Content uploaded by Martin D. Schwartz
Author content
All content in this area was uploaded by Martin D. Schwartz on Jan 16, 2016
Content may be subject to copyright.
Taylor & Francis, Ltd.
is collaborating with JSTOR to digitize, preserve and extend access to
Sociological Focus.
http://www.jstor.org
Gender and Injury in Spousal Assault
Author(s): MARTIN D. SCHWARTZ
Source:
Sociological Focus,
Vol. 20, No. 1 (January 1987), pp. 61-75
Published by: Taylor & Francis, Ltd.
Stable URL: http://www.jstor.org/stable/20831423
Accessed: 16-01-2016 21:51 UTC
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/
info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content
in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship.
For more information about JSTOR, please contact support@jstor.org.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
Gender and
Injury
in
Spousal
Assault
*
SOCIOLOGICAL FOCUS
VoL 20 No.
1
January
1987
VJTender
has
been
an
item
of both
importance
and
some
controversy
in
studies
of
spousal
assault.
Although
most
writers
simply
assume
that
spousal
assault
is
a
problem
of
male
offenders
and
female
victims,
there has
been
some
attempt
to
argue
that
female
abuse
of male
spouses
is
a
serious social
problem.
One
argument
of
this
sort
is that
females
are
just
as
violent
as
males,
both
in
frequency
and seriousness of
assault,
although
for
a
variety
of
reasons
men
may
not
be
as
physically
harmed
as
women
(Straus,
Gelles
and
Steinmetz, 1980).
Another
is
that
although
men
may
not
be
reported
as
victims
of
spousal
assault
as
often
as
women,
they
are
much
more
likely
to
be
injured
by
women
than
they
are
to
injure
women
themselves
(McLeod, 1984).
In
this
article,
data from national
crime
victimization
surveys
will
be used
to
argue
that
neither
position
is correct:
that
spousal
assault is
in
fact
a
problem
of
men
who
beat
women,
and
that
there is
no
difference
in
injury
suffered
by
men
and
women
in
such
assaults. Of
course,
the
latter
finding
can
only
be
read
in
the
context that
virtually
all
spousal
assault
victims
are
women;
one
is
comparing
a
small
number
of
injured
men
to
an
enormous
number of
women.
The
identification of
male
victims,
and the
argument
that
men
are
not
the main
offenders
in
spousal
assault,
are
essential to
the
most
widely
cited
study
to
date
in
the
sociological
literature
on
domestic
violence:
Straus,
Gelles
and
Steinmetz's
Behind
Closed
Doors: Violence
in
the
American
Family
(1980).
This
study
was
based
on
interviews
with
members
of
2,143
American
intact
families
drawn
from
a
national
probability sample.
Respondents
were
asked
to
fit
incidents
to
Straus
et
al.
Conflict
Tactics
Scale,
in which
a
series
of
responses,
including
violence,
are
ranked
in
a
linear fashion.
These researchers
found that
the
number
of
blows
and
threats
by
women
against
men
exceeded
the number
struck
by
men
against
women
by
about
20
percent.
Thus,
Steinmetz
(1977/78:499-508),
was
able
to
proclaim
the existence
of
a
"battered
husband
syndrome"
where she
argued
that
women's
intentions
to
do
harm
are
as
violent
and
frequent
as
men's,
but that
their
inability
to
carry
out
these
intentions
means
that
men
are
less
likely
to
be
physically
harmed than
women.
This
"syndrome"
drew criticisms
The
data
used
in
this
document
were
originally
collected
by
the
Law Enforcement
Assistance
Administration,
and
were
made
available
by
the
Inter-University
Consortium for
Political
and
Social
Research.
Neither
bear
any
responsibility
for
these
analyses
or
interpretations.
An
earlier version
of
this
paper
was
presented
to
the
Academy
of
Criminal
Justice
Sciences,
Orlando,
FL,
in
March,
1986.
Helpful
criticism
has
been
offered
by
Robert
Shelly,
Linda
E.
Saltzman,
and
two
anonymous
reviewers.
61
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
62
SOCIOLOGICAL
FOCUS
pointing
out
that
such
a
simple
counting
of
blows
ignores
relative
power
and
strength,
ignores
the
question
of how
many
of
these blows
were
struck
in
self-defense,
and
does
not
measure
the
consequences
of such
blows,
such
as
injury
(e.g.,
Pagelow,
1984;
Hilberman,
1978;
Pleck, Pleck,
Grossman
and
Bart,
1977/78:680-83;
Fields and
Kirschner,
1978:216-22).
Although
these criticisms
resulted
in
a
few
caveats
in
the full
published
study
(Straus
et
al., 1980:43),
throughout
the
book the
incidence
rate
of
husband
battering
is
presented
as
proof
that violence
is
a
phenomenon
of the entire
family,
not
of
male offenders.
Similarly,
Oswald
(1980:1254),
although
he
misrepresents
his
data,
argues
that
the
present
state
of
knowledge
does
not
allow
a
conclusion that
men are more
harming
than
women.1
A
second
area
which
has
generally
been
ignored by
researchers
has
been
the
extent
of
injury
in
spousal
assault. While
there
have been
a
few
studies where
researchers
have
attempted
to
look
at
some
of the
correlates
of
severity
of
injury
(e.g.,
Dobash,
Dobash
and
Cavanagh,
1985;
Bowker,
1983;
Fagen,
Stewart
and
Hansen,
1983),
theoretical
attention
has
generally
focused
on
the
incidence of
battering
rather than the
severity.
Researchers
have
usually
been
content to
argue
that
there
are
various causative
factors
or
correlates
to
spousal
assult,
with
the
presumption
that
such
spousal
assault
can
cause
injury
(e.g.,
Lincoln
and
Kirkpatrick,
1985;
Breines
and
Gordon, 1983).
