Childhood Attachment, Childhood Sexual Abuse,

and Onset of Masturbation Among Adult

Sexual Offenders

Stephen W. Smallbone
1.3 and Billee-Anne 2
McCabe

Written autobiographies of48 incarcerated adult male sexual offenders (22 rapists,
13 intrafamilial child molesters, and 13 extrafamilial child molesters) were used to
generate retrospective self-report measures of their childhood maternal and pater-
nal attachment, childhood sexual abuse experiences, and onset of masturbation.
Contrary to expectation, the offenders as a combined group more often reported
secure than they did insecure childhood maternal and paternal attachment. There
were no differences between the three offender subgroups with respect to maternal

attachment; however the rapists and the intrafamilial child molesters were more
likely to report insecure paternal attachment than were the extrafamilial child mole-
sters. There were no differences between these offender subgroups in the frequency
with which childhood sexual abuse was reported. However, offenders with insecure
paternal attachment were more likely to report having been sexually abused than
were those with secure paternal attachment. Sexually abused offenders in turn re-

ported earlier onset of masturbation than did those who were not sexually abused.
These results are consistent with contemporary attachment models linking insecure
childhood attachment to childhood sexual abuse, and with traditional conditioning
models linking childhood sexual abuse, early masturbation, and sexual offending.
KEY WORDS: attachment; sexual abuse; masturbation; sexual offenders.

Developmental factors feature in otherwise diverse etiological models of sex-

ual offending behavior. Psychodynamic models have proposed that disruptions
to psychosexual development may, for example, lead to failures to inhibit the
School of Criminology and Criminal Justice, Griffith University, Queensland, Australia.
1
S2chool of Applied Psychology, Griffith University, Queensland, Australia.
To whom correspondence should be addressed at School of Criminology and Criminal Justice, Mt.
3
Gravatt Campus, Griffith University, Queensland 4111, Australia; e-mail: s.smallbone@mailbox.
gu.edu.au.

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

2

expression of infantile sexual interests (Cohen, Seghom, & Calmas, 1969) or of

hostility to women (Groth, Burgess, & Holstrom, 1977). Conditioning and social
learning models tend to rely on the presence of particular developmental events
and sequences of learning, especially with respect to the acquisition of deviant
sexual interests (Laws & Marshall, 1990). The prominent multifactorial models
(Finkelhor, 1984; Malamuth, Heavey, & Linz, 1983; Marshall & Barbaree, 1990)
highlight childhood and adolescent experiences as important elements in the de-
velopment of individual dispositions to engage in sexual offending behavior. In
recent years, the etiological significance of insecure childhood attachment among
sexual offenders has received theoretical consideration (Marshall, 1989; Ward,
Hudson, Marshall, & Siegert, 1995) and tentative empirical support (Smallbone
& Dadds, 1998; Smallbone & Dadds, 2000; Ward, Hudson, & Marshall, 1996).
Given the considerable heterogeneity among sexual offenders, it is likely that
sexual offending behavior may arise from a variety of developmental pathways.
The aim of this study was to explore a particular developmental sequence among
a sample of convicted sexual offenders. The key elements in this proposed path-

way were derived from two distinct theoretical formulations-classical condition-

ing, and attachment theory. Specifically, associations between insecure childhood
attachment, childhood sexual abuse, and onset of masturbation are examined.
Childhood attachment insecurity among sexual offenders has been proposed
to lead to disruptions in empathy, and to the failure to achieve intimacy in adoles-
cence and adulthood (Marshall, 1989; Ward et al., 1995). Smallbone and Dadds

(1998, 2000) have further proposed that early disorganization of the childhood
attachment behavioral system may lead to reduced functional separation between
the adult attachment, sexual, and parenting behavioral systems, allowing other-
wise innocuous internal (e.g., subjective distress) and external cues (e.g., physical
contact with a child) to trigger inappropriate and coercive sexual behavior. Prelim-
inary empirical investigations suggest that childhood attachment insecurity may
be a general risk factor for criminality, and that paternal attachment insecurity may
be a more specific risk factor for sexual offending (Smallbone & Dadds, 2000).
Nonetheless, many sexual offenders report secure attachment relationships with
their parents.
Childhood attachment insecurity has been associated with a wide range of
negative sequelae, including conduct and substance abuse disorders; narcissistic,
borderline, and antisocial personality disorders; and aggressive behavior problems
(Jones, 1996; Lyons-Ruth, 1996). Childhood attachment insecurity has also been
theoretically linked with childhood sexual abuse. In particular, Alexander (1992)
has argued that parental unavailability and rejection leaves the child with reduced
parental supervision and protection, and with fewer psychological resources with
which to resist sexual abuse. Alexander (1992) has also argued that the negative
consequences of sexual abuse for children with insecure attachment relationships
are likely to be more severe than for children with secure attachment relationships,

