0115 966 7955 Today's Opening Times 10:00 - 20:00 (BST)

Understanding Predictors of Sexually Coercive Behaviour

Disclaimer: This dissertation has been submitted by a student. This is not an example of the work written by our professional dissertation writers. You can view samples of our professional work here.

Any opinions, findings, conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of UK Essays.


Purpose: In order to end or at least improve education and reduce sexual coercion, it is crucial to identify the variables that play a part in such behaviour.

Method: One hundred and thirty students (57 female and 73 males) ranging in age from 18 to 26 were randomly approached at one of three university campuses and completed a questionnaire battery.

Results: Overall 91.7% of respondents reported to have never been sexually coercive. Logistic regression analysis, using a dichotomised criterion of coercion, established that frequent pornography and potential sports team membership increased the potential of coercion.



Over the past three decades a plethora of research has concentrated upon the factors involved in and the repercussions of sexual assault, victimisation and perpetration (Abbey, & McAuslan, 2004). Sexual experiences and behaviours are often complex involving; moral and gender stereotyped beliefs, social learning, peer interaction and biological factors. An important part of maturity is developing appropriate sexual beliefs that enable the adolescent to control sexual urges and to manage the consequences when such urges are not controlled. Problems begin to arise when individuals apply pressure or force in a coercive manner in order to obtain sexual intercourse. Sexual coercion is persistent within society and both males and females serve as victims as well as perpetrators (Sigleman, Berry & Wiles, 1984). However, more often than not offenders are young, male and anti-social (Ellis & Walsh, 2000).

A vast spectrum of research (Koss, Gidycz, & Wisniewski, 1987, Forbes & Adams Curtis, 2001, Warkentin & Gidycz, 2007, Martín, Vergeles, Acevedo, Sánchez & Visa, 2005, Lacasse & Mendelson, 2007) also suggests that a large proportion of sexually coercive behaviours occur within university settings. Some suggest that the sexually promiscuous environment promotes sexually coercive behaviours and attitudes resulting in a number of sexual assault and date rape offences (Douglas, Collins & Warren, 1997, Brener, McMahon, Warren, & Douglas, 1999, Koss, Gidycz, & Wisniewski, 1987). Research using student samples also seems to suggest that males who play sports, use pornography, have hyper-masculine sex beliefs and mating strategies will report using and accepting sexually coercive behaviours (Martín, Vergeles, Acevedo, Sánchez & Visa, 2005, Lacasse & Mendelson, 2007).

Further problems arise with what constitutes sexual coercion and when does coercion become attempted rape or rape. Legal definitions are often confusing and broad, the general population usually define rape as vaginal, oral or anal intercourse or attempted intercourse against a person’s will or if the person is unable to give full consent (Humphrey & Kahn, 2000; Abbey, BeShears, Clinton-Sherrod & McAuslan, 2004). Sexual abuse which incorporates sexual coercion is generally defined as behaviour that covers a range of sex acts which includes physically forced sexual contact and verbally coercive intercourse.

Due to the ambiguous definitions police reported prevalence often underestimates the problem (Senn, Desmarais, Verberg, &Wood, 2000). Further to these issues some victims fail to realise that they are in fact themselves victims (Pollard, 1992). If the perpetrator is known to the victim they are often reluctant to label the incident as rape. There are also conflicting views to what constitutes ‘intercourse’. Sonenstein, Ku, Lindberg, Turner, & Pleck, (1998) argued that men attribute vastly different definitions to the word ‘sex’ when compared to women.

Bogart, Cecil, Wagstaff, Pinkerton, & Abramson, (2000) and Sanders, & Reinish (1999) both found that over half of their participants would not define oral sex as sex. Even more participants did not class an act to be ‘sex’ if one or both of those involved did not have an orgasm. As a consequence of these ambiguous definitions some sexual behaviours may not be perceived as sexual acts and both the victim and perpetrator may be under estimating the seriousness of some behaviours.

Research in this subject area has tended to use large representative samples (Abbey et al. 2004). A smaller number of studies have used self reports to estimate the prevalence of sexual coercion. The majority of research has concentrated upon the student population as many studies have shown a generalised problem among university students (Abbey et al. 2004, Martin et al. 2005). Sexual abuse and coercive behaviours will arguably be prevalent in the student population as individuals are still trying to form their identities. Adolescents have powerful motives for sexual intercourse due to pressure from social stereotypes and peers as sex is often viewed as a male conquest. Sexual coercion is said to be a social problem that reflect the stereotypes portrayed by a sexist society (Brownmiller, 1975). Maskay and McCreary-Juhasz (1983) argue that most students will be pressured to engage in sexual activity by either internal or external forces. Arguably males are more likely to hold sexists attitudes and condone sexually coercive behaviours which may be why they are predominantly perpetrators (Monson, Langhinrichsen-Rohling,, & Binderup, 2000).

Female students have continuously been highlighted as being at high risk of coercion in to unwanted sexual behaviour (Mouzon, Battle, Clark, Coleman,& Ogletree, 2005) as females aged 16 to 24 experience rape at a level four times higher than the general female population (Parrot, Cummings, Marchell & Hofher, 1994). In a vast range of research that has been conducted with university students what became apparent is that rape perpetration ranges from 6% to 15%, sexual assault ranges from 22% to 57% and 85% of females reported that their date was a perpetrator of sexual coercion(Abbey et al. 2004).

These varying estimates may be in part due to differing methodologies or the varying definitions of sexual coercion. However what remains clear is that sexual assault and coercion is a serious problem even if we only consider the lower estimates. Coupled with these estimates, 25-60% of male students admitted they would force a female to have sex if they thought they wouldn’t be caught (Russell, 1993, cited in Emmers Sommer, & Allen, 1999).

Despite a wide range of research in this area over the past decade it has become apparent that there are difficulties in eliciting reliable accounts of sexual coercion and therefore intercourse. Some participant may minimise the number of sexual partners where as in many societies young males exaggerate sexual experiences because it as seen as a measure of manhood (Boohene, 1991). As a whole sexual victimisation has been a serious problem on university campuses. Despite efforts from the universities highlighting the risks students put themselves at not much has changed and the prevalence of such behaviours has shown little fluctuation over this time period. It is important to note that sexual aggression varies greatly among different universities (Koss, et al.1987).

