Heliyon 7 (2021) e07421

Contents lists available at ScienceDirect

Heliyon
journal homepage: [Link]/heliyon

Research article

Development and validation of the Male Rape Myth Acceptance

Scale (MRMAS)
Benjamin A. Hine a, *, Anthony D. Murphy b, Jamie S. Churchyard a
a
University of West London, UK
b
University of Birmingham, UK

A R T I C L E I N F O A B S T R A C T

Keywords: Despite growing recognition of male-on-male rape and its related myths, research in this area has been held back
Rape by the lack of a reliable and comprehensive measure or scale. The present work utilises a large and diverse
Male rape participant sample over two studies (Study 1 N ¼ 510, Study 2 N ¼ 527) to validate a new Male Rape Myth
Rape myth acceptance
Acceptance Scale (MRMAS), measuring myths falling under six principle themes: masculinity, sexuality, pleasure,
Validation
Scale
perpetrators, context, and effect. Analysis suggested a two-factor scale, with ‘Blame’ and ‘Minimisation/Exon-
eration’ sub-scales. Both the overall scale and sub-scales demonstrate excellent reliability and construct validity,
and are thus proposed as tools to enable the proliferation of future research on male rape myth acceptance, both in
general and specialist populations, in an attempt to improve the experiences of male rape victims.

1. Introduction 1.1. ‘Traditional’ rape myths and their measurement

Research on the nature, function, and influence of rape myths has Since Martha Burt's pioneering paper in 1980, research on rape myths
typically focussed on incidences involving male perpetrators and fe- has proliferated (Burt, 1980). As described above, rape myths are defined
male victims. Indeed, the term ‘rape myths’ is used so ubiquitously in as beliefs about rape, which serve to minimise men's sexual violence
reference to this type of violence, that definitions have required no towards women in a variety of ways. This is mirrored in the definition
gendered specification. For example, Bohner et al. (2009) describes provided by the UK Crown Prosecution Service (CPS), that a rape myth is
so-called ‘traditional’ rape myths as ‘descriptive or prescriptive beliefs ‘a commonly held belief, idea or explanation that is not true’, further
about rape (i.e., about its causes, context, consequences, perpetrators, stating that: ‘They attempt to explain events, like rape and abuse, in ways
victims, and their interaction) that serve to deny, downplay or justify that fit with our preconceived ideas about the world – they arise from and
sexual violence that men commit against women’. However, following the reinforce our prejudices and stereotypes’ (CPS, 2015, p. 1). Examples of
legal recognition of male rape in the UK in 1994 (see Jamel et al., 2008, rape myths include specific beliefs regarding victims (e.g., if a woman
for review), research on the rape of men by other men, including the wears revealing clothing she is partly responsible for her victimization)
existence of ‘male rape myths’ has increased. Such research has led to and perpetrators (e.g., once men reach a certain level of sexual arousal,
the identification of several genres of myths specific to male rape (e.g., they are unable to control their actions), as well as broad ideas about rape
those surrounding compromised masculinity), and the application of as a crime, such as the ‘real rape stereotype’ (i.e., the belief that legiti-
some ‘traditional’ rape myths to male victims (e.g., increased victim mate rape cases occur suddenly, at night, by an aggressive stranger, with
responsibility through alcohol and/or drug consumption). However, at a weapon, and typically involve visible victim resistance and emotional
present, no reliable, comprehensive measure of male rape myths exists, trauma for the victim; Estrich, 1987; Horvath and Brown, 2013).
despite calls from Struckman-Johnson and Struckman-Johnson over 25 Despite some definitional conflict (Payne et al., 1999), it is widely
years ago ‘to develop a “male rape myth scale”’ (1994, p. 98). This accepted that there are four types of rape myth, those which: attribute
study presents the first, comprehensive ‘Male Rape Myth Acceptance blame to the victim for their rape (e.g., that women who dress scantily
Scale (MRMAS)’ for the accurate measurement of male-specific rape provoke rape), minimise the seriousness of rape itself (e.g., the sugges-
myths. tion that many claims of rape are false), remove blame from the

* Corresponding author.
E-mail address: [Link]@[Link] (B.A. Hine).

[Link]
Received 14 April 2021; Received in revised form 4 June 2021; Accepted 24 June 2021
2405-8440/© 2021 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license ([Link]
B.A. Hine et al. Heliyon 7 (2021) e07421

perpetrator (e.g., the implication that men cannot control their sex drive, fight off an attacker; Groth and Burgess, 1980). Beliefs that ‘real’ men
especially when already aroused), and suggest that rape only happens to would not allow themselves to be raped (McMullen, 1990) or that ‘real’
particular kinds of women (e.g., only promiscuous women get raped; men cannot be raped (Hillman, O'Mara, Tomlinson and Harris, 1991) are
Bohner et al., 2009). Rape myths can therefore be characterised as a also borne from such comparisons. This occurs despite evidence sug-
general cognitive schema that enables negative attributions to be made gesting that men, similarly to women, engage in “passive submission,
about the crime of rape and those involved (Grubb and Turner, 2012). As engendered by an overwhelming sense of disbelief” when attacked (King,
such, rape myths serve several important psychological functions that 1992, p. 3). Moreover, men suffer additional judgment and associated
enable individuals to understand and make sense of negative events in shame for even becoming victims in the first place, as this again con-
their social world, maintain cognitive consistency, avoid the experience tradicts stereotypes of men as powerful, in control, dominating, and
of negative affect, and rationalise problematic behaviour (Bohner et al., strong (Connell, 2002), and thus able to defend themselves from attack
2009). Rape myth acceptance then is characterised as the extent to which (Gonsiorek, 1994; Struckman-Johnson, 1991). Such judgments support
a person adheres to such beliefs, and research demonstrates that rape erroneous beliefs that male victims of rape must therefore be children, or
myths are held by people of all sexes, all ages and across races (Burt, very weak adults (Scarce, 1997), and that if a man does not fight off an
1980; Johnson et al., 1997; McGee, O'Higgins, Garavan and Conroy, attacker, they must have wanted to have been ‘raped’ (Struckman--
2011; Suarez and Gadalla, 2010), and are held by both victims (Peterson Johnson and Struckman-Johnson, 1992). Crucially, such myths predicate
and Muehlenhard, 2004) and perpetrators (Marshall and Hambley, the idea that victimisation should and does result in a loss of, and
1996). compromise to, masculinity, along with a loss of status as a ‘real man’
Importantly, studies demonstrate that rape myths are not only held by (Groth and Burgess, 1980). Moreover, men who have been raped are
the general public (Sussenbach and Bohner, 2011), but by various thought to be weak and vulnerable (Isely, 1991), responsible for the
specialist populations both outside of and within the criminal justice assault (Hickson et al., 1994), and to blame for their victimisation
system, such as counsellors (Feild, 1978), police officers (Murphy and (Myers, 1989).
Hine, 2019; Parratt and Pina, 2017), lawyers and barristers (Temkin, Masculinity myths are strongly related to myths surrounding sexu-
2000; Temkin and Krah
e, 2008), judges (Temkin and Krah
e, 2008) and ality. Indeed, as it has been highlighted that traditional, restrictive and
jurors (see Dinos et al., 2015, for review). Studies have also highlighted regressive masculine ideals often fuel prejudice and discrimination to-
the impact rape myths have on the attribution of blame to victims and wards sexual minority men (E. Anderson, 2009), as they are evaluated as
perpetrators, again both in the general population (Grubb & Harrower, contradictory to those ideals. It is therefore unsurprising that homo-
2008, 2009; Grubb and Turner, 2012), and in specialist populations such phobic assumptions about sexuality are thus made upon victimisation,
as police officers (Hine & Murphy, 2017, 2019; Sleath and Bull, 2017), as and in such ways as to discredit and negatively judge the male victim, as
well as their influence on case investigation and progression (Hohl and masculine perceptions around power, dominance and control are
Stanko, 2015; Sleath and Bull, 2017). violated, and attempts are made to explain this violation by both victims
Such valuable research has been enabled by the existence of several and others. The theme of sexuality myths includes negative and incorrect
reliable measures of rape myth acceptance, developed and refined over beliefs that a man who is sexually assaulted by another male must
nearly 40 years. Indeed, measures including the Rape Myth Acceptance therefore be gay (Stermac et al., 2004), or have been acting in a ‘gay
Scale (RMAS; Burt, 1980), the Attitudes Towards Rape Scale (ATR; Feild, manner’ (Coxell and King, 2010), or that only gay men are raped (Hill-
1978), the Illinois Rape Myth Acceptance Scale (IRMAS; Payne et al., man, O'Mara, Taylor-Robinson and Harris, 1990; Laurent, 1993;
1999), and the Acceptance of Modern Myths about Sexual Aggression Struckman-Johnson, 1991). Confusingly and in complete contradiction,
Scale (AMMSA; Gerger et al., 2007), have all provided the tools with other myths state that gay men cannot be sexually assaulted, as the act of
which to accurately assess and compare myth acceptance across groups, anal penetration itself is ‘homosexual by nature’, and thus cannot be
as well as the predictors and outcomes of rape myth acceptive attitudes. non-consensual (Cotton, 1992), and that gay men constitute ‘willing’
Such findings have proved useful in the development of education and victims (Mezey and King, 1989). These erroneous attitudes are propa-
training programmes which address rape myths, such as those delivered gated and reinforced by still extant homophobia within society, which
to police officers in attempts to improve the experiences of promotes a range of negative responses to both consensual and
victim-survivors, and to challenge societally embedded beliefs which non-consensual sex between men, from disapproval and disbelief, to
enable ‘rape culture’ (Burt, 1980). disgust and violence (Ahmad and Bhugra, 2010). Indeed, such is the
societal prejudice towards sexual minority men, that some myths suggest
1.2. Male rape myths that gay or bisexual men ‘deserve’ their victimisation, as a reward for a
‘deviant’ lifestyle (Turchik and Edwards, 2012) and that they have
It is estimated that similar research on myths relating to male rape, ‘brought it upon themselves’ (Cotton, 1992). Ironically, evidence that
and the measurement of male-specific myths, is approximately 20 years suggests that the vast majority of male rape victims (approximately 80%)
behind that of research on female rape (Davies and Rogers, 2006). identify as heterosexual (Groth and Burgess, 1980; Isley &
Nonetheless, some research detailing male myths is available, with be- Gehrenbeck-Shim, 1997). Though, it must be noted that such individuals
liefs categorised under six central themes; masculinity, sexuality, plea- are often believed to be hiding ‘secret sexual desires’ and have thus
sure, perpetrators, context, and effect, each of which are explored in claimed rape to hide their illicit activities (King, 1992); a myth likely
detail below. It is important to note that so-called ‘male rape myths’ underpinned by broader misconceptions of rape as motivated by sexual
operate similarly to ‘female rape myths’, despite differing in content; in desire, rather than dominance and power (Hickson et al., 1994). Taken
that, all of the myths outlined below seek (a) to blame the victim, (b) together, these myths are mostly representative of broader negative be-
exonerate the perpetrator, (c) minimise the severity of the incident, liefs around sexual minority men; themselves outdated, bigoted, and
and/or (d) suggest that only certain groups or types of men are raped. fundamentally incorrect.
Unsurprisingly, male rape victims appear to be evaluated against Other myths related to sexual desire are those related to pleasure and
stereotypic conceptualisations of masculinity (e.g., hegemonic mascu- men's physiological responses to sexual activity (consensual or other-
linity as described by Connell, 2002), in a process similar to comparisons wise). Specifically, widely held beliefs that physiological reactions to
made of female rape victims against expectations associated with femi- physical stimulation and arousal (e.g., erection and/or ejaculation) imply
ninity (Campbell, 2005). For example, as female victims are punished for pleasure and enjoyment (McMullen, 1990), are problematic for
displays of promiscuous behaviour antecedent to their assault which victim-survivors, as this may lead them to question whether the event
violate virtuous/reputational ideals, men are criticised for behaviour was rape, or to feel as if their body has ‘betrayed them’ (Coxell and King,
which contradicts masculine ideals relating to strength (e.g., failing to 2010; Gonsiorek, 1994; Sarrell and Masters, 1982). Such myths also