Many
writers
further
presume
that
the
greater
physical
strength
of males
means
that
females
are
more
likely
to
suffer
serious
injury
(Saunders,
1986;
Gelles
and
Straus,
1985;
Straus,
Gelles
and
Steinmetz,
1980:43). However,
it
is
the
rare
study
which
reports
more
than
anecdotal data
on
the
extent
of
injury
in
spousal
assault,
and
the rarest
of all
which
reports
this data for
both male and
female
victims.
In
a
recent
study
which
attempted
to
make this latter
comparison,
McLeod
(1984:191)
has
argued:
"Clearly,
violence
against
men
is
much
more
destructive than is
violence
against
women_Male
victims
are
injured
more
often
and
more
seriously
than
are
female
victims.
Medical attention is
necessitated
by
a
large
proportion
of these
injuries.''
The
National
Crime
Survey
(NCS)
incident level data
from
the National
Household
Survey
is useful
in
investigating
some
of
the
questions
raised
by
critics,
although
not
all
of them. For
example,
to
respond
directly
to
the
Straus,
Gelles
and Steinmetz
study
it
would be
useful
if
questions
about
self-defense asked how
many
blows
were
struck
in
direct
response
to
the
attacks of
another. While this is not
asked
in
the
NCS,
one can
still
presume
that
the
wording
of the
screen
questions
would
have
an
effect
on
whether
respondents
would
mention
such blows.
In
the Straus
et
al.
study,
respondents
were
told
that
in
times of
conflict,
people
sometimes do
some
of
the
following
things,
and
were
asked
if
they
had
ever
employed
such
tactics
as
pushing, shoving,
or
hitting.
In
contrast,
in
the
National Crime
Survey,
the
question
was
whether
the
respondent
had been
attacked
by
another
person.
It
does
not
seem
likely
that violent
assaulters who
are
the
recipients
of
some
self-defense blows
would
be
likely
to
report
themselves
as
victims
of
a
violent
or
a
threatened
violent
assault
to NCS
interviewers,
but it does
seem
possible
or even
likely
that
they
would
report
that
dining
a
"conflict situation" their victim
did indeed
strike
some
blows.
In
fact,
in
about
half of
the
families
who
reported
violence
in
the
Straus et
al.
(1980:37)
study,
both
partners
were
reported
as
violent.
This
may
at
least
partially
account
for the
fact
that
in
the
Straus
et
al.
study
women were
reported
as
hitting
men more
often than
men
hit
women,
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
GENDER AND
INJURY
IN
SPOUSAL ASSAULT
63
but
in
the
NCS
survey
data husbands
relatively rarely report
themselves
as
victims of
spousal
assault.
Some further
guidance
on
this
question
can
be
found
in
the
interviews conducted
by
Diana
E. H.
Russell.
Her
study,
based
on
a
random
sample
of
San Francisco
women,
included extensive
probing.
She
found
that
when
women
were
asked
if
they
were
the
victims
of
physical
violence
committed
by
their
husbands,
they
only
responded
affirmatively
if
they
had been
the
victims
of
hitting, kicking,
beating
or
attacks with
weapons.
Most,
she
reported,
distinguished
pushing,
grabbing, shoving,
or
holding
down
and
pinning,
and
did
not
define
these
acts
as
violence
(Russell, 1982:99).
In
other
words,
by including
these acts
in
some
definitions
of
"family
violence," Straus,
Gelles
and
Steinmetz
at
times
used
a
definition broader than
the
victims
themselves
presumably
would
have used.
The
National
Crime
Survey,
on
the
other
hand,
asked
only
about
attacks,
presumably
generally
eliciting
responses
in
the
same
way
as
in
the
Russell
study: by using
the
victim's
own
definition of violence.
Although
the
finding
that
there
is
about
20
percent
more
husband abuse than
wife
abuse
has been
widely
cited and
occasionally
used
to
discredit
efforts
to
help
battered
women
(Pagelow,
1984:267-69;
Pleck
et
al.,
1977/78),
criminal
justice
scholars have
generally
ignored
the
possibility
of battered
men.
Gelles
(1982:207)
has
argued
that
the
reason
most
writers
limit
their
concern
to
women
beaten
by
men
has been
a
sort
of
conspiracy
to
ignore
spousal
abuse
of
men
after
the
political
uproar
which
greeted
the
publication
of Steinmetz's
claims:
"Any
criminal
justice
concern
for
battered
husbands
was
lost
amidst
the
vociferous
claims
by
those
concerned
with
battered
women
that the issue of
battered
men
was a
red
herring."
In
this
study,
NCS
victimization
data
will
be used
to
investigate questions
raised
by
these
researchers.
While
it
has been
previously reported
that
women are
the
overwhelming
majority
of
victims
of
spousal
assault
in
the
National
Crime
Survey
data
(Klaus
and
Rand, 1984),
the
comparative
injuries
incurred
by
men
and
women
has
not
been
reported,
other
than
by
McLeod
(1984).
There is
an
extensive literature
on
the
National
Crime
Survey
data used
in
this article
(e.g.,
Sparks,
1982;
Skogan,
1981;
Block
and
Block, 1984;
Lehnen and
Skogan,
1981;
Saphire,
1984;
Brown and
Woolley,
1985),
as
it has
been
the basis for
a
wide
variety
of
analyses
on
crime
victimization.
Victimization
interviews
have
been
conducted for
the
Bureau
of
Justice
Statistics
at
six month intervals
since
1972
at
about
59,000
housing
units
drawn nationwide from
a
stratified
multistage
cluster
sample.
The
data
here include
all
incidents
from
1973
through
1982 when
a
victim
reported
a
personal
assault
committed
by
a
single
offender
who
was a
spouse
or
exspouse.
Data
were
pooled
across
years,
a
common
procedure
(Nelson,
1980:887;
Shenk
and
Mclnerney,
1981:70),
after checks
were
run
showing
no
significant
differences
by
year
of
reporting
rates
(to
test
for increased
reports
in
recent
years),
no
significant problems
of
memory
decay
over
time
(Biderman,
1981),
or
serious
changes
in
the conclusions
from the
inclusion
or
exclusion
of
multiple
or
series victims
(Sparks,
1982).