It seems viable that in order to attempt to tackle this ongoing problem the factors or social situation that influence or exacerbate such behaviours need to be identified. All men are arguably exposed to similar social experiences, however only a small proportion of these go on to engage in sexually coercive behaviours. The question must be why does this minority conduct these behaviours and accept sexual coercive tactics, what are the mediating factors and how can it be prevented?

The literature examining the use and acceptance of sexual coercion suggests that certain variables remain consistent (Emmers-Sommer, & Allen, 1999). Despite over 47 variables being associated with sexual coercion (Frank, 1989) they generally fall in to one of two categories; psychological and physical pressure (Struckman-Johnson, 1988). Psychological takes the form of verbal pressure, where as physical involves threat of harm to an individual if they refuse to engage in intercourse. Koss (1989) referred to psychological as being implied force and physical as being explicit force. Several studies have concluded that verbal coercion is often the most commonly used coercive tactic (Mouzon et al. 2005). Within a university setting it was apparent in the majority (58.6%) of coercive incidents, the male carried on even though the female had stated ‘No’ (Muehlenhard & Linton, 1986).

Verbal coercion is a negative level of persuasion that can take the form of; blackmail, persuasion, lies, false promises, guilt or threats to end a relationship (Walner-Haugrud, 1995). Many males use verbal coercion as it leaves the victim with a lack of choices to pursue that don’t have severe social or physical consequences (Heise, Moore, & Toubia, 1995). Sexual intercourse may actual seem like the easiest option in order to escape constant pressure or to avoid negative consequences including threats to a relationship (Livingston, Buddie, Testa & VanZile-Tamsen, 2004). The anticipation of a negative reaction from their partner was significantly associated with a woman’s decreased ability to refuse sex and over 40% of females engaged in intercourse due to constant pressure (Koss, 1985) .

However when females are seen to ‘give in’ to intercourse due to the social and emotional pressures it is debated whether this is true victimization (Muehlehard & Peterson, 2004). Regardless of this viewpoint the presence of verbal coercion is consistent in dating relationships, 93% of coercive acts reported by females involved a male that was known to them (Testa & Livingston,1999). With regard to university campuses, social circles constantly overlap thus increasing the likelihood the victim, her friends or acquaintances will now the perpetrator.

Dating situation can often be ambiguous with pressure from societies expectation of gender stereotyped roles and traditional dating scripts. These expectations and pressures can often lead to attitudes of rape justifiability and victim willingness. Males are often viewed by a western society as being dominant and aggressive. Therefore some males want to uphold this view often resulting in hyper-masculinity. Malamuth, Linz, Heavey, Barnes, &Aker (1995) argue the views of these males will be strengthened especially in a university setting.

Their ‘Confluence model’ states that variables such as sexual promiscuity or hyper-masculinity will influence sexually coercive behaviour and attitudes. Some male’s think they are justified in being coercive towards their date especially if they the female accepted the date, they have paid for the date or the female invited him back to her residence (Feltey, Ainslie & Geib, 1991). Such behaviours are justified as being acceptable as certain males believe because the female has ‘allowed’ him to take the ‘lead’ role then she must be willing to engage in intercourse or he is justified in pursuing intercourse without consent as she ‘owes’ him (Parrot, 1990). Surprisingly some females agree that a man is justified in having sex with a female against her will if she willingly goes back to his house (Muehlenhard, 1988).

Even when dating scripts are reversed and the female initiates or pays for the date some males still perceive this as the female’s willingness to engage in intercourse. These males are likely to assume that females who initiate or take control are experienced or ‘easy’. The males described in these situations are more likely to blame the female if sex occurs, are more likely to endorse rape justifiable attitudes and are less likely to perceive that coercion took place (Parrot, 1990). Reassuringly, 98% of university students agreed that using coercion to obtain sex is not acceptable regardless of other dating factors (Cook, 1995).

Some behaviours do not have the ambiguity of whether they are coercive or not for example the use of force, threat or weapons are unequivocal examples of sexual coercion. Koss et al. (1987) argued that over half of their student sample were sexual victimized and 6% stated they had been raped in the past year whilst at university. Fisher, Cullen,& Turner (2000) reported much lower figures of around 11% for sexual victimization and 2% raped in the past 7 months. These differences could indicate a decline in sexually coercive behaviour however it may be more likely due to differing definitions of sexually coercive behaviours and the different time frames used.

Taking even the lower estimates it appears that a significant proportion of students have endured some form of sexual coercion during a relatively small time frame. However research has continually highlighted that the number of reported incidents of sexual coercion by females will always out weight the number of self reports by males (Koss et al.1987). Therefore it appears that there are marked gendered differences in the way both sexes perceive and report sexually coercive behaviours.

The gendered differences in attitudes towards sexual behaviour general highlight that males are more sexually permissive and more interested in casual sex and women tend to have an emotional tie (Forbes, & Adam-Curtis, 2004). Despite some decline in these gender differences, the issue of casual sex remains the same. Males are twice as likely as females to approve of casual sex and four times as likely to think attractiveness is a sufficient enough reason to have sex (Feignbaum & Weinstein, 1995). Males also believed it is ‘normal’ for intercourse to occur after eight dates whereas females argued it would be nearer to 12 (Cohen & Shortland, 1996).

This sample of male students may therefore have unrealistic dating scripts. Expecting intercourse to occur four dates sooner than females may result in the female perceiving the males advances as being sexually coercive whereas the male believe his behaviours are the next ‘logical’ step. Such differences in sexual expectations may be a major contributing factor to the discrepancies between recorded and self reported behaviours. Discrepancies may also be attributed to some men perpetrating multiple acts as recidivism rates are often low. Rubenzahl (1998) reported that within his student sample 29% reported being sexually coercive on more than five occasions.

The underlying differences in attitudes towards sexual coercion may lie with the conditioning of males and females, and their assumed gender identity. Embracing the identity of being masculine or feminine will have an effect upon perceptions of appropriate or inappropriate behaviour, and the decisions regarding engaging in intercourse. As gender identities are diametrically opposed, society views real mean as those who have sex and women, who are ladies, do not (Muehlenhard, & McCoy, 1991).

