Arch Sex Behav
DOI 10.1007/s10508-008-9407-0
O R I G I N A L P A P E R
The Relation Between Sexual Behavior and Religiosity Subtypes:
A Test of the Secularization Hypothesis
Melissa A. Farmer Æ Paul D. Trapnell Æ Cindy M. Meston
Received: 11 October 2006 / Revised: 1 May 2008 / Accepted: 9 May 2008
Ó Springer Science+Business Media, LLC 2008
Abstract
Previous literature on religion and sexual be-
levels of female sexual behavior. The secularization hy-
havior has focused on narrow definitions of religiosity,
pothesis was supported by consistent levels of sexual activity
including religious affiliation, religious participation, or
across affiliations and is contradicted by the differential im-
forms of religiousness (e.g., intrinsic religiosity). Trends
pact of religiosity subtypes on sexual behavior. Findings
toward more permissive premarital sexual activity in the
suggested that the use of religious subtypes to evaluate reli-
North American Christian-Judeo religion support the secu-
gious differences, rather than solely affiliation, may yield
larization hypothesis of religion, which posits an increasing
useful insights into the link between religion and sexual
gap between religious doctrine and behavior. However, the
behavior.
recent rise of fundamentalist and new age religious move-
ments calls for a reexamination of the current link between
Keywords
Sexual behavior Á Religiosity Á Spirituality Á
religion and sexual behavior. The use of dual definitions of
Fundamentalism Á Paranormal belief Á Secularization
religiosity, including religious affiliation and dimensional
subtypes, may further characterize this link. The present
cross-sectional study evaluated patterns of sexual behavior
Introduction
in a young adult sample (N = 1302, M age = 18.77 years)
in the context of the secularization hypothesis using religious
Sexual practices have been regulated by the prescriptive
affiliation and a liberal-conservative continuum of religious
power of religion throughout modern history (DeLamater,
subtypes: paranormal belief, spirituality, intrinsic religiosity,
1981). However, the contemporary relation between religion
and fundamentalism. Results indicated few affiliation dif-
and sexuality is less clear. Two competing theories have been
ferences in sexual behavior in men or women. Sexual
posed to explain the relation between religion and sexuality.
behaviors were statistically predicted by spirituality, funda-
The first theory is the secularization hypothesis, which is a
mentalism, and paranormal belief, and the endorsement of
prominent theory of religious change that was developed
fundamentalism in particular was correlated with lower
throughout the 19th and 20th centuries by philosophers and
later by sociologists of religion. The secularization hypoth-
esis suggests that the modern reliance on reason has resulted
in a lack of faith in religion and has dampened the impact of
M. A. Farmer
Department of Psychology, McGill University, Montreal,
religion on daily life at an individual, social, and institutional
QC, Canada
level (Berger, 1967; Sommerville, 1998). Ultimately, the
secularization hypothesis would predict the demise of reli-
P. D. Trapnell
gious involvement in modern secular life. The second theory,
Department of Psychology, University of Winnipeg, Winnipeg,
MB, Canada
which challenges the secularization hypothesis, was pro-
posed by Stark and Bainbridge (1985) as the market theory of
C. M. Meston (&)
religion. The market theory of religion assumes that the de-
Department of Psychology, University of Texas at Austin,
mand for religion has not diminished with secularization, but
108 E. Dean Keeton, A 8000, Austin, TX 78712, USA
e-mail: meston@psy.utexas.edu
rather it has adapted to a highly secular society by generating
123