A
scale
has
been constructed
here from
a
series
of
questions
asked
in the NCS
about
the
extent
of
injury
in
the assault.
ANALYSIS
The
claim has
been made
by
McLeod
(1984:191)
that
men
are
injured
more
often
and
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
64
SOCIOLOGICAL
FOCUS
more
seriously
than
women
in
spousal
assault,
although
the
process
by
which this
conclusion
is
reached
is
somewhat circuitous. Data
from
Detroit
policie
reports
is
compared
with NCS
data, however,
the
latter
is
generally
limited
to
the
reporting
of
simple
frequencies
on
major
variables for
men
only.
For
example,
McLeod
(1984)
reports
that
of
the
men
in
her
Detroit
sample
who
report
a
spousal
assault,
73
percent report
an
injury,
while in
the
NCS
survey,
77
percent report
an
injury.2
"These statistics
clearly
exceed
estimates
of
the
extent
of
victim
injury
among
female
victims,
generally
documented
as
between
52
and
57
percent"
(McLeod
1984:188).
This
last
comment
is
footnoted
to
an
unpublished
dissertation,
with
no
explanation
of
the
basis for
these
estimates.
In
the
current
data
set,
which includes
the
incidents
examined
by
McLeod
phis
the
incidents
from
two
additional
years,
79.7
percent
of those
men
actually
attacked
reported
an
injury,
but the
equivalent
statistic
for
women
is 84.1
percent.
It
is
possible
that
McLeod
was
confusing
two
different
statistics.
As
can
be
seen
from
Table
1,
if
one
bases the
percentage
of
injuries
on
the total number of
persons
who
report
a
spousal
assault
or an
attempted spousal
assault,
then
53.9
percent
of the
men
Table
I.
Injury
index
by
gender,
for all
victims
of
spousal
assault
or
threatened
spousal
assault
Extent
of
Injury
male
missing
(not
attacked
so
no
possibility
of
injury)
33
32.4%
attacked,
not
injured
14
13.7%
injured,didnTt
see
doctor
36
35.3%
saw
doctor,
no
hospital
1 1
10.8%
emergency
room
treatment
4
3.9%
admitted
to
hospital
4
3.9%
102
5.9%
100%
female
TOTAL
475
28.9%
508
29.
1%
185
11.3%
199
11.4%
639
38.9%
675
38.7%
148
9%
159
9. 1%
158
9.6%
162
9.3%
36
2.2%
40
2.3%
1641
94.
1%
99.9%
1743
100%
X2
p*.29 gamma*.0799
tau
c
p?.
147
Table
does
not
add
to
100% because of
rounding.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
GENDER AND INJURY
IN
SPOUSAL ASSAULT
65
and
59.7
percent
of the
women
report
an
injury.
These numbers
are
similar
to
those McLeod
reports
as
"generally
documented."
However,
as
in
Table
2,
if
the base
population
is
limited
to
those who
were
actually physically
attacked,
then 79.7
percent
of
the
men
and
84.1
percent
of
the
women
report
an
injury.
It
is hard
to
deterinine
just
what
"generally
documented"
estimates
of
female
injury
are
based
upon,
but these data
on
gender
certainly
do
not
allow
a
conclusion that
men
are
injured
more
often
than
women.3
Because of the
manner
in
which the
questions
are
worded
in
the
NCS,
persons
are
asked about
injury
only
for those incidents where
they
were
actually
physically
attacked,
since
presumably
these
are
the
only
cases
where
injury
could result.
As
can
be
seen
in
Table
2,
however,
limiting
the
data
to
these
cases
make
it
even
more
obvious that there
is
no
major
difference
in
injuries by
gender.
While
27.5
percent
of the
men
and
29.2
percent
of
the
women saw
a
doctor
in
some
setting,
women were more
than
twice
as
likely
to
seek
emergency
room
treatment.
A
larger percentage
of the
men were
admitted
to
the
hospital
for
overnight
or
longer,
but
the
small cell
sizes
here make
any
inferences rather
hazardous.
Table 2.
Injury
index
by
gender,
only
for
cases
where victim
actually
physically
attacked.
Extent
of
Injury
male
female
attacked,
not
injured
14
20.3%
185
15.9%
injured,
no
doctor
36
52.2%
639
54.8%
saw
doctor,
no
hospital
1 1
15.9%
148
12.7%
emergency
room
treatment
4
5.8%
158
13.6%
hospital
admission 4
5.
36
3. 1%
TUT
69
100%
1166
100.
1%
X2
p=.2
1
gamma=.0799
tau
c
p=.212
Table
does
not
add
to
100%
because of
rounding.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
66
SOCIOLOGICAL FOCUS
The lack
of
any
difference
in
injury
is
made clearer
when
the
data
are
reduced
into
broad
categories.
In
Table
3,
the
injury
index
is
dichotomized:
no
injuries
at
all,
versus
any
victim-reported
injury
whatsoever,
with
only
those
who
reported
physical
attacks
being
included.
A
greater percentage
of
females
reported
injuries,
although
the
difference
is
not
significant.
Table 3.
Injury
dichotomy
by
gender.
males
females
TOTAL
no
injury
14
185
199
20.3%
15.9%
any
injury
55
981
1036
79.7%
84.1%
69 1166
1235
100%
100%
100%
X2
p?.422
Phi-.028
tau b
p?.166
The
implication
of McLeod's
(1984)
comments,
however,
is that
when
men
are
hurt,
they
are more
seriously
injured
than
women.
While the
small cell
sizes
for
hospital
admissions
make
comparisons
of
the
most serious
injuries
difficult,
it
is
possible
to
rearrange
the data
to
allow
a
comparison
between
those
who
sought
medical
attention
and those
who did
not. In
Table
4,
all
victims
are
included.
The
first
category
includes
all
of
those who
were
not
injured,
whether
attacked
or
not,
and
those
who
claimed
injury
but did
not
seek
any
medical
attention. The second
includes
all those
who
sought
treatment
from
a
private
physician
or
an
emergency
room,
or
were
admitted
to
a
hospital.