2
B.A. Hine et al. Heliyon 7 (2021) e07421

encourage others to assume enjoyment, and thus disbelief in claims of victims; Struckman-Johnson and Struckman-Johnson, 1992). This attri-
rape, in a similar manner to female victims; where vaginal lubrication, bution of shame is similar to that experienced by female rape victims,
and even orgasm as a protective bodily response, is misinterpreted as with additional attributions resulting from male gender role norm vio-
pleasure (Suschinsky and Lalumiere, 2010). This is despite research lations (i.e., showing vulnerability). Subsequently, men are expected to
which states that men frequently achieve an erection and/or ejaculation be able to cope on their own following victimisation (Struckman-John-
during an assault and that such reactions may be a generalised bodily son, 1991). Thus, taken together, whilst more research on male rape
reaction to extreme emotional turmoil (Sarrell and Masters, 1982). myths is needed, previous studies have at least provided some informa-
Arousal myths and sexuality myths also heavily related, as the supposed tion on the nature of such myths.
‘enjoyment’ of anal or oral penetration, as implied by an erection or
ejaculation, trigger associated homophobic beliefs. 1.3. The influence of male rape myths
A specific subset of myths relate to the behaviour and nature of per-
petrators, many of which mirror sexuality myths relating to victims. As knowledge of male rape myths has increased, a limited number of
Specifically, as with male victims, assumptions are made that men who studies have emerged exploring their effect on judgments towards male
rape other men are gay (Groth and Burgess, 1980; McMullen, 1990; victims and perpetrators, most commonly through attribution of re-
Mezey and King, 1987; Struckman-Johnson, 1991) and are acting upon sponsibility ratings. For example, several studies demonstrate that gay
either secret or overt sexual desires (Coxell and King, 2010). Again, this is victims are attributed higher levels of responsibility for their victim-
despite evidence suggesting that approximately 90% of perpetrators isation than heterosexual victims (see Davies and Rogers, 2006, for re-
identify as heterosexual (Isley & Gehrenbeck-Shim, 1997). These view). Subsequently, it has been reasoned that participants are informed
perpetrator specific myths are again most likely explained by mis- by homophobic attitudes and sexuality male rape myths, in making their
understandings concerning motivations for rape (e.g., that male rape is judgments (M. Davies and Rogers, 2006). Moreover, several studies have
motivated by sexual desire, rather than agreed upon dominance expla- demonstrated how myths interact, for example by finding that partici-
nations) and male sexuality (e.g., that only exclusively gay men are pants worryingly believe that gay male victims find their assault more
interested in sexual interactions with other men). pleasurable, and that they suffer from less trauma (Michelle Davies,
Context myths broadly relate to disbelief concerning the existence of Pollard and Archer, 2001; Mitchell et al., 1999). The influence of mas-
male rape at all, a phenomenon exacerbated by a lack of visibility for culinity myths are shown through studies exploring resistance, where
male victims within political, social and academic spheres. Indeed, the higher responsibility attributions are awarded to men who ‘fail’ to resist
myth that male rape is exceptionally rape, if it occurs at all (Mezey and their attacker (Howard, 1984a, 1984b). More broadly, several studies
King, 1989; Scarce, 1997; Struckman-Johnson and Struckman-Johnson, show that men are attributed greater responsibility for their victimisation
1992), is informed by all other myths outlined above which seek to than women (I. Anderson and Quinn, 2009), particularly when judged by
maximise disbelief in its occurrence and likelihood. Such myths are also other men (White and Kurpius, 2002). Such results suggest that partici-
contrary to the information and data available on the prevalence of male pants are more blaming of men due to the effects myths which present
rape, both within the UK (~12,000 per annum, Office for National Sta- male rape as unlikely, and a consequence of victim behaviour. Male rape
tistics, 2018), and countries worldwide (e.g., the US, ~131,000 per myths have also been shown to be related to a number of demographic
annum, Stemple and Meyer, 2014). Other context myths seek to suggest variables (i.e., males have higher adherence; see Walfield, 2018 for re-
that male rape, when it does occur, only happens in particular contexts, view), and several proximal attitudes, including measures of traditional
and that this occurrence is typical or inevitable. These include prisons, rape myths, traditional gender roles, negative attitudes towards sexual
the military, LGBTQ þ venues, and male societies (e.g., sports clubs or minority men, stereotypes about masculinity and male inexpressiveness,
fraternities; Garnets et al., 1990; Kaufman, 1984; Lacey and Roberts, and the endorsement of traditional male gender roles (Chapleau et al.,
1991; Scarce, 1997; Turchik and Edwards, 2012). It is principally within 2008; M. Davies, Gilston and Rogers, 2012; Kassing et al., 2005; Mel-
this genre of myth that ‘traditional’ rape myths are applied to men. For anson, 1999; Nalavany and Abell, 2004; Walfield, 2018).
example, men, similarly to women, are judged to be more responsible for However, whilst some information on the influence of male rape
their victimisation if they have consumed alcohol or drugs, thus myths is available, the research outlined above remains limited, as the
compromising their ability to control and consciously participate in effects of only some male rape myths are explored. Moreover, informa-
sexual interactions (Sleath and Bull, 2010). It is worth noting however tion regarding the prevalence of male rape myth acceptance within so-
that applications of myths are never identical, as, whilst some elements ciety, and our understanding of the relationship between male rape myth
are similar (i.e., consumption of alcohol having a detrimental effect on acceptance and judgments given, is currently extremely inadequate.
memory in both men and women), other elements are shaped by Moreover, levels of MRMA in specialist populations who directly interact
sex-specific expectations and norms (i.e., men's consumption of alcohol with male victims (such as counsellors, service providers, police officers
as compromising their strength and ability to fight off an attacker). etc.) is also yet to be comprehensively measured. The principal reason for
Finally, several myths relate to the effect (or lack thereof) of rape on the scarcity of literature in this area is the lack of a reliable and targeted
men. Specifically, myths articulate that men are not psychologically or scale measuring said myths.
physically affected by rape, principally because they are men and can
‘take it’, and that they are not as affected by the incident as women 1.4. Current measures of male rape myths
(Struckman-Johnson and Struckman-Johnson, 1992). This is despite
growing literature exploring and highlighting the profound psychologi- At present, two scales for the measure of male rape myth acceptance
cal and physical effects of rape on men, and their need for belief and care exist. The first is a 12-item scale developed by Struckman-Johnson and
(Goyer and Eddleman, 1984). Indeed, research has demonstrated that Struckman-Johnson, 1992. This scale is constructed of six items each
men who have had sexually coercive experiences as an adult are more with a male and female perpetrator version, with two items measuring
likely (than those who have not) to experience a range of psychological each of these three general beliefs: (a) male rape does not happen (e.g.,
problems, such as lower self-esteem, increased depressive symptoms, “it is impossible to rape a man”), (b) rape is the victim's fault (e.g., “men
suicidal ideation and self-harm behaviours, anxiety and post-traumatic are to blame for not escaping”), and (c) men would not be traumatised by
stress symptoms, substance abuse and dependence issues, sexual rape (e.g., “men do not need counselling after being raped”). Re-
dysfunction and identity confusion (Turchik and Edwards, 2012). spondents self-rate their beliefs on a 6-point Likert scale ranging from 1
Alongside masculinity myths described above, it is within effect myths (strongly disagree) to 6 (strongly agree).
that notions of shame are most explored (i.e., that men should feel shame Several issues with this scale present. First, questions relating to in-
following their victimisation, for ‘allowing’ themselves to ‘become’ cidents involving female perpetrators are included. This is problematic as

