Conscious Regulation of Sexual Arousal in Men

14 Pages·2009·0.18 MB·English

JOURNALOFSEXRESEARCH,46,1–14,2009
Copyright#TheSocietyfortheScientificStudyofSexuality
ISSN:0022-4499print=1559-8519online
DOI:10.1080/00224490902754103
Conscious Regulation of Sexual Arousal in Men
Jason Winters, Kalina Christoff, and Boris B. Gorzalka
Department of Psychology, University of British Columbia
Thegoalsofthisstudyweretoexaminetheeffectiveness ofemotionalreappraisalinregula-
ting male sexual arousal and to investigate a set of variables theoretically linked to sexual
arousal regulationsuccess. Participantsfirst completed a series ofonlinesexuality question-
naires.Subsequently, they were assessedfor their success in regulating sexual arousal in the
laboratory.Resultsshowedthattheabilitytoregulateemotionmaycrossemotionaldomains;
those men best able to regulate sexual arousal were also the most skilled at regulating their
level of amusement to humorous stimuli. Participants, on average, were somewhat able to
09 regulate their physiological and cognitive sexual arousal, although there was a wide range
0
2 ofregulationsuccess.Whereassomemenwereveryadeptatregulatingtheirsexualarousal,
y
r
a others became more sexually aroused while trying to regulate. Age, sexual experience, and
u
r
eb sexual compulsivity were unrelated to sexual arousal regulation. Conversely, sexual excita-
F
2 tion,inhibition,anddesirecorrelatedwithsexualarousalregulationsuccess.Increasedsexual
2

7 excitationanddesirewereassociatedwithpoorerregulatoryperformance,whereasapropen-
4
:
01 sity for sexual inhibition due to fear of performance consequences was related to regulatory

t: success.
A

]
a
i
b
m
u
l
o
C

h
s A small body of research indicates that men have some Earlyevidenceshowedthatarousalcouldbesuppressed
i
t
ri voluntary control over sexual arousal, as assessed using (Abel, Blanchard, & Barlow, 1981; Freund, 1963, 1965,
B
f penileplethysmography(PPG; Abel,Blanchard,&Bar- 1967; Quinsey & Bergersen, 1976; Quinsey & Carrigan,
o

y low, 1981; Adams, Motsinger, McAnulty, & Moore, 1978). However, sample sizes were very small, and no
t
i
rs 1992; Freund, 1963, 1965, 1967; Golde, Strassberg, & techniques were used to control for distraction. It was
e
iv Turner, 2000; Henson & Rubin, 1971; Laws & Rubin, quite possible that, when instructed to try to suppress
n
U
[ 1969; Mahoney & Strassberg, 1991; McAnulty & arousal, participants merely distracted themselves from

:
By Adams, 1991; Quinsey & Bergersen, 1976; Quinsey & the stimuli by looking away from them, closing their

ed Carrigan, 1978). This research was largely motivated eyes, or focusing on sexually repulsive thoughts. This
d
oa byconcernamongforensicpractitionersthatsexualpre- oversight was corrected in later, well-controlled studies
l
n
ow ference testing using PPG may be vulnerable to faking (Henson&Rubin,1971;Laws&Rubin,1969;Mahoney
D
by some sexual offenders. Sexual preference testing for &Strassberg,1991;McAnulty&Adams,1991).Various
sexual offenders is an essential component of compre- techniqueswereusedtoensurethatparticipantsfocused
hensive offender management, as inappropriate sexual on the sexual stimuli presented. Such techniques
preference (i.e., preference for sexual violence or under- included embedded signal detection tasks (e.g., button
age targets) is a strong predictor of risk for reoffence pressing in response to an embedded flashing dot), tests
(e.g., Hanson & Morton-Bourgon, 2004). The PPG, for stimulus content memory, and ongoing descriptions
which is applied on the assumption that degree of erec- of sexual stimuli during presentation.
tion is a valid peripheral indicator of central sexual Findings from those studies revealed that men can
arousal (e.g., Geer & Head, 1990), is currently the best suppress physiological and self-reported sexual arousal
measure of inappropriate sexual preference. topreferredstimulibutareunabletoenhancearousalto
In past PPG faking studies, participants were asked non-preferred stimuli. Average suppression rates range
to either suppress penile responses to preferred stimuli, from26%to38%maximumerection,withsomemenable
maximize responses to non-preferred stimuli, or both. toentirelysuppresstheirsexualarousalandothersunable
tosuppresswhatsoever(Adamsetal.,1992;Goldeetal.,
2000;Mahoney&Strassberg,1991;McAnulty&Adams,
Correspondence should be addressed to Jason Winters, Depart- 1991). According to results reported by McAnulty and
mentofPsychology,UniversityofBritishColumbia,2136WestMall, Adams, men are more successful at suppressing
Vancouver,BC,CanadaV6T1Z4.E-mail:[email protected]
WINTERS,CHRISTOFF,ANDGORZALKA
cognitive than physiological arousal. McAnulty and normally follow the emotional experience is inhibited.
Adams proposed that this was the result of ‘‘emotional Suppressionismorecognitivelytaxingthanreappraisal,
distancing’’ (p. 574), and that participants processed astheexpressivebehaviormustbemutedwhiletheemo-
thestimuli ascognitivelyarousingbutwereabletosup- tional experience remains unchanged. Although labora-
press physiological arousal. Similarly, they in the study tory studies have shown that both reappraisal
performed by Adams et al. claimed that, although they (Beauregard, Le´vesque, & Bourgouin, 2001; Jackson,
wereunabletocontrolcognitivearousal,theydidexperi- Malmstadt, Larson, & Davidson, 2000; Le´vesque et al.,
ence a sense of control over penile response. During 2003;Ochsner,Bunge,Gross,&Gabrieli,2002;Ochsner
debriefing, participants in the Mahoney and Strassberg etal.,2004)andsuppression(Colby,Lanzetta,&Kleck,
study were asked to describe techniques they used to 1977; Gross, 1998a; Gross & Levenson, 1993, 1997) are
suppress, and most indicated that they tried to view the effective in regulating emotional response, reappraisal
stimuli inas detached a way as possible. appears to be a far more robust strategy in reducing
From these comments, it seems that the tactic most emotional experience.
effectivelyusedtominimizephysiologicalsexualarousal In the decade preceding the emergence of emotion
is emotional detachment (Mahoney & Strassberg, 1991; regulation in the literature, Everaerd (1989) suggested
McAnulty & Adams, 1991). This is not surprising, as a similar regulatory process with respect to sexual
emotional detachment is fundamental to emotion regu- arousal. He proposed that voluntary control of sexual
9 lation (Gross, 1998a, 1998b, 1999, 2002; Ochsner & arousal is achievable when an individual can subdue
0
0
2 Gross,2005),andvariousresearchersandtheoristshave emotional responses to sexual stimuli while still cogni-
y
ar suggested that sexual arousal can be best understood tively attending to them. In the only study to address
u
r
b within the rubric of emotional function (Everaerd, this proposition directly, participants either attempted
e
F
2 1989; Everaerd, Laan, Both, & Spiering, 2001; Frijda, to inhibit sexual responses by detaching or distancing
2
7 1986; Geer, Lapour, & Jackson, 1993; Janssen & themselves from the sexual stimuli (i.e., reappraisal)
4
:
1 Everaerd, 1993;Janssen, Everaerd, Spiering, & Janssen, or made no attempt to inhibit their sexual responses
0