Once
again,
there
is
no
statistical
significance
in
the data
regarding
the difference
between
men
and
women
on
the
seriousness of
injury
received
in
spousal
assaults;
although
to
the
extent that there
is
any
difference
at
all,
it
is
that
women
are more
likely
to
seek
medical
care.
It would
be difficult
under these
circumstances
to
affirm
a
finding
that
men
are
more
seriously
injured
in
spousal
assaults,
at
least
using
the
NCS
victimization
reports.
Women make
up
94.1
percent
of
all
spousal
assault
victims
in
the
survey,
and
94.4
percent
of
those victims
who
are
actually physically
attacked,
showing
that
there
is
no
difference
by
gender
between
assaults
which
do
not
result
in
a
physical
attack
(e.g.,
threats with
a
knife)
and
those
which
do.
It
might
be
useful
to
look
at
the
relative
rates
at
which
men
and
women
call the
police.
Men
reported
they
called
the
police
in
67.2
percent
of
the
cases
where
they
were
physically
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
GENDER
AND
INJURY
IN
SPOUSAL
ASSAULT
Table 4.
Doctor's
care
dichotomy
by
gender.
67
male
female TOTAL
no
injury
or
minor
83
81.4%
1299
79.2%
1382
doctor's
care
19
18.6%
342
20.8%
361
102
100%
1641
100%
1743
100%
X2
p-.682
Phi?.013
tau
b
p=.296
attacked,
while
women
called
56.8
percent
of
the time.
Where
there
was
"any
injury,"
men
called
73.6
percent
of
the
time,
while
women
called
57.5
percent
of
the
time.
In
other
words,
we
could
argue
that
police
statistics
on
injury
overrepresent
men
in
spousal
assault!
Of
course,
if
must
be
pointed
out
that
the
entire
question
of
comparative
rates
of
injury
between
men
and
women
is
a
somewhat
esoteric and academic
question.
Even
if
the data
were
different
?
even
if
men were
much
more
seriously injured
in
spousal
assaults
than
women
?
the
most
important
statistic
is
that
women
make
up
94.1
percent
of
all
spousal
assault
victims.
No
matter
what
the
injury
rates,
the
most
important
finding
is
that
spousal
assault
is
primarily
a
problem
of
male
offenders
and
female
victims. To
choose
one
category
in
the
above
analysis
for
an
example,
a
higher
percentage
of
men
saw a
private
doctor
out of
a
hospital
setting
than did
women:
15.9
percent
of
the attacked
men,
as
opposed
to 12.7
percent
of the
attacked
women.
However,
it
must
always
be
kept
in
mind that
what
we
are
comparing
here
are
11
men
and
148
women.
Even
though
the
rate
for
men
is
higher,
there
are
still
more
than
13 times
as
many
women
seeking
medical
care
from
a
private physician
for
injuries
received
in
a
spousal
assault.
While it
can
hardly
be
suggested
that the National
Crime
Survey
is free from
methodological
flaws,
some
confidence
for
the
finding
that
approximately
95
percent
of
the victims
are
female
can
be obtained
from
smaller
studies,
where
findings
are
similar.
Although
there
are
no
similar
victimization
data drawn
from random
samples,
Minnesota
statistics
drawn
from
a
variety
of
sources
show
males
as
about four
percent
of
spousal
assault victims
(Watkins,
1982:32).
In
Ohio
studies,
Bell
(1985b)
found
that
in
an
examination
of all
police
reports
on
domestic
violence
wives
were
the
victims
in
about
92
percent
of those
complaints
which involved
spousal
assault,
while
Quarm
and Schwartz
(1985)
found that male
victims of
female
assault
accounted
for
about
five
percent
of
cases
actually
filed
in court.
Dobash
and Dobash
(1978)
gathered
both
police
and
court
data
in
Scotland,
and
came
to
the
same
conclusion:
males
were
less than
two
percent
of the
spousal
assault
victims
they
studied.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
68
SOCIOLOGICAL FOCUS
The
use
of
more
complex
multivariate
statistics
to test
the
relationship
between
gender
and victimization
does
not
change
any
of
the above conclusions.
First,
a
disciminant
analysis
was run on
the
major
predictive
variables discussed
in
the
literature:
family
income,
race
and
sex.
The
groups
used
were
those attacked
but
not
injured,
and
those
with
any
injury.
Table
5.
Discriminant
analysis
between
groups
not
injured,
and
any
injury.
Method:
Wilks
Model
Significance
After
Function
0: 0.024
standard
canonical
variable
wilks1
lambda
significance
function
coefficients
Family
Income
.993
.015
-.406
Race
.994
.008
.814
Sex
.992
.024
-.332
FUNCTION:
eigenvalue
?
.008 canonical
correlation
=
.091
eta2
-
.008
GROUP
CENTROIDS:
not
injured
.207
injured
-.040
PERCENT
CORRECTLY
CLASSIFIED:
66.2%
This
seems
to
be
an
excellent
argument
either
for
setting
the
acceptable
significance
level
for models
to
a
higher
.01
level,
or
else
a
good
demonstration
of
how
a
model
can
achieve
statistical
significance
while
being
of almost
no
substantive
value.
The
eigenvalue,
showing
the
power
of the
function
to
cfocriminate
between
the
two
groups,
is
virtually
at
zero,
as
in
the
eta2.
While
the
percentage
correctly
classified is
high,
this statistic is
often
as
misleading
as
it
is
helpful.
Correctly
classifying
66.2
percent
is
valuable,
however,
it
is
worthwile
to
put
this into
some
context
by noting
that
in this
particular
analysis
84
percent
of
all
cases
were
injured.
In
other
words,
simply guessing
that
all
cases were
in
this
second
category
would result
in
a
roughly
27
percent
increase
in
predictive
power
over
that
provided
by
the discriminant function.
