Content uploaded by Kathryn C. Seigfried-Spellar
Author content
All content in this area was uploaded by Kathryn C. Seigfried-Spellar on Jan 31, 2018
Content may be subject to copyright.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/authorsrights
Author's personal copy
Does deviant pornography use follow a Guttman-like progression?
Kathryn C. Seigfried-Spellar
a,
⇑
, Marcus K. Rogers
b
a
The University of Alabama, Tuscaloosa, AL 35487, USA
b
Purdue University, West Lafayette, IN 47907, USA
article info
Article history:
Keywords:
Pornography
Guttman-like progression
Age of onset
Child pornography
Bestiality
Internet crime
abstract
This study investigated whether deviant pornography use followed a Guttman-like progression in that a
person transitions from being a nondeviant to deviant pornography user. In order to observe this progres-
sion, 630 respondents from Survey Sampling International’s (SSI) panel Internet sample completed an
online survey assessing adult-only, bestiality, and child pornography consumption. Respondents’ ‘‘age
of onset’’ for adult pornography use was measured to determine if desensitization occurred in that indi-
viduals who engaged in adult pornography at a younger age were more likely to transition into deviant
pornography use. Two hundred and 54 respondents reported the use of nondeviant adult pornography,
54 reported using animal pornography, and 33 reported using child pornography. The child pornography
users were more likely to consume both adult and animal pornography, rather than just solely consuming
child pornography. Results suggested deviant pornography use followed a Guttman-like progression in
that individuals with a younger ‘‘age of onset’’ for adult pornography use were more likely to engage
in deviant pornography (bestiality or child) compared to those with a later ‘‘age of onset’’. Limitations
and future research suggestions are discussed.
Ó2013 Elsevier Ltd. All rights reserved.
1. Introduction
Research suggests child pornography collections not only con-
tain sexualized images of children, but other genres of pornography
both deviant and socially acceptable in nature (c.f., Quayle & Taylor,
2002, 2003). In fact, interviews with child pornography consumers
have suggested some offenders move ‘‘thorough a variety of pornog-
raphies, each time accessing more extreme material’’ (Quayle &
Taylor, 2002, p. 343) as a result of desensitization or appetite satia-
tion, which led to collecting and discovering other forms of deviant
pornography (Quayle & Taylor, 2003). Also, some consumers stated
they downloaded the images simply because they were available
and accessible, making the behaviors primarily a result of compul-
sivity rather than a specific sexual interest in children (Basbaum,
2010). However, prior analyses rely on case studies of convicted
child sex offenders and child pornography users. If a more broadly
based representative sample (as utilized here) were employed, then
researchers may have a more congruent and complete understand-
ing of the collections of child pornography users.
Some child pornography consumers exhibit a complex array of
sexual interests, which may be representative of a more general le-
vel of paraphilic tendencies rather than a specific sexual interest in
children (c.f., Wolak, Finkelhor, Mitchell, & Ybarra, 2008). In a
study conducted by Endrass et al. (2009), the collection of images
from 231 men charged with child pornography use also revealed
other types of deviant pornography. Specifically, nearly 60% of
the sample collected child pornography and at least one other type
of deviant pornography, such as bestiality, excrement, or sadism,
with at least one out of three offenders collecting three or more
types of deviant pornography (Endrass et al., 2009). This research
suggests the majority of Internet child pornography consumers
are collecting a wider range of deviant pornography, which may re-
flect a general level of sexual deviance rather than a specific para-
philia, such as pedophilia. In other words, some child pornography
consumers may be dissidents within the normal population who
exhibit a wider range of sexual interests or curiosity.
Although case studies exist, few quantitative research studies
have assessed the question of whether individuals who use nonde-
viant forms of pornography (e.g., adult pornography) are at a great-
er risk for consuming deviant forms of pornography (e.g., animal
and child pornography). In other words, does deviant pornography
use follow a Guttman-like progression (c.f., Hollinger, 1988) with
age of onset being a key factor in whether a person transitions from
being a nondeviant to deviant pornography user? Regarding age of
onset, the majority of research focuses on the emotional conse-
quences of unwanted exposure to pornography at a young age
(c.f., Flood, 2009). For example, Mitchell, Wolak, and Finkelhor
(2007) found 10% of 10–17 year olds described themselves as being
‘‘very or extremely upset’’ by unwanted exposure to pornography.
On the other hand, McKee (2007) interviewed 46 Australians,
0747-5632/$ - see front matter Ó2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.chb.2013.04.018
⇑
Corresponding author. Address: 425 Farrah Hall, Box 870320 (mailing),
Tuscaloosa, AL 35487, USA. Tel.: +1 205 348 5489 (O).
E-mail address: kseigspell@as.ua.edu (K.C. Seigfried-Spellar).
