSHIP BETWEEN HGSHS AND SHSS:C
339
high vs. medium groups, and medium vs. low groups, both differ sig-
nificantly
( p
<
.005;
p
<
.05,
respectively).
The correlation between the two clusters of items (containing only
two items each) is .43. The passive suggestion items in HGSHS:A and
SHSS:C correlate .36 while the challenge clusters correlate
.65.
The hallucination items do not discriminate well between subgroups.
Both Fly and Taste Hallucinations fail to discriminate between me-
dium and high groups. The number of
8s
passing the
Box
and Voice
Hallucination is too small to make meaningful comparisons.
If
these
effects are functions of item difficulties, i t would be expected that
only
some high
Ss
would pass these items. This is not
so."
Discussion
Relationship between HGSHS :A and
SHSS
:
C
The adequacy with which HGSHS :A performance predicts
SHSS
:
C
scores depends
on
the degree of accuracy required, and the range of
scores being considered. However, the rather disappointing predictive
relationship between the two scales does
not
imply the psychometric
inadequacy of one of them.
As instruments to be used conjointly to select subgroups for research
in hypnosis, the scales have some limitations. None of the
20
Ss
scoring
4
or
less
on
HGSHS:A scored higher than 6 on SHSS:C. Within this
range the predictive value of HGSHS:A seems satisfactory.
In
con-
trast, only
9
of the
20
Ss
classified as high on HGSHS:A, and 10 of the
20
Ss
classified as medium, were similarly classified on
SHSS:
C
scores.
The stability of low
Ss
occurs because the only items they tend to
pass are the passive suggestibility items. Although these items have
high reliability (Evans, 1966; Eysenck
&
Furneaux,
1945),
they do
not discriminate between levels of susceptibility. The low
Ss
fail the
challenge suggestibility items, which
do
consistently discriminate be-
tween levels of susceptibility in both scales.
aIt
is
also meaningful to divide Ss into subgroups according to
SHSS:C
scores
and conduct a
similar
analysis. This examines the discrimination power
of
each
item rather than the ability
of
items
to
discriminate when subgroup allocation
i s
determined independently.
As
in the analysis of
HGSHS:A
above, classifica-
tion based
on
SHSS:C
scores capitalizes
on
chance measurement error.
ClaSaitying
Ss
on
SHSS:C
using
the same cutting pointa,
15
Ss have high
scores,
25
have me-
dium
and
20
low.
Chi
square ditlerences between
ss
passing itema
in
each
mb-
group were: item
9:
Anosmia to Ammonia,
p
=
.129;
item
6:
Dream,
p
=
.085;
item
10:
Hallucinated Voice,
p
=
.030;
item
3:
Fly
Hallucination,
p
=
remaining eight items,
p
<
.001.
Downloaded by [University of Macedonia] at 04:28 30 March 2012