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

 3

because insecure children will have a less functional family support system avail-
able to them. Thus, childhood attachment insecurity may both increase the risk of
sexual abuse, and lead to more negative outcomes.
Although the prevalence and significance of childhood sexual abuse among
sexual offenders remains controversial (see e.g., Freund, Watson, & Dickey, 1990),
empirical evidence from correctional surveys and clinical studies suggests that sex-
ual offenders are more likely than both nonoffenders and nonsexual offenders to
report having been sexually abused (Bureau of Justice Statistics, 1997; Dawhan
& Marshall, 1996). Psychodynamic explanations linking childhood sexual abuse
and later sexual offending typically involve the operation of complex, unobserv-
able intrapsychic processes, such as victim identification with the offender and
psychological resolution of sexual trauma (Groth, 1979). Learning models are po-
tentially more easily tested, although evidence is scarce. Laws and Marshall (1990)
have described a range of classical conditioning, operant conditioning, and social
learning processes that may be involved in the acquisition of deviant sexual inter-
ests, illustrated with reference to early sexual experiences. One prominent model
involves the role of masturbation in facilitating learned associations between de-
viant sexual imagery and sexual arousal (McGuire, Carlisle, & Young, 1965). This
model suggests that childhood sexual abuse may serve to introduce the victim to
deviant sexual images that are subsequently used during masturbation. It is thus
plausible that childhood sexual abuse may trigger premature masturbation, which
may in turn provide a means to acquire and maintain deviant sexual interests.
Several testable hypotheses may be derived from the attachment and con-
ditioning theory formulations discussed above. Specifically, the following three
hypotheses were tested in a sample of convicted sexual offenders:
1. Sexual offenders will be more likely to have experienced insecure than
secure childhood attachment.
2. Sexual offenders who have experienced insecure childhood attachment
will be more likely to have been sexually abused than will those who have
experienced secure childhood attachment.
3. Sexual offenders who have been sexually abused will have begun mastur-
bating earlier than will those who have not been sexually abused.

METHOD

Participants
Written autobiographies of 48 incarcerated adult male sexual offenders
(22 rapists, 13 intrafamilial child molesters, and 13 extrafamilial child molesters)
were used to generate measures of childhood attachment, childhood sexual abuse,
and onset of masturbation. The offenders had submitted their autobiographies as a

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

4

requirement of their participation in a comprehensive cognitive-behavioral treat-

ment program operating in a maximum security prison in Queensland, Australia.
To be eligible for inclusion in the program an offender must have been convicted
of at least one sexual offense and to be willing to engage in the program. Offend-
ers are excluded if there is an active psychiatric disorder, if their IQ is formally
assessed as being below 85, or if they categorically deny having committed a sex-
ual offense. Offenders who killed their victim are also excluded. Offenders who
participate in the program are required to sign a consent form, which among other
things indicates their permission for information provided by them to be used for
the purposes of research.
Offenders’ ages at the time of their current conviction ranged from 18 years
to 67 years (M 34.0, SD 10.63). There were no significant differences in
= =

age between the intrafamilial child molesters, extrafamilial child molesters, and
rapists.
Forty percent (40%) of the offenders had previously been convicted of at
least one sexual offense, and 25% had previously been convicted of a nonsexual
violent offense. Extrafamilial child molesters (69.2%) were more likely than either
the intrafamilial child molesters (30.8%) or the rapists (27.3%) to have previously
been convicted of a sexual offense, X (2, N = 48) 6.59, p < .04. There were
=

no differences between the subgroups in the likelihood of previous convictions for

nonsexual violent offenses.