Computers in Human Behavior 29 (2013) 1997–2003
Contents lists available at SciVerse ScienceDirect
Computers in Human Behavior
journal homepage: www.elsevier.com/locate/comphumbeh
Author's personal copy
regarding their exposure to pornography at a young age, who de-
scribed their pre-pubescent exposure to pornography as ‘‘funny’’
and with ‘‘little interest’’ whereas their post-pubescent exposure
was a ‘‘right of passage’’ (p. 10). In addition, research has suggested
a relationship between pornography use at a young age and vari-
ous sexual behaviors. Specifically, Johansson & Hammaré, 2007
found young pornography users were more likely to have had sex-
ual intercourse and a one-night stand, and young consumers of
violent pornography were more likely to exhibit sexually aggres-
sive attitudes and behaviors (c.f., Flood, 2009).
Overall, previous research has mainly focused on the emotional
impact of unwanted exposure to pornography for young people.
The current study focused on the ‘‘age of onset’’ for intentional
use, rather than unwanted exposure, of nondeviant and deviant
pornography. Since the current study sampled respondents from
the United States, definitions of nondeviant and deviant pornogra-
phy were based on the current obscenity laws in the United States.
In the United States, adult pornography is protected by the First
Amendment (although there are exceptions); however, child por-
nography and animal pornography (bestiality) are obscene, there-
fore, illegal forms of expression. Thus, adult pornography was
operationalized as nondeviant, whereas, animal and child pornog-
raphy were labeled as deviant forms of pornography.
Despite the formal social controls (laws) regulating pornogra-
phy use, all three genres of pornography remain readily available
on the Internet. Therefore, this study explored at what age individ-
uals first knowingly searched for, downloaded, and exchanged/
shared the following pornography genres: adult-only, animal (bes-
tiality), and child pornography. By examining the interrelations
among the self-reported age and pornography use variables, the
authors hoped to understand how nondeviant pornography use
either facilitated or predicted the probability of engaging in more
deviant forms of pornography.
Three primary objectives were the focus of the current study. The
first aim of this study was to determine whether or not the age of on-
set was a risk factor for engaging in deviant pornography. In other
words, are individuals who engage in nondeviant pornography use
at an earlier age more likely to engage in deviant forms of pornogra-
phy use compared to late onset users? The second aim of this study
determined whether female respondents were consuming Internet
child pornography. Previous research suggests the majority of child
pornography users are male; however, the majority of these samples
are from the forensic or clinical populations (c.f., Babchishin,
Hanson, & Hermann, 2011). In addition, Internet-based research
studies suggest women may be engaging in child pornography more
than previously expected (c.f., Seigfried, Lovely, & Rogers, 2008;
Seigfried-Spellar & Rogers, 2010). Thus, the current study specifi-
cally assessed the prevalence of female child pornography use in a
sample of Internet users rather than a forensic or clinical sample,
in order to provide a broader conceptualization of female users of
child pornography (non-convicted and self-reported).
Finally, the third aim of this study explored the frequency of por-
nography use by collapsing the respondents into pornography cate-
gories: none, adult-only, animal-only, child-only, adult-animal,
adult-child, animal-child, and adult-child-animal. This methodolog-
ical analysis allowed assessment of whether self-reported child por-
nography users were more likely to self-report adult and animal
pornography behaviors compared to the other categories of users.
Few research studies have specifically assessed the variety of genres
collected by Internet child pornography users (c.f., Seigfried-Spellar,
in press). Specifically, if child pornography use followed a Guttman-
like progression then there should be no ‘‘exclusive consumers’’ of
only child pornography; instead, child pornography users should re-
portengaginginotherformsof deviant and nondeviant pornography.
This study was exploratory in nature since no previous research
has assessed whether individuals who reported a younger ‘‘age of
onset’’ for adult pornography use were more likely to engage in
deviant pornography use compared to individuals who reported
a later ‘‘age of onset’’. The expectation is to find no relationship be-
tween ‘‘age of onset’’ for adult pornography and later deviant por-
nography use. However, the modest amount of research on child
pornography use indicates child pornography collections include
both deviant and nondeviant pornographic images. Therefore, it
is hypothesized child pornography consumers will be more likely
to consume adult-only and bestiality pornography and less likely
to be sole consumers of child pornography. Finally, the authors ex-
pect to find a sex difference; specifically, men will be more likely to
self-report the use of child pornography (e.g., Babchishin et al.,
2011). Uniquely, there will be a higher prevalence of female child
pornography use in this Internet-based research study due to the
difference in sampling methodology.
2. Methods
2.1. Participants
The current study utilized Survey Sampling International (SSI),
which provided a panel Internet sample of both male and female
respondents, who were at least 18 years of age or older, from the
United States. Rather than snowballing the Internet in order to
identify respondents, these clients or respondents have already
gone through the SSI’s quality control and verification system in
order to identify individuals who are at risk of lying on a survey
just to qualify or claim any rewards or incentives (Survey Sampling
International, 2009). In addition, SSI prevents the same person
from being able to take the survey multiple times (Survey Sam-
pling International, 2009). Most importantly, these clients or
respondents were more likely to be confident in the reliability
and confidentiality of this study, as well as comfortable and trust-
ing in the research process itself, which is essential when examin-
ing attitudes and behaviors toward socially sensitive topics.