At
least
theoretically,
problems
could
be caused
by
the lack of
variation between
the
groups,
and
because
two
of the
independent
variables
and the
dependent
variable
are
dichotomies
(Aldrich
and
Nelson, 1984).
Therefore,
multiple
logistic regression
which does
not
require
the
mathematical
assumptions
of
discrirriinant
analysis
(AfM
and
Clark, 1984),
was
used
on
the
same
data.
Unfortunately,
it is
not
any
more
helpful.
As
with the discriminant
analysis,
it
would
be difficult
to
imagine
a
significant
outcome
with less
predictive
power.
Only
race
entered
the
equation,
showing
that
whites
are
very
slightly
more
likely
to
be
injured
in
spousal
assaults.
A
much better
conclusion, however,
given
the
multiple
R2,
is
to
reject
even
race as a
predictive
variable.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
GENDER
AND
INJURY
IN
SPOUSAL
ASSAULT
69
Table
6.
Stepwise
logistic
regression
analysis using injury
dichotomy
as
the
dependent
variable.
Model
Likelihood
Ratio
X2
p=.0108
variable
mean
beta
X2
p
partial
r
intercept
1.138 32.38 TOOO
Race
Dummy
.88
.578 6.98
.008
.070
Not
entered:
Family
Income,
Sex
Multiple
R=.067
Multiple
R2=.004
Fraction of concordant
pairs
=
.160
rank
correlation
of
probability
and
response
=.070
WEAPON
USE
It
is
not
uncommon
to assert
that
one reason
for
differences
in
the
severity
of
injury
would
be differences
in
the
use
of
weapons.
Many
authors have
argued
that
when
women
do strike
men,
whether
as
aggressors
or
in
self-defense,
they
are
more
likely
than
men
to
use
weapons
(Pagelow,
1984:272;
Mitchell,
1978;
Wolfe, 1979).
McLeod
(1984:185)
asserts
that
weapons
use causes
men
to
be
injured
more
than
women
in
spousal
assaults,
while
Steinmetz
(1977/78:505)
suggests
that
women
and
men are
equally
injured
when
weapons
are
used.
Although
it has
already
been
seen
that
in
general
men
are
not
injured
more
than
women
in
such
assaults,
the
relative
frequency
of
weapons
use
has
not
yet
been
examined.
There
is
no
question
but that males
are more
likely
to
report
that
a
weapon
was
used
against
them.
Whether this
is
due
to
an
actual difference
in
behavior,
or
a
difference
in
perception
of when
an
assault
has
taken
place,
is
impossible
to
answer
from these
data.
For
example,
a
man
who believes that
a
smaller
woman
cannot
hurt
him
without
a
weapon
might
not
take
a
physical
threat
without
a
weapon
seriously,
and
will
not
define
it
as
an
assault
(Saunders, 1986:49-50).
However,
approximately equal
numbers
of
the
assaults
reported
resulted
in
an
actual
physical
attack: 69
percent
of the
men
and
71.7
percent
of
the
women
were
physically
attacked.
We
can
compare
injuries
resulting
from these
attacks
with the
weapons
used.
Within each
sex,
there is
no
relationship
between
the
weapon
used,
or
in
fact,
whether
a
weapon
was
used
at
all,
and the
percentage
of
"any
injury."
However,
one
can
see
that
except
for
knife
attacks,
women are
injured
more
often than
men.
Of
course,
this
dichotomy
lumps
all
injuries
together,
and
does
not
differentiate
serious
from
minor
injuries.
While
the
5-point
ordinal scale of
extent
of
injury
by
weapons
use
was run
for each
sex,
these
data
have been reduced below
to
show,
first,
the
percentage
of
assault victims who
sought
any
medical
attention,
and
second,
the
percentage
who
sought
hospital
treatment
(either
emergency
room or
inpatient
treatment).
Of
course,
the
second
group
is
a
subset
of
the
first.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
70
SOCIOLOGICAL FOCUS
Table
7.
"Any injury"
dichotomy
by weapon
used
by
gender.
men women
no
mj
any
injury
no
injury
any
injury
no
weapon
7
22 151 764
24.1%
75.9%
16.1%
83.5%
gun
2 6 8 47
25%
75%
14.5%
85.5%
knife
18 10 59
11.1%
88.9% 14.5%
85.5%
other,
4
19
14
100
unknown
17.4%
82.6% 12.3%
87.7%
14
55
183 970
20.3% 79.7% 15.9% 84.1%
X2
p=.8
1
X2
p=.67
Cramer's
V=.118
Cramer's
V=.037
As
discussed
earlier,
the
claim has been
made
that this data
set
supports
a
contention
that
men are
more
seriously injured
than
women.
Certainly
there
is
no manner
in
which
the tables above could be
interpreted
in
this
manner.
Of
course,
it
would
also be
difficult
to
argue
that
the
extent to
which
women
are
injured
more
than
men
is
statistically
and
substantively significant.
Thus,
the conservative conclusion
from
these data would
be
that there is
no
difference
in the
seriousness of
injury by gender.
However,
the
higher
percentage
of
women
who
receive
more
serious
injuries
than
men
does
not
merit
this
extreme
caution:
one
can
safely
argue
that
women are more
seriously
injured
in
spousal
assault
cases
than
are men.
Substantively,
the fact that
women
make
up
94.1
percent
of all
spousal
assault
victims
means
that
no
matter
what
the
results
are
of
the above
analysis,
the
problem
of
injury
in
spousal
assault is
essentially
a
problem
for
women.
MURDER
Unfortunately,
the
nature
of
victimization
surveys
precludes
the
use
of
this
technique
to
investigate
violence
which
ends
in
death:
one
cannot
ask
potential
victims
if
they
have
been killed
recently. Despite
the absence of
any
comprehensive
surveys
on
the
circumstances of
spousal
killings,
we can
look
to
several
indicators
to
investigate
whether
the
pattern
seen
here of
injury
holds
true
for
killings
also.