3
B.A. Hine et al. Heliyon 7 (2021) e07421

it can be argued that whilst there are some areas of overlap (e.g., vic- 1.5. The present study
timisation as a threat to masculinity, misinterpretation of physical
arousal as pleasure), many myths about men raped by other men and As detailed above, since the early 1990s, interest and research in the
about men sexually assaulted by women, are distinct and should be existence of male rape myths has significantly increased. Such research has
measured as such. Take, for example, myths around sexuality which only detailed the nature of said myths and has begun to examine the influence
exist in relation to male-on-male assaults, or myths around perpetrators of such attitudes on reactions and judgments toward male victims. How-
which by nature must be sex-specific. A further example is the common ever, at present, research is limited by the lack of a reliable, comprehen-
myth concerning female-perpetrated sexual assault, that ‘men always sive, and targeted tool to measure acceptance of male rape myths, as exists
want sex’ (Clements-Schreiber and Rempel, 1995), which would carry for ‘traditional’ rape myths. The necessity for such a scale is outlined well
very different conceptualisations in the context of a man raped by in Struckman-Johnson and Struckman-Johnson, 1992 paper:
another man. It is thus argued that there is a strong theoretical basis for
“The ultimate goal of research on this topic… is the investigation of
the separate investigation and measurement of male-on-male and
the relationship between male rape myth acceptance and under-
female-on-male myths, as they evoke very different affective and cogni-
reporting of male sexual assault… If rape myth acceptance can be
tive interpretations and reactions. Indeed, Walfield (2018) argues that a
documented, one can then determine if beliefs are indeed related to
gender neutral approach to these kind of measurements is to be avoided,
reporting, to treatment, and to justice received by male victims. At a
as it invites the risk of erasure of gender-specific experiences. Indeed,
minimum, research will stimulate awareness of the problem and
there are several dedicated organisations across the UK which specif-
encourage development of programs to counteract cultural mis-
ically focus on providing support to men assaulted by other men, to
understandings of male rape” (p. 98)
whom research utilising such a scale would be beneficial (though it
should be noted that an additional scale/sub scale assessing Arguably, this need is most exemplified within criminal justice con-
female-on-male rape/sexual assault would be beneficial in the future). texts, as increasing our understanding of how specialist populations (e.g.,
Additionally, it is important to note that traditional rape myth measures police officers) interact with male victims is critical in improving victim
(e.g., IRMAS) specifically assess attitudes towards acts of male-on-female engagement and satisfaction with justice processes, and case success.
rape only (i.e., they do not focus on female perpetrated sexual assault). The aim of the present research therefore was to produce and validate
Thus, as a starting point, a measure constructed to measure myths sur- the first Male Rape Myth Acceptance Scale (MRMAS), measuring myths
rounding rape of males would, in the UK, have a justifiably similar focus under six themes: masculinity, sexuality, pleasure, perpetrators, context,
of only male-on-male incidents. and effect. Study 1 details the formulation, analysis, and selection of male
The second issue is that the range of myths measured is too narrow rape myth items through a variety of reliability measures including
and does not capture the variety of beliefs regarding male rape. As out- exploratory factor analysis. Study 2 describes validation of a revised 44-
lined above, there are (at least) six principal male rape myth ‘themes’; item scale, including assessment of validity in relation proximal
this scale only covers three (with only two items each). The final issue is constructs.
that measures of scale reliability are unavailable. Cronbach's alpha values
are not reported in the original publication (Struckman-Johnson and 2. Study 1
Struckman-Johnson, 1992) or in most subsequent studies utilising the
scale (e.g., Chapleau et al., 2008). 2.1. Method
The second available scale is a 22-item questionnaire, developed as
part of a doctoral dissertation programme by Melanson (1999), designed Full ethical approval for this study was granted by the University
to measure false, stereotypical, or prejudicial beliefs about male rape. Research and Ethics Committee at the University of West London.
Sample items include “male rape is usually committed by homosexuals”
and “a man who has been raped has lost his manhood”. As above, par- 2.1.1. Scale development
ticipants answer using a 6-point Likert scale ranging from 1 (strongly As described by Clifton (2019), when approaching scale develop-
disagree) to 6 (strongly agree). ment, an inevitable trade-off between validity and reliability must
Melanson's scale improved upon the Struckman-Johnson and occur. Such battles are fought in a number of areas, ranging from the
Struckman-Johnson measure in two ways. First, several reliability ana- wording and content of items, to their administration and scoring. As
lyses are available, with this scale demonstrating excellent reliability such, Clifton recommends that authors are forthright in their priorities,
scores (.90 in the original research, and between .85 and .91 when uti- so that reviewers may hold their processes to account (2019). As such,
lised in subsequent studies, e.g., Kassing et al., 2005). Second, a greater we make it clear here that our goal was to produce a highly reliable
variety of questions assess all but one of the themes outlined above. measure, that preserved content validity wherever possible. As such, as
However, despite these advancements, five significant issues remain. outlined below, decisions regarding item generation were designed to
First, as stated, not all the myth themes described above are measured in ensure maximum validity, with decisions taken during item adminis-
detail (i.e., there are no questions that ask about the actions of perpe- tration and analysis then designed to maximise reliability, which was
trators, and only two questions measure beliefs about the effect of rape on subsequently prioritised. These choices were made as this study was not
male victims). Second, this scale still includes questions regarding female designed to explore which myths exist (i.e., theoretically exploratory),
perpetrators, which, as argued above, is problematic. Third, the sample but rather to develop a robust system of measurement (i.e., methodo-
used to validate this measure is both too small (only 303 participants logically exploratory). Thus predictive, rather than content validity was
were utilised to assess and validate 80 potential items) and too homog- prioritised (Clifton, 2019).
enous (only undergraduate students were utilised). Fourth, this scale was To formulate scale items, the authors engaged in over 60 h of liter-
still developed over 20 years ago, when attitudes towards male victims ature review and formulated discussions. Some deliberations were be-
were likely different to the present day, and, as such, would potentially tween just the authors, whilst others were alongside research assistants
have been phrased differently. Finally, the publication detailing this scale and departmental colleagues, undergraduate students, or with family and
has not been peer-reviewed (though it should be noted that subsequent friends (acting as representative of the general public). Discussions were
studies utilising the scale have been), and it can thus be argued that the either theory-led, drawing from existing literature on male rape myths
process for development and determination of reliability of this scale has (i.e., to measure one of the six myth ‘themes’), or more informal, for
not been subjected to rigorous academic appraisal. In summation, whilst example by asking others what they thought about male rape (e.g.,
some measures of male rape myth acceptance do exist, they require “When you think of male rape, what comes to mind?”). During item
improvement in a number of key areas. generation, careful consideration of the balance between reliability and

4
B.A. Hine et al. Heliyon 7 (2021) e07421

Table 1. Descriptive statistics for each MRMAS item – study 1.

Item Code Mean SD Item-Total correlation Skewness

1. If a man is raped it does not mean he is weak 1M.C.A.R 2.69 2.18 .283 1.036
2. A man who is raped must have been behaving in a 1S.B.A 1.84 1.35 .568 1.856
way that made him appear homosexual
3. I would be less inclined to believe a man who said 1PL.B.A 2.39 1.59 .616 1.026
he had been raped if he got an erection during the
incident
4. Heterosexual men are more traumatized by their 1E.C.A. 3.17 1.81 .421 0.365
experience of being raped than women
5. Male rape is very rare, if it occurs at all 1C.T.C. 3.15 1.73 .534 0.449
6. Heterosexual men who commit rape against other 1P.B.C.R. 3.91 1.69 .251 0.292
men do so to assert their dominance
7. I find it difficult to believe one man could sexually 2M.C.A. 2.26 1.59 .649 1.362
overpower another man
8. Male on male rape only happens to homosexual 2S.C.C. 1.92 1.30 .651 1.849
men
9. A male victim who ejaculates during the incident 2PL.B.C. 2.13 1.43 .672 1.335
has not been raped
10. Male victims of rape are not traumatized by 2E.C.A. 1.55 1.03 .624 2.529
the incident
11. Almost all male rape occurs in institutions such 2C.T.C. 2.85 1.58 .539 0.622
as prisons or the military
12. A homosexual man who rapes other men does so 2P.B.C. 3.60 1.67 .443 0.052
out of sexual desire
13. Most men would be able to fight off a male 3M.C.C. 2.96 1.66 .678 0.613
sexual attacker
14. Rape is an accepted risk of a ‘homosexual 3S.B.A. 1.96 1.38 .682 1.509
lifestyle’
15. During a sexual attack it is reasonable for the 3PL.B.A. 2.09 1.44 .687 1.225
victim's erection to be viewed as consent
16. Without physical trauma, I would be less 3E.C.A. 2.09 1.39 .636 1.407
included to believe a man had been raped
17. The idea of a man being raped is somewhat 3C.C.A 1.76 1.45 .562 2.010
amusing
18. Heterosexual men who commit rape do so to act 3P.C.C. 3.44 1.69 .497 0.121
upon secret homosexual desires
19. In ‘real’ cases of male rape, there will be some 4M.B.C. 3.60 1.81 .506 0.064
evidence of physical resistance
20. Heterosexual men ‘cry rape’ to hide their 4S.B.A. 2.72 1.54 .649 0.456
homosexual activities
21. Even if force is used to initiate sex, the victim's 4PL.B.C. 2.47 1.56 .635 0.889
erection can be interpreted as pleasure
22. I would expect heterosexual victims of rape to 4E.C.A. 2.74 1.89 .659 0.758
be more traumatized than homosexual victims
23. Coercive sexual practices between men (e.g., 4C.T.C. 2.73 1.55 .555 0.343
forced oral sex) form a legitimate part of group
initiations such as those used in fraternities or
sporting societies
24. A man would not rape another man if he was 4P.B.C. 2.57 1.71 .670 0.908
sexually fulfilled elsewhere
25. A man who fails to escape a sexual attack is 5M.B.A. 1.64 1.21 .670 2.245
partially responsible for his rape
26. Just because a man is raped does not mean he is 5S.C.C.R 2.08 1.70 .225 1.794
homosexual
27. If a man is being sexually attacked, his 5PL.B.A. 2.22 1.46 .699 1.040
ejaculation is proof he found the experience
somewhat pleasurable
28. Men should feel ashamed as a result of being 5E.C.A. 1.97 1.62 .596 1.789
raped
29. Most cases of male rape include the use of a 5C.T.C. 3.54 1.35 .330 -0.182
weapon
30. Male rape is only perpetrated by homosexual 5P.C.C. 2.18 1.41 .741 1.286
men
31. For a man, not resisting a sexual attack from 6M.B.A. 4.94 1.65 -.154 -0.378
another man, is a reasonable response
32. A man who is raped must be homosexual 6S.C.A. 1.74 1.25 .679 1.900
even if he claims to be heterosexual