Based on the desire to sample respondents from the ‘‘general
population of Internet pornography users,’’ rather than a sample
from the clinical or forensic population, and the need to increase
the respondent’s confidence in self-disclosure, this sampling meth-
odology best met the needs of the current study. As shown in Table 1
and 630 respondents completed the online survey; 502 (80%) were
women and 128 (20%) were men (Note: This gender disparity will
be discussed later in the paper). Overall, the majority of the sample
was white (n= 519, 82.4%), between the ages of 36–55 years
(n= 435, 69%), married (n= 422, 67%), and 68% (n= 427) of the
respondents had completed some college or post-graduate work.
2.2. Measures
The respondent’s Internet pornography behavior and age of on-
set were measured using a shorted version of the Online Pornogra-
phy Survey (OPS; Seigfried, 2007;Seigfried-Spellar, 2011). The
original OPS included 54 questions, which assessed the respon-
dents’ pornography behaviors including intentional searching,
accessing, downloading, and exchanging of sexually explicit Inter-
net images. Adult pornography was defined as pornographic
images ‘‘featuring individuals over the age of 18 years’’, whereas
child pornography was defined as pornographic materials ‘‘featur-
ing individuals under the age of 18 years’’. Animal pornography or
bestiality was defined as pornographic images ‘‘featuring individu-
als over the age of 18 years with an animal’’.
Only 15 items from the Online Pornography Survey, which fo-
cused on the respondent’s age of onset for online pornography
use, were included in this study. All 15 questions employed the
same answer format. The following is an illustrative sample
1998 K.C. Seigfried-Spellar, M.K. Rogers / Computers in Human Behavior 29 (2013) 1997–2003
Author's personal copy
question related to age of onset from the OPS: ‘‘How old were you
the first time you knowingly accessed a website in order to view
pornographic materials featuring individuals under the age of
18 years?’’ The respondents’ choices for age of onset items were:
does not apply to me, under 12 years of age, 12 to under 16 years
of age, 16 to under 19 years of age, 19 to under 24 years of age,
24 years of age or older, and decline to respond. Based on item
endorsement, the respondents were classified as either users or
non-users of adult, animal (bestiality), and child pornography.
Finally, the respondents’ basic demographic information was
self-reported via an online questionnaire, which included items such
as sex, age, and marital status. The demographics survey appeared at
the beginning of the study for all of the respondents. The current
study advertised as assessing ‘‘attitudes toward adult websites’’,
and by placing the demographics questionnaire prior to the more so-
cially sensitive questions regarding pornography use, this method
increased the accuracy of self-reported sex for this study
(c.f., Birnbaum, 2000). Also, all survey items were forced-choice,
but the respondents were able to select ‘‘decline to respond’’ to
any item, as required by the Institutional Review Board (IRB).
Further, all respondents were treated in accordance with the ethical
standards set forth by the American Psychological Association (APA).
2.3. Procedure
This study was conducted electronically using an Internet-
based survey. This method of conducting research via the Internet
has seen increasing used by researchers due to the accessibility of
respondents and the perceived anonymity and increased willing-
ness to self-disclose socially unacceptable or controversial behav-
iors or attitudes (Mueller et al., 2000). Once the respondents
accessed the website, the home page explained the study while
acting as a consent form to which the respondents had to agree
or decline to participate. If the prospective respondents agreed,
they had to click on the ‘‘I Agree’’ button in order to participate.
After clicking on the ‘‘I Agree’’ button, the respondents were asked
to complete the questionnaires, which took approximately 15 min
to complete.
At no time were the respondents asked for any identifying
information (e.g., name). In order to protect the respondent’s
Table 1
Demographic information.
Variable Male Female Total
(n= 128) (n= 502) (N= 630)
Age (years) 18–25 4 (3.1) 26 (5.2) 30 (4.8)
26–35 13 (10.2) 80 (15.9) 93 (14.8)
36–45 29 (22.7) 186 (37.0) 215 (34.1)
46–55 47 (36.7) 173 (34.5) 220 (34.9)
56 or older 33 (25.8) 33 (6.6) 66 (10.5)
Decline 2 (1.5) 4 (0.8) 6 (0.9)
Income Less than $15,000 4 (3.1) 43 (8.5) 47 (7.5)
$15,000 – under $25,000 8 (6.3) 50 (9.9) 58 (9.2)
$25,000 – under $35,000 17 (13.3) 59 (11.8) 76 (12.1)
$35,000 – under $50,000 14 (10.9) 89 (17.7) 103 (16.3)
$50,000 – under $75,000 33 (25.8) 114 (22.7) 147 (23.3)
$75,000 – under $100,000 15 (11.7) 55 (11.0) 70 (11.1)
$100,000 or more 29 (22.7) 55 (11.0) 84 (13.3)