---

Arch Sex Behav
new religious movements (cults) and leading revivals of
frequent sexual intercourse, fewer lifetime sexual partners,
traditional religions (sects). The result is a market of com-
and delayed age of first intercourse (Barkan, 2006; Cochran
peting religious groups that drives a cultural renewal of
& Beeghley, 1991; Cochran, Chamlin, Beeghley, & Fen-
religion and maintains the influence of religion on daily life.
wick, 2004; Davidson, Darling, & Norton, 1995; Davidson,
These competing market religious groups are thought to re-
Moore, & Ullstrup, 2004; Jensen, Newell, & Holman, 1990;
flect a continuum of religious demand, including religiosity
Lefkowitz et al., 2004; Miller & Olson, 1988; Studer &
that is ultra-liberal, liberal, moderate, conservative, strict,
Thornton, 1987; Visser, Smith, Richters, & Rissel, 2007).
and ultra-strict (Finke & Stark, 2001). The current study was
However, it is unclear whether attendance is an accurate
designed to evaluate whether the secularization hypothesis or
indicator of religiosity across faith groups, as members of
the market theory best characterizes the current relation be-
religious affiliations differ dramatically in the expectations of
tween religion and adult premarital sexual behavior across a
religious participation. Furthermore, religious attendance
variety of religious affiliations. For the sake of simplicity, we
does not reflect specific religious attitudes and motivations
will limit our discussion to Judeo-Christian religious trends
that are associated with religiosity, nor does it account for the
and recent literature on religion in adult populations.
external influences that determine attendance (e.g., social
Much of our knowledge about the influence of religion on
influence of peers and family). These limitations challenge
adult premarital sexual behavior is based on studies that de-
the assumption that religious attendance reliably reflects
fine religiosity in terms of religious affiliation. However, the
religiosity across affiliations.
measurement of religious affiliation is problematic due to the
In an effort to measure religiosity that transcends religious
ideological variation within religious affiliations and the
affiliation, religiosity has also been described through
increasing number of affiliations that span the liberal-con-
dimensional subtypes, including intrinsic religiosity, spiri-
servative continuum (Robbins & Dick, 1990). Indeed, Kin-
tuality, fundamentalism, and paranormal belief. Allport’s
sey, Pomeroy, Martin, and Gebhard (1953) first reported that
(1950) intrinsic religiosity, which describes the extent of
variation in devoutness within religious groups was more
infusion of religious meaning into daily life, has been con-
important than affiliation in determining patterns of sexual
ceptually paired with religious commitment and importance
behavior, such as masturbation to orgasm, premarital petting,
(Donahue, 1985). Intrinsic religiosity in men and women is
and premarital coitus. With some exceptions (Studer &
associated with less willingness to engage in casual sex,
Thornton, 1987; Thornton & Camburn, 1989), past research
fewer desired sexual partners, and decreased likelihood of
has indicated that religious affiliations may be linked with
having engaged in a sexual relationship (Rowatt & Schmitt,
unique patterns of sexual behavior. In adult populations,
2003), and it is negatively correlated with frequency of sexual
Catholics have reported fewer sexual partners than Protes-
intercourse in the previous year (Zaleski & Schiaffino, 2000).
tants (Lefkowitz, Gillen, Shearer, & Boone, 2004) yet sur-
These findings indicate that intrinsic religiosity is inversely
prisingly have comparable levels of sexual activity compared
related to a variety of premarital sexual behavior in adults.
to conservative Protestants and fundamentalist Protestants
Spirituality is a subtype of religiosity that centers on an
(Bearman & Bruckner, 1999; Beck, Cole, & Hammond,
internalized subjective experience of belief in a divine or
1991). No difference in sexual experience between Jewish
superordinate being that is distinct from beliefs held by
and Christian participants has been consistently supported
specific religions (Hyman & Handal, 2006; Marler & Had-
(Lottes & Kuriloff, 1994; Pluhar, Frongillo, Stycos, &
away, 2002). ‘‘Embodied spirituality,’’ or a spiritual per-
Dempser-McClain, 1998). Nonreligious individuals (who
ception of sexual experiences, was found to positively
endorse ‘‘no religion’’) report more sexual experience (Plu-
correlate with a general measure of level of comfort with the
har et al., 1998) and greater frequency of masturbation than
body, intimacy, and sexual expression (Horn, Piedmont,
conservative Protestants, Catholics, and Jewish individuals
Fialkowski, Wicks, & Hunt, 2005). Belief in spiritual sexual
(Laumann, Gagnon, Michael, & Michaels, 1994; Leiblum,
experiences in heterosexual college students was associated
Wiegel, & Brickle, 2003). Other than the finding that non-
with increased sexual intercourse and a higher frequency and
religious individuals are generally more sexually active than
variety of sexual activities (Murray-Swank, Pargament, &
other affiliations, the use of religious affiliation has provided
Mahoney, 2005). These data suggest that spirituality is an
inconsistent information about religious differences in sexual
understudied subtype of religiosity that may explain patterns
behavior. Few solid conclusions can be drawn from the sole
of sexual behavior.
use of religious affiliation, indicating that supplementary
Fundamentalism, or the rigid belief in absolute religious
measures of religiosity are needed.
authority and strict adherence to religious texts, has not been
Like religious affiliation, religious participation has been
studied in relation to sexual behavior. Fundamentalism is a
widely used to evaluate the relation between religion and
nondenominational trait (Altemeyer & Hunsberger, 1992;
sexual behavior. Frequent attendance of religious services is
Hunsberger, 1996) and is associated with more conservative
related to greater sexual abstinence, less sexual activity, less
attitudes about homosexuality and gender roles (e.g., Bang,
123

---

Arch Sex Behav
Hall, Anderson, & Willingham, 2005; Hunsberger, Owusu,
The market theory of religion would predict a corre-
& Duck, 1999). The emphasis on traditional rituals and
spondence between religious beliefs and sexual behavior
religiously guided behavior suggests that individuals who
such that individuals who adhere to conservative religions
endorse fundamentalism would likely use their beliefs to
would report more restricted sexual behavior and individuals
dictate sexual behaviors as well.
who identify with liberal religions would report more unre-
A lesser-studied subtype of religiosity is paranormal be-
stricted sexual behavior. Accordingly, the following hypo-
lief, which includes religious paranormal beliefs (e.g., life
theses were tested:
after death, belief in angels or devils) and nonreligious para-
1.
As the most conservative subtype of religiosity, funda-
normal beliefs (e.g., superstitions or ‘‘new age’’ beliefs; see
mentalism will be negatively associated with frequency
Bainbridge, 2004; Irwin, 1993). Adherents of traditional reli-
and variety of premarital sexual behavior.
gions endorse more religious paranormal beliefs compared to
2.
As a moderately conservative subtype of religiosity,
nonreligious paranormal beliefs, and females endorse all
intrinsic religiosity will also be negatively associated
types of paranormal beliefs more than males (Emmons &
with frequency and variety of premarital sexual behav-
Sobal, 1981). A study using structural equation modeling
ior.
found that the motivation to seek short-term sexual expe-
3.
As a moderately liberal subtype of religiosity, spiritu-
riences was associated with a latent factor that included
ality will be positively associated with frequency and
paranormal interests (Weiss, Egan, & Figueredo, 2004). No
variety of premarital sexual behavior.
additional research has evaluated and the link between a pro-
4.
As an ultra-liberal form of spiritual belief, paranormal
clivity toward paranormal beliefs and sexual behavior.
belief will be positively associated with frequency and
In order to address the significance of religion in adult
variety of premarital sexual behavior.
sexual behavior, the current study was designed to evaluate
whether the secularization hypothesis or the market theory
Whereas the secularization hypothesis would predict a
of religion best characterize the relation between religiosity
desynchrony between religious beliefs and sexual practices
and sexual behavior. The literature on religion and sexuality
across all religious groups, the market theory would support
has typically interpreted sexual behavior as restricted versus
low correlations between religiosity and sexual behavior.
unrestricted (i.e., nonpermissive versus permissive, or con-
This is the first study we are aware of to represent in a study
servative versus liberal), and thus we selected to examine
of religion and sexuality the full range of the Finke and Stark
religiosity as it maps onto a conservative-liberal continuum.
(2001) liberal-fundamentalist faith continuum.
To accomplish this objective, we adopted two methods of
analyzing religiosity. First, we evaluated sexual behavior
differences in religious affiliation so that our sample could
Method
be compared to previous studies that relied on this method of
analysis. Second, we examined sexual behavior differences
Participants
in four subtypes of religiosity—intrinsic religiosity, funda-
mentalism, spirituality, and paranormal belief—across the
A total of 1,303 undergraduates (975 females, 470 males)
entire sample to determine whether the patterns of sexual be-
from a large public university participated in this question-
havior in our sample supported the secularization hypothesis
naire study for course credit. Participants were enrolled at the
or the market theory of religion.
university during either Fall (August–December) or Spring
To categorize religious affiliations along a conservative-
(January–May) semesters between 2000 and 2004. The sam-
liberal continuum, we used Smith’s (1990) classification,
ple age ranged from 18 to 25 (M age, 18.77 years). Married
which designates affiliations (e.g., ‘‘Assembly of God’’) as
participants were excluded from the sample because the
predominantly ‘‘liberal’’ (L), ‘‘moderate’’ (M), or ‘‘funda-
hypotheses were aimed at assessing patterns in premarital
mentalist’’ (F) based on surveys of denominational members,
sexuality. Demographic characteristics of the current sam-
surveys of denominational clergy, theological beliefs of
ple are shown in Table 1.
denominations, and prior related classifications, such as the
Participants were allowed to write in a response to the
General Social Survey. To further verify our classifications,
question, ‘‘What is your religion?’’ Two levels of analysis
we also consulted a number of publications where similar
were conducted: (1) the evaluation of general religious
classifications, or denomination characteristics relevant to
affiliation differences so that the current sample could be
such classifications, have been reported (e.g., Beck et al.,
compared to past research, and (2) the examination of reli-
1991; Laumann et al., 1994; Pluhar et al., 1998; Robbins &
gious subtype differences according to our central hypothe-
Dick, 1990; Studer & Thornton, 1987; Thornton & Camburn,
ses. For general religious affiliation groups, responses were
1989).
coded according to Smith’s (1990) classification system.
123