t: 2000; Lambie & Marcel, 2002; Rosen & Beck, 1988). (Beauregard et al., 2001). Subsequently, participants
A
] Emotions are distinct from moods in that they are were asked to self-report their sexual arousal.
a
i
b incited by specific triggers. They are characterized by During stimulus presentation, functional MRI identi-
m
u
l experiential, behavioral, and physiological changes fied regions of the brain implicated in regulation of
o
C
h (Cacioppo, Berntson, Larsen, Poehlmann, & Ito, sexual arousal. Participants self-reported 60% less
s
ti 2000). They may be unlearned responsesto stimuli with sexual arousal when they attempted to inhibit sexual
i
r
B intrinsicemotionalvalueorlearnedresponsestostimuli responses, and inhibition of sexual arousal was asso-

f
o withacquiredemotionalsignificance.Multipleappraisal ciatedwithincreasedactivationincorticalregionsthat
y
it processes can be involved in determining the reward have been associated with regulation of other emo-
s
r
e value of emotion-inducing stimuli (Scherer, Schorr, & tions (Le´vesque et al., 2003; Ochsner et al., 2002;
v
i
Un Johnstone, 2001). Sexual arousal appears to meet these Ochsner etal.,2004). Beauregard etal.didnotinclude
[
: criteria. a physiological measure of sexual arousal, so it is
y
B
According to Gross (1998b), emotion regulation is unclear how penile response was affected by sexual
d
e
ad themeansbywhichindividualscaninfluencetheiremo- arousal regulation.
o
nl tional responses. Through emotion regulation, indivi- Although it has been established in the research lit-
w
o
D duals are able to exert control on which emotions they eraturethatmen,onaverage,havesomevoluntarycon-
have and how they experience and express those trolovertheirphysiologicalsexualarousal,andthatthe
emotions.Emotionregulationcan beautomatic or con- tacticusedmostsuccessfullytoregulatearousalappears
trolled and conscious or unconscious, and can occur to be reappraisal (i.e., emotional detachment), no pre-
during processing of emotional cues or after responses vious studies have provided participants with reapprai-
are activated. Gross (2002) suggested that two distinct sal strategy instructions while also including measures
processes may be at play: reappraisal and suppression. of both self-reported and physiological sexual arousal.
He defined reappraisal asthe process by which a poten- One of the goals of our study was to address this issue.
tially emotion-eliciting situation is reframed in non- Because there appear to be individual differences in the
emotionalterms.Thiscanbeaccomplishedbydetaching ability to regulate other emotions (e.g., Jackson et al.,
oneself from, or reappraising the meaning of, an emo- 2000; Le´vesque et al., 2003; Ochsner et al., 2004), we
tion-eliciting stimulus. Lambie and Marcel (2002) predicted that men would also vary in their capacity to
described a similar process whereby an individual can regulate sexual arousal. We hypothesized that men’s
regulate his or her emotional response to an emotion- abilities to regulate sexual arousal would be related to
evoking stimulus by taking an objective perspective. their general emotion regulation capability. In other
Emotional suppression, on the other hand, does not words, those men best able to regulate sexual arousal
change the emotional experience, but does affect its would be most adept at regulating other emotions.
expression (Gross, 2002). The behavior that would Given that self-reported sexual arousal correlates
2
CONSCIOUSREGULATIONOFSEXUALAROUSAL
reasonably well with physiological arousal (Haywood, 2004). This may culminate in distress sufficient to
Grossman,&Cavanaugh,1990;Sakheim,Barlow,Beck, instigatetreatment-seekingbehavior,aspersonal,social,
& Abrahamson, 1985), we expected that the two or occupational life is negatively affected. Research has
measures would remain related when men attempted linked sexual compulsivity with sexual behavior that is
to regulate sexual arousal. This would provide evidence illegal (e.g., Bradford, 2001; Kafka, 2003) or carries an
that sexual arousal regulation, as an application of increased risk for sexually transmitted infections
emotion regulation, can affect both cognitive and (Benotsch, Kalichman, & Kelly, 1999; Benotsch,
physiological sexual response. Kalichman, & Pinkerton, 2001; Dodge, Reece, Cole, &
We also predicted that other factors associated with Sandfort, 2004; Kalichman & Cain, 2004; Kalichman,
sexualrespondingmightinfluenceaman’sabilitytoreg- Greenberg, & Abel, 1997a,b; Semple, Zians, Grant, &
ulate his sexual arousal. Bancroft and Janssen (2000) Patterson, 2006). Although we were unable to dis-
proposedthatsexualresponseiscontrolledbytwoinde- tinguish sexual compulsivity from sexual desire in a
pendent neurophysiological systems: sexual excitation previous study (Winters et al., in press), it is possible
and sexual inhibition. Together, they modulate the that sexual compulsivity is related to a deficit in sexual
affective, physiological, and behavioral experiences that arousal regulation.
accompany sexual arousal. A strong sexual excitation Anecdotal evidence suggests that as men get older
system would contribute to robust sexual responding, and gain sexual experience, they become better able to
9 whereas a strong sexual inhibition system would reduce control their sexual response. For this reason, we mea-
0
0
2 sexual response. Janssen, Bancroft, and their colleagues sured sexual experience and age as variables that may
y
ar constructedtheSexualInhibitionandSexualExcitation relate to sexual arousal regulation.
u
r
b Scales (SIS=SES) to measure the strength of the sexual Basedontheresearchreviewedandtheresultingpre-
e
F
2 excitation and inhibition systems (Janssen, Vorst, Finn, dictions, we formulated four hypotheses:
2
7 & Bancroft, 2002a,b). Items were created to describe
4
:
1 situationsthatwouldincreaseordecreasesexualarousal
0 H1: Self-reported sexual arousal will correlate with
t: and penile response. Those items clustered to form
A physiological sexual arousal, as measured by
] three subscales: (a) propensity for sexual excitation, (b)
ia PPG, when sexual arousal is both unregulated
b propensity for sexual inhibition due to threat of perfor-
um and regulated.
l mance failure (i.e., erectile failure), and (c) propensity
Co H2: Men will exhibit a range of physiological and
h for sexual inhibition due to threat of performance
s self-reported sexual arousal regulation success.
ti consequences (i.e., risk of sexually transmitted infec-
ri H3: Sexual arousal regulation success will correlate
B tions, pregnancy, or legal repercussions). Research has
f positivelywithage,sexualexperience,andsexual
o shown that sexual excitation is associated with
y inhibition and negatively with sexual excitation,
it increased sexual responsivity in the laboratory, a
rs sexual desire, and sexual compulsivity.
e greater frequency of sexual behaviors, and increased
iv H4:Thosemenwhoarebestatregulatingtheirsexual
Un partnered and solitary sexual desire (SSD; Bancroft &
[ arousalwillalsobethebestatregulatinganother
: Vukadinovic, 2004; Janssen et al., 2002a,b; Winters,
By emotional response, amusement.
Christoff, & Gorzalka, 2008). Sexual inhibition due to
d
e
ad threat of performance consequences is inversely related
o
nl to frequency of unprotected intercourse and partnered To test these hypotheses, we designed a two-part
w
o
D and SSD, and is positively associated with sexual study.Menfirstcompletedaseriesofsexualityquestion-
restrictiveness. Based on these associations, we pre- nairesthatmeasurethefactorsofinterestdescribedear-
dicted that increased sexual excitation and decreased lier.Thequestionnaireswerecompletedonline,asonline
sexual inhibition would be related to poorer sexual surveysaremoreconvenientandmayresultinincreased
arousal regulation performance. Similarly, we hypothe- disclosure (Schroder, Carey, & Vanable, 2003). Online
sized that heighteneddyadicsexual desire (DSD) would measures appear to be as valid as, and can perform in
also be related to decreased regulation success. An a similar manner to, traditional pencil-and-paper mea-
increased appetitive sexual drive and propensity for sures (Dixon & Turner, 2007; Meyerson & Tryon,