Decline 8 (6.2) 37 (7.4) 45 (7.2)
Ethnicity Caucasian/white 109 (85.2) 410 (81.7) 519 (82.4)
Black 2 (1.6) 31 (6.2) 33 (5.2)
Hispanic 4 (3.1) 24 (4.8) 28 (4.4)
Biracial 2 (1.6) 16 (3.2) 18 (2.9)
Asian/pacific islander 4 (3.1) 9 (1.8) 13 (2.1)
Native American 3 (2.3) 3 (0.6) 6 (0.9)
Other 1 (0.8) 2 (0.4) 3 (0.5)
Decline 3 (2.3) 7 (1.3) 10 (1.6)
Religion Protestant 32 (25.0) 90 (17.9) 122 (19.4)
Roman catholic 24 (18.8) 100 (19.9) 124 (19.7)
ND Christian 20 (15.6) 130 (25.9) 150 (23.8)
Jewish 5 (3.9) 11 (2.2) 16 (2.5)
Muslim 1 (0.8) 0 (0) 1 (0.1)
Other 14 (10.9) 85 (16.9) 99 (15.7)
None 24 (18.8) 71 (14.1) 95 (15.1)
Decline 8 (6.2) 15 (3.1) 23 (3.7)
Marital status Single 11 (8.6) 29 (5.8) 40 (6.3)
Married 90 (70.3) 332 (66.1) 422 (67.0)
Sig other 8 (6.2) 37 (7.4) 45 (7.1)
S, D, W 18 (14.1) 103 (20.5) 121 (19.2)
Decline 1 (0.8) 1 (0.2) 2 (0.4)
Highest completed education level <12 years of H.S. 1 (8) 10 (2.0) 11 (1.8)
H.S. diploma, GED 13 (10.2) 112 (22.3) 125 (19.8)
Vocational/training 11 (8.6) 52 (10.4) 63 (10.0)
Some college 47 (36.7) 191 (38.0) 238 (37.8)
Bachelors 37 (28.9) 102 (20.3) 139 (22.1)
Post-graduate Degree 19 (14.8) 31 (6.2) 50 (7.9)
Decline 0 (0) 4 (0.8) 4 (0.6)
Values represent frequencies with percentages in parentheses.
ND Christian = non-denomination Christian, sig other = living with a partner or significant other, S = separated, D = divorced, W = widowed, H.S. = high school, GED = high
school equivalency diploma, decline = decline to respond.
K.C. Seigfried-Spellar, M.K. Rogers / Computers in Human Behavior 29 (2013) 1997–2003 1999
Author's personal copy
anonymity and confidentiality, the respondents were provided
with an ID number so responses to the questionnaires could not
be linked or matched to any particular individual.
2.4. Statistical analyses
After data collection, statistical analyses were conducted using
the Statistical Package for the Social Sciences (SPSSs) version 19. Sta-
tistical significance was set at the alpha level of .05 prior to any anal-
yses. The Fisher–Freeman–Halton exact test tested for significant
relationships between age of onset, sex, and pornography type. This
decision was made for the following reasons: expected cell fre-
quency counts were small due to the study assessing rare occur-
rences (i.e., child pornography use), it approximates the chi-square
test as sample size (N) increases, and the Fisher–Freeman–Halton
Exact Test extends the Fisher’s exact test to the RxCcase (c.f., Free-
man & Halton, 1951). Finally, a backward stepwise (Wald) logistic
regression was conducted in order to determine if sex and ‘‘age of
onset’’ for adult pornography use predicted group membership for
nondeviant versus deviant Internet pornography use. Logistic
regressions are appropriate for exploratory analyses, for they are
more robust with fewer violations of assumptions, such as small
and unequal sample sizes (Tabachnick & Fidell, 2007).
3. Results
As shown in Table 2, 5.2% (n= 33) of the respondents self-re-
ported the use of Internet child pornography. 16 (12.5%) of the
male respondents were child pornography users, and 17 (3.4%) of
the female respondents were child pornography users. Of the 630
respondents, only 8.6% (n= 54) of the respondents self-reported
the use of bestiality pornography, yet nearly half (n= 254, 40.3%)
of the respondents reported the use of adult-only pornography.
As shown in Table 3, the respondents were further categorized
based on their use of adult-only, bestiality, and child pornography.
In support of the study’s premise, no respondents reported the
sole use of child pornography. Only 1 female respondent reported
only consuming bestiality pornography. In addition, 9.8% (n= 60)
of the respondents consumed some mixture of nondeviant and
deviant pornography compared to only .5% who reported consum-
ing only deviant pornography (bestiality and child).
Since the descriptive data suggested there was a relationship
between adult, animal, and child pornography use (see Table 3),
a zero-order correlation was conducted to determine the direction
of the relationship. Based on item responses, a dichotomous vari-
able was created for each pornography category: adult, animal,
and child. The respondents were coded as either non-users (0) or
users (1) for each category of pornography. As shown in Table 4,
there was a statistically significant relationship between adult por-
nography and bestiality use, r
/
(635) = .36 with p< .01, and adult
pornography and child pornography use, r
/
(635) = .27 with
p< .01. There was a significant positive relationship for individuals
who self-reported engaging in adult pornography, animal/bestial-
ity, and child pornography. In addition, men were significantly
more likely to self-report the use of adult, r
/
(630) = .28 with
p< .01, animal/bestiality, r
/
(630) = .18 with p< .01, and child
pornography, r
/
(630) = .17 with p< .01 (see Table 4).