It
is
a
reasonable
presumption
that the
problem
of
nonreporting
of crimes
applies
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
GENDER
AND INJURY
IN SPOUSAL
ASSAULT
71
Table
8.
Percentage
of
victims who
sought
any
medical
care,
by
offender1s
weapon
use
All Cases Subset of
All
Cases
Any
doctor's
care
hospital
treatment
men women
men
women
no
weapon
7/29
244/915 1/29
133/915
24.1%
26.7%
3.4%
14.5%
gun
2/8
18/55 2/8 9/55
25%
32.7%
25%
]6.4%
knife
5/9 25/69
2/9 19/69
55.6%
36.2%
22.2%
27.5%
other,
5/23
52/114 3/23
30/114
ukn
type
21.7%
45.6%
13% 26.3%
19769
339/1153
"
8769
191/1
153
27.5% 29.4%
11.6%
16.6%
full
table
statistics:
men:
X2
p=.26
Cramer's
V=.266
women:
X2
p=.009
Cramer's
V=.088
less
to
murder
than
to
any
other
crime;
dead bodies tend
to
come
to
the
attention
of
the
police
at
a
greater
rate than
battered bodies.
The
nation's
major
index
of
reported
murders
in
America,
the Uniform
Crime
Reports,
shows
that wives
killed
by
their
husbands
make
up
61.9
percent
of
all
spousal
murders and
nonnegligent manslaughters,
while what the
FBI
terms
"girlfriends"
make
up
60
percent
of
all
"girlfriend/boyfriend"
killings.4
Once
again,
this
does
not
take
the issue of
self-defense
into account. There
are numerous case
studies
available
of
women
who
killed their husbands
only
after
extreme
battering
(Jones,
1985),
and
at
least
one
study
reported
that
in 40
percent
of
the
cases
of
women
jailed
for
murdering
their husbands chronic
physical
battering
was
a
factor
in
the
killing
(McCormick, 1976).
Given
the small
number
of
spousal
killings reported
by
the Uniform Crime
Reports,5
it is
unlikely
in
any
event
that the number
of
cases
which
end
in
death could
strongly
influence
any
findings
of
whether,
on
the
average,
men or
women
are
more
injured
in
spousal
assaults;
even
the
most
conservative
estimates
would situate murder
as a
tiny
proportion
of
battering
incidents in America.
However,
the
pattern
of
spousal
assault
clearly
shows
that
women
are
more
likely
to
be
killed
by
men
than the
reverse.
It
is
a
reasonable
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
72
SOCIOLOGICAL FOCUS
supposition
that
if
self-defense
cases
could
be removed
from these
statistics,
the
overwhelming majority
of
spousal
murder
victims
would be
women.
CONCLUSION
The
conclusion
to
be drawn
then,
is that the data
support
those
whom
Gelles
(1982)
was
quoted
earlier
as
attacking
for
arguing
that
the
issue of
male
spousal
assault
victimization is
a
"red
herring/' Actually,
Gelles has
recently
refined
his
critique
(Gelles
and
Cornell,
1985:80),
in
sensitivity
to
the
strong
attacks which
have
been
mustered
on
the
original
Straus,
Gelles
and
Steinmetz data:
It is
quite
clear that
men are
struck
by
their wives. It is also clear that because
men
are
typically
larger
than
their
wives
and
usually
have
more
social
resources
at
their
command,
that
they
do
not
have
as
much
physical
or
social
damage
inflicted
on
them
as
is
inflicted
on
women.
Data
from
studies
of
households
where the
police
intervened
in
domestic
violence,
clearly
indicate
that
men are
rarely
the victims
of
'battery'.
. . .
Thus,
although
the
data
. . .
show similar
rates of
hitting,
when
injury
is
considered,
marital violence is
primarily
a
problem
of victimized
women.
In
this
interesting
argument,
the
line
of
reasoning
seems
to
be that
because
the
criminal
justice
system
records few
cases
of males
injured
in
spousal
assaults
by
their
wives
we
can assume
that
men are
rarely
injured
in
such
attacks.
Convoluted
as
this
logic
may
be,6
the data
here
support
these
arguments.
Men
are even more
likely
to
call the
police
if
injured,
and
are
about
equally likely
to be
injured
in
spousal
assaults.
If
anything
male
assault victims
should be
overrepresented
in official
crime statistics. The fact that
they
remain
a
small
percentage
of
victims
in
official
data
is
supported
by
the
simple
explanation
that
they
are
a
small
percentage
of victims.
While
there
are
no
important
differences
in
the
injury
rates
for male
and
female
victims
of
spousal
assault,
the fact
that
women
make
up
about
95
percent
of
all
victimizations
in
the
National Crime
Survey
data
base
makes
it obvious that
the
ultimate conclusion
of
Gelles
and
Cornell
above
is
correct:
"marital violence
is
primarily
a
problem
of
victimized women."
However,
unlike
their
conclusion,
which
implies
that while
marital
violence
is
spread equally
across men
and
women,
but that
the
problem
can
be
seen as one
for
women
because
women
are
more
injured,
this
analysis
makes
it
plain
that
at
any
level of
injury
or
threat
(whether
or
not
physically
attacked),
spousal
assault
"is
primarily
a
problem
of victimized women" because
women are
virtually
all of
the victims.
As
Berk, Berk,
Loseke
and
Rauma
(1983:210)
conclude:
While
there
are
certainly
occasional instances
of
husbands
being
battered,
it is
downright
pernicious
to
equate
their
experiences
with
those
of
the
enormous
number of
women
who
are
routinely
and
severely
victimized.
FOOTNOTES
1. A
record of his
data
shows
that in
interviews
with
female
para-suicide
hospital
admissions,
12
percent
of
those
previously
involved
in
any
serious
violence said
they
were
the sole
offenders,
while 46.5
percent
said
their
husbands
were
the sole
offenders.
Both
were
violent,
including
self-defense
cases,
41.5
percent
of the
time.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
GENDER
AND INJURY
IN SPOUSAL ASSAULT
73
2.