---

Arch Sex Behav
Table 1 Demographic, religious, and sexual characteristics by gender
Table 1 continued
Item
Females
Males
Item
Females
Males
(N = 975)
(N = 470)
(N = 975)
(N = 470)
%
n
%
n
%
n
%
n
Ethnicity
Sexually activea, by religious affiliation
62.3 607
44.8 214
Caucasian
63.2 621
61.1 287
Fundamentalist
57.9 190
43.6
68
African American
4.4
43
3.8
18
Monotheist Christian
64.9 237
51.6
80
Hispanic
14.8 143
14.7
69
Jewish
55.6
25
26.3
5
Asian American
13.8 133
16.6
78
No religion
74.3
55
54
27
Other
2.8
27
3.8
18
Atheist
81.3
13
41.2
7
Religious affiliation and ethnicity
Agnostic
71.4
15
28.6
2
Fundamentalist
38.4 328
38.3 156
Spiritualist
83.3
5
33
1
Caucasian
65.4 214
64.9 100
Sexual Orientationb
African American
11.3
37
9.7
15
Heterosexual with no homosexual fantasies
67.7 653
83.8 394
Hispanic
8.3
27
9.1
14
Heterosexual with an occasional
23.8 230
11.7
55
Asian American
13.1
43
14.3
22
homosexual fantasy
Other
1.8
6
1.9
3
Heterosexual with some homosexual
7
68
2.3
11
experience
Monotheist Christian
42.7 365
38.1 155
Homosexual with occasional heterosexual
0.8
8
1.1
5
Caucasian
64.3 234
59.4
92
fantasies or experience
African American
1.1
4
1.9
3
Homosexual
0.6
6
1.1
5
Hispanic
27.5 100
28.4
44
a Sexually active status was based on FSFI scores indicating an indi-
Asian American
6.6
24
10.3
16
vidual had engaged in sexual activity in the previous month
Other
0.5
2
0
0
b Measured with a revised Kinsey Heterosexual–Homosexual Rating
Jewish
5.3
45
4.7
19
Scale (fantasy and experience)
Caucasian
100
45
84.2
16
Hispanic
0
0
5.3
1
Other
0
0
10.5
2
This classification scheme was adopted so that specific
No religion
8.7
74
12.3
50
affiliations could be grouped into similar groups for analysis
Caucasian
71.6
53
71.4
35
of religious subtypes. In order from most conservative to
African American
1.4
1
0
0
most liberal affiliation, the current sample included Christian
Hispanic
6.8
5
6.1
3
fundamentalist (including Christian, Baptist, Pentecostal),
Asian American
17.6
13
20.4
10
Monotheist Christian (Catholic, Protestant, Methodist, Lu-
Other
2.7
2
2.0
1
theran, Presbyterian, Episcopal), Jewish, and nonreligious
Atheist
1.9
16
4.2
17
(Atheist, Agnostic, and Spiritualist, including Pagans and
Caucasian
68.8
11
82.4
14
Wiccans) participants. The decision to collapse Protestants
Hispanic
6.3
1
0
0
and Catholics into a single ‘‘Monotheist Christian’’ category
Asian American
18.8
3
17.6
3
was based on the growing evidence of interdenominational
Other
6.3
1
0
0
differences due to an explosion of subgroups within mono-
Agnostic
2.5
21
1.7
7
theistic religions and, based on past evidence, Protestants and
Catholics did not significantly differ in their endorsement
Caucasian
85.7
18
28.6
2
of any religiosity subtype we measured. Furthermore, due to
Hispanic
3
14.3 28.6
2
the large percentage of Hispanic students in the sample
Asian American
0
0
28.6
2
(16.1% of males, 17% of females), we did not want to con-
Other
0
0
14.3
1
found Catholicism with Hispanic ethnicity. Additionally,
Spiritualist
0.7
6
0.7
3
Agnostic and Atheist/nonreligious data were analyzed sep-
Caucasian
66.7
4
66.7
2
arately due to evidence that these groups may differ in sex,
Hispanic
16.7
1
33.3
1
rates of cohabitation, sociability, and extent of religious iden-
Asian American
16.7
1
0
0
tification (Bainbridge, 2005; Hayes, 2000). In sum, the sam-
Relationship characteristics
ple consisted of 1% Spiritualist, 2% Agnostic, 3% Atheist,
Long-term relationship in past 12 month?
65.7 631
57.1 269
10% no religion, 5% Jewish, 41% Monotheistic Christian,
Living with dating partner?
4.9
46
4.1
19
and 38% Christian Fundamentalist participants.
123