Next, respondents were categorized as either: adult-only
(adult-only) or adult and child/animal (adult + deviant) pornogra-
phy users. The ‘‘age of onset’’ was then compared between the
two groups to determine if ‘‘age of onset’’ for adult pornography
use was related to later use of deviant pornography. Based on
the Fisher–Freeman–Halton exact test (p< .01), adult + deviant
pornography users reported a significantly younger ‘‘age of onset’’
compared to adult-only pornography users. As shown in Table 5
and 29% of the adult + deviant pornography users reported an
‘‘age of onset’’ between 12 and 18 years of age compared to only
10% of the adult-only respondents. Instead, the majority (89%) of
the adult-only pornography users reported an age of onset of
19 years of age or older compared to 69% for the adult + deviant
pornography users (see Table 5).
Based on the significant findings from the zero-order correlations
and Fisher–Freeman–Halton exact test, the authors conducted a
backward stepwise (Wald) logistic regression to determine if ‘‘age
of onset’’ and sex were significant predictors of adult-only versus
adult + deviant pornography use. As shown in Table 6, the best pre-
dictive model for adult-only versus adult + deviant pornography use
included both variables, sex (W= 7.69, p< .01) and age of onset
(W= 5.16, p< .02). Individuals with a younger ‘‘age of onset’’ for
adult pornography use were .8 times more likely to engage in devi-
ant pornography. In addition, men were .4 times more likely to be a
deviant pornography user. The Hosmer and Lemeshow test was non-
significant,
v
2
(4) = 6.42 with p= .17, indicating the final model fit
the data. In addition, variance inflation factors (VIF) and condition
index values were calculated in order to test for multicollinearity,
all of which indicated no cause for concern (sex, VIF = 1.00; age of
onset, VIF = 1.00; condition index < 30).
Based on these analyses, the authors were able to achieve their
aims of determining if ‘‘age of onset’’ and sex significantly pre-
dicted adult-only versus adult + deviant pornography users. Over-
all, the hypothesized expectation that child pornography users
would be more likely to consume both adult and animal pornogra-
phy, rather than just solely consuming child pornography, was
Table 2
Percentage of non-deviant and deviant pornography use by sex.
Porn type Sex Total (N= 630)
Male 128 (20) Female 502 (80)
Non-deviant
Adult-only 86 (67.2) 168 (33.5) 254 (40.3)
Deviant
Animal 24 (18.8) 30 (6.0) 54 (8.6)
Child 16 (12.5) 17 (3.4) 33 (5.2)
Note: Values represent frequencies with percentages in parentheses.
Table 3
Classification of respondents by self-reported use of adult, animal, and child
pornography.
Porn user Sex Total
Male Female
None 37 (30.6) 326 (66.5) 363 (59.4)
Adult-only 55 (45.5) 132 (26.9) 187 (30.6)
Animal-only 0 1 (0.2) 1 (0.2)
Child-only 0 0 0
Adult and animal 13 (10.7) 15 (3.1) 28 (4.6)
Adult and child 6 (5.0) 5 (1.0) 11 (1.8)
Animal and child 1 (0.8) 1 (0.2) 2 (0.3)
Adult, animal, and child 9 (7.4) 10 (2.0) 19 (3.1)
Note: Due to missing values, N= 611.
Values represent frequencies with percentages in parentheses.
Table 4
Zero-order correlation for sex, adult, animal, and child pornography use.
Sex Type of Pornography
Adult Animal Child
Sex 1.0 0.28
*
0.18
*
0.17
*
Adult 1.0 0.36
**
0.27
**
Animal 1.0 0.44
**
Child 1.0
*
p< .01 (1-tailed).
**
p<.01 (2-tailed).
2000 K.C. Seigfried-Spellar, M.K. Rogers / Computers in Human Behavior 29 (2013) 1997–2003
Author's personal copy
supported. In addition, the postulation that men were more likely
to engage in child pornography use was supported as well as the
expectation of a higher prevalence of female child pornography
use in this Internet-based sample.
However, the authors’ expectation of no difference between the
‘‘age of onset’’ for adult pornography use between adult-only and
adult + deviant pornography users was not supported. Based on
the Fisher–Freeman–Halton exact test and logistic regression,
adult + deviant pornography users reported a significantly younger
‘‘age of onset’’ for adult pornography use compared to the adult-only
pornography users. In other words, deviant pornography users en-
gaged in adult pornography at a significantly younger age compared
to those who engaged in only nondeviant pornography.
4. Discussion
The current study was the first to assess whether ‘‘age of onset’’
for nondeviant pornography use (i.e., adult-only) was related to la-
ter use of deviant pornography (i.e., bestiality, child) using a large
Internet-based sample. This study represents an improvement over
prior case studies, which rely on samples of convicted offenders. As
such, the current study moved away from the clinical or forensic
population of child pornography users to child pornography user
from the ‘‘general population of Internet users.’’ In addition, this
study assessed whether child pornography users collected both
deviant and nondeviant pornography or whether they self-re-
ported only consuming child pornography. Overall, significant dif-
ferences emerged between nondeviant and deviant pornography
users for ‘‘age of onset’’ and sex.