sexual excitation, in conjunction with muted sexual 2003; Roberts, 2007). Participants were subsequently
inhibition, should theoretically make it more difficult assessedforarousalregulationsuccessinthelaboratory.
to regulate sexual arousal when one is confronted with Regulation instructions obtained from the emotion reg-
sexually arousing stimuli. ulation literature were provided. Two stimulus condi-
Another factor that we hypothesized should be tions, erotic and humorous, were crossed with two
relatedtosexualarousalregulationissexualcompulsiv- instruction conditions, experience or regulate, to pro-
ity.Sexualcompulsivity,orcompulsivesexualbehavior, duce four possible trial types. Psychophysiological
is characterized by disinhibited or undercontrolled sex- and self-reported arousal across the trial types were
ual thoughts, feelings, and behaviors, as identified by compared and correlated with scores on the survey
the individual (Coleman, 2003; Kalichman & Cain, measures.
3
WINTERS,CHRISTOFF,ANDGORZALKA
Method DemographicsandGeneralInformationForm(DGIF),
which always appeared first, the survey measures were
Participants randomly presented. The set of questionnaires took
approximately 45min to complete. A more detailed
Forty-nine sexually functional men who were free of
description of the survey procedure can be found in
medication that may affect sexual response participated
Winters et al. (in press).
inthestudy.Theiraverageagewas27.7(SD¼10.1)and
ThesecondpartofthestudywasconductedataUBC
ranged from 18 to 67. Their median and modal age
laboratory.Uponparticipants’arrivalatthelaboratory,
was 24. Sixty-five percent of the participants were
the proceduresand instructions were explained in detail
Caucasian,20%wereEastAsian,4%wereSouthAsian,
by the research technician, and participants were given
4%wereLatinAmerican,and6%wereofanotherethni-
the opportunity to examine the PPG apparatus and
city.Themajorityofparticipantsidentifiedasheterosex-
ask questions. Before testing began, participants were
ual (n¼44; 90%), although there was a small group of
required to sign a consent form. They were also asked
men who identified as bisexual (n¼5; 10%). More than
to provide basic demographic information (age, ethnic
onehalfoftheparticipantswereinexclusivesexualrela-
identity, English as first language, city of birth, and
tionships (n¼28; 57%); the rest were in non-exclusive
undergraduate student status) so that each participant’s
sexual relationships (n¼4; 8%) or were not in a sexual
survey data could be linked with his laboratory data.
relationshipatthetimeofthestudy(n¼17;35%).Most
9 In the laboratory, we set up a private testing room
00 participants were single (n¼35; 71%). Others were
2 with a lounge chair placed 4 feet back from the video
y cohabiting (n¼3; 6%), married (n¼6; 12%), or sepa-
ar presentation television. Headphones for audio stimuli
ru rated or divorced (n¼5; 10%). Slightly less than one
b and a numeric keypad for self-report responses were
Fe half the sample was comprised of undergraduate
2 connected to the testing laptop, located outside the
2 students (47%).
7 testing room. Two clean towels were provided for each
:4 All participants except one reported masturbating at
1 participant,onetositonandtheothertobeplacedover
0 least once per week. The sample’s average weekly mas-
t: his lap. An inflatable seat pad, connected to the data
A turbation frequency over the preceding 3 months was
] acquisition laptop, allowed us to monitor participants’
ia 6.0 (SD¼6.9). The majority (87.7%) of the sample
b movements during data acquisition. Participant move-
um reported viewing pornography on at least a weekly
l ment can tug the strain gauge lead creating spikes in
Co basis. The average amount of time devoted to viewing
h thePPGpenilecircumferencedata.Duringpostproces-
s pornography per week over the preceding 3 months
ti sing, data spikes that were artefacts of participant
ri was 2.4hr (SD¼2.0). A small minority of the sample
B movement were removed.
f (n¼4; 8.2%) had never experienced any partnered sex-
o Afterwritteninformedconsentwasobtained,partici-
y ual activity, although three of those participants viewed
it pants were asked to enter the PPG testing room, pull
rs pornography and masturbated at least once per week.
e their pants down around their ankles, be seated, fit the
v
i
Un gauge, cover themselves with a towel, and put on the
[
: headphones. For each participant, the technician
By Procedure
visually inspected the incoming pre-testing PPG data
d
e
ad Participants were recruited by three means. A link to insure that the gauge was seated properly.
o
nl providedattheendoftheonlinesurveyusedinWinters Participants viewed 16randomlyordered videoclips:
w
o
D et al. (in press) briefly described the study and provided 8 erotic and 8 humorous (control). Before each of the
contactinformationforthoseinterestedinparticipating. clips was presented, either ‘‘Experience’’ or ‘‘Regulate’’
Wealso posted astudy advertisementontheUniversity was displayed on the television screen. These acted as
of British Columbia (UBC) Department of Psychology task cues, corresponding to instructions borrowed from
SubjectPoolPsychologyResearchParticipationSystem. the emotion regulation literature (Beauregard et al.,
As a final means of recruitment, advertisements were 2001; Gross, 2002; Jackson et al., 2000; Le´vesque et al.,
posted around Vancouver and the UBC campus. Parti- 2003).Forexperiencetrials,participantswereinstructed
cipants were given $30 remuneration upon completion to become immersed in the video stimuli as they nor-
of the entire study. Undergraduate students who were mally would. For the erotic and humorous regulate
eligible for course credit were offered a choice of either trials, participants were instructed to detach or disen-
twocoursecreditsorthe$30remuneration;only1chose gage themselves from the stimuli by taking a distanced
the credits. or objective point of view. The instruction cues were
Both the online survey and the laboratory testing randomly ordered; however, they were balanced across
were approved by the UBC Behavioural Research stimulus conditions so that one half of both the erotic
EthicsBoard.Theonlinesurveyincludedanonlinecon- and humorous clips were experience and the other half
sent form, a demographics and general information were regulate. To insure that participants did not
questionnaire, four sexuality measures, and a results manipulate their responses during the regulate trials by
and debriefing page. With the exception of the closingtheireyes,lookingaway,orimaginingsomething
4
CONSCIOUSREGULATIONOFSEXUALAROUSAL
that would reduce their responses, they were told that makingiteasytoedit3-minclipsfromhisperformances.
they would be asked to recount various aspects of the As with the erotic clips, CD compilations of 12 Mitch
video scenarios. Hedberg clips were distributed to 20 volunteers who
At theend ofeach trial, participantswereinstructed, rated each clip on a scale ranging from 1 (not at all
bytextmessagingonthetelevisionscreen,toself-report amusing)to9(maximallyamusing).Basedontheratings,
maximum level of sexual arousal (SRMA), erection 8clipswereselectedthatdidnotsignificantlydifferfrom
(SRME), and amusement (SRA). Responses for sexual each other.
arousal ranged from 0 (not sexually aroused at all) to 9
(maximally sexually aroused). Responses for degree of Measures
erection ranged from 0 (no erection at all) to 9 (maxi-
mally erect). Responses for amusement ranged from 0
DGIF. The DGIF was based on measures used in
(not at all amused) to 8 (maximally amused). After each
online sexuality studies at Indiana University’s Kinsey
erotic trial, time was given for penile tumescence to
Institute (http://www.kinseyinstitute.org/research/sur-
returntobaselinebeforethenexttrialbegan.Oncetest-
veylinks.html). Although it is comprised of 22 items,
ingwascomplete,participantsweredebriefed andgiven
datafrom only8items wereofinterest forthepurposes
a chance to ask questions about the study.
of this study. Those items assessed age, gender,
language, ethnicity, sexual experience, sexual identity,
9
00 Stimuli relationship status, and undergraduate status.
2