Males who adopt this gender identity have more traditional values and expectation. As discussed previously, these males will ask for the date, pay for the date, initiate sex and ultimately endorse sexually coercive strategies. Such males are also more likely to believe that females engage in ‘Token Resistance’, in that they say no in order to preserve their feminine identity, even though they are willing to engage in sexual intercourse. Others believe that token resistance may actually be a change of intention rather than a self protective factor (Shortland, & Hunter, 1995). Around 40% of college women admitted they had engaged in token resistance, even though they planned to have sexual intercourse (Shortland, & Hunter, 1995) and 17% make regular practice of it (Muehlenhard & Hollabaugh, 1988).Therefore, even if the female truly means ‘No’, a minority of males will still pursue intercourse as they perceive it to be the female being coy or bluffing to protect her reputation.

Given that males and females are expected, by the rules that govern society, to act in opposing ways and that some males believe because of these expectations, females say ‘no’ when they in fact mean ‘yes’, it is hardly surprising that coercive behaviours become problematic. Confusion for males will obviously arise when females they have previously engaged in sexual intercourse with have used token resistance. Problems will arise when the same male engages with other females, yet uses his previous experience as a rule of thumb. Some women use verbal and non verbal cues, along with the physical resistance to refuse sexual advances from a male. As discussed verbal cues are often confusing especially as ‘no’ can often mean ‘yes’. Consent is also usually signalled by doing nothing during sexual advances, therefore further issues may arise for the male in that lack of consent is not necessarily a spoken ‘no’. Men are also more likely to perceive a wide variety of non verbal, verbal and situational variables as cues for a female’s sexual interest (Koukounds, & Letch, 2001). Non verbal cues are said to be the least direct and effective methods of resistance (VanWie, & Gross, 2001).

Environmental factors such as a college setting and, “perpetrator characteristics, personality and previous sexual experience enhance a man’s willingness to act upon his coercive beliefs and attitudes” (Berkowitz, Burkhart, & Bourg, 1994 p.6). White, Donat, & Humphrey’s (1995) suggest that it is not specific cognitive beliefs but negativity towards women that is associated with sexual coercion. Therefore holding supportive attitudes in the absence of other variables such as hostility may not contribute to sexual coercion. Attitudes of a coercive nature however do not just appear but they are in part due to socialisation. Attitudes are modelled and developed at home, school, by the media and our peers; they are also in part influenced by our own personality. Hostility towards females and sensation seeking is such aspects of personality that is linked towards sexual coercive behaviours (Berkowitz, 1992).

Much of the sexual coercion literature has focused upon fraternity and sports team members with the majority suggesting that those who individuals who are associated with either group are more likely to commit sexual assault. This viewpoint has gained momentum within the popular media (Martin, & Hummer, 1989) however from an academic perspective the results are mixed. Koss and Gains (1993) argue that both athletic membership and living in a fraternity were significantly related to the severity of the incident but only athletic membership was a significant predictor. However, Lackie and DeMan (1997) argues the opposite in that fraternity membership not athletic membership predicted sexual assault. One of the many reasons for this is that, some sports teams especially on modern university campuses are mixed sex or if they are single sex affiliated teams share transport and training resources.

Also many fraternities do not represent the stereotypical view of hyper-masculine, alcohol fuelled environments and they actually promote equal rights and community service. However a number studies, (Boeringer,1999, Sawyer, Thompson, & Chicorelli, 2002, Forbes, Adams-Curtis, Pakalka, & White, 2006) have highlighted relationships between group memberships, acceptance of rape myths, objectification of women and admiration of violence. Each of these variables has been constantly related to sexual aggression and coercion. It is important however to understand that such attitudes are developmental and aggression towards females does not just manifest during college. Such attitudes develop throughout the life course and participation is sports, college experiences, the pressure from peers or male gender roles may reinforce these existing attitudes (Forbes, & Adam-Curtis, 2001).

Peers will provide justification and support for each other’s sexually aggressive tactics. Athletes were over responsible for 19% of sexual assaults despite only accounting for 3% of the overall university population (Crosset, & Benedict, 1995). On the other hand self report questionnaires highlighted that sexual assault by members of athletic teams or fraternity memberships were no different than non members (Jackson, 1991). Closely tied in with the theory of masculinity is the role of males peers in socialisation and initiation of sexual beliefs (Brooks-Gunn, & Furstenberd, 1989). ‘Frat’ membership is good reflection of the pressure applied to peers to hold narrow masculine views (Senn, Desmarais, Verberg, &Wood, 2000). The link between ‘frat’ houses, sports team and sexual coercion may be a reflection of the psychological characteristics of the group such as hostility and dominance over women not the purpose of the group itself that is associated with sexual coercion.

Further variables such as pornography usage have been linked to sexual assault and sexual coercion. Within some student populations and arguably some ‘frat’ houses or university halls of residence social affairs involve pornographic entertainment which may further reinforce sexually aggressive behaviour. The concern surrounding pornography usage is that substantial research (Williams, Cooper, Howell, Yuille, & Paulhus, 2009) argues it promotes sexual misbehaviour. Such effects are arguably stronger when the exposure is self induced (e.g. DVD’s and internet based searches) rather than involuntary (e.g. pop ups and spam emails).

Experimental studies (Williams et al. 2009) seem to suggest that exposing males to pornography increases fantasies, willingness to rape and acceptance of rape myths. Malamuth, Neil, & Koss, (2000) argue that there is an association between pornography usage and sexually aggressive behaviours in that those who had higher levels of viewing were at a greater risk for being sexually aggressive. However they also suggest that those males who are already aggressive in nature are at a far greater risk than those who are not aggressive. Boeringer (1994) had previously concluded that males exposed frequently to violent pornography were six times more likely to report rape supporting attitudes and behaviours compared to a sample with low exposure.

Exposure to pornography may reinforce the notion that having multiple partners is normal and having a steady relationship in undesirable. Pornography also fails to deal with real life relationship problems. Some issues such as affection, communication, dating scripts or the consequences of promiscuous sexual behaviour are avoided. Viewing material of this nature especially on a frequent basis exposes the individual to perverted sex acts that are otherwise unimaginable and instead are regarded as exciting thus desensitising the individual to the content. Some 12% of males imitated pornography during a sexually abusive incident (Bergen, Raquel Kennedy, 2000).