It is
always dangerous,
of
course,
to
attempt
to
project
national trends from
a
single
police department's
data,
particularly
in the field of
spousal
assault,
where the
discrepancies
from
city
to
city
are
well
known
(Loving,
1980;
Pagelow,
1984).
For
example,
one
might
compare
to
Detroit
(a)
figures
obtained
by recalculating
tables in Bell
(1985b:32):
in
Ohio
smaller
towns
and
cities,
21.7
percent
of
husbands
and
47
percent
of
wives
were
injured;
or
(b)
Berk,
Berk,
Loseke and Rauma's
(1983:199)
single
California
police department,
where
7
percent
of the
men
and
43
percent
of the
women were
injured
in
"domestic
disturbance" incidents.
The
latter
study
included
cases
where both
were
injured
after
a
fight.
In
incidents
where
only
one
person
was
injured,
4
percent
of
the
men
and 39
percent
of
the
women were
injured
(Bert
et
al.,
1983:204).
3.
An
examination
of the
data make
it
obvious
that
McLeod based
her
77
percent
figure
on
men
who
were
actually
attacked.
An
attempt
to
limit the
figures
in
Table
1
to
only
the
period
covered
by
McLeod's
analysis
resulted in
figures
of
76
percent
of
men
and
84.6
percent
of
women
actually
attacked
reporting
any
injury.
4.
Data
reconstructed
from
tables
in
Federal Bureau of
Investigation
(1985:11).
Although
1984
data
were
used
to
calculate the
percentages
here,
these
figures
have been rather stable
in recent
years.
5.
The Uniform Crime
Reports
does
not
report
this
number,
and does
not
report
the number of
cases
included
in
the table which
gives
the
percentage
of total
cases
which
are
spousal
murder.
Using,
for
1984,
the
same
table
N
as
in other
similar
tables
(16,689
murders),
the 8.4
percent
which
are
spousal
murder
would
mean
1,401
cases.
McGarrell
and
Flanagan
(1985:419)
report
1983 UCR
data,
using
a
total
N of
18,673
for the
table. Since 9.4
percent
of
1983
murders
were
spousal,
that would
mean
1,755
cases.
Girlfriend/boyfriend
murder made
up
4
percent
of 1984
murders,
and 3.7
percent
of
1983 murders.
6. After
conducting
an
expensive federally
funded
survey
of
potential
spousal
assault victims
on
the
grounds
that official criminal
justice
agency
statistics
are
incomplete
and
invalid,
the
argument
is
now
made
by
Gelles
and
Cornell
(1985)
that
one
of
the
most
important
findings
of
these
surveys
must
be
modified
in
light
of
the
superior
official
statistics
of
the
one
police
agency
cited
by
Berk, Berk,
Loseke and Rauma
(1983).
REFERENCES
Afifi,
A. A.
and
Virginia
Clark
1984
Computer-Aided
Multivariate
Analysis.
Belmont,
CA:
Wadsworth.
Aldrich,
John H.
and
Forrest
D.
Nelson
1984 Linear
Probability, Logit,
and
Probit
Models.
Beverly
Hills,
Sage.
Bell,
Daniel J.
1985a
"Domestic violence:
Victimization,
police
intervention,
and
disposition."
Journal
of
Criminal
Justice
13:525-34.
1985b
'
'Domestic Violence
in
Rural
Areas,''
paper
resented
at
the annual
meetings
of the
Academy
of
Criminal Justice
Sciences,
Las
Vegas,
Nev.
(March).
Berk,
Richard
A.,
Sarah Fenstermaker
Berk,
Donileen
R.
Loseke
and David
Rauma
1983
"Mutual combat and other
family
violence
myths." Pp.
197-212 in
David
Finkelhor,
Richard
J.
Gelles,
Gerald
T.
Hotaling
and
Murray
A.
Straus
(eds.),
The
Dark
Side
of
Families:
Current
Family
Violence
Research.
Beverly
Hills,
CA:
Sage.
Biderman,
Albert
D.
1981 "Sources of data for
victimology."
Journal
of
Criminal
Law
and
Criminology
72:789-817.
Block,
Carolyn
Rebecca
and Richard L.
Block
1984
"
Crime
definition,
crime measurement and
victim
surveys."
Journal
of
Social Issues
40:137-60.
Bowker,
Lee H.
1983
Beating
Wife
Beating. Lexington,
MA:
Lexington
Books.
Breines,
Wini
and Linda Gordon
1983
"The
new
scholarship
on
family
violence."
Signs:
Journal
of Women
in
Culture and
Society
8:490-531.
Brown,
Steven E. and Thomas W.
Woolley
1985 "The
national
crime
survey
program:
Prob
lems
in
sample
selection
and
data
analysis."
Social
Science
Quarterly
66:86-93.
Dobash,
R.
Emerson
and
Russell
P.
Dobash
1978
"Wives:
The
'appropriate'
victims of
marital
violence."
Victimology
2:426-42.
Dobash,
R.
Emerson,
Russell
P.
Dobash
and
Katherine
Cavanagh
1985 "The contact
between
battered
women
and
social
and
medical
agencies."
Pp.
142-65
in
Jan
Pahl
(ed.),
Private
Violence and
Public
Policy:
The
Needs of
Battered Women
and
the
Response
of
the
Public Services.
Lon
don:
Routledge
and
Kegan
Paul.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
74
SOCIOLOGICAL
FOCUS
Fagen,
Jeffrey
A.,
Douglas
K.
Stewart
and Karen
V.
Han
sen
1983 "Violent
men or
violent
husbands?
Background
factors
and
situational
cor
relates."
Pp.
49-68
in
David
Finkelhor,
Richard J.
Gelles,
Gerald
T.
Hotaling
and
Murray
A.
Straus
(eds.),
The
Dark
Side of
Families:
Current
Family
Violence
Research.
Beverly
Hills,
CA:
Sage.
Federal
Bureau of
Investigation
1985 Crime
in
the
United States
?