y
ar Weusedatwo-stageprocesstoselecttheeroticvideo
u
br clips.First,wehad75malevolunteersselecttheirtop10 Sexual compulsivity scale (SCS). The SCS
e
2 F preferences from a list of 41 sexual behaviors and actor (Kalichman et al., 1994; Kalichman & Rompa, 1995,
2
7 traits (e.g., attractiveness of face and physique, breast 2001) is a 10-item measure of sexual compulsivity. The
4
1: size, etc.) that are typical of commercial pornography. SCS items (e.g., ‘‘I sometimes get so horny I could lose
0
t: We summed those preferences and then used the 7 control,’’ ‘‘I feel that sexual thoughts and feelings are
A
] most-frequently endorsed to guide selection of erotic stronger than I am,’’ and ‘‘I have to struggle to control
a
bi videos. Those 7 preferences were attractiveness of the my sexual thoughts and behavior’’) capture sexual pre-
m
lu female actor (body), attractiveness of the female actor occupations and undercontrolled sexual thoughts and
o
C
h (face), female actor exhibiting sexual pleasure, vaginal feelings that are core to the current understanding of
s
ti sex(femaleonhandsandknees;i.e.,‘‘doggystyle’’),oral sexual compulsivity. The responses for each item,
i
Br sex (male recipient), male ejaculating on the female’s ranging from 1 (not at all like me) to 4 (very much like

f
o face (i.e., ‘‘facial cum shot’’), and vaginal sex (female me), are summed and divided by 10 to give an overall
y
it on top facing male; i.e., ‘‘cowgirl’’). Over 200 videos sexual compulsivity score. The SCS has good internal
s
er weredownloadedfromanonlinecommercialpornogra- consistency, with Cronbach’s alpha coefficients ranging
v
i
Un phy links site. The videos were vetted for content and from .82 to .95, and is the only measure of sexual
[
: quality. Eighteen were selected, based on the 7 prefer- compulsivity that has been both well-validated and
y
B ences,tobeeditedinto3-minclips.Theamountoftime widely used in previous research (Dodge et al., 2004;
d
ade devoted to each type of sexual behavior was balanced Kalichman et al., 1994; Kalichman & Rompa, 1995,
o
nl across the18videoclips.The videoclips were dispersed 2001).
w
o
D as randomly ordered sets, saved onto two CDs, to 20
male volunteers. Volunteers rated each video clip on a SIS=SES. The SIS=SES (Janssen et al., 2002a,b) is
scale from 1 (not at all arousing) to 9 (maximally arous- a 45-item measuredesigned toassess thestrength ofthe
ing)andthenreturnedtheirratingstoourlaboratoryby sexual excitation and inhibition systems under various
mail,inself-addressedenvelopesthatwereprovided.We circumstances. Responses for each SIS=SES item range
averaged ratings for each video clip across volunteers from1(stronglyagree)to4(stronglydisagree)and,after
and then used repeated-measures analysis of variance reverse keying some items, responses are summed to
(ANOVA) to determine which 8 video clips would be form three subscale scores: (a) propensity for sexual
usedfortheexperiment.Theratingsforthe8videoclips excitation (SES; range¼20–80), (b) propensity for
that were chosen did not differ significantly from each sexual inhibition due to threat of performance failure
other. Comedy clips were selected in a similar fashion. (SIS1;range¼14–56),and(c)propensityforsexualinhi-
We perused various Internet comedy sites and noted bitionduetothreatofperformanceconsequences(SIS2;
names of stand-up comics that were rated most amus- range¼11–44). Internal consistency for the three
ing. Video clips of those performers were screened for subscales is good (Cronbach’s as¼.88, .82, and .66,
content.AcomediannamedMitchHedbergwaschosen respectively; Janssen et al., 2002a). Scores on the scales
fortworeasons.First,hisjokes donotcontainanysex- appear to be stable over time and normally distributed
ualcontent,whichwasanecessarycriterionforthecon- (to date, over 2,500 men have been tested; Bancroft &
trol condition stimuli. Second, his jokes are short, Vukadinovic, 2004).
5
WINTERS,CHRISTOFF,ANDGORZALKA
Duringanalysis,weincludedscoresfromtheSESand Becker, & Djenderedjian, 1981; Kuban, Barbaree, &
SIS2onlybecauseinhibitionduetofearofperformance Blanchard, 1999). Based on the recommendation of
failure, as captured by SIS1, measures sexual dysfunc- Kuban et al., we used a threshold of 3-mm changes to
tion, which was not related to any of our hypotheses. signify interpretable arousal.
Sexual Desire inventory–2 (SDI–2). The SDI–2 Data Analysis
(Spector, Carey, & Steinberg, 1996) is a 14-item, self-
Physiologicalandself-reportresponseswereaveraged
report test of interest in partnered sexual activity (e.g.,
overtrialsofthesametype(i.e.,eroticexperience,erotic
‘‘How strong is your desire to engage in sexual activity
regulate,humorousexperience,andhumorousregulate).
with a partner?’’) and solitary sexual activity (e.g.,
For the erotic stimuli, the two instruction conditions
‘‘How important is it for you to fulfill your desires to
(i.e., experience and regulate) were crossed with the
behave sexually by yourself?’’). Each item is scored on
three possible responses (PB, SRMA, and SRME) to
a 9-point scale, and responses are summed to produce
produce six outcome variables. For the humorous
an overall score ranging from 0 to 112. Scoring the
stimuli,thetwoinstructionconditionswerecrossedwith
SDI–2 also produces two subscale scores: DSD and
self-reported amusement (SRA) to produce two out-
SSD. Cronbach’s alphas for the two factors are .86
come variables. Paired-samples t tests were conducted
and .96, respectively. As there is no theoretical reason
9 to determine differences in outcome variables between
00 to believe that SSD is related to sexual arousal regula-
y 2 tion, we included only DSD in analyses. the instruction conditions. Small, medium, and large
ar effect sizes were interpreted according to Cohen’s
u
r
b recommended cutoffs of 0.2, 0.5, and 0.8, respectively
e
2 F Derogatis Sexual Functioning Inventory–sexual (Cohen,1992).Todetermineifsexualarousalregulation
2
7 Experiences Subtest (DSFI–SE). The DSFI–SE is improved, became worse, or remained stable over time,
4
1: one of 10 self-report subtests of the DSFI (Derogatis arepeated-measuresANOVAwasusedtoassesschange
0
t: & Melisaratos, 1979). It lists 24 sexual behaviors (e.g., in PB PPG scores and self-reported sexual arousal
A
] deepkissing,oralstimulationofyourpartner’sgenitals, across the four individual sexual arousal regulation
a
bi andintercourse—youinsuperiorposition),andtheindi- trials.Thesameprocedurewasusedforamusementreg-
m
lu vidual being assessed indicates which of those he or she ulation across amusement regulation trials.
o
C
h has experienced ever, and experienced in the preceding Regulation indexes were calculated by dividing
s
ti 60 days. Items endorsed for each period are summed average response during regulate trials by average
i
Br to create 2 scores out of 24. Internal consistency for response during experience trials. The resulting values