A small body of research suggests the majority of Internet child
pornography users are collecting a wider range of deviant pornog-
raphy (c.f., Endrass et al., 2009). In the current study, none of the
respondents self-reported the sole consumption of Internet child
pornography. Instead, the majority of child pornography users
were also collecting other forms of pornography including nonde-
viant adult pornography and bestiality pornography. Of the 32
child pornography consumers, 60% (n= 19) also collected both
nondeviant adult and animal pornography, 34% (n= 11) consumed
just nondeviant adult pornography, and only 6% (n= 2) had just
animal pornography (see Table 3). These findings support the
Seigfried (2007) study, which observed no sole consumers of Inter-
net child pornography. Overall, child pornography users are engag-
ing in a wide range of sexual content and future research should
assess whether these collections provide information as to their
offline intentions (e.g., hands-on contact offending) as well as per-
sonality characteristics (e.g., violent individuals collect violent por-
nography; Rogers & Seigfried-Spellar, 2012; Seigfried-Spellar, in
press).
Consistent with previous research, men continue to be more
likely to engage in Internet child pornography use. However, the
current study suggests women may be consuming child pornogra-
phy more than previously suggested by research samples from the
clinical for forensic population. For example, Babchishin et al.
(2011) conducted a meta-analysis of 27 articles, which included
samples of online offenders. The results of the meta-analysis sug-
gest the majority of the child pornography offenders are male,
and of the 27 articles, only five studies include female offenders.
Thus, less than 3% of the entire sample of online offenders was wo-
men (Babchishin et al., 2011). However, previous research includ-
ing samples from the general population of Internet users, rather
than the clinical or forensic population, has reported higher per-
centages of female consumers of child pornography. For example,
the Seigfried et al. (2008) study found 10 of the 30 self-reported
child pornography users from an Internet-based research study
to be women. In addition, the Seigfried-Spellar (2011) study re-
ported 20% of the self-reported child pornography users were wo-
men. Finally, 17 of the 33 (52%) child pornography consumers were
women in the current study. Future research should assess why
there is a difference in the prevalence of child pornography use
for women from different sampling populations.
Along with the variable sex, ‘‘age of onset’’ was significantly re-
lated to deviant pornography use. Respondents who reported a
younger ‘‘age of onset’’ for nondeviant pornography use were more
likely to engage in deviant pornography use compared to those
individuals who reported a later ‘‘age of onset’’. As shown in
Table 5, the adult + deviant pornography users were twice as likely
to self-report an ‘‘age of onset’’ between 12 and 18 years of age
compared to the adult-only pornography users. Finally, the logistic
regression suggested the best predictive model for deviant pornog-
raphy use included variables, sex and ‘‘age of onset’’. That is to say,
men were significantly more likely to engage in deviant pornogra-
phy compared to women. In addition, individuals who started
engaging in adult pornography use at a young age were more likely
to use deviant pornography compared to those who engaged in
adult pornography at a later age.
The findingsof the current study suggestInternet pornography use
may follow a Guttman-like progression. In other words, individuals
who consume child pornography also consume other forms of por-
nography, both nondeviant and deviant. For this relationship to be a
Guttman-like progression,child pornographyuse must be more likely
to occur after other forms of pornography use. The current study at-
tempted to assess this progression by measuring if the ‘‘age of onset’’
for adult pornography use facilitatedthe transition from adult-only to
deviant pornography use. Based on the results, this progression to
deviant pornography use may be affected by the individuals ‘‘age of
onset’’ for engaging in adult pornography. As suggested by Quayle
and Taylor (2003), child pornography use may be related to desensi-
tizationor appetite satiationto which offendersbegin collecting more
extreme and deviant pornography. The current study suggests indi-
viduals who engage in adult pornography use at a younger age may
be at greater risk for engaging in other deviantforms of pornography.
If child pornography use follows a Guttman-like progression, then fu-
ture research should assess the relationship between age of onset for
nondeviant pornography and future appetite satiation leading to
other deviant forms of pornography.
4.1. Limitations
Although this study sampled from the ‘‘general population of
Internet users’’, there is no claim that the findings are representative
Table 5
Adult-only versus adult and deviant pornography use by age of onset.
Age of onset Type of porn user
Adult-only Adult and deviant
w= 176 n=59
<12 0 (0%) 1 (1.7%)
12 < 16 8 (4.5%) 8 (13.6%)
16 < 19 11 (6.3%) 9 (15.3%)
19 < 24 28 (15.9%) 15 (25.4%)
24 < 129 (73.3%) 26 (44.1%)
r=.135.
p< .01.
Table 6
Exploratory backward (Wald) logistic regression for pornography use.
Variables BSEB Exp (B)
Step l
Sex 0.86 0.31 0.42
**
Age of onset 0.26 0.12 0.77
*
*
p< .05.
**
p<.01.
K.C. Seigfried-Spellar, M.K. Rogers / Computers in Human Behavior 29 (2013) 1997–2003 2001
Author's personal copy
of the entire population of Internet users. While sampling respon-
dents from the same country (United States) limits external validity,
the authors were able to increase control over certain confounds,
such as the legality of child pornography and animal pornography
use. The current methodology targets Internet users who were living
in a country where child pornography and animal pornography are
illegal. For example, the self-reported Internet child pornography
users in the current study were engaging in illegal child pornogra-
phy behaviors, and legality of child pornography use could be a con-
found if individuals are sampled from countries where child
pornography use is legal (e.g., Russia, Japan, Thailand; see Interna-
tional Centre for Missing, 2010).
