Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
1996
1996 The Guilford Press
PSYCHOPATHY, GENDER,
Hamburger, PhD,
Hogben, MA
that antisocial
proposed
personality
disor
der (ASPD) and histrionic personality disorder (HPD) are sextyped manifestations of the same underlying predisposition,
namely psychopathy.
males, 90 females),
pathy
tween
In
we
sample
of 180
tested three
undergraduates (90
hypotheses: (1) psycho
moderated
by biological gender,
to exhibit
with
psychopathic
males tend
individuals
by gender roles,
possessing stereotypically "mas
psychopathic
tending to exhibit an antisocial pattern and
individuals
possessing stereotypically "feminine"
psychopathic
features tending to exhibit a histrionic pattern. The latter hy
pothesis was tested both across and within gender. Structural
equation modeling applied to self-report measures of psycho
pathy, gender roles, and ASPD and HPD traits provided sup
port for the first two hypotheses but not the third. Implications
of these findings for the differential expression of psychopathy
culine'' features
in males and
are
females,
as
well
as
discussed.
(Grove &
41
42
Carson, 1993) to call into question the assumption that these disorders
represent independent entities. Second, a number of personality disorders
exhibit sex differences in prevalence (Ford & Widiger, 1989; Kass, Spitzer,
& Williams, 1983). Some authors (e.g., Kaplan, 1983; Tavris, 1992) have
conjectured that these sex differences are attributable to gender biases in
iors
ASPD and HPD also appear to exhibit sex differences in prevalence. ASPD
has consistently been found to be more common among males than females;
population prevalence
data indicate
rate of
approximately
3% among
males and
1994).
among females than males (Ford & Widiger, 1989; Kass et al., 1983; cf.
Reich, 1987), particularly in clinical samples (APA, 1994). Some dispute
remains, however, concerning the sex distribution of HPD in the population
at
large;
the
disorders be
HAMBURGER, LILIENFELD.
AND HOGBEN
43
in terms of
affect, and sexual provocativeness (APA, 1994). HPD has been reported
to
be
GENDER ROLES
1974).
Other researchers have focused
extreme
gender
role
stereotypes.
on
identifying
colleagues (e.g.,
gender
construct
Mosher &
reflecting
roles
(i.e.,"hypermasculinity").
behavioral style, enjoy power,
HYPOTHESES
Because
we
have utilized
operationalize
nonclinical
we
(i.e.,
44
ses.
METHOD
SUBJECTS
Data
were
collected from 96
men
Six males
were
were
the Deviant
validity scales
final sample thus
(M
MEASURES.
Subjects completed
described below.
battery
of
self-report
measures,
which
are
the
appears
many
45
These
constructed to
respectively.
Rosnick. 1990).
DSM-III MMPI
Personality
Disorder Scales
for
pool by means
targeted to assess
validity
have been
reported
more
the
empirical strategy.
Their items
promising pattern
of
with the standard MMPI clinical scales (Morey et al., 1985) and
to disciminate clinical diagnoses of ASPD and HPD, respectively,
criminant
In this
were
study,
content-pure
were
used to
provide
to assess
Inventory (BSRI). The BSRI (Bern, 1974) is designed
and
masculine
of
and
consists
roles,
feminine
and
masculine
traditional
gender
The
feminine trait descriptors selected on the basis of sex-typed social desirability.
both
on
BSRI yields a fourfold classication: masculine, feminine, androgynous (high
on both masculinity and
masculinity and femininity), and undifferentiated (low
1974) indicate that the masculine and
of the BSRI
femininity). Analyses
(Bern,
that
feminine subscales are essentially orthogonal. Bern (1981, 1984) reported
at
better
were
recalling sex-typed
masculine and feminine sex-typed individuals
in a test of recall, sex-typed
words than non-sex-typed individuals. Moreover,
-masculine" statements to
individuals were more likely to erroneously attribute
The BSRI is used to test
males and "feminine" statements to females (Bern, 1984).
are not large enough
sizes
Hypothesis 3 across both biological genders (the sample
each
within
classification
gender).
to permit a meaningful fourfold
The HMI (Mosher & Sirkin, 1984), which is
Hypermasculinity
designed
for male
Inventory (HMI).
respondents only,
is
forced-choice
questionnaire designed
to
46
within each
biological gender.