It is important however to remember that a number of individuals watch pornography on a regular basis and no not engage in sex crimes. It is important however to argue that the relationship between pornography and sexual coercion is not simply due to levels of exposure and moderating factors must be considered. The combination of peers with attitudes that justify coercion, pornography usage and hyper-masculinity and an appreciation of violence often found in some sports teams may all interact and contribute to sexual coercion.

The age at which individuals have their first sexual experience and the number of partners they have had may lead to more coercive attitudes and behaviours. Factors such as impersonal sex and lack of intimacy are associated with both promiscuity and sexual coercion. A number of studies (Forbes et al. 2001, Senn at al. 2000, Martin et al. 2005) have found those that had admitted to endorsing or perpetrating sexual coercion had more sexual partners and early age of first intercourse than a non coercive sample. Those males that are sexually coercive and have had a high number of partners will be active daters and the frequency of their sexual contact can increases their likelihood to commit sexually coercive acts (Byers & Eno, 1991). In summary, “the more sexually the young man, the longer he is active and the greater number of sexual partners the more likely some assaultive behaviour will occur at least once” (White & Koss, 1993 p.144 ).

It is apparent that a number of different variables can contribute to the acceptance and reinforcement of sexually coercive behaviours. The complexity of this relationship can arguably only be understood by taking human evolution and natural selection into account. In line with previous research evolutionary theory argues sexual coercion occurs in aggressive males who are more eager to mate, are sexually assertive and are less discriminative when choosing a mate (Thornhill & Palmer, 2000). Those male with a high sex drive that are thwarted by a female when they are sexually aroused may respond in a sexually coercive manner (Kanin, 1969).

Females on the other hand have different reproductive strategies as they are more selective because they contribute more long term effort to the reproduction of the offspring (McKibbin, Shackelford, Goetz, & Starratt, 2008). Such difference may be attributed to the gender differences in attitudes towards casual sex and some males desire to have multiple partners. Evolutionary psychologists propose a number of different explanations of individual difference in sexual coercion. Those males that have low parental investment due to the nature of their unstable environment such as university may trigger development patterns leading to a high mating effort.

These individuals may have multiple partners due to their desire for greater investment in reproduction. This potential for reproduction ‘pays off’ as a greater number of partners may result in a greater number of offspring. Such behaviour patterns are supported by risk taking strategies that are apparent in sexually coercive behaviour. Adaption theory suggests that sexual coercion is a conditional strategy. Its main argument is that those males who have low mating value and have failed with non sexually coercive tactics use conditional mating strategies such as sexual coercion in order to improve their number of sexual partners because these behaviours are now adaptive (Gladden, Sisco, & Figueredo, 2008).

Further arguments surround the notion that sexual coercion is a by-product of trait differences between genders to ensure that males do not miss sexual opportunities. Selection for a greater variety in sexual encounters (casual sex) by males and for some to act aggressively could result in sexual coercion as a side effect of these traits. From an evolutionary perspective those individuals that adapt behaviour to context adaptive problems (domain specific adaptive mechanisms) tend to out compete rivals that pursue invariant behaviour tactics. It may be in these specific context cues trigger coercive behaviours such that a male who was coercive as a by product of an interest in casual sex maybe also be high in aggression and be willing to use force as they lack normal inhibitions.

On the contrary those males that use coercion because they have low status may have normal inhibitions against forcing women however may alter these inhibitions when the female acts inappropriately according to gender stereotypes i.e. initiates the date or invites him back to her home. Based on this argument it appears that different predictors of sexual coercion will correlate preferentially with different types of offences. By expanding this research to consider more fundamental evolutionary constructs such as mating effort and sex strategies, it may be possible to think about sexual coerciveness theoretically.


It is predicted that a higher score on a measure of self-reported sexually aggressive tactics will be positively correlated with greater impulsivity, more stereotyped sexual attitudes, active sports team participation, greater pornography use, higher levels of mating effort and residing within university halls of residence. It is further predicted that males will believe using sexual coercive behaviours is more acceptable than females and they will admit to higher levels or perpetration than females.



The study will be a within-subjects correlational study, and examines the use and acceptance of sexually coercive tactics among university students. The study will use higher self-rated scores on measures of sexual coercion as the criterion variable and sexual experience, impulsivity-sensation seeking, sexual beliefs and whether the participant lives in halls, is a member of a sports team, if they use pornography and mating effort as the predictor variables.


Participants were recruited via opportunistic sampling around various public areas of a University campus and online via a social networking website. Participants in this study composed of University students ranging from first years to postgraduate level. A total of 130 students participated in the study 73 male and 57 female. The participants included in the analyses were aged between 18 and 26 the overall mean age was 22.2 years (SD= 1.58).

Furthermore about 43 % were said to be in a long term relationship, 43 % dated casually, 15 % didn’t date and only2 % were married or engaged. The vast majority of participants (96 %) indicated they were heterosexual, with 2 % being homosexual and 1 % being bi sexual.


Within the questionnaire battery participants were asked to provide demographic information which included; gender, age, the age at which participants willingly has sexual intercourse, the number of sexual partners they have had, whether they lived in halls of residence, whether they were part of a single or mixed sex sports team, and whether they used pornography.

Further measures include:

The SES Scale (sexual experiences scale, Koss & Oros, 1982).

A 10 item scale was used to measure the participant’s involvement in acts of sexual coercion. From this measure it is possible to collect data regarding four different types of sexual aggression (sexual contact, sexual coercion, attempted rape and rape). As this scale is currently devised for male participants a revised version for females will created simply by changing the use of male to female. This will be altered as it may be possible for some females to use sexually coercive behaviours.

The respondents answer ‘yes’ of ‘no’ to a series of items in a self report format. With regard to rape it will not be possible for women participants to respond ‘yes’ however the scale will be able to identify other sexually coercive behaviours The measure is used to detect sexual assault even when the perpetrator is unaware that his behaviour constitutes a crime. The SES is one of the most frequently used measures of assessing sexually aggressive behaviours and reliability and validity are very good. The SES was found to be of good internal consistency reliability with a Cronbach alpha =.89, and has re-test reliability of .93 (Koss & Gidycz, 1985) and validity of .61(Koss & Gidycz, 1985). The scale has been used with women were the internal consistency was found to be .74(Koss & Gidycz, 1985).