(continued on next page)

5
 B.A. Hine et al. Heliyon 7 (2021) e07421

Table 1 (continued )
Item Code Mean SD Item-Total correlation Skewness
33. A homosexual man who has been raped 6PL.C.A. 1.88 1.31 .748 1.617
probably enjoyed the experience to some extent
34. Homosexual men are more traumatized by their 6E.C.A. 2.55 1.52 .476 0.618
experience of being raped than women
35. A man is more responsible for his own rape if he 6C.B.A. 2.05 1.42 .791 1.279
frequents a known homosexual area or
establishment
36. Only men who are big and strong are able to 6P.C.C. 2.13 1.39 .692 1.274
rape other men
37. I would find it difficult to consider a man a ‘real 7M.C.A. 1.89 1.42 .731 1.733
man’ if he said he had been raped
38. If a man has already had consensual sex with 7S.B.A. 2.06 1.48 .765 1.468
other men, I would not believe his claims of rape
39. A man who is raped is not as traumatized by the 7E.C.A. 2.22 1.52 .646 1.132
experience as a woman
40. If a man is drunk or taking drugs he is accepting 7C.B.A 2.26 1.63 .684 1.108
rape as a possible risk
41. Men who commit rape are naturally more 7P.C.C. 3.40 1.64 .424 0.057
aggressive in their day to day lives
42. It is acceptable for a ‘real man’ to show fear 8M.B.A. 2.57 1.81 .301 1.149
during a sexual attack by another man
43. A man who claims to have been raped probably 8S.B.A. 2.29 1.42 .714 0.983
just changed his mind after initially consenting to
sex
44. A male victim's reaction to rape is more likely to 8E.B.C. 3.54 1.75 .356 -0.002
be practical than emotional (e.g., obtaining a HIV
test rather than seeking support)
45. A male victim of rape must have behaved in a 8C.B.A. 2.02 1.38 .762 1.335
way that invited the assault
46. Raping another man is not a sign of mental illness 8P.C.A.R. 4.66 1.87 -.040 -0.365
47. A heterosexual man who had been raped would 9M.C.A.R. 2.75 1.62 .303 0.934
still be desirable to women
48. Male rape is a homosexual act 9S.C.C. 2.91 1.85 .574 0.639
49. If a man has been raped he should be able to 9E.C.A. 1.99 1.33 .667 1.383
cope on his own
50. I would find it difficult to believe a man had 9C.B.A.R. 2.40 1.59 .741 1.038
been raped if he had previously consented to sex
with the same man
51. Regardless of how they identify themselves, I 9P.C.A 2.91 1.81 .590 0.636
believe that men who rape other men are
homosexual
52. ‘Real men’ cannot be raped 10M.C.A. 1.76 1.34 .757 2.123
53. I would expect a man to be ‘matter of fact’ and in 10E.B.A. 2.36 1.53 .694 1.012
control of his emotions when reporting a rape
54. A man who has been raped did not set sexual 10C.B.C. 2.45 1.51 .657 0.761
limits understood by the perpetrator
55. Male victims of rape have very little emotional 11E.C.A. 1.85 1.36 .698 1.735
trauma to cope with

Note: Skewness SE ¼ 0.108.

aCode is comprised of a) a number denoting the position of that item within each theme, b) a letter denoting the theme (M ¼ Masculinity, S ¼ Sexuality, PL ¼ Pleasure,
E ¼ Effect, C ¼ Context, P ¼ Perpetrator), c) whether the item is Behavioral (B), Characterological (C) or Typology (T), and d) whether the item is Affective (A) or
Cognitive (C). If the item also has an R at the end of its code, this means the question is reverse scored.
bItems shown in bold were identified for elimination due to low item means.
cItems shown in italics were identified for elimination due to weak item-total correlations.
dUnderlined items were eliminated due to feedback suggesting participant lack of understanding of terminology.

validity was considered, as outlined in Clifton (2019). Specifically, to constituted the second letter in the item label. An additional letter was
maximise validity: item content was kept as diverse as possible, whilst added to item labels based on whether they represented a cognitive (C) or
still being theoretically informed; item difficulty was moderate to ensure affective (A) evaluation. Each of the six principal categories had between
even levels of accessibility; and items were constructed using varied six and eleven questions (with numbers of items determined by authors
terminology, again, to ensure even levels of accessibility (Clifton, 2019). judgement of appropriate theoretical coverage). Some items were also
Items were initially specifically generated to measure one of the six initial reversed to guard against acquiescence bias, though this was not exces-
male rape myth categories, and were labelled appropriately: Masculinity sive so as to preserve validity, but not compromise reliability (as outlined
(M), Sexuality (S), Pleasure (PL), Effect (E), Context (C), and Perpetrators by Clifton, 2019).
(P). They were also classified as to whether they represented a behav- The process outlined above resulted in the first iteration of the Male
ioural (B), characterological (C) or Typological (T) judgment, and this Rape Myth Acceptance Scale (MRMAS). This constituted 55 items

6
B.A. Hine et al. Heliyon 7 (2021) e07421

covering six areas of male rape myths, as outlined above, including: 1979; and Velicer, 1976; all cited in Revelle, 2017) to help determine the
Masculinity (e.g., “I find it difficult to believe one man could sexually number of factors to extract. The Psych package was also used to calculate
overpower another man”), Sexuality (e.g., “Male on male rape only reliability estimates (Cronbach's Alpha, Hierarchal Omega and Omega
happens to homosexual men”), Pleasure (e.g., “A male victim who ejac- Total).
ulates during the incident has not been raped”), Effect (e.g., “Male vic-
tims of rape are not traumatised by the incident”), Context (e.g., “Almost 2.2.1. Descriptive statistics and reliability estimates
all male rape occurs in institutions such as prisons or the military”), and The MRMAS full scale mean (average across all 55 items) for indi-
Perpetrators (e.g., “A man would not rape another man if he was sexually vidual participants ranged from 1.22 to 5.91 (Range ¼ 4.69), with a
fulfilled elsewhere”). Participants answered on a 7-point Likert scale sample MRMAS full scale mean of 2.59. A skewness value of 0.77 sug-
between 1 (Strong Disagree) and 7 (Strongly Agree), as used in other gests moderate positive skew in the MRMAS full scale mean. This is
popular rape myth scales (i.e., the AMMSA). Items were presented in the unsurprising, as scales of this nature, which ask questions on sensitive or
same order to all participants, and in unidimensional blocks (as reliability emotive topics such as rape, often produce lower means. A kurtosis value
was prioritised during administration; Clifton, 2019). of 0.26 in the MRMAS full scale mean suggests a normal level for a
sample of this size (Field, 2009). The Cronbach's Alpha for all 55 items
2.1.2. Participants was .97, while the Hierarchal Omega/Omega total coefficients were
Participants were 510 undergraduate students (M ¼ 24.08, SD ¼ .88/.97 respectively when calculated using Maximum Likelihood esti-
6.63, min ¼ 18.00, max ¼ 58.00, 295 female) from two universities in the mation with Pearson correlations. A Promax rotation was applied in this
UK; one in London, and one in the North West (though this was not analysis to account for the possibility of more than one factor existing in
recorded during participation). Participants’ identified gender was pre- the data. An oblique rotation was chosen because any latent factors
dominantly as a woman (53.1%) or a man (33.7%), with male (6.9%) and should still be representative of rape myths, and therefore should be
female (4.3%) constituting the next highest percentages. Other responses significantly related. As the response options for each item in the measure
included agender, cisgender, fourth gender and genderqueer (all be- can be viewed as continuous ordered categories (Schmidt, 2011), this
tween .2 and .6%). Most participants identified as heterosexual (81.4%), analysis was also repeated using Polychoric correlations instead of
with 8.4% identifying as bisexual, 5.1% as homosexual, and 4.5% Pearson correlations to estimate the reliability coefficients (Schmidt,
preferring not to say. The sample also included a variety of ethnic 2011). This suggested the Cronbach's Alpha for all 55 items was .98,
backgrounds, with White participants constituting 47.1%, Black/Black while the Hierarchal Omega/Omega total coefficients were .75/.98
British providing 21.8%, and Asian/Asian British making up 16.3%. respectively. These demonstrate overall reliability for a single general
Other and Mixed groups constituted the remaining 14.9%. Participants factor solution, although the discrepancy in the Hierarchal Omega and
were recruited through opportunity sampling across the two institution Omega total coefficients suggests some of the variance in the 55 items
campuses. All participants were offered the chance to be entered into a could be attributable to a multiple factor structure or specific item
prize draw to win a £25 Amazon voucher, and psychology students were variances.
also awarded two research credits, as part of departmental initiatives to Individual item descriptive statistics and item-total correlations are
encourage research participation. shown in Table 1. Overall, item means sat towards the middle to lower
part of the Likert scale, and a number of moderate to high positive
2.1.3. Procedure skewness values were observed suggesting the data were somewhat
Participants were recruited using two principal methods. Some were skewed. However, most item-total correlations demonstrated that,
recruited through advertisement in lectures at their institution and were despite several ‘lower’ mean scores, most items correlated strongly to the
provided a link to the study should they wish to take part. Others were total. Thus, considering the commonality of positive skew for scales of
approached by a research assistant whilst they were alone and in a quiet this type, when identifying extreme outlier candidates for elimination
space (such as a library) and were asked to either complete the survey on rather than set the arbitrary threshold for elimination at mean < 2, the
a tablet or take a leaflet to do so in their own time. Participants were decision was taken to identify just the three items with the lowest mean
made aware of the general context of the study (i.e., that they would be score (shown in bold). The two items which showed weak item-total
answering questions pertaining to male-on-male rape), before being correlations (shown in italics) were also identified for elimination. One
asked if they would like to participate. If they clicked on the link pro- final candidate item was identified after feedback from research assis-
vided, they were presented with an information sheet outlining the study tants suggesting a lack of understanding of terminology from participants
in more detail, before being asked to provide informed consent – which (specifically, the word ejaculation; item underlined), with no alternative
was obtained from all participants. They were then asked to provide basic phrasing possible. All such decisions were taken to maximise reliability at
demographic information, before being presented with brief instructions the analysis stage (Clifton, 2019).
on how to complete the MRMAS, followed by the questionnaire. Once
complete, participants were presented with a debrief explaining the 2.2.2. Exploratory factor analysis
purpose and aims of the study and directing them to appropriate support All items were subjected to a Principal Axis Factoring exploratory
services should they be needed. Participants could also give their contact factor analysis (EFA), to assess the underlying structure of the scale, and
information (stored separately to their data) to be entered into the prize to identify further candidate items for elimination (under recommenda-
draw. This study was approved by the University Research and Ethics tions from Costello and Osborne, 2005; Fabrigar et al., 1999; Russell,
Committee (UREC) at the University of West London. 2002). Principal Axis factoring was used due to the positive skew dis-
played by a number of items measured in the scale (see Table 1). The use
2.2. Results of this extraction technique has been advised in cases where a number of
the items display a non-normal distribution, as the estimation technique
Using SPSS Version 25 descriptive statistics and item-to-total corre- does not require normal distribution assumptions to be met (Costello and
lations were produced to allow for assessment of sample distribution, and Osborne, 2005; Fabrigar et al., 1999).
to provide guidance on which items required elimination for Study 2. Inspection of the correlation matrix revealed the presence of many
Exploratory factor analysis was performed using SPSS to assess the latent coefficients of .30 and above. The Kaiser-Meyer-Olkin value of .97
structure of the scale and further identify items which did not load well, exceeded the recommended value of .5, and is described as ‘great’
and thus could also be eliminated. The Psych package in R (Revelle, (Kaiser, 1970, 1974) suggesting that the sample size was sufficient.
2017) was used to perform the Scree test, Parallel analysis, the Very Furthermore Bartlett's Test of Sphericity (Bartlett, 1954) reached sta-
Simple Structure (VSS) and MAP tests (Horn, 1965; Revelle and Rocklin, tistical significance, supporting the factorability of the correlation