Hyperfemininity Scale (HFS). The HFS (Murnen & Byrne, 1991), which is designed
for female respondents only, is a forced-choice scale patterned after the HMI.
Murnen and Byrne (1991) found that HFS scores were correlated with less negative
perceptions of, and
more
numerous
coercion. Further
manner
and
was
desirability
influenced the relations among measures. The MCSD is highly correlated with other
self-report measures of social desirability (Paulhus, 1984). In addition, the MCSD
has been
reported
a measure
Byrne, 1991).
Analyses. Although we report both zero-order and partial (corrected for social
desirability) correlations among measures, our principal hypotheses were tested by
path analysis using Linear Structural Relations (LISRELJoreskog & Sorbom,
1993: see also Loehlin. 1992). LISREL is a program based on structural
that enables researchers to construct models from correlational data in
similar to
LISREL and
MRA
measurement
The
equations
a manner
are
specified
in
Hypotheses
a manner
categorical,
role classification
feasible.
according
involving
as
analogous
is the
case
moderator variables,
are
continuous,
to moderated MRA
with
an
(Stone, 1988).
biological gender
and
gender
to the
Four indices of
generally accepted
level for
.90
with
higher.
RESULTS
Descriptive Statistics
tions for the
thu
n
variables dimensionally,
in Table 1 although
nrlS PSSible l fcore f'"of these threer'e
"teratUre ^ B' 1984)* The lack f
multiple
S andZ yP!f th^mder P3thS '" the lambda matriCeS <t0 denne metric
for
as seen
SP?C'fy
he unHerlvin
>
?
' made U imPossible to construct the interaction terms
m
T g outlined
f
congruent with
the limitations
in Joreskog and Sorbom
th!
if
(1993)
HAMBURGER, LILIENFELD,
AND HOGBEN
Variable
Overall
Women
Men
PPI
378.83
369.57
388.09
(40.27)
(41.46)
(37.01)
ASPMMPI
HPMMPI
ASPPDQ
HPPD9
BSRIMAS
BSRIFEM
HMI
HFS
MCSD
Note. Overall N
11.36
10.22
12.50
(3.57)
(2.94)
(3.78)
14.66
1466
15.08
1508
14.23
,423
(3.60)
(3.37)
(3.80)
3.13
3.x
3
2.08
208
4.19
4
19
(3.34)
(2.49)
(3.74)
3.09
3.40
2.78
(2.07)
(2.09)
(2.02)
101.59
98.67
104.52
(13.04)
(13.55)
(11.89)
94.92
99.61
90.23
(11.44)
(11.75)
(8.97)
11.76
11.76
(5.90)
(5.90)
7.31
7.31
(4.30)
(4.30)
14.94
16.01
ie.01
13.87
(5.58)
(5.55)
(5.43)
Sample
and
.47
.07
.70
.10
24
.04
6g
.10
.30
.04
.46
.07
.91
.14
.39
.06
the sum of the feminine Items of the Bern Sex Role Inventory; HMI
inventory; BSRIFEM
Marlowe-Crowne Social Desirability
Hyperfemininity Scale; MCSD
Hypermascullnity Inventory; HFS
Sex Role
Scale.
on
are
the
q3-
ft
2
CO
IS
^h
-<
<<
>I
<
CO
"
<
-a -a
o
Ofl 4)
2" 12 2
0. a.
Su
slip
00
~*
~*
tf
CO
00
T?
*
CN
CN
"
1^-
O)
to
-I
Si
o a.
G l
S a
00
0)
3500
c
B"i
^^
6 E~
Xl
rt
SpoS
^ujll
5
00
-<
-h
CO
Tf
CO
-i
t*
cn
o
-
i?0
rt
C cu
in
CN
in
m
O)
CN
co
o
05
O
CO
O
tO
co
co
o>
co
co
-H
CD
'
in
in
-i
CO
CO
-h
7.
If s
oi a.
-i
CO
CD
CN
in
i-h
O)
in
CO
-i
I?