Measures

Offenders were provided with an autobiography outline prior to beginning

treatment, and asked to respond to each of 64 questions about their experiences of
childhood and adolescence, sexual development, adult employment and relation-
ships, sexual and nonsexual offending, and previous counseling or treatment. For
the purposes of this study, answers to the following questions were examined:

1. In what ways did your mother/father show that she/he accepted you?
2. In what ways was she/he rejecting of you?
3. Was your mother/father inconsistent (sometimes warm, sometimes not)?
4. Was your mother/father ever abusive or violent toward you?
5. When did you start to masturbate?
6. Before you were 16 years old, did you ever have sexual contact with an
adult? What were the circumstances of this?

Measures of maternal and paternal childhood attachment were derived by

examining offenders’ answers to questions 1-4 (above). Three raters (the second
author and two graduate psychology students who were familiar with attachment
theory and associated concepts) examined answers separately. The two indepen-
dent raters were blind to the purposes of the study, and to any information other

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

 5

than offenders’ verbatim answers to questions 1-4. Each of the 48 autobiogra-

phies was rated as indicating either secure or insecure attachment. Ratings were
made separately for maternal and paternal attachment. There was complete agree-
ment (r =
1.0) between the three raters on whether the autobiography responses
indicated secure or insecure maternal attachment. There was also very strong
agreement (r =
.95) on whether the autobiography responses indicated secure or
insecure paternal attachment. Data reported below are based on the ratings of the
second author.
Onset of masturbation was measured by simply recording the age at which
offenders indicated their first masturbation experiences (question 5, above). Al-
though varying amounts of qualitative information were available, only the age of
onset was used in this study.
Childhood sexual abuse was measured by examining offenders’ answers to
question 6 (above). Again, although qualitative information was available, only
the presence or absence of childhood sexual abuse was used in this study.

RESULTS

Attachment Insecurity
The first hypothesis predicted that the sexual offenders would be more likely
to report insecure than secure childhood attachment. Almost one third of the of-
fenders (31.3%) had experienced insecure maternal attachment, and 40.5% had
experienced insecure paternal attachment. Thus, as a combined group, the offend-
ers in this study were not more likely to have experienced insecure than secure
maternal or paternal attachment. There were no differences between the extrafa-
milial child molesters, intrafamilial child molesters, and rapists in maternal at-
tachment security. However, the intrafamilial child molesters (58.3%) and rapists
(47.4%) were more likely than were the extrafamilial child molesters (9.1 %) to
report insecure paternal attachment, X (2, N = 42) = 6.46, p < .05.
Because approximately 38% of the general population is likely to have ex-
perienced insecure childhood attachment (Campos, Barrett, Lamb, Goldsmith, &
Stenberg, 1983), the present sample of incarcerated sexual offenders cannot as a
whole be considered to have experienced more childhood attachment insecurity
than the general population.

Attachment Insecurity and Childhood Sexual Abuse

Almost half (45.8%) of the combined group reported sexual contact with
an adult prior to 16 years of age. No differences were found between the three
offender subgroups. For those who had been sexually abused, the average age at
which the abuse occurred was just over 10 years. Using the combined group, two

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

6

chi-square analyses were used to compare the frequency of childhood sexual abuse
among secure maternal (N 33) versus insecure maternal attachment (N = 15),
=

and secure paternal (N = 25) versus insecure paternal attachment (N = 17). No

association was found between insecure maternal attachment and childhood sex-
ual abuse. However, offenders who reported insecure paternal attachment were
significantly more likely to have been sexual abused (61.1 %) than were those with
secure paternal attachment (25.0%), X2 (1, N 42) 4.45, p < .04.
= =

Childhood Sexual Abuse and Onset of Masturbation

The third hypothesis that childhood sexual abuse would be associated with
early onset of masturbation was tested by conducting an independent groups t test.
Presence (N 21 ) or absence (N 26) of childhood sexual abuse was used as the
= =

independent variable, and age of onset of masturbation was used as the dependent
variable. Results showed that those offenders who had been sexually abused had
begun to masturbate significantly earlier (11.0 years) than had those who had not
been sexually abused (13.4 years), t (45) 2.46, p < .02.
=