1984.
Washington,
D.C.:
U.S. Government
Printing
Office.
Fields,
Marjory
D. and
Gioghan
M. Kirschner
1978 "Battered
women
are
still
in
need."
Vic
timology
3:216-22.
Gelles,
Richard J.
1982 "Domestic criminal
violence."
Pp.
201-35 in
Marvin
E.
Wolfgang
and
Neil Alan
Weiner
(eds.),
Criminal
Violence.
Beverly
Hills,
CA:
Sage.
Gelles,
Richard
J.
and
Claire Pedrick
Cornell
1985
Intimate Violence
in
Families.
Beverly
Hills:
Sage.
Gelles,
Richard
J.
and
Murray
A.
Straus
1985
"Violence
in
the
American
family." Pp.
88-110
in Alan
Jay
Lincoln
and
Murray
A.
Straus
(eds.),
Crime
and
the
Family.
Springfield,
IL:
Charles
C.
Thomas.
Hilberman,
Elaine
1978
"Response."
Pp.
157-59
in
U.
S. Commis
sion
on
Civil
Rights,
Battered
Women:
Issues of
Public
Policy. Washington,
D.C.:
U.S.
Commission.
Jones,
Ann
1985
Everyday
Death.
New York:
Holt,
Rinehart
and Winston.
Klaus,
Patsy
A. and Michael
R. Rand
1984
Family
Violence.
Washington,
D.C.:
Bureau
of
Justice
Statistics.
Lehnen,
Robert
G. and
Wesley
G.
Skogan
(eds.)
1981
The National
Crime
Survey: Working
Papers,
Vol.
I,
Current
and
Historical
Perspectives.
Washington,
D.C.: U.S.
Dept.
of Justice.
Lincoln,
Alan
Jay
and
John
T.
Kirkpatrick
1985
'
'Criminological
theory
and
family
crime.''
Pp.
46-63
in
Alan
Jay
Lincoln
and
Murray
A.
Straus
(eds.),
Crime
and
the
Family.
Springfield,
IL:
Charles
C.
Thomas.
Loving, Nancy
1980
Responding
to
Spouse
Abuse
and
Wife
Beating:
A
Guide
for
Police.
Washington,
D.C.:
Police Executive
Research
Forum.
McCormick,
C.
1976
Battered
Women
?
The Last
Resort. Un
published
manuscript.
McGarrell,
Edmund
F. and
Timothy
J.
Flanagan
1985
Sourcebook
of
Criminal Justice
Statistics
-
1984.
Washington,
D.C.:
U.S.
Govern
ment
Printing
Office.
McLeod,
Maureen
1984
"Women
against
men:
An examination
of
domestic
violence based
on
an
analysis
of
official
data and
national
victimization
data." Justice
Quarterly
1:171-94.
Mitchell;
Marilyn
Hall
1978
"Does
wife abuse
justify
homicide?"
Wayne
Law
Review 24:1705-31.
Nelson,
James
F.
1980
"Multiple
victimization
in American
cities:
A
statistical
analysis
of
rare
events."
American
Journal
of
Sociology
85:870-91.
Oswald,
Ian
1980
"Domestic violence
by
women."
Lancet
(II)
8206:1253-54.
Quarm, Daisy
and
Martin
D.
Schwartz
1985
"Domestic violence
in
criminal
court.''
Pp.
29-46
in
Claudine
Schweber
and
Clarice
Feinman
(eds.),
Criminal
Justice
Politics
and Women:
The
Aftermath
of
Legally
Mandated
Change.
New York:
Haworth.
Pagelow,
Mildred
Daley
1984
Family
Violence.
New
York:
Praeger.
Pleck,
Elizabeth,
Joseph
H.
Pleck,
Marlyn
Grossman
and
Pauline
B. Bart
1977/78
"The battered
data
syndrome."
Vic
timology
2:680-83.
Russell,
Diana
E.
H.
1982
Marital
Rape.
New York:
Macmillan/
Collier.
Saphire,
Diane
Griffin
1984 Estimation
of
Victimization Prevalence
Using
Data From
the National Crime
Survey.
New
York:
Springer-Verlag.
Saunders,
Daniel
G.
1986
"When battered
women use
violence:
Husband-abuse
or
self
defense." Victims
and
Violence
1:47-60.
Shenk,
Fred
and William
Mclnerney
1981
"Analytic
limitations
of
the national
crime
survey."
Pp.
70-75
in
Robert
G.
Lehnen
and
Wesley
J.
Skogan
(eds.),
The
National
Crime
Survey: Working
Papers,
Vol.
I,
Current
and
Historical
Perspectives.
Washington,
D.C.:
U.S.
Bureau
of
Justice
Statistics.
Skogan,
Wesley
J.
1981
Issues
in
the
Measurement of
Victimiza
tion.
Washington,
D.C.: U.S.
Govt.
Printing
Office.
Sparks,
Richard
F.
1982 Research
on
Victims
of
Crime:
Accomplishments,
Issues
and New
Direc
tions.
Rockville,
MD:
Dept.
of
Health
&
Human
Services.
Steinmetz,
Suzanne
K.
1977/78
"The
battered husband
syndrome."
Vic
timology
2:499-509.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
GENDER
AND
INJURY
IN
SPOUSAL
ASSAULT
75
Straus,
Murray
A.,
Richard
J. Gelles
and Suzanne
K.
Steinmetz
1980
Behind Closed
Doors.
Garden
City,
NY:
Anchor Books.
Watkins,
Carol
R.
1982
Victims,
Aggressors
and the
Family
Secret:
An
Exploration
into
Family
Violence.
St.
Paul:
Minnesota
Department
of
Public
Welfare.
Wolfe,
Nancy
1979
"Victim
provocation:
The battered
wife
and
legal
definition
of
self-defense."
Sociological
Symposium
25:98-118.
This content downloaded from 198.91.37.2 on Sat, 16 Jan 2016 21:51:36 UTC
All use subject to JSTOR Terms and Conditions