CN
Tf
CN
O)
M
CO
r-
Tf
CN
CD
t-
CO
O)
-h
in
111
CO
48
CD
y
SL.
c
rt
S 2 fc
litis
S g fc1?
i>
cu
<
cn
<u
O J3
lac
JS
13 -S
49
between the PPI and the indicators of ASPD and HPD traits. The correlation
between scores on the HPPDQ and scores on the PPI, however, was relatively
low. A
two
significant,
measures
of ASPD traits
of HPD traits
measures
on
cantly positively
Scores
on
correlated with
masculinity
scores were
the HPMMPI
or
HPPDQ,
with
negatively correlated
significant positive
correlation
HPMMPI. Scores
the HFS
on
by significant
scores
on
or
significantly
HPPDQ. Instead,
observed between
BSRIMAS and
significantly positively
correlated with
was
were
the BSRIMAS
the HPMMPI
desirability
as
assessed
by
the relations
on
(i.e., the
equal 1.0)
measures (PPITOT1
constrained to
items
excluded. There
PPITOT1
the
or
are
no
prima facie
reasons
for
choosing
either
construct; for
analyses reported
ASPPDQ
with DR scale
HPPDQ
for HPD traits because these subscales (unlike the ASPMMPI and HPMMPI)
disorders.
assess the DSM criteria for their respective
directly
1 model showed
significant departure
(X2(7)
fit
from
x2
hypothesized
size
97
(Hays, 1988).
and AGFI
pathy,
on
the
were
satisfactory:
path coefficients.
GFI
.96, CFI
coefficients between
one
2.UsingPPITOT2
statistics and
on
the other,
were
psycho
.74 and
50
/ Biological v,
\ Gender /
-
'
at p
<
.73 (both
significant
predictor
is a
strong
predicted by Hypothesis 1.
of the Hypothesis 1 model using
as
In addition,
<
.05, and
traits
were
traits
not
significantly
estimates
were
significant atp
generated from
zero-order correlations.
31.21,
the basic model form for each group, resulting in x (14)
statistic is significant
.96. Again, although the
p < .01; GFI = .95; CFI
at conventional levels, the other fit indices are satisfactory and suggest that
specifying
%2
samples
fit the
same
model form.
gender
is shown in
followed the
same
tors instead of
of the
Figure
across
males and
Substituting
yielded roughly
correlations
either BSRIMAS
similar
were
found.
or
BSRIFEM for
goodness-of-fit
Indices to those
HAMBURGER, LILIENFELD,
AND HOGBEN
51
ASPD traits is
chopathy
The
same
classification
according
to the
as
moderators.
The four
psychopathy and ASPD and HPD traits for either psychopathy0.64, NS). The two
2.76, NS) or psychopathy-HPD (x2(3)
(x [3)
categories of the HMI also had equal model form (x2(14) 28.92, p < .001;
between
ASPD
.92; GFI
.91). The HMI did not moderate the relation between
psychopathy and ASPD and HPD traits for either psychopathy- ASPD
0.30, ns). The two categories of
0.06, ns) or for psychopathy-HPD
.92; CFI
19.76, ns; GFI
the HFS again fit into equal model form
CFI
(x2(l)
(x2(l)
(x2(14)
=
and ASPD
traits
3 received
no
DISCUSSION
52
that two
our
our use
Second,
bility
our
of
an
exclusive reliance
that at least
of
on
self- report
measures
raises the
possi
such
as
some
our
direct interview,
file data,
interview
and observer
measures
of
ratings. In
personality
Psychopathy
findings indicating that the PDQ-R has a very high false positive rate and is best viewed as a
screening, rather than diagnostic, measure for personality disorders (Hyler, Skodol. Olham,
Kellman. & Doidge, 1992). Because of the relatively low base rates of DSM-III-R defined ASPD,
however, analyses operationallzing HPD at the diagnostic level were not conducted, as
comparisons between DSM-III-R defined ASPD and HPD could not be performed. Because the
DSM-III MMPI Personality Disorder scales do not
map directly onto DSM-III criteria, these
scales cannot be used to derive
diagnoses
of either ASPD
or
HPD.
HAMBURGER, LILIENFELD.
53
AND HOGBEN
significant
and
measures
ratings from
others.