f
o theDSFI–SEisexcellent(Cronbach’sa¼.97;Derogatis were each multiplied by 100 to create four percentage
y
it & Melisaratos, 1979). regulation success indexes: sexual arousal regulation
s
r
e success index–PPG peak-base (SAI–PB), SAI–
v
i
Un PPG. Physiological sexual arousal was measured self-reported maximum arousal (SAI–SRMA), SAI–
[
: with a PPG purchased from Limestone Technologies self-reported maximum erection (SAI–SRME), and
y
B
(Kingston, Ontario). The PPG assesses the change in amusement regulation success index–self-reported
d
e
ad penile tumescence, which corresponds to the degree to amusement (AMI–SRA).
o
nl which a man is sexually aroused. Penile circumference To address H3 and H4, we calculated Pearson corre-
w
o
D is measured using a mercury-in-rubber strain gauge lation coefficients for the variables of interest. To inter-
placedtwothirdsofthewaydowntheshaftofthepenis. pret the strength of those correlations, we adhered to
As tumescence increases, the mercury column in the Cohen’s (1992) suggestion that coefficients of 0.1, 0.3,
strain gauge is stretched thinner, changing its cross- and 0.5 indicate the lower bounds of small, medium,
sectionalcircumference. Electrical resistance ofmercury andlargecorrelationeffectsizes,respectively.Aftercor-
is directly related to its cross-sectional area; therefore, relationcoefficientshadbeencalculatedforthevariables
any change in tumescence results in a concomitant of interest, we partialled out the effects of sexual desire
change in electrical resistance. The mercury-in-rubber andsexualexcitationfromthecorrelationsbetweensex-
PPG strain gauge is calibrated for precise measurement ualcompulsivityandthethreeSAIstoinsurethatsexual
before each use, and small changes in resistance can be desireandsexualexcitationwerenotaccountingforany
translated into millimeter changes in penile circumfer- possiblerelationbetweensexualcompulsivityandsexual
ence. Data from the strain gauge are relayed, via a arousal regulation success.
transducer, to a testing laptop computer. Limestone
provided us with software that displays, records, and
tabulates the incoming time-sequenced PPG data. Peak Results
minus baseline (peak-base [PB]) scores were used to
determine maximum millimeter changes in circum- All participants exhibited increased penile tumes-
ference during each trial (Abel, Blanchard, Murphy, cence (i.e., greater than 3-mm changes in penile
6
CONSCIOUSREGULATIONOFSEXUALAROUSAL
Table 1. SurveyMeasure DescriptiveStatistics Table 3. DescriptiveStatisticsforRegulationSuccessIndexes
(percentage RegulationSuccess)
Variable M SD Range
Index M SD Minimum Maximum
SCS 1.7 0.6 1–3.8
SES 58.7 6.5 42–70 SAI-PB 75.2 25.4 17.6 118.6
SIS2 28.2 5.2 17.5–39 SAI-SRMA 79.2 19.5 45.5 133.3
SDI–2—DSD 43.4 7.5 21–62 SAI-SRME 78.1 26.0 38.5 180.0
DSFI—Sexualexperiences 19.6 6.2 0–24 AMI-SRA 83.6 26.2 33.3 200.0
DSFI—Sexualexperiences 14.3 8.7 0–24
past60days Note. SAI–PB¼sexual arousal regulation success index–penile
plethysmograph peak-base; SAI–SRMA¼sexual arousal regulation
Note. SCS¼Sexual Compulsivity Scale; SES¼Sexual Excitation success index–self-reported maximum arousal; SAI–SRME¼sexual
Scale;SIS2¼SexualInhibitionScale,duetofearofperformancecon- arousal regulation success index–self-reported maximum erection;
sequences; SDI–2¼Sexual Desire Inventory–2; DSD¼dyadic sexual AMI–SRA¼amusement regulation success index–self-reported
desire;DSFI¼DerogatisSexualFunctioningInventory. amusement.
circumference) to the erotic stimuli and no sexual regulationindexesscoresdidnotdiffersignificantlyfrom
response to the humor stimuli. Therefore, all assess- eachother,F(1)¼1.69,p¼.20.Therewasalargevaria-
9 ments were deemed valid, and data from the 49 partici- tion in all four regulation index scores. The lowest PB
0
0
2 pantswereincludedinanalyses.Descriptivestatisticsfor regulationindexscorewas17.6%,indicatingthatthepar-
y
ar thesurveymeasuresarepresentedinTable1.Theresults ticipant who was best able to regulate his physiological
u
r
b ofpaired-samplesttests,withcorrespondingeffectsizes, responseexhibitedan82.4%decreaseinerectileresponse
e
F
2 for regulate versus experience trials can be found in duringerotic-regulatetrials.Thelowestregulationindex
2
7 Table 2. scoresforSRMA,SRME,andSRAwere45.5%,38.5%,
4
:
1 Participants, on average, were able to regulate their and 33.3%, respectively. No single participant scored
0