---

Arch Sex Behav
Measures
spiritual belief (e.g., ‘‘I’m completely skeptical toward any
type of religious or spiritual belief’’). The natural language
There is currently no measure of spiritual orientations
referent ‘‘spiritual’’ appears in all but two of the eight items.
explicitly designed to represent a full range of liberal-fun-
The expanded referent ‘‘religious or spiritual’’ is used in
damentalist positions along an ultra-left to ultra-right con-
some negation items where such an expansion properly
tinuum of faith orientations such as that proposed by Finke
serves the above definition of SP. Reliability of this eight-
and Stark (2001). Their theory was not originally developed
item index of spirituality was appropriate (a = .88).
in a psychological research context, but rather in a socio-
logical one addressing trends in religious affiliations and the
Paranormal/New Age Beliefs
emergence and transformations of new religious movements.
In the absence of a measure of the Finke–Stark continuum, we
Finke and Stark (2001) refer to New Age beliefs as ‘‘Ultra-
administered measures which approximated the liberal to
liberal’’, and supporting this characterization are numerous
conservative regions of the continuum.
studies that link paranormal, New Age, and similarly ‘‘eso-
teric’’ ontological beliefs with openness to experience
(Saucier, 2000; Saucier & Skrzypinska, 2006), a broad per-
Fundamentalism
sonality dimension strongly associated with liberal values,
beliefs, and political preferences (for a review, see McCrae,
To represent the ultra-right region of the liberal-conservative
1996). In contrast to New Age beliefs, fundamentalist reli-
religious spectrum, we administered a short form version of
gious beliefs tend to be strongly negatively correlated with
Altemeyer and Hunsberger’s (1992) Religious Fundamen-
openness (Saroglou, 2002; Saucier & Skrzypinska, 2006). To
talism scale (RF), consisting of RF items 1, 2, 4, 7, 10, 11, 15,
represent the ultra-left region of the Stark–Finke spectrum,
16, and 19 (a = .87), with equal numbers of affirmatively
we consequently used a balanced, 10-item index of para-
worded items (e.g., ‘‘To lead the best, most meaningful life,
normal/New Age beliefs constructed by Trapnell (2005) that
one must belong to the one, true religion’’) and negations
mainly targets the Psi, Precognition, and Spiritualism para-
(e.g., ‘‘No single book of religious writings contains all the
normal belief factors described by Tobacyk and Milford
important truths about life’’). The original RF scale has
(1983), which includes content similar to ‘‘New Age’’ belief
demonstrated excellent reliability in Christian as well as non-
markers commonly administered in public opinion surveys
Christian Canadian populations (Altemeyer & Hunsberger,
(e.g., Bainbridge, 2004; Orenstein, 2002; Rice, 2003). For
1992; Hunsberger, 1996).
example, Bainbridge (2004) refers to telepathy, astrology,
precognition, extrasensory perception, and psychic powers
Intrinsic Religiosity
as markers of New Age beliefs. Although New Age spiritu-
ality can be clearly distinguished from paranormal psychol-
To represent the center-right region of the Stark–Finke
ogy (Rao, 1984), the paranormal beliefs associated with each
spectrum, we administered the Age-Universal Intrinsic
are so strongly intercorrelated in practice that we will treat
Religiosity Scale (IR) of Gorsuch and Venable (1983). This
them as equivalent for current purposes. The 10-item para-
measure included questions such as, ‘‘I have often had a
normal measure used here included items targeting extra-
strong sense of God’s presence,’’ and, ‘‘My whole approach
sensory perception, psychic powers, astrology, precognition,
to life is based on religion.’’ One IR item, ‘‘I enjoy reading
past lives, out of body experiences, as well as a generalized
about my religion’’ was inadvertently omitted due to an
belief in ‘‘supernatural,’’ ‘‘magical,’’ and ‘‘paranormal’’ phe-
apparent mislabeling of this item in Gorsuch and Venable
nomena. Five items were affirmatively worded (e.g., ‘‘I
(1983). Reliability of the current eight-item IR measure was
am receptive to magical experiences’’), and five were nega-
adequate (a = .91).
tively worded (e.g., ‘‘No unexplained phenomenon can be
explained by a supernatural force’’). Reliability in the cur-
Spirituality
rent sample was adequate (a = .80). Saucier and Skrzy-
pinska’s (2006) operational distinction between ‘‘traditional
To represent the center-left region of the Stark–Finke con-
religiousness’’ (TR) and ‘‘subjective spirituality’’ (SS) clo-
tinuum, we administered a short index of Spiritual Belief (SP)
sely resembles the operational distinction here between ultra-
constructed by Trapnell (2005) to operationalize a minimal
left (New Age and paranormal beliefs) and ultra-right (fun-
definition of SP as ‘‘belief in a superordinate being, or an
damentalist beliefs) faith orientations.
intelligent energy or force, perceived to have a cosmic scope
These four measures were administered on the same page
or significance’’. The measure consists of four affirmations
of the survey using an identical 5-place Likert scale strongly
(e.g., ‘‘I tend to have deeply spiritual beliefs and feelings that
disagree (1), disagree (2), neutral or neither (3), agree (4),
are profoundly important to me’’) and four negations of
strongly agree (5). The paranormal measure appeared first,
123