Third, the
construct
validity of BSRI
question by some
authors. Helmreich, Spence, and Wilhelm (1981), for example, have argued
that BSRIMAS and BSRIFEM are better construed as indices of "instrumen
and "expressivity," respectively, than as omnibus measures of gender
roles. This interpretation of BSRIMAS may help to explain a puzzling finding
in the correlational analyses: the significant positive correlation between
tality*'
(although
between BSRIMAS and HPPDQ was close to zero). Given that histrionic
individuals are often noted for their manipulative and controlling qualities
tion of BSRIMAS
as a measure
of
REFERENCES
Akiskal. H. S., Chen, S. E.. Davis. G. C,
Puzantian, V.
R..
Kashgarian,
M.. &
Nebraska Press.
Author.
Association
(1987).
Di
Psychiatric
agnostic and statistical manual of men
American
tal
disorders
ed.,
(3rd
rev.).
DC: Author.
Washington.
Psychiatric Association (1994). Di
men
agnostic and statistical manual of
tal disorders (4th ed.). Washington. DC:
American
chological androgyny.
Journal
of
Con
42,
(1981).
545-551.
Gender schema
theory:
and
hysterical
of Psy
155-162.
Bern. S. L.
Carson,
hysterical personality,
Author.
Bern, S. L.
Cloninger.
C.
Mosby.
R.
54
soclopathy: An Integrative
pathogenesis based on clini
cal, genetic, and neurophysiological ob
hysteria
and
can
servations. In H.
Psychiatric diagnosis: Ex
plorations of biological predictors (pp.
189-218). New York: Spectrum.
Cloninger. C. R. (1987). A systematic method
for clinical description and classifica
tion of personality variants. Archives of
General Psychiatry. 44. 573-588.
Crowne. D. P. & Marlowe, D. (1960). A
desirability independent
of psychopathology. Journal of Consult
ing Psychology. 24. 349. 354.
Ford. M. R. & Widiger. T. A. (1989). Sex bias
in the diagnosis of histrionic and anti
social personality disorders. Journal of
Consulting and Clinical Psychology, 57,
301-304.
of Personality
Disor
ders. 5. 31-42.
of Consulting
chology. 53.
and Clinical
rated.
W. L.
in
(1988). Statistics.
Psychiatric
Institute.
soma
review
and
assessment
of
psychopathy.
17-38.
antisocial
personality
and
nal
of Psychiatry.
143. 718-722.
bidity"'
in
Clinical
Psychology:
concept of "comor
psychopathology
research.
tice. 1, 71-83.
Validity
the
Clinical
Hyler,
disorders:
Hyler,
association be
personality and
Psychology. 1.
the
211-218.
Hays,
(1983).
and Clinical
M.
American
7-16.
chological
Consulting
147, 1043-
Kaplan.
Psy
of Psychiatry,
new
scale of social
orders. Journal
Hyler.
S. Akiskal & W. L.
Webb (Eds.).
Journal
1048.
model of
Personality Diagnostic
Questionnaire-Revised: Comparison
with two structured interviews. Ameri
American Journal
in
of Psychiatry,
131. 518-521.
Morey.
L. C,
& Jewell, J.
HAMBURGER. LILIENFELD,
DSM-III
personality
liminary
validation
Clinical
L. C.
Morey,
K.
disorders: A pre
Journal of
Pope.
R.
sonality disorders:
correlates. Journal
of Personality
As
and
nual Review
247.
personality
study.
(1985).
55
AND HOGBEN
macho
Journal
personality
of Research
in
constellation.
Personality,
18.
150-163.
disorders
in
psychiatric
ceptual
and
search
in
Resources
analysis of con
issues. Re
methodological
Personnel
and
Management, 6,
Human
191-229.
of woman.
Tellegen,
A.
consis
in
of
ences.
Journal
of
Sex Research,
26,
85-106.
of
Journal
tency
542-545.
and
An
diagnosis of antisocial
hysterical personality disorders:
example of sex bias. Journal of
845.
Widiger,
T. A.
dimensional models
IV. Journal
386-398.
Widiger. T.
A., &
Spitzer,
R. L.
1-22.
Zimmerman, M. (1994).
ity disorders:
Diagnosing personal
Psychiatry. 51.
225-245.
of General
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.