DISCUSSION
1. &dquo; ,

Contrary to our expectations, the sexual offenders in this study were on the
whole no more likely to have experienced insecure than they were to have ex-

perienced secure childhood attachment. Moreover, the frequency with which the
present sample reported insecure childhood attachment was no greater than would
be expected in the general population (see e.g., Campos, et al., 1983). There were,
however, significant differences between subgroups of offenders, with rapists and
intrafamilial child molesters more likely than extrafamilial child molesters to report
insecure attachment relationships with their fathers. The finding that extrafamilial
child molesters experienced secure maternal and paternal attachment is consistent
with previous findings that this subgroup tends to report positive childhood attach-
ment experiences (Smallbone & Dadds, 1998). It is not clear why this should be the
case, because extrafamilial child molesters are often, as in this study, shown to be
more chronic offenders (Marshall & Barbaree, 1990), and are more likely to have a

diagnosable sexual disorder (Barbaree & Seto, 1997), than rapists and intrafamil-
ial child molesters. Further research is clearly needed to identify developmental
antecedents among extrafamilial child molesters. In particular, measures that are
relatively free from subjective biases may be needed to identify developmental
problems in this subgroup.
The importance of insecure paternal attachment has been suggested by previ-
ous studies, which have found rapists to report more disturbed relationships with

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

 7

their fathers than do nonoffenders, nonsexual offenders, and other types of sexual
offenders (Smallbone & Dadds, 1998), and insecure paternal attachment to pre-
dict coercive sexual behavior in nonoffenders (Smallbone & Dadds, 2000). In the
present study, insecure paternal attachment, but not insecure maternal attachment,
was found to be associated with childhood sexual abuse. Less is known about the

influence of paternal attachment than is known about the influence of maternal

attachment, because most developmental studies have been based on observations
of interactions between infants and their mothers. It may be that rejecting or un-
available fathers fail to provide adequate supervision and protection from potential
abusers. Such fathers may leave their sons more prone to obtaining attention and
affection from other males, thus increasing opportunities for sexual abuse to occur.
In any case, given these repeated findings in relation to paternal attachment, the
quality of father-son relations clearly warrants further theoretical and empirical
attention in sexual offender research.
Almost half of the offenders in this study reported having been sexually
abused. Those who had been sexually abused had begun masturbating on average
more than 21/2 years earlier than had those who had not been sexually abused.

Among the sexually abused group, onset of masturbation was on average less than
1 year after the sexual abuse had occurred. It is thus plausible that images associated
with sexual abuse experiences may be introduced into early masturbatory fantasies.
Through the process of second-order classical conditioning (see e.g., Laws &
Marshall, 1990), these fantasies may be partly responsible for the acquisition of
deviant sexual interests in some boys who are sexually abused.
Taken together, the results of this study suggest that many (but by no means
all) sexual offenders experience insecure childhood attachment, that offenders with
insecure paternal attachment are more likely to have been sexually abused, and that
sexually abused offenders tend to have engaged prematurely in masturbation. This
sequence of childhood and early adolescent experiences, which was predicted from
a combination of established attachment and conditioning theory formulations,

may represent an important developmental pathway by which some sexual offend-

ers become motivated to engage in inappropriate and coercive sexual behavior.

Although the hypotheses tested in this study were theoretically derived, and
were in a number of respects supported by the results, the above conclusions should
be considered tentative. First, the small numbers of offender autobiographies avail-
able for the present investigation provided relatively low statistical power. This was
a more serious problem for examining differences between offender subgroups;

adequate power was available for analyses of differences within the combined
offender group. Second, given that the data were self-reported in a treatment set-
ting, substantial biases are likely to have confounded the results. Collecting data
in a treatment setting will also, of course, limit the extent to which the find-
ings can be generalized. Third, although the offender autobiographies represent a
unique and potentially rich source of data, the measures derived from this source

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

8

lack standardization, and provide only face validity. Finally, the pattern of results
found in this study needs to be replicated before more confident conclusions can
be drawn.
These limitations notwithstanding, this study provides empirical evidence for
positive associations between insecure paternal attachment and childhood sexual
abuse, and between childhood sexual abuse and early onset of masturbation. Be-
cause the testing of developmental models has clear and important implications for

prevention and early intervention, we would hope that this study provides a sound
basis for further theoretical and empirical exploration of developmental factors in
sexual offending. In particular, differences in maternal and paternal influences on
the development of sexual offending warrant further attention.