t: sexual arousal according to all three outcomes (i.e., highestonmorethanoneindex.
A
] PB, SRMA, and SRME). On average, they were also Despite the fact that, on average, participants were
a
i
b abletoregulatetheiramusementduringhumor-regulate able to regulate their arousal, some participants
m
u
ol trials. The effect sizes for PB, SRME, and SRA paired- reported and demonstrated increased sexual arousal
C
h samples t tests were moderate, whereas that for the during erotic-regulate trials. The penile responses of 8
s
i
t SRMA regulation comparison was large. participants (16.3%) were greater during erotic-regulate
i
r
B When responses were compared across the four trials than during erotic-experience trials. The least suc-
f
o individual sexual arousal regulation trials, there were cessful regulator was, on average, 18.6% more respon-
y
it no statistically significant differences in PB PPG scores, sive during regulate trials. The self-reported sexual
s
r
ve F(1, 47)¼1.01, p¼.32; or SRMA, F(1, 47)¼0.44, arousalresponsesfor5participants(10.2%)weregreater
i
Un p¼.73. In other words, SRMA regulation did not during erotic-regulate trials than during erotic-experi-
[
: change (i.e., improve or become worse) as testing encetrials,withthelowestscoringparticipantreporting
y
B
d progressed. The same was true for SRA across 33.3% more sexual arousal in the regulate condition.
e
ad amusement-regulate trials, F(1, 47)¼1.94, p¼.17. The self-reported maximum erection responses for 6
o
nl Descriptive statistics for the regulation success participants (12.2%) were greater during erotic-regulate
w
o
D indexes can be found in Table 3. Lower index values trials than during erotic-experience trials. Similarly, the
indicate increased regulation success. The mean least successful regulator self-reported 80.0% greater
Table 2. DescriptiveStatistics andPaired-samples t Testsfor Experience VersusRegulate Trials
OutcomeVariable M SD Minimum Maximum t(48) p Cohen’sd
EE–PB 27.7 12.2 4.67 55.4 5.39 <.001 0.55
ER–PB 21.0 12.3 2.79 49.9
EE–SRMA 5.6 1.7 2.3 8.5 7.33 <.001 0.81
ER–SRMA 4.3 1.5 1.3 7.0
EE–SRME 5.5 2.1 1.3 8.5 6.46 <.001 0.73
ER–SRME 4.1 1.7 1.3 8.3
HE–SRA 5.0 1.6 1.8 7.5 5.99 <.001 0.67
HR–SRA 4.0 1.4 1.5 7.5
Note.EE–PB¼eroticexperience–penileplethysmograph(PPG)peak-basemillimeterscircumferencechange;ER–PB¼eroticregulate–PPGpeak-
base millimeters circumference change; EE–SRMA¼erotic experience–self-reported maximum arousal; ER–SRSA¼erotic regulate–self-reported
maximumarousal;EE–SRME¼eroticexperience–self-reportedmaximumerection;ER–SRME¼eroticregulate–self-reportedmaximumerection;
HE–SRA¼humourousexperience–self-reportedamusement;HR–SRA¼humourousregulate–self-reportedamusement.
7
WINTERS,CHRISTOFF,ANDGORZALKA
Table 4. Correlation Coefficients for Erotic Experience Table 7. Correlation Coefficients for Survey Measures and
Sexual ArousalResponses Regulation SuccessIndexes
Variable EE–PB EE–SRMA Variable SAI–PB SAI–SRMA SAI–SRME AMI–SRA
EE–SRMA .562(cid:1) — Age .077 .217 .254 .047
EE–SRME .604(cid:1) .923(cid:1) p¼.600 p¼.139 p¼.081 p¼.747
DSFI–SE .153 .118 .245 .023
Note. EE–PB¼erotic experience penile plethysmograph (PPG) p¼.294 p¼.425 p¼.093 p¼.876
peak-base millimeters circumference change; EE–SRMA¼erotic DSFI–SE60 (cid:2).021 (cid:2).094 .120 (cid:2).110
experience–self-reported maximum arousal; EE–SRME¼erotic p¼.888 p¼.524 p¼.418 p¼.451
experience–self-reportedmaximumerection. SDI–2—DSD .091 .332(cid:1) .375(cid:1)(cid:1) .245
p¼.533 p¼.021 p¼.009 p¼.089
Table 5. CorrelationCoefficientsforEroticReg- SES .253 .289(cid:1) .301(cid:1) (cid:2).047
p¼.079 p¼.047 p¼.037 p¼.748
ulateSexual ArousalResponses
SIS2 (cid:2).506(cid:1)(cid:1) (cid:2).273 (cid:2).205 (cid:2).175
Variable ER–PB ER–SRMA p<.001 p¼.061 p¼.161 p¼.230
SCS .132 .216 .326(cid:1) .143
ER–SRMA .598(cid:1) — p¼.367 p¼.140 p¼.024 p¼.328
ER–SRME .685(cid:1) .873(cid:1)
Note. SAI–PB¼sexual arousal regulation success index–penile
9
200 Note.ER–PB¼erotic regulate–penile plethysmograph plethysmograph peak-base; SAI–SRMA¼sexual arousal regulation
y (PPG) peak-base millimeters circumference change; success index–self-reported maximum arousal; SAI–SRME¼sexual
r
ua ER–SRMA¼erotic regulate–self-reported maximum arousal regulation success index–self-reported maximum erection;
ebr arousal;ER–SRME¼eroticregulate–self-reportedmax- AMI–SRA¼amusement regulation success index–self-reported amu-
2 F imumerection. sement; DSFI–SE¼Derogatis Sexual Functioning Inventory–Sexual
2 *p<.001. Experiences subtest; DSFI–SE60¼Derogatis Sexual Functioning
7
4 Inventory–Sexual Experiences past 60 days; SDI–2¼Sexual Desire
:
01 Inventory–2; DSD¼dyadic sexual desire; SES¼Sexual Excitation

t: Table 6. Correlation Coefficients for Regulation Success Scale; SIS2¼Sexual Inhibition Scale, due to fear of performance
A
] Indexes consequences;SCS¼SexualCompulsivityScale.
ia (cid:1)p<.05.(cid:1)(cid:1)p<.01.
b
m
u Variable SAI–PB SAI–SRMA SAI–SRME
l
o
C