Also, sex representation was disproportionate in the current
study. According to the United States Census Bureau (2009a),
50.7% of the United States population was women. When consider-
ing only those individuals who had Internet access either inside or
outside their household (N= 197,871), 48.6% were women (United
States Census Bureau, 2009b). Based on the current panel demo-
graphics for Survey Sampling International (personal communica-
tion, 2012), 56% of the United States Internet panel is women. It is
possible the sex disparity in this study was related to the respon-
dents’ employment status. In the current study, men were signifi-
cantly more likely to be employed full-time and part-time whereas
the women were more likely to be homemakers,
v
2
(9) = 73.82,
p< .00. Previous research cites respondents who are employed
full-time and are ‘‘busy’’ are less likely to complete online surveys
(Cavallaro, 2012). So, the sex disparity may have been due to
employment status in that the female respondents who were
homemakers had more time to complete the online survey. When
controlling for employment status, there was a still a significant
relationship between ‘‘age of onset’’ and adult-only versus
adult + deviant pornography use, r
ab +c
=.28, p< .01.
Although, the proportion of women to men in the current study
was not representative of the United States Internet population, it
did sample individuals outside of the clinical or forensic popula-
tion. In addition, the current study suggests this methodology
may reveal more women who are consumers of Internet child por-
nography compared to other research designs (i.e., clinical or
forensic population; Seigfried-Spellar & Rogers, 2010).
Although there was a sex disparity in the current study, the
relationship between adult-only versus adult + deviant pornogra-
phy use and ‘‘age of onset’’ was still significant when controlling
for sex, r
ab +c
=.30 with p< .01. When only assessing the male
respondents, men who engaged in adult + deviant pornography re-
ported a significantly younger ‘‘age of onset’’ for adult pornography
use compared to the men who engaged in adult-only pornography,
Fisher–Freeman–Halton exact test = 15.79 with p< .01. When only
assessing the female respondents, women who engaged in
adult + deviant pornography also reported a significantly younger
‘‘age of onset’’ for adult pornography use compared to the women
who engaged in adult-only pornography, Fisher–Freeman–Halton
exact test = 7.36 with p< .05.
Finally, a recent study using the same Internet-based research
design but with a snowball sample of Internet respondents repli-
cated the findings of this study in that individuals who self-re-
ported a younger age of onset for adult pornography use were
more likely to engage in deviant pornography (Seigfried-Spellar,
2013).
5. Conclusion
There is a debate in the literature regarding the effects of un-
wanted exposure to pornography by young children; however,
few studies assess the age of intentional use of nondeviant and
deviant pornography. Despite attempts of monitoring, filtering,
or deleting images or websites on the Internet, nondeviant and
deviant of pornography will continue to be accessible, affordable,
and anonymous (c.f., Rogers & Seigfried-Spellar, 2011; Seigfried-
Spellar, Bertoline, & Rogers, 2012). Growth in the number of devi-
ant pornography users (i.e., child pornography) will only increase
as the current 2.45 billion of the world’s population (35%) with
Internet access continues to increase (International Telecommuni-
cation Union., 2011). This growth will only add importance to
understanding ‘‘why’’ some people view, download and exchange
deviant pornography when others do not. This exploratory study
suggests ‘‘age of onset’’ for nondeviant pornography use is related
to later deviant pornography use. In addition, women are engag-
ing in child pornography, but men are still more likely to be con-
sumers of child pornography. As suggested by Quayle and Taylor
(2003), desensitization may put an individual at risk for progres-
sion from nondeviant to deviant pornography behaviors. Future
research should assess whether individual differences (e.g., open-
ness to experience, consciousness, extraversion, agreeableness,
and neuroticism; see Seigfried-Spellar & Rogers, submitted for
publication) are related to this Guttman-like progression for devi-
ant (i.e., child) pornography use.
References
Babchishin, K. M., Hanson, R. K., & Hermann, C. A. (2011). The characteristics of
online sex offenders: A meta analysis. Sexual Abuse: A Journal of Research and
Treatment, 23(1), 92–123.
Basbaum, J. P. (2010). Sentencing for possession of child pornography: A failure to
distinguish voyeurs from pederasts. Hastings Law Journal, 61, 1–24.
Birnbaum, M. H. (Ed.). (2000). Psychological experiments on the Internet. San Diego,
CA: Academic Press.
Cavallaro, K. (2012). Revealing the answer to life’s most difficult market research
question: Why don’t people join panels? Retrieved from Survey Sampling
International’s website <www.surveysampling.com>.
Endrass, J., Urbaniok, F., Hammermeister, L. C., Benz, C., Elbert, T., Laubacher, A.,
et al. (2009). The consumption of Internet child pornography and violent and
sex offending. BMC Psychiatry, 9(43), 1–7.
Flood, M. (2009). The harms of pornography exposure among children and young
people. Child Abuse Review, 18, 384–400.