REFERENCES

Alexander, P. (1992). Application of attachment theory to the study of sexual abuse. Journal of
Consulting and Clinical Psychology, 60, 185-195.
Barbaree, H. E., & Seto, M. C. (1997). Pedophilia: Assessment and treatment. In D. R. Laws, &
W. O’Donohue (Eds.), Sexual deviance: Theory, assessment, and treatment (pp. 175-193). New
York: Guilford.
Bureau of Justice Statistics. (1997). Sex offenses and offenders: An analysis of data on rape and sexual
assault. Washington, DC: U.S. Department of Justice.
Campos, J. J., Barrett, K. C., Lamb, M. E., Goldsmith, H. H., & Stenberg, C. (1983). Socioemotional
development. In M. M. Haith &J. J. Campos (Eds.), Handbook of child psychology: Vol. 2. Infancy
and psychobiology (pp. 783-915). New York: Wiley.
Cohen, M. L., Seghorn, T., & Calmas, W. (1969). Sociometric study of sex offenders. Journal of
Abnormal Psychology, 94, 249-255.
Dawhan, S., & Marshall, W. L. (1996). Sexual abuse histories of sexual offenders. Sexual Abuse: A
Journal of Research and Treatment, 8, 7-15.
Finkelhor, D. (1984). Child sexual abuse: New theory and research. New York: Free Press.
Freund, K., Watson, R., & Dickey, R. (1990). Does sexual abuse in childhood cause pedophilia: An
exploratory study. Archives of Sexual Behavior, 19, 557-568.
Groth, A. N. (1979). Men who rape: The psychology of the offender. New York: Plenum.
Groth, A. N., Burgess, A. W., & Holstrom, L. L. (1977). Rape: Power, anger, and sexuality. American
Journal of Psychiatry, 134, 1239-1243.
Jones, E. E. (1996). Introduction to the special section on attachment and psychopathology: Part 1.
Journal of Consulting and Clinical Psychology, 64, 5-7.
Laws, D. R., & Marshall, W. L. (1990). A conditioning theory of the etiology and maintenance of
deviant sexual preference and behavior. In W. L. Marshall, D. R. Laws, & H. E. Barbaree (Eds.),
Handbook of sexual assault: Issues, theories, and treatment of the offender (pp. 209-229). New
York: Plenum.
Lyons-Ruth, K. (1996). Attachment relationships among children with aggressive behavior problems:
The role of disorganized early attachment patterns. Journal of Consulting and Clinical Psychology,
, 64-73.
64
Malamuth, N. M., Heavey, C. L., & Linz, D. (1993). Predicting men’s antisocial behavior against
women: The interaction model of sexual aggression. In G. C. N. Hall, R. Hirschman, J. R. Graham,
& M. S. Zaragoza (Eds.), Sexual aggression: Issues in etiology, assessment and treatment (pp. 63-
97). Washington, DC: Taylor and Francis.
Marshall, W. L. (1989). Intimacy, loneliness and sexual offenders. Behaviour Research and Therapy,
27,491-503.
Marshall, W. L., & Barbaree, H. E. (1990). An integrated theory of the etiology of sexual offending.
In W. L. Marshall, D. R. Laws, & H. E. Barbaree (Eds.), Handbook of sexual assault: Issues,
theories, and treatment of the offender (pp. 257-275). New York: Plenum.

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016

 9

McGuire, R. J., Carlisle, J. M., & Young, B. G. (1965). Sexual deviations as conditioned behavior: A
hypothesis. Behavior Research and Therapy, 2, 185-190.
Smallbone, S. W., & Dadds, M. R. (1998). Childhood attachment and adult attachment in incarcerated
adult male sex offenders. Journal of Interpersonal Violence, 13, 555-573.
Smallbone, S. W., & Dadds, M. R. (2000). Attachment and coercive sexual behavior. Sexual Abuse: A
Journal of Research and Treatment, 12,
3-15.
Ward, T., Hudson, S. M., & Marshall, W. L. (1996). Attachment style in sex offenders: A preliminary
study. The Journal of Sex Research, 33, 17-26.
Ward, T., Hudson, S. M., Marshall, W. L., & Siegert, R. J. (1995). Attachment style and intimacy
deficits in sexual offenders: A theoretical framework. Sexual Abuse: A Journal of Research and
Treatment, 7, 317-335.

Downloaded from sax.sagepub.com at PENNSYLVANIA STATE UNIV on May 12, 2016