---

Arch Sex Behav
followed by the spirituality measure, the intrinsic religious-
the SOI items was adequate (Cronbach’s alpha = .83 and .84
ness measure, and the fundamentalism measure. Scores for
for females and males, respectively).
each scale were computed as the mean of item responses for
each scale after first reverse scoring responses to reverse
Sexual Orientation
keyed items for each scale, if there were any (e.g., Paranor-
mal, Spirituality, and Fundamentalism scales). Scale scores
A short form of the Kinsey Heterosexual-Homosexual Rating
for all four religious measures therefore represent the mean
Scale (Kinsey, Pomeroy, & Martin, 1948) was used to eval-
item score for each scale, with an absolute range of 1 to 5.
uate sexual orientation. This revised scale enables responders
to endorse one of five sexual orientation categories that de-
scribe gradations in heterosexual-homosexual fantasy and
Sexual Behavior
experience. Categories range from exclusively heterosexual
(1) to exclusively homosexual (5).
Six items from the Derogatis Sexual Functioning Inventory
(DSFI; Derogatis, 1979) Experience Scale and 10 additional
Procedure
items were used to measure sexual behavior experiences. For
each behavior item, participants indicated whether or not
All participants completed questionnaires in same-sex
they had ever experienced that behavior (e.g., ‘‘kissing or
groups and were administered questionnaires by a same-sex
petting with a female’’) by responding yes or no. Three
research assistant. Testing sessions lasted up to 1 h each.
composite behavior experience scales were constructed by
Informed consent was obtained and same-sex researchers
summing responses to subsets of the behavior items repre-
provided instructions and answered any questions that arose
senting three levels of sexual involvement petting (6 items;
before and during testing. Each participant was randomly
a = .86), oral sex (5 items a = .87), and intercourse (4 items
assigned a code number to ensure confidentiality, and all data
a = .92). A fourth composite was constructed to represent
was solely connected to this code. Participants were informed
nonvirgin status. Participants who indicated an age at which
that should they feel uncomfortable with the personal subject
they first engaged in intercourse and/or endorsed any one of
matter in the questionnaires, they could discontinue the
the four intercourse experience items were classified as
study. No students declined participation. Questionnaires
nonvirgins. Participants having null or ‘‘no’’ responses to all
were inserted into a large ‘‘drop box’’ as participants left the
of these items were classified as virgins, whether or not they
testing room. Collected questionnaires and consent forms
endorsed any oral experience items. Reliability of the sexual
were stored separately in locked file cabinets. All data were
experience subtypes was adequate (Cronbach’s alpha = .86
entered into a password-protected data file. The research was
and .80 for females and males, respectively).
approved by the University of Texas Institutional Review
Board each year during the 2000–2004 time intervals. No
Unrestricted Sexual Behavior
students declined participation.
Due to the Judeo-Christian emphasis on sexual behavior
within the context of committed relationships, the Socio-
Results
sexual Orientation Inventory (SOI; Simpson & Gangestad,
1991) and two additional items were used to measure will-
To determine whether ethnicity played a role in religious
ingness to participate in uncommitted sexual behavior and
affiliation differences in sexual behavior, we performed a
fantasy. The SOI consists of a behavior component (number
one-way ANOVA. Ethnic differences were found in reli-
of sexual partners in the past year, and number of one-time
gious affiliation, F(4, 1254) = 8.66, p \ .01, and thus eth-
sexual encounters), a desire component (desired number of
nicity was entered as a covariate into all analyses. Because
sexual partners in the next 5 years), a fantasy component
pronounced gender differences are commonly observed both
(frequency of other partner fantasy during sex), and an atti-
in sexuality self-reports and on religious and spirituality
tude component measured by strength of endorsement of two
measures (Miller & Stark, 2002; Stark, 2002), we also report
attitudes, ‘‘Sex without love is okay,’’ and ‘‘I can imagine
results separately for men and women.
myself being comfortable and enjoying ‘casual sex’ with
As a check of validity, we mapped religious affiliation
different partners,’’ on a 5-point scale ranging from strongly
against the four religiosity subtypes (paranormal belief,
disagree (1) to strongly agree (5). The SOI total score is
spirituality, intrinsic religiosity, and fundamentalism) in or-
computed as the unit-weighted sum of z-scores representing
der to gain empirical support for our designations of religi-
each of the five components. The two additional SOI-related
osity subtype as conservative versus liberal in the current
items were number of foreplay partners during the previous
sample. Due to unequal group sizes and unequal variances,
year, and lifetime number of sexual partners. Reliability of
Games-Howell post-hoc tests were used to discern specific
123