The Impulsive Sensation Seeking subscale of the Zuckerman-Kuhlman Personality Questionnaire (ZKPQ-ImpSS; Zuckerman, Kuhlman, Joirement, Teta, & Kraft, 1993).

A 19 item impulsivity sensation seeking subscale requires participants to answer true or false to whether they felt the statements were descriptive of themselves. Higher scores indicate participants who have higher levels of sensation seeking. Research has found the scale has a Cronbach alpha of .77 and a test retest reliability of .80 (Zuckerman, 2002).

The Sexual Strategies Questionnaire (SSQ Struckman- Johnson and Struckman-Johnson, 1991).

Participants are asked to consider 10 different items in relation to a date scenario vignette. Item are devised to measure acceptability of different sexually coercive tactics. Within the scale there are 10 levels of sexual coercion, the least severe being use of isolation (item 1) ranging to moderately severe, alcohol used as a specific tactic (item 6) and most severely the use of weapon (item 10). Participants are require to answer on a 7 point Likert scale ranging from the behaviour being ‘always unacceptable’ to ‘always acceptable’. The scale demonstrated good internal consistency with a Cronbach alpha =.87 and a good test re-test reliability of .82 (Struckman-Johnson, Struckman-Johnson, 1992).

Sexual Beliefs Scale (SBS; Muehlenhard & Felts, 1998). A 20 item questionnaire aimed to measure five beliefs related to rape. The five dimension subscale measures; token refusal, leading on justifies force, women like force, men should dominate and no means stop. Each item is scored on a 4 point Likert scale which ranges from ‘disagree strongly’ to ‘agree strongly’. Those who score higher are said to have a higher agreement with the subscale it is measuring. The scale demonstrated a good internal consistency with a Cronbach alpha =.72 (Martin et al. 2005).

The mating effort scale (MES; Rowe, Vazsonyi & Figueredo, 1997).
The MES is a 10 item scale used to examine intra-sexual competition in males and females. It measures aspects such as the effort (e.g. time and energy) devoted towards obtaining and maintain sexual partners. Reponses range from ‘strongly agree’ to ‘strongly disagree’ on a 5 point Likert scale. The scale has demonstrated a good internal consistency with a Cronbach alpha =.71 (Gladden, Sisco, & Figueredo, 2008).

The Paulhus Deception Scale (PDS; Paulhus, 1998). The PDS is a 40 item self report scale used to examine a participant’s tendency to give socially desirable answers. The PDS captures two forms of socially desirable answering, measured via two subscales; Self-Deceptive Enhancement and Impression Management. Respondents are asked to rate each item on a 5 point Likert scale ranging from ‘Not true’ to ‘Very true’ choosing an item that best represents their view point. The scale has demonstrated a good internal consistency with a Cronbach alpha =.89 (Cash, Phillips, Santos & Hrabosky, 2004).


Ethical approval was obtained before any primary data collection and the terms of the ethical protocol were faithfully observed in the conduct of the project. A pilot study (n=15) was conducted to test the suitability of the questionnaire battery’s layout and structure. Results were pleasing and the battery was deemed suitable. Each participant was asked whether they would participate in a study investigating relationships at university.

All participants received a copy of the consent form. Once individuals agreed to participation they signed the consent form and were given a copy of the questionnaire battery and asked to complete each section in the order they appeared. They were also asked to create a pseudonym with which they could later indentify themselves with if they wished for their data to be removed from the study. Once the completed questionnaire had been completed all participants received a copy of the debriefing form.

Those who participated online were asked to follow an electronic link to the questionnaire. Participants will first be taken to a page containing the consent form. To agree to participation participants clicked the appropriate ‘agree’ link and were taken to the questionnaire battery. Once online participants completed the questionnaire they were taken to the debriefing page which they could print out if they wished to do so. It was ensured that no scores were linked to individual participants as SPSS calculated the total scores when all the data had been collected.

Ethical Issues

Given the sensitivity of the research, the terms of the ethical protocol were faithfully observed in the conduct of the research. Informed consent was obtained prior to testing, the researcher requested that participants sign a consent form. Participants were informed that they are under no compulsion to engage, and even if they did they weren’t obliged to continue if they do not wish to. To ensure anonymity consent forms and questionnaires were kept separate and participants were asked to create a pseudonym. Participants were also informed that only those with a direct involvement in the research would be able to view the data, as the numbers were aggregated so not link to any particular individual.

As the questionnaire battery contained potentially sensitive material which could have potentially aroused negative emotions participants were prior warned of the frankness of the questioning when they are approached to complete the questionnaire. Given some participants could have felt troubled by memories the questionnaires might ask, participants who completed the study were provided with the contact details of support organisations such as Victim Support, The Samaritans, sexual health services and alcohol dependency services. After the questionnaire had been completed participants were told that they may withdraw their data from the study up to a month after the day they participated.

Table 1

Variables and Measures

Variable Measure

Acceptance of sexually aggressive tactics Sexual Strategies Questionnaire

Sexual assault perpetration Sexual Experience Scale

Sensation seeking Zuckerman-Kuhlman Personality Questionnaire

Acceptance of rape tactics Sexual Belief Scale

Mating effort Mating Effort Scale

Social desirability The Paulhus Deception Scale

Halls of residence Demographics

Athletic team membership Demographics

Pornography use Demographics


Table 2, presents means, standard deviations and internal reliabilities for all the measures used within the study. Reliabilities within the measure were mostly satisfactory, impulsivity and sexual belief were bordering on satisfactory whereas mating effort had low reliability (α =0.45). As a result of its low reliability, it was decided that mating effort was unsuitable for use. The internal reliability remained strong when both sexual strategy and social desirability were broken down into their subscales. The sexual strategies subscales of covert and overt tactics were both highly reliable (0.62 and 0.82 respectively). The subscales of social desirability, Self-Deceptive Enhancement (SDE) and impression management (IM) also demonstrated reasonably high internal reliability (0.68 and 0.68). The subscales of sexual beliefs also had good internal reliability ranging from, 0.53 to 0.93.

Table 2.