7
B.A. Hine et al. Heliyon 7 (2021) e07421

Table 2. Principal Axis Factoring factor loadings for each MRMAS item, following Promax rotation – study 1.

Item Factor 1 – Blame Factor 2 – Minimisation/Exoneration

26. Just because a man is raped does not mean he is homosexual .830 -.614
42. It is acceptable for a ‘real man’ to show fear during a sexual attack by another man .816 -.518
55. Male victims of rape have very little emotional trauma to cope with .755
32. A man who is raped must be homosexual even if he claims to be heterosexual .750
10. Male victims of rape are not traumatized by the incident .733
25. A man who fails to escape a sexual attack is partially responsible for his rape .729
52. ‘Real men’ cannot be raped .724
45. A male victim of rape must have behaved in a way that invited the assault .712
47. A heterosexual man who had been raped would still be desirable to women .700 -.404
38. If a man has already had consensual sex with other men, I would not believe his claims of rape .683
33. A homosexual man who has been raped probably enjoyed the experience to some extent .639
37. I would find it difficult to consider a man a ‘real man’ if he said he had been raped .633
49. If a man has been raped he should be able to cope on his own .621
35. A man is more responsible for his own rape if he frequents a known homosexual area or establishment .586
14. Rape is an accepted risk of a ‘homosexual lifestyle’ .564
8. Male on male rape only happens to homosexual men .554
43. A man who claims to have been raped probably just changed his mind after initially consenting to sex .542
28. Men should feel ashamed as a result of being raped .530
1. If a man is raped it does not mean he is weak .511
39. A man who is raped is not as traumatized by the experience as a woman .500
36. Only men who are big and strong are able to rape other men .480
2. A man who is raped must have been behaving in a way that made him appear homosexual .466
53. I would expect a man to be ‘matter of fact’ and in control of his emotions when reporting a rape .446
30. Male rape is only perpetrated by homosexual men .434
9. A male victim who ejaculates during the incident has not been raped .423
40. If a man is drunk or taking drugs he is accepting rape as a possible risk .421
18. Heterosexual men who commit rape do so to act upon secret homosexual desires .850
19. In ‘real’ cases of male rape, there will be some evidence of physical resistance .773
12. A homosexual man who rapes other men does so out of sexual desire .696
29. Most cases of male rape include the use of a weapon .657
20. Heterosexual men ‘cry rape’ to hide their homosexual activities .654
22. I would expect heterosexual victims of rape to be more traumatized than homosexual victims .643
13. Most men would be able to fight off a male sexual attacker .633
5. Male rape is very rare, if it occurs at all .617
48. Male rape is a homosexual act .611
51. Regardless of how they identify themselves, I believe that men who rape other men are homosexual .594
41. Men who commit rape are naturally more aggressive in their day to day lives .594
11. Almost all male rape occurs in institutions such as prisons or the military .583
44. A male victim's reaction to rape is more likely to be practical than emotional (e.g., obtaining a HIV test rather than seeking .550
support)
21. Even if force is used to initiate sex, the victim's erection can be interpreted as pleasure .512
24. A man would not rape another man if he was sexually fulfilled elsewhere .506
23. Coercive sexual practices between men (e.g., forced oral sex) form a legitimate part of group initiations such as those used .496
in fraternities or sporting societies
54. A man who has been raped did not set sexual limits understood by the perpetrator .495
50. I would find it difficult to believe a man had been raped if he had previously consented to sex with the same man .439
27. If a man is being sexually attacked, his ejaculation is proof he found the experience somewhat pleasurable .435
3. I would be less inclined to believe a man who said he had been raped if he got an erection during the incident .415
4. Heterosexual men are more traumatized by their experience of being raped than women .412

16. Without physical trauma, I would be less included to believe a man had been raped .412
6. Heterosexual men who commit rape against other men do so to assert their dominance
7. I find it difficult to believe one man could sexually overpower another man
15. During a sexual attack it is reasonable for the victim's erection to be viewed as consent
17. The idea of a man being raped is somewhat amusing
31. For a man, not resisting a sexual attack from another man, is a reasonable response
34. Homosexual men are more traumatized by their experience of being raped than women
46. Raping another man is not a sign of mental illness

Note Items shown in bold were identified for elimination due to a lack of substantial loading on either factor.

8
B.A. Hine et al. Heliyon 7 (2021) e07421

Table 3. Demographic information for participant groups in study 2.

Undergraduate Students General Population

Age Mean 22.97 35.39
SD 5.00 10.95
Min 18.00 19.00
Max 49.00 68.00
Income Median N/A £30,000-£40,000
N % N %
Sex (%) Male 132 38.2 74 40.2
Female 214 61.8 110 59.8
Identified Gender (%) Woman 196 56.6 100 54.3
Man 114 32.9 61 33.2
Female 17 4.9 6 3.3
Male 9 2.6 11 6.0
Other 10 3 6 3.2
Sexuality Heterosexual 271 78.3 134 72.8
Homosexual 11 3.2 15 8.2
Bisexual 32 9.2 28 15.2
Asexual 9 2.6 1 0.5
Prefer Not to Say 23 6.6 6 3.3
Ethnicity White 196 56.6 166 90.2
Black/Black British 40 11.6 5 2.7
Asian/Asian British 60 17.3 7 3.8
Mixed 25 7.2 4 2.2
Other Ethnic Group 25 7.2 2 1.1
Employment Status Employed – Full time N/A 102 55.4
Employed – Part time or Zero Hours Contract 25 13.6
Self-Employed 31 16.8
Stay at Home Caregiver 6 3.3
Unemployed 20 10.9
Profession Manager 28 18.2
Professional 76 49.4
Technician or Associate Professional 12 7.8
Clerical Support Worker 15 9.7
Skilled Agricultural 12 7.8
Craft or Related Trade Worker 1 0.6
Plant or Machine Operator 5 3.2
Elementary Occupation 1 0.6
Armed Forces 2 1.3