---

Arch Sex Behav
1246) = 56.18, p \ .01, indicated that Spiritualist, Agnos-
tic, Atheist, and nonreligious participants reported less fun-
damentalism compared to Jewish, Monotheist Christian, and
Fundamentalist participants, with Fundamentalists reporting
the greatest fundamentalism.
Religious Affiliation
Religious affiliation differences in the sexual behavior of
women and men are presented in Tables 2, 3. Bonferroni
corrections were calculated for each set of comparisons to
reduce chance of Type I error. Sexual activity was catego-
rized as masturbation, petting, oral sex, intercourse, unre-
stricted sociosexuality, and sexual history behaviors.
Significant differences emerged in sexual behavior across
Fig 1 A comparison of paranormal belief, spirituality, intrinsic religi-
religious affiliation in women for masturbation, petting, oral
osity, and fundamentalism across members of seven religious affilia-
sex, and intercourse behaviors. Religious differences were
tions. Higher scores indicate stronger endorsement of each religiosity
found in rates of solitary masturbation, F(6, 738) = 4.13,
subtype
p \ .01. Agnostics reported greater frequency of masturba-
tion than Jewish participants, and Jewish participants re-
group differences in all subsequent analyses. As shown in
ported less masturbation than Monotheist Christians. No
Fig. 1, this assumption was moderately supported. Religious
religious differences were found in the composite petting
affiliation differences were found in paranormal belief, F(6,
score, but overall affiliation differences were found in the
1251) = 6.64, p \ .01, with Spiritualists reporting greater
item regarding kissing/petting with a woman, F(6, 738) =
paranormal belief compared to all other affiliations and
2.49, p \ .01. Post-hoc analyses did not indicate specific
Monotheist Christians reporting greater paranormal belief
group comparisons responsible for this difference. No sig-
than Fundamentalists. These findings suggest that paranor-
nificant religious differences were found in the composite
mal belief is preferentially endorsed by more liberal religious
oral sex score; however, affiliation differences were found in
affiliations (e.g., Spiritualists). Differences in spirituality,
the item regarding mutual oral stimulation of the genitals,
F(6, 1250) = 86.06, p \ .01, indicated that Fundamental
F(6, 734) = 2.82, p \ .01. Specifically, nonreligious par-
participants endorsed higher spirituality compared to all
ticipants reported more frequent mutual oral stimulation
other affiliations, Spiritualists reported more spirituality than
compared to Fundamentalist women. Finally, affiliation dif-
Atheists, Agnostics, and nonreligious participants, Jewish
ferences were found in the composite intercourse score, F(6,
participants reported less spirituality than Monotheist
739) = 3.73, p \ .01. Nonreligious participants reported
Christians and Fundamentalists and more spirituality than
greater frequency of intercourse behaviors compared to
Agnostic, Atheist, and nonreligious participants, and Atheist
Jewish, Monotheist Christian, and Fundamentalist partici-
participants reported less spirituality than nonreligious par-
pants. Of the intercourse behavior items, affiliation differ-
ticipants. Therefore, spirituality was more frequently en-
ences were found in vaginal intercourse (man on top), F(6,
dorsed in more conservative affiliations, with Jewish,
739) = 5.63, p \ .01, vaginal intercourse (woman on top),
Monotheist Christian, and Fundamentalist (e.g., conserva-
F(6, 739) = 3.71, p \ .01, and vaginal intercourse (entry
tive) participants generally more spiritual than Agnostic,
from behind), F(6, 736) = 4.60, p \ .01. For each of these
Atheist, and nonreligious (e.g., liberal) participants. Reli-
items, Jewish, Monotheist Christian, and Fundamentalist
gious affiliation differences in intrinsic religiosity, F(6,
women were less likely to report engaging in the vaginal
1247) = 85.34, p \ .01, indicated that Agnostic, Atheist,
intercourse positions compared to nonreligious women. No
and nonreligious participants endorsed less intrinsic religi-
significant religious affiliation differences were found in
osity than Jewish, Monotheist Christian, and Fundamentalist
measures of unrestricted sociosexuality or sexual history
participants, and additionally Fundamentalists reported more
behaviors. In men, no religious affiliation differences were
intrinsic religiosity than all other affiliations expect for
found for any of the behavioral measures.
Spiritualists. These findings suggest that intrinsic religiosity
is highly endorsed in more conservative affiliations (e.g.,
Religiosity Subtypes
Jewish, Monotheist Christian, and Fundamentalist partici-
pants, with Fundamentalists being most intrinsically reli-
Significant gender differences were found for paranormal
gious). Finally, religious differences in fundamentalism, F(6,
belief, with women reporting higher paranormal belief as
123

---

Arch Sex Behav
df
538
440
362
843
837
826
836
830
838
1
F
1.23
1.08
1.66
1.15
1.48
1.82
1.47
\
1.29
2 g
.01
.02
.03
.01
.01
.01
.01
.01
.01
D
S
2.22
1.66
1.44
3.62
2.01
3.79
4.31
2.10
1.21
df
738
739
738
738
736
737
729
739
740
740
738
737
738
734
739
739
739
736
739
280–284
=
4.13*
1
2.49*
1
1
1
1.14
1
1.75
1
1.69
2.16
1.16
2.82*
3.73*
5.63*
3.71*
4.60*
1
F
Fundamentalist
M
15.27
16.58
16.69
3.31
1.52
2.87
3.30
1.11
1.31
n
\
\
\
\
\
\
\
2 g
.03
.00
.02
.00
.01
.01
.01
.01
.01
.00
.01
.02
.01
.02
.03
.04
.03
.04
.00
SD
2.13
1.61
1.50
4.60
2.21
4.52
4.53
2.54
1.16
christian
305–309
5
7
6
Fundamentalist
64
94
1
88
27
50
81
85
30
50
65
0
44
68
54
43
45
12
=
Monotheist
M
15.12
16.65
16.56
4.03
1.93
3.43
3.94
1.31
1.40
n
D
S
1.38
1.43
1.34
3.22
1.80
4.02
5.20
2.54
1.35
christian
41
=
Jewish
M
15.03
16.40
17.25
4.02
1.53
4.39
4.30
1.88
1.37
n
5
9
7
Monotheist
62
95
1
88
30
52
86
90
31
59
75
51
74
61
52
50
14
SD
2.15
1.62
1.72
4.21
1.94
3.23
4.63
2.47
1.06
64–67
5
5
5
=
Jewish
53
98
90
24
46
88
83
27
59
76
39
59
54
37
39
12
None
M
14.50
16.26
16.38
3.66
1.89
3.45
4.76
1.52
1.71
n
shown
1
D
None
82
97
20
89
31
58
91
14
93
31
65
80
1
67
89
88
65
75
17
S
0.76
1.25
1.37
3.44
2.60
3.70
4.86
2.97
1.46
ethnicity
16
responses
for
6
9
=
yes
88
38
81
44
50
94
44
38
81
1
33
88
75
69
69
19
Atheist
M
15.71
17.71
17.67
3.69
2.13
4.60
4.50
1.44
1.44
n
Atheist
100
100
of
controlled
SD
1.23
1.10
1.17
2.77
1.87
4.44
3.23
1.75
1.02
5
5
20
Agnostic
95
95
25
95
35
65
90
1
90
40
60
80
1
55
80
75
70
50
20
percentage
=
Agnostic
M
15.67
16.73
17.11
3.43
2.00
5.05
4.29
1.43
1.62
n
with
participants
affiliation,
by
0
0
75
50
25
75
50
75
75
75
75
50
75
25
D
Spiritual
100
100
100
100
100
1.23
0.58
1.53
1.94
1.76
2.17
2.42
0.64
0.52
format,
religious
4–6
endorsed
by
=
Spiritual
MS
16.00
17.50
17.33
2.17
1.50
2.50
3.33
0.67
1.50
n
female
male
response
female
male
b
b
with
with
a
a
b
partners
behavior
year
b
by
by
male
yes/no
a
b
once
a
anus
top)
behind)
years)
past
years
lifetime
on
sexual
sexual
orgasm
orgasm
top)
5
male
female
on
year
in
in
just
of
to
to
a
a
genitals
on
from
from
(in
years)
next
female
male
of
past
stimulated
stimulated
to
to
based
female
a
a
vagina
partner’s
(in
in
b
(male
(female
(entry
years)
number
in
partners
partners
partners
genitals
genitals
of
of
a
items
a
with
with
(in
of
of
orally
orally
intercourse
caress
intercourse
sex
partners
sex
sex
the
a
partner
a
stimulation
stimulation
stimulation
partners
on
alone
anal
oral
oral
oral
petting
petting
sexual
petting
petting
genitals
genitals
oral
oral
oral
subscore
intercourse
intercourse
intercourse
sexual
orgasm
sexual
or
or
or
Differences
penetration
penetration
subscore
experience
represent
2
subscore
first
first
first
foreplay
cheated
.01
sex
of
of
of
Kissing,
Kissing,
Mutual
Mutual
Finger
Finger
Having
Having
Giving
Giving
Mutual
Vaginal
Vaginal
Vaginal
Receiving
\
p
Sexual
Values
Table
Item
Masturbating
Petting
Oral
Intercourse
Item
Age
Age
Age
Sexual
Intercourse
Anticipated
Intercourse
Intercourse
Times
*
a
b
123