sh SAI–SRMA .515 — — and SRME indexes. These correlations were of a med-
i
t p<.001
i ium effect size.
r
B SAI–SRME .515 .846 —
of p<.001 p<.001 The correlation results for the survey measures and
ty AMI–SRA .226 .368 .329 regulation success indexes are presented in Table 7.
i
rs p¼.119 p¼.010 p¼.022 The PB index correlated negatively with sexual inhibi-
e
iv tion due to fear of performance consequences. The
[Un Note.SAI–PB¼sexualarousalregulationsuccessindex–penileplethys- correlation coefficient was of a large effect size. There
: mograph peak-base; SAI–SRMA¼sexual arousal regulation success
y was a trend toward significance for the correlations
B index–self-reported maximum arousal; SAI–SRME¼sexual arousal
ed regulation success index–self-reported maximum erection; AMI– between sexual inhibition and both the SAI–SRMA
d
oa SRA¼amusementregulationsuccessindex–self-reportedamusement. and the SAI–SRME. The SAI–PB did not significantly
l
n
w correlate with any other variables of interest. The
o
D
erectileresponseduringtheregulatetrials.Asimilarpat-
ternwasevidentforthehumorcondition.Seven(14.3%)
Table 8. Zero-order and Partial Correlation Coefficients for
participants reported more amusement, on average, Sexual Compulsivity and Sexual Arousal Regulation Success
during the humor-regulate trials than during the Indexes
humor-experience trials. The participant least able to
Variable SAI–PB SAI–SRMA SAI–SRME
regulatereported100%moreamusementintheregulate
condition. SCSa .132a .216a .326a
Across both erotic-experience and erotic-regulate p¼.367 p¼.140 p¼.024
conditions, PB, SRMA, and SRME were all signifi- SCSb .087b .069b .183b
cantly and positively intercorrelated (see Tables 4 p¼.564 p¼.648 p¼.224
and 5). The effect sizes for all correlations were large.
Note. SAI–PB¼sexual arousal regulation success index–penile
As is shown in Table 6, the SAI–PB correlated with
plethysmograph peak-base; SAI–SRMA¼sexual arousal regulation
boththeSAI–SRMAandtheSAI–SRME.Thosecorre- success index–self-reported maximum arousal; SAI–SRME¼sexual
lations were of large effect size. SRMA and SRME arousal regulation success index–self-reported maximum erection;
indexes also correlated very strongly with each other. SCS¼SexualCompulsivityScale.
aZero-ordercorrelationcoefficients.
The PB index did not correlate with the SRA index;
bPartial correlation coefficients controlling for the effects of sexual
however, the AMI index correlated with both SRMA
desireandsexualexcitation.
8
CONSCIOUSREGULATIONOFSEXUALAROUSAL
SAI–SRMA and the SAI–SRME correlated with DSD Unlike the correlations between the two self-report
and sexual excitation; these correlations were all of a sexualarousalmeasuresandamusementregulationsuc-
moderate effect size. The AMI–SRA did not correlate cess,thecorrelationbetweentheSAI–PB andtheAMI–
with any of the sexuality variables. Age and sexual SRA did not reach statistical significance. This is likely
experiences did not correlate with any of the regulation due to imperfect concordance between cognitive and
success indexes. physiological sexual arousal. Previous research has
Only the SAI–SRME correlated with sexual compul- shown that concordance between self-reported sexual
sivity. The correlations for sexual compulsivity with the arousal and penile response is good, at best (Haywood
other two SAIs did not reach statistical significance but et al., 1990; Sakheim et al., 1985).
were in the predicted direction. When we partialled out Meninoursamplewere,onaverage,abletoregulate
the effects of sexual desire and sexual excitation, the their physiological sexual arousal when instructed todo
strength of the correlations all considerably decreased so. During erotic-regulate trials, they exhibited a 25%
(see Table 8), and the correlation between the SAI– reduction in erectile response. This is consistent with
SRME and sexual compulsivity dropped below statisti- success rates from previous,well-controlled PPG faking
cal significance. studies in which success rates range from 26% to 38%
(Adams et al., 1992; Golde et al., 2000; Mahoney &
Strassberg, 1991; McAnulty & Adams, 1991). Some
9 Discussion meninthetwostudiesperformedbyMcAnulty,Adams,
0
0
2 and their colleagues were able to wholly suppress their
y
ar The overallpurposeofthis study was toexamine the penile response, whereas all of the participants in our
u
r
b effectivenessofemotionalreappraisalinregulatingmale study and the study by Mahoney and Strassberg exhib-
e
F
2 sexual arousal. Results showed that men, on average, ited some physiological arousal during regulate trials.
2
7 were somewhat able to regulate their physiological and Wesuspect thatstimulus modalitymayaccount forthis
4
:
1 cognitive sexual arousal, although there was a wide discrepancy. McAnulty, Adams, and their colleagues
0

t: range of regulation success. Whereas some men were used slides and accompanying audio vignettes, whereas
A
] very adept at regulating their sexual arousal, others both we and Mahoney and Strassberg used video sti-
a
i
b became more sexually aroused while trying to regulate. muli. Video stimuli are more arousing than slides or
m
u
l Further, the results indicate that the ability to regulate audio stimuli (Abel, Blanchard, & Barlow, 1981; Julien
o
C
h emotion may cross emotional domains; those men best & Over, 1988; Sakheim et al., 1985), likely increasing
s
ti abletoregulatesexualarousalwerealsothemostskilled the difficulty of sexual arousal regulation.
i
r
B at regulating amusement. Age, sexual experience, and Participants in our study self-reported 21% less sex-