matrices. Using Kaiser's criteria of eigenvalues of >1 alone to retain A two-factor solution demonstrated near simple structure with most
factors was not appropriate here, as variable communalities (after items loading onto one of the two factors. Upon closer inspection, it
extraction) did not consistently reach above .60 (Field, 2009). Using the appears that, whilst some overlap occurs, most of the items loading onto
scree plot, parallel analysis, the MAP and VSS tests, and inspecting factor one are principally representative of attitudes which a) directly
cumulative variance explained, this suggested either a two factor (scree blame the victim, and/or b) a suggest that only certain groups of men (in
plot, parallel analysis) or three factor (MAP, VSS) solution. Both solu- this case, gay or weak men) are raped. Items which load onto factor two
tions were run, with the two factor model explaining 41.66% of the appear to be primarily representative of attitudes which a) exonerate the
variance in the 55 items measured, while the three factor model perpetrator, and/or b) minimise the incident itself. The grouping of items
explained 43.70%. A Promax rotation was then applied to the factors. in this way suggests that, whilst they are all representative of male rape
Promax rotation specifically was used upon recommendations from the myths, there may be a latent delineation between two distinct subsets of
EFA literature (Fabrigar et al., 1999; Russell, 2002), including the myth; Blame versus Minimisation/Exoneration. The two factors were found
technique being more efficient in determining an optimal rotated so- to be strongly positively correlated in the two factor solution (r ¼ .76).
lution through fewer iterations in large scale datasets. As the MAP and Seven items did not load above .40 onto either factor (6, 7, 15, 17, 31,
VSS tests indicated a very limited impact of the addition of the third 34, 46), suggesting that they were a) answered significantly differently to
factor, which only explained 2% of the variance with only three items other items, b) measuring an unrelated concept, or both. These items
loading out of 55 (26, 42 and 47, which all tapped into maintaining the were therefore identified for elimination for Study 2 and the final scale.
heteronormativity of the victim), it was decided the two factor solution In addition, any items loading below .42 onto a factor (items 3, 4 and 16)
would be the most appropriate to ensure the stability of the Study 1 data were identified for elimination to ensure only securely (not borderline)
solution. Factor loadings for each item are presented in Table 2. As loading items were taken forward for the study 2 scales(s). Finally the
recommended by Field (2009), a suppression threshold of .4 was chosen remaining four items identified for elimination through the descriptive
as this was an exploratory study. statistics and item-total correlations as above (items 10, 25, 27, 32) were

9
B.A. Hine et al. Heliyon 7 (2021) e07421

Table 4. Descriptive statistics for each MRMAS item – study 2.

Item Code Mean SD Item-Total Skewness raw Skewness SQRT Skewness LG10
Correlation item scores transformed transformed
1. If a man is raped it does not mean he is weak 1M.C.A.R 2.28 1.85 .437 1.410 1.059 0.757
2. A man who is raped must have been behaving in a 1S.B.A 1.80 1.28 .755 1.603 1.261 1.006
way that made him appear homosexual
3. Male rape is very rare, if it occurs at all 1C.T.C. 2.67 1.51 .568 0.600 0.242 -0.096
6. Male on male rape only happens to homosexual 2S.C.C. 1.82 1.21 .774 1.438 1.111 0.858
men
7. A male victim who ejaculates during the incident 1PL.B.C. 1.80 1.19 .685 1.549 1.151 0.863
has not been raped
8. Almost all male rape occurs in institutions such as 2C.T.C. 2.68 1.48 .547 0.589 0.208 -0.145
prisons or the military
9. A homosexual man who rapes other men does so 2P.B.C. 3.47 1.58 .418 0.057 -0.369 -0.786
out of sexual desire
10. Most men would be able to fight off a male 3M.C.C. 2.66 1.45 .653 0.588 0.224 -0.131
sexual attacker
11. Rape is an accepted risk of a ‘homosexual 3S.B.A. 1.80 1.23 .743 1.518 1.193 0.939
lifestyle’
14. Heterosexual men who commit rape do so to act 3P.C.C. 3.19 1.56 .478 0.163 -0.226 -0.579
upon secret homosexual desires
15. In ‘real’ cases of male rape, there will be some 4M.B.C. 3.03 1.72 .519 0.413 0.072 -0.256
evidence of physical resistance
16. Heterosexual men ‘cry rape’ to hide their 4S.B.A. 2.40 1.39 .681 0.631 0.329 0.065
homosexual activities
17. Even if force is used to initiate sex, the victim's 3PL.B.C. 2.03 1.35 .647 1.203 0.869 0.601
erection can be interpreted as pleasure
18. I would expect heterosexual victims of rape to 1E.C.A. 2.32 1.61 .646 1.069 0.728 0.443
be more traumatized than homosexual victims
19. Coercive sexual practices between men (e.g., 4C.T.C. 2.16 1.36 .513 0.817 0.575 0.371
forced oral sex) form a legitimate part of group
initiations such as those used in fraternities or
sporting societies
20. A man would not rape another man if he was 4P.B.C. 2.11 1.40 .635 1.256 0.859 0.534
sexually fulfilled elsewhere
21. Just because a man is raped does not mean he is 5S.C.C.R 2.05 1.69 .492 1.604 1.289 1.007
homosexual
22. Men should feel ashamed as a result of being 2E.C.A. 1.74 1.29 .678 1.852 1.462 1.181
raped
23. Most cases of male rape include the use of a 5C.T.C. 3.36 1.30 .345 -0.394 -0.762 -1.115
weapon
24. Male rape is only perpetrated by homosexual 5P.C.C. 2.07 1.34 .731 1.145 0.808 0.525
men
25. A homosexual man who has been raped 4PL.C.A. 1.71 1.19 .768 1.173 1.389 1.132
probably enjoyed the experience to some extent
26. A man is more responsible for his own rape if he 6C.B.A. 1.77 1.24 .832 1.592 1.284 1.037
frequents a known homosexual area or
establishment
27. Only men who are big and strong are able to 6P.C.C. 1.95 1.27 .791 1.316 0.962 0.661
rape other men
28. I would find it difficult to consider a man a ‘real 5M.C.A. 1.76 1.25 .806 1.694 1.351 1.091
man’ if he said he had been raped
29. If a man has already had consensual sex with 6S.B.A. 1.73 1.14 .830 1.637 1.261 0.974
other men, I would not believe his claims of rape
30. A man who is raped is not as traumatized by the 3E.C.A. 1.84 1.31 .752 1.595 1.228 0.953
experience as a woman
31. If a man is drunk or taking drugs he is accepting 6C.B.A 1.84 1.26 .685 1.506 1.162 0.882
rape as a possible risk
32. Men who commit rape are naturally more 7P.C.C. 3.30 1.60 .272 0.160 -0.231 -0.611
aggressive in their day to day lives
33. It is acceptable for a ‘real man’ to show fear 6M.B.A.R 2.38 1.87 .414 1.208 0.917 0.642
during a sexual attack by another man
34. A man who claims to have been raped probably 7S.B.A. 1.87 1.16 .801 1.244 0.916 0.652
just changed his mind after initially consenting to
sex
35. A male victim's reaction to rape is more likely to 4E.B.C. 3.14 1.70 .285 0.247 -0.092 -0.400
be practical than emotional (e.g., obtaining a HIV
test rather than seeking support)

(continued on next page)

10
 B.A. Hine et al. Heliyon 7 (2021) e07421

Table 4 (continued )
Item Code Mean SD Item-Total Skewness raw Skewness SQRT Skewness LG10
Correlation item scores transformed transformed
36. A male victim of rape must have behaved in a 7C.B.A. 1.74 1.13 .826 1.625 1.241 0.946
way that invited the assault
37. A heterosexual man who had been raped would 7M.C.A.R. 2.75 1.78 .519 0.842 0.485 0.130
still be desirable to women
38. Male rape is a homosexual act 8S.C.C. 2.80 1.76 .528 0.709 0.336 0.004
39. If a man has been raped he should be able to 5E.C.A. 1.79 1.24 .757 1.652 1.261 0.965
cope on his own
40. I would find it difficult to believe a man had 8C.B.A.R. 2.01 1.35 .769 1.343 0.963 0.653
been raped if he had previously consented to sex
with the same man
41. Regardless of how they identify themselves, I 8P.C.A 2.69 1.71 .588 0.794 0.408 0.073
believe that men who rape other men are
homosexual
42. ‘Real men’ cannot be raped 8M.C.A. 1.61 1.16 .773 2.149 1.721 1.403
43. I would expect a man to be ‘matter of fact’ and in 6E.B.A. 1.98 1.35 .745 1.377 1.010 0.708
control of his emotions when reporting a rape
44. A man who has been raped did not set sexual 9C.B.C. 2.02 1.36 .681 1.366 0.950 0.636
limits understood by the perpetrator
45. Male victims of rape have very little emotional 7E.C.A. 1.63 1.16 .753 1.997 1.617 1.335
trauma to cope with

Note: Skewness SE ¼ 0.106. Code is comprised of a) a number denoting the position of that item within each theme, b) a letter denoting the theme (M ¼ Masculinity, S ¼
Sexuality, PL ¼ Pleasure, E ¼ Effect, C ¼ Context, P ¼ Perpetrator), c) whether the item is Behavioral (B), Characterological (C) or Typological (T), and d) whether the
item is Affective (A) or Cognitive (C). If the item also has an R at the end of its code, this means the question is reverse scored.