---

Arch Sex Behav
df
346
312
239
391
392
389
392
392
395
1
2.33
1
1
1
1
1
1.12
1
F
\
\
\
\
\
\
\
2 g
.01
.04
.01
.01
.01
.01
.01
.02
.02
D
df
278
278
277
278
275
278
270
277
278
277
278
278
277
278
278
278
278
272
278
242
S
2.61
1.62
2.29
4.94
3.69
9.64
11.59
3.21
1.37
96–99
1.39
1.24
1.07
1
1.67
1
2.16
1.15
1.75
1.48
1
1.02
1.37
1.08
1
1.52
2.11
1
1.03
1
=
F
\
\
\
\
\
\
Fundamentalist
M
14.55
16.02
16.25
4.15
1.95
5.20
5.69
1.71
1.54
n
2 g
.03
.03
.02
.02
.04
.01
.05
.03
.04
.03
.01
.02
.03
.02
.02
.03
.05
.01
.02
.01
SD
2.41
1.83
2.47
5.60
3.82
6.29
4.53
3.95
1.33
christian
107–111
5
8
2
Fundamentalist
88
94
77
1
65
10
75
15
87
67
10
14
57
42
65
48
47
40
=
Monotheist
M
14.75
15.96
16.43
4.14
2.30
5.54
3.94
1.64
1.70
n
D
S
1.59
1.41
1.00
4.60
2.50
8.39
3.05
0.99
1.05
christian
11
=
Jewish
M
14.53
16.35
17.00
4.21
1.58
8.78
3.21
0.74
1.26
n
2
9
4
Monotheist
91
98
80
2
78
10
84
19
93
79
11
20
65
49
74
60
60
49
SD
3.27
2.90
3.31
4.19
1.78
5.86
5.58
1.49
1.28
37
9
9
0
0
9
0
0
91
91
73
73
64
64
64
64
64
40
Jewish
100
100
100
=
None
M
14.56
15.85
16.57
3.49
1.59
5.58
4.43
1.00
1.84
n
shown
5
7
5
5
0
0
97
92
11
78
92
2
89
87
76
57
76
73
76
57
None
100
D
S
3.93
1.17
1.30
4.99
2.53
4.92
12.24
11.12
1.18
responses
ethnicity
15
for
7
0
0
yes
87
87
13
64
13
67
2
73
67
13
20
67
27
53
47
40
40
=
Atheist
100
Atheist
M
14.87
16.50
16.56
3.12
1.47
3.84
4.56
3.63
1.47
n
of
SD
1.29
1.38
1.00
3.65
2.79
4.73
7.57
1.10
1.51
controlled
0
0
0
0
0
0
0
75
75
75
25
50
50
50
25
4–7
Agnostic
100
100
100
100
100
percentage
=
Agnostic
M
16.00
16.71
17.00
3.60
2.40
5.00
5.83
1.00
1.67
n
with
participants
affiliation,
by
0
0
0
0
0
0
50
50
50
50
50
50
50
D
Spiritual
100
100
100
100
100
100
100
S
3.22
0.71
1.41
0.00
0.58
0.58
3.22
1.53
0.00
format,
3
religious
=
endorsed
by
Spiritual
M
13.33
11.50
15.00
1.00
0.67
2.33
3.67
1.67
1.00
n
female
male
response
female
male
b
b
with
with
a
a
b
partners
male)
year
b
behavior
by
by
male
yes/no
a
b
once
a
anus
top)
behind)
past
years
lifetime
sexual
orgasm
orgasm
top)
5
on
male
female
on
year
in
in
just
of
sexual
to
to
a
a
genitals
on
from
from
next
female
male
of
past
stimulated
stimulated
to
to
penetrating
based
male
a
a
vagina
partner’s
in
b
(male
(female
(entry
number
in
partners
partners
partners
genitals
genitals
of
of
a
(male
items
a
with
with
of
of
orally
orally
intercourse
caress
intercourse
sex
partners
sex
sex
the
a
partner
a
stimulation
stimulation
stimulation
partners
on
alone
anal
oral
oral
oral
petting
petting
sexual
petting
petting
genitals
genitals
oral
oral
oral
subscore
intercourse
intercourse
intercourse
sexual
orgasm
sexual
or
or
or
Differences
penetration
penetration
subscore
intercourse
experience
represent
3
subscore
first
first
first
foreplay
cheated
sex
of
of
of
Kissing,
Kissing,
Mutual
Mutual
Finger
Finger
Having
Having
Giving
Giving
Mutual
Vaginal
Vaginal
Vaginal
Receiving
Anal
Sexual
Values
Table
Item
Masturbating
Petting
Oral
Intercourse
Item
Age
Age
Age
Sexual
Intercourse
Anticipated
Intercourse
Intercourse
Times
a
b
123