Means, standard deviations and reliabilities of all measures and subscales

Females Males

n Mean SD n Mean SD Cronbach’s α

Sexual experiences

Sexual experience and group membership

Table 3, summarises data regarding sexual experiences, coercive behaviours and group membership. The majority (61%) of students lived within a rented house or flat, however, all of those (15%) who were in their first year of university lived within halls of residence. Only a minor number (1.2%) of instances were perpetrated by individuals who lived within halls of residence. A significant proportion (78%) of males played for a sports team however only 34% of females played in similar teams. Sports team membership was significantly related to sexual coercion, t(126)=6.35, p<.001 suggesting that those who played sport endorsed more sexually coercive attitudes.

The average male student in this sample lost their virginity at 16 years (SD=1.59) and had a mean number of 10 sexual partners (SD=7.44). Females were slightly younger at 15 years (SD=8.25) but had a similar mean number of 10 partners (SD=8.25).

With regard to pornography, females overall usage was significantly at the lower level of viewing with16% stating ‘never’ and 17% stating they had, ‘once but not in the past year’. Males on the other hand were significantly at the frequent levels of usage, with 29% stating they used pornography ‘weekly’ and over 2 % using it ‘daily’. Pornography usage was also significantly related to sexual coercion, t (129)= 23.29, p<.001 meaning those who frequently used pornography had more sexually coercive attitudes. Those high in pornography use frequency had higher scores on impulsivity (M=12.6, SD=2.6) compared to lower frequency users (M=11.7, SD=2.6). Higher users had also engaged in sexual intercourse more frequently (M=10.8, SD=7.3) than less pornography frequent users (M=9.3, SD=8.1).

With regard to perpetration of sexually coercive behaviours the majority of perpetrators were male they also lived outside of halls of residence, half

played sport and all of these males used pornography on a frequent basis. Only one female admitted to engaging in sexually coercive contact behaviours, she lived in halls of residence yet did not play sport or use pornography. Males held significantly more sexually coercive ideas when compared to women, t (129)=12.85, p<.001.

Data reduction: Sexual strategies questionnaire (SSQ)

As the majority of respondents reported less than average scores on the sexual strategies questionnaire, it was felt results would be more promising if they were transformed using a square root transformation, which were then used for the analyses. The suitability for factor analysis was assessed using the correlation matrix (Table 4). An inspection of the correlation matrix revealed a significant number of coefficients of aroud 0.3 and over. Such sufficient coefficients levels indicate that the variables have a sufficient level of inter relationships between them. The Kaiser-Meyer-Olkin was .69 therefore above the recommended value of .6, indicating that the factors share a large proportion of the common variance.

A principle component analysis revealed the presence of two Eigenvalues exceeding one, explaining 49.5% and 23.2% of the variance. Further inspection of the scree plot concluded that there was a clear break after the second component and these should be retained for further inspection.

A varimax rotation was performed to aid the interpretation in the rotated solution. In the rotated solution (table 5) both components show strong loadings with all variables only loading on to one component. Component one, labelled covert tactics consisted of alcohol use for both victim and perpetrator, as well as, being a specific tactic, use of isolation, silence and promise of positive consequence. Component two, labelled overt tactics consisted of use of drugs, weapons, force, alcohol and threat of a negative consequence. The two factor solution explained a total of 72.5% of the variance with component one (covert) accounting for 23,3% and component 2 (overt) 49.3%.

Relationship between overt and covert tactics

To examine whether the use of tactics was contextually based, the components extracted from the factor analysis of the SSQ were also applied to the SES. Therefore, the two tactics of overt and covert were matched to items on the SES. The items on the SES that were labelled as being a covert tactics included the use of arguments or authority. The items labelled as overt tactics included the use of physical force, alcohol, drugs. A correlation between the two tactics reported a significant negative relationship, r. =-.46, p<.001, indicating that those who preferred to use an overt tactic did not also use a covert tactic. It May be possible to assume that individuals do not switch between tactics regardless of the situation context.

Table 6, presents the correlations between the main measures used within the study. As the majority of perpetrators were male it was assumed that it would be more suitable for the remainder of the study to just concentrate upon this population. From the data presented in the table, it seemed important to discuss certain findings. As expected, subscales within scale were significantly correlated (p <.001) to each other. Not surprisingly there were also clear and significant correlations between SSQ, impulsivity and sexual beliefs (all subscales except, men should dominate and no means no).

SSQ was also negatively correlated with the self-deceptive enhancement subscale of the social desirability measure. Importantly, covert tactics significantly correlated more often with other variables more so than overt tactics. However, the only measure overt did highly correlate with was impulsivity. Covert tactics significantly correlated with impulsivity along with social desirability, (total and self-deceptive enhancement but not impression management) and all subscales of sexual beliefs except no means no.

Social desirability was significantly correlated in a negative direction with sexual beliefs (men should dominate) and token refusal. Any measure significantly correlated with social desirability is questionable to the point whether participants could be lying about their sexual beliefs. As social desirability and men should dominate were negatively correlated it could indicated participants who scored higher on socially desirability deemed it necessary to accept that men shouldn’t dominate women, in order to keep in line with a socially desirable appearance.

Regression Analysis

A multiple linear regression was employed to help determine which of the factors could be used to predict sexual coercion. Homoscedasticity was examined via scatter plots and these indicated reasonable consistency of spread through the distributions. Skew and kurtosis were found to be within the normal range as no values were greater than one, suggesting normal distribution. The association between the criterion and the explanatory variable was moderately weak (R² adj= 0.50) Together sexual belief, social desirability, sexual strategy and impulsivity-sensation seeking accounted for only 11% of the variation in sexual coercion. The overall relationship was not significant (f (11, 61) = 7.45 p>.05).

Logistic Regression

As a result of not having a significant relationship it was decided to run a logistic regression to further test the variable predictability of sexual coercion. Logistic regression was deemed the most suitable method to use once some variables had been dichotomised. The categories of pornography usage, sports team and halls of residence membership were dichotomised. A logistic regression procedure was run with sexual coercion as the criterion (dichotomised) variable.

The overall model of, pornography use, sports team and halls membership along with, sexual strategy, social desirability, impulsivity and sexual belief showed no significant improvement over a constant only model, χ²(9, N=73) = 14.54, p<.05. In the full model, only pornography (P <.05) appeared to contribute to the prediction of coercion when other variables were held constant.