f
o sexual compulsivity were unrelated to sexual arousal ual arousal and 22% less perceived erectile response
y
it regulation. Conversely, sexual excitation, inhibition, during erotic-regulate trials. These results are also
s
r
e anddesirecorrelatedwithsexualarousalregulationsuc- within the range of results reported in previous PPG
v
i
Un cess. Increased sexual excitation and desire were asso- studies. However, they are substantially different from
[
: ciated with poorer regulatory performance, whereas a those described by Beauregard et al. (2001). Their
y
B
propensity for sexual inhibition due to fear of perfor- sample self-reported a 60% reduction in sexual arousal
d
e
ad mance consequences was related to regulatory success. during regulate trials, despite the fact that video sti-
o
nl Ashypothesized,theregulationsuccessindexesforboth muli were utilized. This disparity may reflect the dif-
w
o
D self-reported sexualarousal and perceived degreeof erec- ferent testing environments, as participants were
tionwerepositivelyassociatedwithamusementregulation lying inside the bore of an MRI scanner. The scanner
success. This finding suggests that the ability to regulate environment is quite uncomfortable, and the consider-
emotion is generalized across emotional domains. To the able noise during scanning, which can reach 130 to
best of our knowledge, this has not been tested before. 140 decibels, is distracting, even with hearing protec-
Thatemotionregulationabilityappearstocrossemotional tion. Discomfort and distraction may have made it
domains is consistent with emotion regulation theory, as easier to regulate arousal.
well as findings from neurophysiological emotion regula- Men in our sample exhibited a very wide range of
tion research (Ochsner & Gross, 2005). Functional brain regulationsuccessacrossallfourresponsetypes:physio-
imaging studies have consistently identified a single emo- logical sexual response, self-reported sexual arousal,
tionregulatorysystemthatisimplicatedintheregulation perceived degree of erection, and amusement. Surpris-
of the various emotions tested (e.g., Beauregard et al., ingly, some men self-reported and exhibited increased
2001;Le´vesqueetal.,2003;Ochsneretal.,2002;Ochsner sexual arousal and penile response during the regulate
et al., 2004). Although we only examined the association trials. The same was true for SRA during humor-
between the regulation of two positively valenced emo- regulatetrials.Weconsideredtwopossibleexplanations
tions, we would expect that regulation ability for other for this increased responding: regulatory depletion
emotions, including those that are negatively valenced and anxiety’s potentially augmenting effect on sexual
(e.g.,sadness,anger,etc.),wouldbesimilarlyrelated. arousal.
9
WINTERS,CHRISTOFF,ANDGORZALKA
Ourstimuli,whichwereeach3-minlongtoallowfor Thistypeofresponseamplificationmayberelatedto
full sexual response, were of substantially longer dura- the well-established link between sympathetic arousal
tion than those used in previous emotion regulation and increased sexual interest and response (Bancroft,
research (e.g., Beauregard et al., 2001; Jackson et al., Janssen, Strong, Carnes, et al., 2003; Bancroft, Janssen,
2000; Ochsner et al., 2002; Ochsner et al., 2004). It is Strong, & Vukadinovic, 2003; Barlow, Sakheim, &
possible that emotion regulation, for some individuals, Beck, 1983; Dutton & Aron, 1974; Extona et al., 2000;
is only effective over a shortperiod of time, after which Heiman & Rowland, 1983; Kru¨ger et al., 1998; Kru¨ger
emotion regulation resources become depleted. The et al., 2006; Meston & Gorzalka, 1995, 1996; Meston
results from a study of sexual self-restraint and regula- & Heiman, 1998; Palace & Gorzalka, 1990; Wolchik
tory depletion by Gailliot and Baumeister (2007) pro- et al., 1980). Increased anxiety can have an augmenting
vide some support for this explanation. They found effect, via the sympathetic nervous system, on sexual
that participants had more difficulty with sexual response, particularly for women. In men, results are
restraint (i.e., inhibiting sexual thoughts and behaviors) moremixed(Barlowetal.,1983;Farkas,Sine,&Evans,
following regulatory tasks compared to control tasks. 1979; Hale & Strassberg, 1990; Heiman & Rowland,
Our results, however, did not support the regulatory 1983;Lange,Wincze,Zwick,Feldman,&Hughes,1981),
depletion explanation. Penile responses and self- with some men exhibiting increased sexual responses
reportedsexualarousalduringsexualarousalregulation when anxious, specifically those with a strong propensity
9 trialsremainedstableastestingprogressed.Ifregulatory for sexual excitation and a low propensity for sexual
0
0
2 depletion was happening, sexual arousal responses inhibition(Bancroft,Janssen,Strong,Carnes,etal.,2003;
y
ar wouldhaveincreased.Depletionofregulatoryresources Bancroft,Janssen,Strong,&Vukadinovic,2003).
u
r
b also does not explain why, for some men, responses Bancroft and colleagues (Bancroft, Janssen, Strong,
e
F
2 were greater during regulate trials when compared to Carnes, et al., 2003; Bancroft, Janssen, Strong, &
2
7 experience trials. Vukadinovic, 2003) suggested that transference of
4
:
1 Some of the men who exhibited increased respond- arousal (more generally known as excitation transfer-
0

t: ing in the regulate condition reported that they ence; Zillmann, 1983), arising from anxiety and the
A
] became more enmeshed in the stimuli while trying to concomitant increase in autonomic arousal, could
a
i
b regulate. This suggested that something other than augment sexual response, particularly in those indivi-
m
u
l regulatory depletion was happening. The other possi- duals with a strong propensity for sexual excitation
o
C
h ble explanation for increased responding is based on and low propensity for sexual inhibition. In our
s
ti findings from research on emotional control and study, participants may have experienced some anxi-
i
r
B thought suppression. In one of the first studies of ety while trying to perform the regulation task. If

f
o thought suppression, Wegner, Schneider, Carter, and attempted regulation can actually increase emotional
y
it White (1987) instructed participants not to think of response, especially when regulation fails (Wegner
s
r
e a white bear and then monitored their thoughts over et al., 1990), and if perceived poor performance height-
v
i
Un the following 5min. Initially, all participants were ens anxiety and general autonomic arousal, increased
[
: unable to rid their minds of a white bear. As the trial arousal may explain why some individuals responded
y
B
proceeded, however, some participants were able to more strongly while trying to regulate. In future
d
e
ad stop the thoughts, whereas others were not. In a fol- work, measures of state and trait anxiety may help
o
nl low-up study, participants were asked to try not to clarify the possible relation among anxiety, sexual
w
o
D think of emotionally charged sexual thoughts arousal, and sexual arousal regulation.
(Wegner, Shortt, Blake, & Page, 1990). They found Based on previous research (McAnulty & Adams,
that attempted thought suppression and intrusive sex- 1991; Nobre et al., 2004; Rosen & Beck, 1988), we had
ual thoughts, arising after initial thought suppression predicted that there would be good concordance
success, increased sympathetic arousal. Based on their between physiological sexual arousal and both self-
findings, they concluded that ‘‘suppression of exciting reportedsexualarousalandperceiveddegreeoferection
thoughts can undermine the process of emotional con- duringerotic-experiencetrials.Wealsoexpectedconcor-
trol’’ (Wegner et al., 1990, p. 415). They posited that dance to be good during regulate trials, as emotion reg-
the mere act of trying to suppress exciting thoughts ulationappearstomodulatethecognitive,affective,and
increases excitement, which then intensifies the initial physiological changes that accompany an emotional
emotional response. The cycle of attempted suppres- response (Jackson et al., 2000). Results confirmed our
sion, increased emotional response, followed by more predictions: Correlations among the PPG and two self-
attempts at suppression causes the emotional response reports within each instruction condition were all
to become more robust. Thus, attempted control has strong. In addition, the correlations among the three
the exact opposite effect than intended. For a small SAIs were all statistically significant. These results indi-
minority of our participants, being attuned to and cate that sexual arousal regulation, when effective,
attempting to regulate sexual arousal and humor seems to affect cognitive, affective, and physiological
actually increased responses. aspects of sexual response in an equivalent manner.
10