removed. Again, all such decisions were taken as reliability was priori- Sexuality, Pleasure, Effect, Context, and Perpetrators. Importantly,
tised during analysis (Clifton, 2019). This left 41 items to be taken over following analysis from Study 1, this scale now has two identified sub-
into study 2 to assess the validity and reliability of the measure in a scales – a 23 item Blame subscale, and an 18 item Minimisation/Exon-
second sample. The reliability estimates for the 41 item model using eration subscale. Participants answered all items on a 7-point Likert scale
Maximum Likelihood estimation with Pearson (Polychoric) correlations between 1 (Strongly Disagree) and 7 (Strongly Agree). The reliability
and a Promax rotation were Alpha ¼ .96 (.97), Hierarchal omega ¼ .89 estimates for the 41 item model in this sample using Maximum Likeli-
(.71), Omega total ¼ .97 (.98). hood estimation with Pearson (Polychoric) correlations and a Promax
rotation were Alpha ¼ .97 (.97), Hierarchal omega ¼ .79 (.74), Omega
3. Study 2 total ¼ .97 (.98). The Cronbach's alpha and hierarchal omega demon-
strates reliability for a general factor solution, however the discrepancy
A second study was then conducted to provide reliability and validity in the Hierarchal omega and Omega total coefficients suggests the vari-
measurement for the new, refined MRMAS scale, utilising a more ance in the 41 items is likely to be attributable to a multiple factor
representative population. This included measurement of the relation- structure or specific item variances. Cronbach's alpha values for the study
ship between the new MRMAS scale, and related, proximal constructs 1 version of the subscales 1 (Blame) and 2 (Minimisation/Exoneration)
previously shown to have an association with male rape myth accep- were .96 and .91 respectively.
tance, including both demographic (i.e., age) and attitudinal variables An additional measure of male rape myths was also administered.
(i.e., homophobia; see Walfield, 2018 for review). Namely, the six items from Struckman-Johnson and Struckman-Johnson
(1992) rape myth measure, which describe a man victimised by another
3.1. Method man. Two items measured myths surrounding trauma (e.g., “Most men
who are raped by a man do not need counselling after the incident”), two
3.1.1. Participants measured denial (e.g., “It is impossible for a man to rape a man), and two
528 individuals (who did not take part in Study 1) took part in study 2 measured blame (e.g., “Most men who are raped by a man are somewhat
as part of two groups. First, undergraduate students (N ¼ 346, M ¼ to blame for not being more careful). This measure uses a 6-point Likert
22.97yrs, SD ¼ 5.00, min ¼ 18.00, max ¼ 49.00, 214 female) from two scale (1 ¼ strongly disagree, 6 ¼ strongly agree), with higher scores
universities in the UK; one in London, and one in the North West (though indicating more endorsement of these rape myths. The Cronbach's alpha
this was not recorded during participation). Second, members of the in the present sample was .76 (.88), Hierarchal omega ¼ .57 (.82) and
general population (N ¼ 182, M ¼ 35.39yrs, SD ¼ 10.95, min ¼ 19.00, Omega total ¼ .87 (.93), suggesting acceptable unidimensional reliability
max ¼ 68.00, 110 female). Information on participants sex, identified when estimated using Polychoric correlations.
gender, sexuality, ethnicity, employment status, and profession are
detailed in Table 3. All participants were offered the chance to be entered [Link]. Female rape myths. The Acceptance of Modern Myths About
into a prize draw to win a £25 Amazon voucher, and psychology students Sexual Aggression (AMMSA) scale (Gerger et al., 2007) provided a
were also awarded two research credits, as part of departmental initia- measure of female rape myth acceptance. Participants answered 30
tives to encourage research participation. questions, using a Likert scale ranging from ‘strongly disagree’ to
‘strongly agree’. Examples include “When it comes to sexual contacts,
3.1.2. Materials women expect men to take the lead” and “If a woman invites a man to her
home for a cup of coffee after a night out this means that she wants to
[Link]. Male rape myths. A revised 41-item version of the Male Rape have sex”. This measure was selected as it overcomes many of the pitfalls
Myth Acceptance Scale (MRMAS) from Study 1 was administered to associated with historical measures of rape myth acceptance (e.g. Rape
participants. These items again assessed beliefs relating to Masculinity Myth Acceptance Scale – RMAS; Burt, 1980; Attitudes Towards Rape

11
B.A. Hine et al. Heliyon 7 (2021) e07421

Table 5. Exploratory factor loadings using the Principal Axis Factoring technique for each MRMAS item – study 2

Item Factor 1 – Blame Factor 2 – Minimisation/Exoneration

42. ‘Real men’ cannot be raped .928
45. Male victims of rape have very little emotional trauma to cope with .912
21. Just because a man is raped does not mean he is homosexual .878 -.409
2. A man who is raped must have been behaving in a way that made him appear .874
homosexual
33. It is acceptable for a ‘real man’ to show fear during a sexual attack by another .833 -.443
man
36. A male victim of rape must have behaved in a way that invited the assault .818
6. Male on male rape only happens to homosexual men .807
37. A heterosexual man who had been raped would still be desirable to women .785
26. A man is more responsible for his own rape if he frequents a known .768
homosexual area or establishment
39. If a man has been raped he should be able to cope on his own .768
29. If a man has already had consensual sex with other men, I would not believe his .767
claims of rape
28. I would find it difficult to consider a man a ‘real man’ if he said he had been .746
raped
25. A homosexual man who has been raped probably enjoyed the experience to .743
some extent
43. I would expect a man to be ‘matter of fact’ and in control of his emotions when .728
reporting a rape
22. Men should feel ashamed as a result of being raped .687
11. Rape is an accepted risk of a ‘homosexual lifestyle’ .672
34. A man who claims to have been raped probably just changed his mind after .657
initially consenting to sex
30. A man who is raped is not as traumatized by the experience as a woman .665
27. Only men who are big and strong are able to rape other men .598
44. A man who has been raped did not set sexual limits understood by the .590
perpetrator
40. I would find it difficult to believe a man had been raped if he had previously .546
consented to sex with the same man
1. If a man is raped it does not mean he is weak .543
7. A male victim who ejaculates during the incident has not been raped .529
31. If a man is drunk or taking drugs he is accepting rape as a possible risk .511
24. Male rape is only perpetrated by homosexual men .416
14. Heterosexual men who commit rape do so to act upon secret homosexual .801
desires
9. A homosexual man who rapes other men does so out of sexual desire .768
15. In ‘real’ cases of male rape, there will be some evidence of physical resistance .749
41. Regardless of how they identify themselves, I believe that men who rape other .681
men are homosexual
32. Men who commit rape are naturally more aggressive in their day to day lives .665
23. Most cases of male rape include the use of a weapon .595
16. Heterosexual men ‘cry rape’ to hide their homosexual activities .586
20. A man would not rape another man if he was sexually fulfilled elsewhere .568
38. Male rape is a homosexual act .550
18. I would expect heterosexual victims of rape to be more traumatized than .517
homosexual victims
10. Most men would be able to fight off a male sexual attacker .480
17. Even if force is used to initiate sex, the victim's erection can be interpreted as .444
pleasure
8. Almost all male rape occurs in institutions such as prisons or the military .423
3. Male rape is very rare, if it occurs at all
19. Coercive sexual practices between men (e.g., forced oral sex) form a legitimate part
of group initiations such as those used in fraternities or sporting societies
35. A male victim's reaction to rape is more likely to be practical than emotional (e.g.,
obtaining a HIV test rather than seeking support)

Note aItems shown in bold loaded onto different factors in Study 2 than in Study 1. bItalicised items did not load on either factor at .42, and were thus eliminated from the
scale and further analysis.

12
B.A. Hine et al. Heliyon 7 (2021) e07421

Scale – ATR; Feild, 1978; Illinois Rape Myth Acceptance Scale – IRMAS;

Chi square/df, good CFI (standard), TFI, good RMSEA (standard), SRMR

transformed, LG10 ¼ Log 10 transformed scores, df ¼ degrees of freedom, CFI ¼ Comparative fit index, TLI ¼ Tucker-Lewis fit index, RMSEA ¼ Root Mean Square Error of Approximation, SRMR ¼ Standardized Root Mean
Note: Error variances were set to be correlated between the items falling within each of the six item type criteria (Masculinity, Sexuality, Pleasure, Effect, Context, Perpetrator). Abbreviations: SQRT ¼ Square root
Payne et al., 1999), (for details see (Bohner et al., 2009; Gerger et al.,
2007; Hinck and Thomas, 1999; Payne et al., 1999). The Cronbach's

Chi square/df, acceptable CFI, good RMSEA (standard), SRMR

Alpha of the AMMSA in the present sample was .94 (.95), Hierarchal
omega ¼ .77 (.85) and Omega total ¼ .95 (.96), in keeping with previous

Chi square/df, CFI, TFI, good RMSEA (standard), SRMR

Alpha levels of .92 (Gerger et al., 2007).

Chi square/df, CFI, TFI(standard), RMSEA, SRMR

[Link]. Homophobia. A modified 27-item version of the LGB-KASH
(Worthington et al., 2005) was administered to measure attitudes to-
ward LGB individuals. It contained the following five subscales: Hate (6
items), LGB Knowledge (5 items), LGB Civil Rights (4 items), Religious
Chi square/df, CFI, RMSEA, SRMR
suggesting at least acceptable fit

Conflict (7 items), and Internalised Affirmativeness (5 items). One LGB

Chi square/df, RMSEA, SRMR

Civil Rights item relating to provision of healthcare for same-sex cou-

ples was removed, as, in the UK, healthcare is free to all at the point of
delivery. Items relating to LGB knowledge were also modified to
Model fit criteria

represent knowledge of UK LGB history. All items are rated on a

Likert-type scale ranging from 1 (very uncharacteristic of me or my
views) to 6 (very characteristic of me or my views). Higher scores on
each subscale are indicative of higher levels of the construct measured
N/A

N/A

N/A

by that subscale. Examples of items include the following: “I think

marriage should be legal for same-sex couples” (LGB Civil Rights);
SRMR

“Feeling attracted to another person of the same sex would not make me
.062
.057

.060
.055

.057
.052

uncomfortable” (Internalised Affirmativeness); “I am knowledgeable

about the history and mission of the PFLAG organization” (LGB
RMSEA (Robust)

Knowledge); “I sometimes think about being violent toward LGB peo-

ple” (Hate); and “I keep my religious views to myself in order to accept
Table 6. A summary of the fit indices of the confirmatory factor analysis models run using the 38 items identified across both studies.