A test of the significant predictor met the met the criteria for Homer and Lemeshow’s (1989) goodness of fit χ²(9, N=73) = 11.25, p <.05. and were significantly different from the constant only model, χ²(9, N=73) =26.73, p<.05. The predictor of pornography usage assisted in the differentiation between coercive and non coercive males. Table 7 shows the regression coefficients and odds ratio for the significant coefficient and those close to being significant (sports team membership).

The odds ratio demonstrated that the risk of coercion decreases as pornography usage decreases by half a standard deviation. As sports membership was almost significant it would have indicated, sexually coercion increases by six standard deviations for those participants who are sports team members. It was decided to investigate further the sports team membership variable by dichotomising further variables than are often associated with both sports team membership and sexual coercion.

The variables of promiscuity, determined by sexual partners (above average and below average) and age at first sexual experience (under the legal age or over the legal age) were dichotomised and added to the logistic regression. Once these variables were added to the regression both pornography (p.<.05) and sports team membership (p.<05) increased in significance. The regression coefficient and odds ratio for the significant coefficient for these are shown in table 8.

These revised ratios indicated that once the variables of promiscuity had been added the likelihood of coercion decreases further for those who use pornography less frequently. Sexual coercion, however, increases by ten standard deviations for those who are sports team members. However, caution should be drawn when making any inferences based upon these results as the results are rather speculative given that they generally don’t integrate with the other measures. It should be noted that sexual belief, sexual strategies, impulsivity and social desirability did not contribute to the model, contrary to popular research.


The aim of the present study was to test a model of predictors of sexually coercive behaviour in a sample of university students. As all of the data was based upon self reports and the questions were sexual in nature then it is important to consider the difference between knowing males are coercive and knowing their own behaviour that they are reporting, is coercive. It is further documented that males often under report their coercive behaviours (Koss, 1992) and over reporting is generally not thought to be an issue.

With such research in mind it may be wise to consider that some of the males within this study under reported, resulting in an underestimation of the level of coercion in this sample. Therefore, it is possible to assume that some of those males and potentially females who were labelled as being ‘non coercive’, have in fact perpetrated some coercive behaviours.

The proportion of participants who reported perpetrating sexually coercive behaviours (8.3%) is a slightly lower rate that previous studies who reported prevalence of around 25% (Koss et al. 1987). Within this sample no males reported committing an act that could legally be described as rape, opposed to around 4% reported in similar studies (Koss et al.1987, Brener et al, 1999). It may be realistic to assume that such differences occurred as a result of the sampling within the study and obtaining responses from a lager or more diverse population would result in higher levels of perpetrated sexual coercion.

However, it is more likely that the discrepancy occurs due to the cultural differences between dating in the UK and the USA. Despite our data highlighting prevalence at a lower rate than comparable research, the very fact that 8.3% admitted perpetration and considering that victimisation by women is often higher than male self reports (Sonenstein, Ku, Lindberg, Turner, & Pleck, 1998) there may well be a significant problem. Therefore, such results are sufficient enough for us to justify that this research area warrants attention especially from an educational and preventative point of view

With regard to gender differences only one female admitted to perpetrating sexual coercion. Therefore, opposing some research (Sigleman, Berry & Wiles, 1984) that female coercion is becoming increasingly more common. All of the male participants within this study, regardless of whether than had admitted be to being coercive or not, had higher reported scores on sexual strategies (both overt and covert), sexual beliefs and impulsivity compared to the female sample. Even the single female, who admitted to coercion, had lower scores on all of the measure compared to males who had admitted to coercion. These results support previous sexual coercion research which demonstrates that all these factors are correlated with sexually coercive beliefs and behaviours (Emmers-Sommer, & Allen, 1999).

Of those that admitted to perpetration of sexually coercive behaviours, 53% had willingly engaged in sexually intercourse with over eleven partners. Therefore, the majority of those individuals who had admitted to being coercive had also had intercourse with a greater number of people. This apparent link may be due to the fact that these individuals have had more sexual experience therefore have more exposure to the potential of sexual coercion. However, it could be argued that these same individuals are sexually coercive in nature which drives them towards higher levels of sexual experiences in order to satisfy their urges.

Despite, sexual beliefs, sexual strategies, social desirability and impulsivity being significantly correlated with sexual coercion, none of these variables were found to be significant in the regression model. We suggest that within the university context that we were sampling these factors are not present in any obvious way. If, as it seems, these factors are not especially important then this could account for the observed absence of this positive association with the SES scores. Further explanation for the lack of a significant relationship could due to the variables being too highly correlated with other variables to be found significant, thus affecting the predictability within the model.

Neither is it possible to rule out that because self reports and attitudes are not good predictors of behaviour it somehow confounded our results. It should not be overlooked that sexual strategies and sexual beliefs are significantly associated dominance and control over women because these attitudes can exert their impact influence on coercive sexual behaviours in a less than obvious way.

The logistic regression analysis produced somewhat more pleasing results, however, only pornography usage way found to be related to coercion. It was found that as pornography usage decreases sexual coercion decreases, albeit by half a deviation it is still an interesting result. Pornography usage alone however, is not likely to be the reason for acting in a sexually coercive manner. Those that choose to use pornography frequently may be more motivated than less frequent users, to seek new, more varied partners and wish to engage in ever more unconventional sexual activities.

The heterogeneity of the relationship between pornography and sexual coercion may arise because attitudes and behaviours manifest out of interactions with peers and existing personality factors, and characteristics such as impulsivity and promiscuity (Emmers-Sommer, & Allen, 1999).Within this study both promiscuity (number of sexual partners) and impulsivity scores were higher for those participants who watched pornography frequently.

Therefore, it is possibly to argue that pornography and sexually coercive behaviours are related to third parties such as, promiscuity and impulsivity. This does not mean that pornography is not harmful but rather that it is not the origin of harmful effects. Those with higher levels of promiscuity and porn usage have been previous linked to higher score of mating effort (Gladden, Sisco, & Figueredo, 2008). Unfortunately, due to the low reliability of the mating effort measure, we were unable to explore the suggestion that porn may have an effect on sexually coercive behaviour through its effect on mating effort.