---

Arch Sex Behav
compared to males, F(1, 1433) = 29.45, p \ .001. Women
intrinsic religiosity when fundamentalism was controlled for.
endorsed significantly more spirituality items than men, F(1,
Additionally, paranormal belief, spirituality, and funda-
1432) = 43.59, p \ .001. No gender differences for intrinsic
mentalism were entered into a regression analysis to predict
religiosity or fundamentalism were found. Paranormal be-
variance in each sexual behavior. Intrinsic religiosity was
liefs showed a weak correlation with spirituality, r(1444) =
excluded from the regression due to high collinearity and
.21, p \ .001, and fundamentalism, r(1439) = -22, p \
potential construct overlap with fundamentalism. In Tables 4
.001. Intrinsic religiosity and fundamentalism were some-
and 5, behaviors are listed in order of the magnitude of var-
what strongly correlated, r(1435) = .64, p \ .001. Spiritu-
iance predicted by the regression model.
ality was moderately correlated with fundamentalism,
In women, paranormal religiosity was significantly posi-
r(1439) = .41, p \ .001, and strongly correlated with intrin-
tively correlated with all domains of sexual behavior,
sic religiosity, r(1438) = .75, p \ .001.
including masturbation, petting, oral sex, intercourse, unre-
Tables 4 and 5 display correlation and regression results
stricted sociosexuality, and sexual history behaviors. Spiri-
for women and men, respectively. The first three columns list
tuality, fundamentalism, and intrinsic religiosity were nega-
simple Pearson correlations between sexual behaviors and
tively correlated with the majority of sexual behaviors,
paranormal belief, spirituality, and fundamentalism, whereas
although the magnitude of correlations with intrinsic religi-
the last two columns list the simple and partial correlations
osity was somewhat reduced when fundamentalism was
between sexual behaviors and intrinsic religiosity and
controlled for. The regression model containing paranormal
Table 4 Association of female sexual behavior with paranormal, spiritual, and intrinsic religiosity
Para
Spir
Fund
Adj. R2b
Intrin
Intrin fundc
Vaginal intercourse (female on top)
.08*
-.16**
-.22**
.06**
-.25**
-.15**
Vaginal intercourse (male on top)
.05
-.20**
-.19**
.06**
-.26**
-.18**
Vaginal intercourse (entry from behind)
.07*
-.17**
-.20**
.05**
-.24**
-.15**
Masturbating alone
.12*
-.10**
-.18**
.04**
-.19**
-.10**
Giving oral stimulation to a male
.07*
-.10**
-.19**
.04**
-.16**
-.05
Times cheated on partnera
.13**
-.12**
-.16**
.04**
-.17**
-.10**
Finger penetration of vagina
.06
-.11**
-.17**
.03**
-.17**
-.08*
Mutual oral stimulation of genitals
.09*
-.10**
-.16**
.03**
-.17**
-.09**
No. anticipated sexual partners in next 5 yearsa
.10**
-.12**
-.16**
.03**
-.19**
-.12**
Age of first sexual intercourse
-.09
.04
.16**
.03**
.13**
-.04
No. intercourse partners in lifetimea
.12**
-.10**
-.15**
.03**
-.20**
-.14**
Kissing, petting with a female
.08*
-.07*
-.12**
.02**
-.14**
-.09**
Age of first sexual caress
-.12**
.06
.12**
.02**
.10*
.03
Having genitals orally stimulated by female
.09*
.01
-.12**
.02**
-.08*
-.01
Mutual petting of genitals to orgasm with female
.11**
.02
-.12**
.02**
-.08*
.00
Having genitals orally stimulated by male
-.04
-.13**
-.03
.02**
-.08*
-.08*
No. sexual foreplay partners in past yeara
.08*
-.07*
-.11**
.02**
-.13**
-.08*
No. sexual intercourse partners just oncea
.08*
-.06
-.12**
.02**
-.14**
-.08*
No. intercourse or oral sex partners in past yeara
.07*
-.08*
-.13**
.02**
-.16**
-.10**
Kissing, petting with a male
.02
-.02
-.10**
.01*
-.07*
-.01
Mutual petting of genitals to orgasm with male
-.02
-.11**
-.04
.01*
-.08*
-.07
Receiving anal intercourse from male
.02
-.08*
-.09*
.01*
-.11**
-.07*
Finger penetration of partner’s anus
-.01
-.09*
-.10**
.01*
-.10**
-.05
Giving oral stimulation to female
.03
-.08*
-.02
.01
-.07
-.07*
Age of first orgasm
.00
.06
.05
.01
.05
.02
* p \ .05, ** p \ .01
a Analyses on participants who first engaged in sexual intercourse after age 12. Abbreviations are as follows: Paranormal belief
(Para), Spirituality (Spir), Fundamentalism (Fund), Intrinsic religiosity (Intrin)
b Refers to a regression model containing only paranormal, spirituality, and fundamentalist beliefs
c Intrinsic religiosity, controlling for fundamentalism
123

---

Document Outline

• The Relation Between Sexual Behavior and Religiosity Subtypes: A Test of the Secularization Hypothesis
  • Abstract
  • Introduction
  • Method
    • Participants
    • Measures
      • Fundamentalism
      • Intrinsic Religiosity
      • Spirituality
      • Paranormal/New Age Beliefs
      • Sexual Behavior
      • Unrestricted Sexual Behavior
      • Sexual Orientation
    • Procedure
  • Results
    • Religious Affiliation
    • Religiosity Subtypes
  • Discussion
  • References

---