Freeman, G. H., & Halton, J. H. (1951). Note on exact treatment of contingency,
goodness of fit and other problems of significance. Biometrika, 38, 141–149.
Hollinger, R. C. (1988). Computer hackers follow a Guttman-like progression.
Sociology and Social Research, 72(3), 199–200.
International Centre for Missing & Exploited Children (2010). Child pornography:
Model legislation and global review (6th ed.). <www.icmec.org>.
International Telecommunication Union. (2011). The world in 2011: Facts and figures.
<www.itu.int/ict>.
Johansson, T., & Hammaré, N. (2007). Hegemonic masculinity and pornography:
Young people’s attitudes toward and relations to pornography. Journal of Men’s
Stutides, 15, 57–70.
McKee, A. (2007). Saying you’ve been at dad’s porn books I apart of growing up:
youth, pornography and education. Metro Magazine, 155, 112–118.
Mitchell, K., Wolak, J., & Finkelhor, D. (2007). Trends in youth reports of sexual
solicitations, harassment, and unwanted exposure to pornography on the
Internet. Journal of Adolescent Health, 410, 116–126.
Mueller, J., Jacobsen, D., & Schwarzer, R. (2000). What are computing experiences
good for?: A casestudy in online research. In M. H. Birnbaum (Ed.), Psychological
experiments on the internet (pp. 195–216). New York, NY: Academic Press.
Quayle, E., & Taylor, M. (2002). Child pornography and the Internet: Perpetuating a
cycle of abuse. Deviant Behavior: An Interdisciplinary Journal, 23, 331–361.
Quayle, E., & Taylor, M. (2003). Model of problematic Internet use in people with a
sexual interest in children. Cyberpsychology and Behavior, 6(1), 93–106.
Rogers, M. K. & Seigfried-Spellar, K. C. (2012, February). Applied predictive behavioral
modeling: The role of behavioral sciences in digital forensics. Presentation at the
American academy of forensic sciences 64th annual scientific meeting, Atlanta,
GA.
Rogers, M., & Seigfried-Spellar, K. (2011). Internet child pornography: Legal issues
and investigative tactics. In T. Holt (Ed.), Crime On-Line: Correlates, Causes, and
Context. Durham, NC: Carolina Academic Press.
Seigfried, K. (2007). Self-reported online child pornography behavior: A
psychological analysis. Unpublished master’s thesis, John Jay College of
Criminal Justice, New York.
Seigfried, K., Lovely, R., & Rogers, M. (2008). Self-reported consumers of Internet
child pornography: A psychological analysis. International Journal of Cyber
Criminology, 2(1), 286–297.
Seigfried-Spellar, K. C. (2011). The role of individual differences in predicting the
type of images collected by internet child pornography consumers. Unpublished
Dissertation. Purdue University, West Lafayette, IN.
2002 K.C. Seigfried-Spellar, M.K. Rogers / Computers in Human Behavior 29 (2013) 1997–2003
Author's personal copy
Seigfried-Spellar, K. C. (2013, February). Replicating the Seigfried-Spellar and Rogers
(2011) study on deviant pornography use by age of onset and sex. Presentation at
the American academy of forensic sciences 65th annual scientific meeting.
Washington, D.C.
Seigfried-Spellar, K. C. (in press). Measuring the preference of image content for
self-reported consumers of child pornography. In M. Rogers & K. Seigfried-
Spellar (Eds.), Lecture notes of the institute for computer sciences, social informatics
and telecommunications engineering: Digital forensics and computer crime. New
York: Springer.
Seigfried-Spellar, K. C. & Rogers, M. K. (submitted for publication). Discriminating
self-reported Internet child pornography users by individual differences.
Seigfried-Spellar, K. C., Bertoline, G. R., & Rogers, M. K. (2012). Internet child
pornography, US sentencing guidelines, and the role of Internet service
providers. In P. Gladyshev & M. Rogers (Eds.), Lecture notes of the institute
for computer sciences, social-informatics and telecommunications engineering:
vol. 88. Digital forensics and cyber crime (pp. 17–32).
Seigfried-Spellar, K., & Rogers, M. (2010). Low neuroticism and high hedonistic
traits for female Internet child pornography consumers. Cyberpsychology,
Behavior, and Social Networking, 13(6), 629–635.
Survey Sampling International (2009). Verification in the online World. Retrieved from
Survey Sampling International’s Web site. <http://www.surveysampling.com>.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics (5 ed.). Boston,
MA: Pearson Education, Inc.
United States Census Bureau (2009a). State and county quick facts. Retrieved from
United States Census Bureau’s website: <http://www.quickfacts.census.gov>.
United States Census Bureau (2009b). Computer and Internet use in the United
States: October 2009. Retrieved from United States Census Bureau’s website.
<http://www.census.gov>.
Wolak, J., Finkelhor, D., Mitchell, K., & Ybarra, M. (2008). Online ‘‘predators’’ and
their victims: Myths, realities, and implication for prevention and treatment.
American Psychologist, 63(2), 111–128.
K.C. Seigfried-Spellar, M.K. Rogers / Computers in Human Behavior 29 (2013) 1997–2003 2003