The failure of halls of residence membership to significantly contribute to acceptance of sexually coercive tactics was in hindsight not a surprising finding. This is likely to be due to the vast cultural differences that have already been highlighted as potential reason for not finding higher rates of perpetration. Much of the ‘frat’ culture and research is based upon North American student samples. What, if anything can be deduced from this finding of non significance is that, the organisation of fraternities in the USA and hall of residence in the UK are not comparable and much of this literature is nor relevant to the UK.

Within the UK, halls of residence are a short term housing solution for all first year students, just for the first year of study. More often than not ‘halls’ are mixed and thus the hyper-masculine attitudes and peer pressure to conform to sexist attitudes and coercion (Forbes, Adams-Curtis, Pakalka, & White, 2006) may just not be evident in UK student accommodation. After their first year, many students go on to rent private housing with small groups of friends they have made whilst being in ‘halls’.

Another contribution to the apparent cultural difference in the relationship between type of student accommodation and sexual coercion may be a confounding factor such as alcohol, that was not considered in the study. Within the UK, the significant majority of students are legally allowed to consume alcohol when they begin university at 18 years. The majority of UK students therefore drink within city centre or student establishments and nightclubs. Within North America due to the legal drinking age being 21 years those students starting university at 18 cannot legally drink, therefore consume alcohol at ‘frat’ parties. The environment at these parties lacks the security of UK nightclubs and more often than not they are seen to promote excessive drinking and behaviours (Sawyer, Thompson, & Chicorelli, 2002) that could get the same individual removed from a nightclub in the UK. The intimate environment of these parties exacerbates isolation, a factor often associated with sexual aggression (as indicated on the SSQ).

Despite being found to be significant in the logistic regression model (table 8) and data suggesting that those individuals who participate in sports did report higher levels of coercive behaviour compared to non sports players (3.8% and 3.7% respectively) caution needs to be drawn surrounding this result. As sport team membership only became significant after the inclusion of the promiscuity variables the result is questionable due to the tentative nature of the relationship between promiscuity and sports team membership. Neither the less, levels of coercion only differed slightly between sports players and non sports players. These results contrast existing research by Koss and Gains, (1993) but are more in line with Lackie and DeMan (1997) who questioned the link between sports team membership and sexual coercion.

Cultural differences, discussed previously with regard to this study’s findings of lower prevalence rates of coercion and the hall of residence situation, could also be attributed to the seemingly little difference between coercion in sport members and non members. Within the USA, university sport plays a much larger role in university life. Many students obtain university scholarships and funding based upon their sporting ability. The fixtures themselves attract a large number of student fans, on a scale unheard of in the UK except in professional sporting events.

Within the UK, emphasis of this kind is not placed upon sports teams and membership for many is a way to extend their social circles. Arguably in the USA, sport members are extremely loyal and competitive not only due to the pressure from fans, coaches, rival universities but also the threat of losing a university place if their playing standard drops or the pressure of obtaining a professional sporting contract which many of the top players achieve. Such pressures may arguably result in greater levels of hyper-masculinity and due to loyalty to their team and fellow team members; peer pressure may play a much more significant role.

Within the UK there doesn’t appear to be this kind of pressure or comradely, due to funding restraints many sports teams are affiliated with a female team in order to reduce transport and training costs. Such a move may remove feelings of hyper-masculinity. However, these possible relationships are speculative and would require an in depth level of research to investigate these apparent cultural differences further.

Before considering the implications of the current research, we should first consider the limitations of this study. The fact that the sample was solely comprised of a homogenous group of young university students may limit the generalisability of the study. This lack of diversity among participants may further limit attempts to apply findings to groups that have homosexual or bi-sexual orientation. However, within this study unlike a number of student based samples, postgraduate students were included.

It was hoped that the inclusion of more mature students would overcome sampling issues as older students have had more dating experience and more potential exposure to sexual coercion. Efforts to increase robustness and achieve a greater variability in responses would require a larger sample size as the current study only used three university locations which were all in a similar geographical region.

There are good reasons to believe that the sample may be biased towards an oversampling of males and females who already have a vested interested in the sexuality topic, or those who are concerned about the issue. These kinds of biases in participant selection are common in this area of research. Interestingly though scores on the social desirability measure did not predict the reporting of sexual coercion in the model. This suggests that biases of this type may not limit the findings.

Finally, we must comment on the limitations caused by the measurement of coercion. Because of our moderate sample size, we could not discern between types of coercion, specific levels of pornography usage or type of sports teams membership. This left us in the fairly constrained situation of predicting membership in a dichotomy that oversimplifies the real world of coercive behaviours. Arguablya, to gain a better understanding of the casual relationship between variables a longitudinal design is beneficial; however, due to the time constraints of this study such a design was not possible.

Implications exist in terms of educational programmes and interventions for university students. It is imperative that students especially females understand that certain people will hold certain attitudes and assumptions, and that certain situation may exacerbate their risk of being coerced. If the tentative link between sports team and coercion is found to be a significant one, it is up to those who manage such clubs to challenge the group norms, beliefs and attitudes of its members. Intervention programmes that are aimed at groups rather than the individual should have more of an impact.

With regard to pornography, exposure can never be truly monitored or prevented. As those choosing to watch pornography are become increasingly younger, education surrounding sexual coercion should begin in high school. Such interventions should aim to highlight that the actions depicted in pornography are merely fantasy and should remain as such. There is arguably nothing wrong watching pornography but it should be highlighted that it does become problematic when viewer make fantasy a reality through coercion.

Te perpetration of sexual coercion is never acceptable however it is important to learn more about the characteristics of perpetrators and other confounding variables in order to understand such behaviour and to develop treatment and preventative measures. Future research should concentrate upon investigating cultural differences further. Advances in this area of research would be beneficial to the UK sexual coercion research as much of it is currently based upon the USA and it appears it is not applicable. Further and more stringent measures of peer influence and hyper-masculinity could help to clarify the possi

To export a reference to this article please select a referencing stye below:

Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.
Reference Copied to Clipboard.

Request Removal

If you are the original writer of this dissertation and no longer wish to have the dissertation published on the UK Essays website then please click on the link below to request removal:

More from UK Essays

Get help with your dissertation
Find out more