.060 (.073)
.054 (.065)

.061 (.071)
.054 (.064)

.060 (.069)
.053 (.061)

LGB people” (Religious Conflict). Worthington et al. (2005) reported

evidence for the validity of the scale via findings of hypothesised re-
lationships with social dominance orientation, sexual identity devel-
opment, homophobia and biphobia, age, gender, and sexual orientation
TLI (Robust)

identity. They report internal consistencies from .73 to .94 for the
.86 (.86)
.89 (.89)

.89 (.88)
.92 (.90)

.92 (.89)
.93 (.92)

various subscales in three separate studies. In this sample the Cron-

bach's alpha estimate for all items in the scale overall was .91 (.94),
while the Hierarchal Omega coefficient was .52 (.64) and the Omega
total was .93 (.96) suggesting a multivariate structure was appropriate
CFI(Robust)

as previously determined by Worthington et al. (2005). Alpha Scores

.89 (.89)
.91 (.91)

.92 (.90)
.93 (.92)

.93 (.91)
.95 (.93)

varied for the five subscales: Hate (.82), LGB Knowledge (.85), LGB Civil
Rights (.85), Religious Conflict (.78), and Internalised Affirmativeness
(.35) subscales.
Robust Chi square/df

3.2. Results

Descriptive statistics, including item means, item distribution statis-

tics and item-to-total correlations were produced. An exploratory factor
14.70

19.17
2.87
2.51

2.94
2.54
2.52
2.90
2.48

analysis was also computed to ratify the presence of the two-factor so-
lution identified in Study 1. The aim was to then use Confirmatory factor
analyses to compare the general factor, two factor (Blame and Mini-
Robust df

misation/Exoneration) and six factor solution (Masculinity, Sexuality,

703
556
555
703
556
555
703
556
555

Effect, Context, Pleasure, Perpetrator) in the Study 2 data to see how each
solution fits. Group differences were assessed through the computation of
Robust Chi square

several t-tests. Correlations were also conducted between the MRMAS

measures (in models displaying good fit) and other proximal constructs.
10332.918

13478.129

16533.831
1597.99
1395.65

1633.71
1411.70

1610.63
1376.61

3.2.1. Descriptive statistics

Item means, skewness statistics and item-total correlations for the
revised MRMAS are shown in Table 4. Overall, item means were still
towards the middle to lower part of the Likert scale, showing that data
General factor model, SQRT scores

General factor model, LG10 scores

General factor model, raw scores

remained positively skewed (supported by the raw item skewness sta-

Two factor model, SQRT scores

Two factor model, LG10 scores

Two factor model, raw scores
Baseline model, SQRT scores

Baseline model, LG10 scores

tistics). However, item-total correlations again demonstrated that most

Baseline model, raw scores

items correlated strongly to the total.

Square Residual.

3.2.2. Exploratory factor analysis

All of the parameters identified in Study 1 (e.g., coefficients values,
Kaiser-Meyer-Oklin values etc.) were met in this sample also. Using the
Model

Scree plot, Parallel Analysis, inspecting cumulative variance explained,

and running the VSS test suggested a two-factor extraction was

13
B.A. Hine et al. Heliyon 7 (2021) e07421

Table 7. Descriptive statistics for all scales – study 2.

N Range Minimum Maximum Mean SD Skewness Kurtosis

Male Rape Myth Acceptance Scale (MRMAS) – 38 items Men 204 4.50 1.03 5.53 2.63 1.01 0.559 -0.352
Women 324 4.39 1.00 5.39 1.94 0.74 1.263 1.614
Students 346 4.53 1.00 5.53 2.40 0.99 0.780 -0.093
General Pop 182 2.45 1.00 3.45 1.83 0.61 0.807 -0.174
Total 528 4.53 1.00 5.53 2.20 0.92 1.013 0.559
Male Rape Myth Acceptance Scale (MRMAS) – 25 items Blame subscale Men 204 4.76 1.00 5.76 2.36 1.13 0.701 -0.456
Women 324 4.32 1.00 5.32 1.63 0.77 1.673 2.618
Students 346 4.76 1.00 5.76 2.14 1.09 0.888 -0.168
General Pop 182 2.20 1.00 3.20 1.47 0.52 1.326 1.090
Total 528 4.76 1.00 5.76 1.91 0.99 1.264 0.887
Male Rape Myth Acceptance Scale (MRMAS) – 13 items Minimisation/Exoneration subscale Men 204 5.23 1.08 6.31 3.15 1.03 0.067 -0.532
Women 324 4.54 1.00 5.54 2.53 0.91 0.418 -0.314
Students 346 5.31 1.00 6.31 2.91 1.02 0.258 -0.442
General Pop 182 4.15 1.00 5.15 2.51 0.92 0.417 -0.525
Total 528 5.31 1.00 6.31 2.77 1.01 0.330 -0.453
Acceptance of Modern Myths of Sexual Aggression (AMMSA) – 30 items Men 197 5.27 1.10 6.37 3.22 1.01 0.236 -0.035
Women 315 4.07 1.00 5.07 2.57 0.91 0.330 -0.875
Students 336 5.37 1.00 6.37 2.91 0.99 0.190 -0.409
General Pop 176 4.87 1.17 6.03 2.65 0.99 0.671 0.047
Total 512 5.37 1.00 6.37 2.82 0.99 0.344 -0.367
Lesbian, Gay, and Bisexual Knowledge and Attitudes Scale for Heterosexuals (LGB-KASH) – 27 items Men 192 3.52 1.19 4.70 2.96 0.74 -0.320 -0.418
Women 318 3.93 1.22 5.15 2.43 0.74 0.582 -0.372
Students 337 3.96 1.19 5.15 2.80 0.77 -0.040 -0.868
General Pop 173 344 1.19 4.63 2.29 0.68 0.779 0.456
Total 510 3.96 1.19 5.15 2.63 0.78 0.226 -0.816
Struckman-Johnson & Struckman-Johnson Rape Myth Scale (SJ-SJ RMS) – 6 items Men 200 4.50 1.00 5.50 1.99 0.96 0.690 -0.463
Women 318 3.83 1.00 4.83 1.57 0.74 1.389 1.626
Students 341 4.50 1.00 5.50 1.91 0.94 0.774 -0.284
General Pop 177 2.33 1.00 3.33 1.38 0.52 1.488 2.12
Total 518 4.50 1.00 5.50 1.73 0.86 1.107 0.518

appropriate. The Velicer's MAP test suggested a four factor structure, A two-factor solution demonstrated near simple structure with most
however upon review of the indices the third and fourth factors added items loading onto one of the two factors. All items (bar two, shown in
little to the model, so we proceeded with the two factor extraction. bold) loaded on the same factors as in Study 1, further supporting the
Exploratory factor analysis was again conducted using the Principal selection of a two-factor solution; victim blaming versus exonerating
Axis Factoring (PAF) technique with a promax (oblique) rotation then perpetrator/minimisation. Two items did not load onto either factor in
applied due to the degree of positive skewness observed in item re- this sample (items 3 and 19), while another two did not load at .42 (items
sponses. The two factor model explained 50.75% of the variance. A 24 and 35). These items were identified for elimination. As the two factor
suppression threshold of .42 was chosen to ensure secure factor load- structure has been maintained through a second PAF analysis, all of the
ings onto a factor. Factor loadings for each item are presented in 41 item scores were transformed towards greater normality of distribu-
Table 5. tion in preparation for running the CFA models. Square root and Log 10

Table 8. Construct validity Pearson correlations.

1 2 3 4 5 6 7 8 9 10 11 12
1. MRMAS-Total -
2. MRMAS-Blame .96*** -
3. MRMAS-Exon .86*** .68*** -
4. Age -.21*** -.22*** -.14** -
5. Coded sexuality -.22*** -.19*** -.22*** -.01 -
6. LGB-KASH hate .53*** .54*** .40*** -.18*** -.16*** -
7. LGB-KASH knowledge .56*** .52*** .52*** -.14** -.40*** .25*** -
8. LGB-KASH Rights .67*** .71*** .46*** -.17*** -.20*** .49*** .52*** -
9. LGB-KASH religious .24*** .21*** .25*** -.10* -.13** .44*** .15** .13** -
10. LGB-KASH IA .50*** .47*** .47*** -.09 -.43*** .29*** .62*** .53*** .24*** -
11. AMMSA .55*** .47*** .59*** -.07 -.15** .48*** .32*** .27*** .33*** .35*** -
12. SJ-SJ MRS .69*** .75*** .43*** -.20*** -.12** .53*** .38*** .54*** .25*** .33*** .41*** -

Note: *p < .05. **p < .01. ***p < .001.

Categorical variables included for complete output only.

14
B.A. Hine et al. Heliyon 7 (2021) e07421

transformations were conducted separately on the raw item scores due to raw or normalised score transformation, with the two factor structure
the trend towards positive skew in the raw item data. The skewness providing a better fit to the data than the general factor model. Although
statistics suggest that the Log 10 transformation was more effective at under conditions of normalised transformation of scores (Square root or
normalising the distribution of the data (see Table 4). When correlations Log 10 transformation) the general factor model starts to demonstrate
were performed between the raw item scores and the two different acceptable levels of fit across two of the indices outlined above. The Log
normalised transformations of the same item, the raw item scores were 10 transformed versions of the scores demonstrate a better degree of
found to be correlated at .99 with the square root transformed scores, and model fit to the data than both the square root transformed scores, and
at .97 with the Log 10 transformed scores. The same two factor structure the raw scores, with the Two factor model estimated using Log 10
was maintained in the transformed for normality data (both the Square transformed scores meeting good levels of fit.
root and Log 10 analyses separately), although item 24 maintained a Following on from the confirmatory factor analyses supporting the
loading above .42 in both of these transformations. Therefore items 3, 19 existence of the general factor (under normal transformations) and the
and 35 only were eliminated from the scale and subsequent analyses, as two factor solution, descriptive statistics were run on the total scores
the raw item loading of item 24 was .416. This leaves a two factor so- calculated for all measures in the study (presented in Table 7).
lution containing 38 items overall (25 Blame, 13 Minimisation/Exoner- For the MRMAS measures these displayed low mean scores as would
ation). The reliability estimates for the raw 38 item model in this sample be expected, although there was a general trend towards the Mini-
using Maximum Likelihood estimation with Pearson (Polychoric) corre- misation/Exoneration subscale means displaying a slightly higher
lations and a Promax rotation were Alpha ¼ .97 (.97), Hierarchal omega average when compared to the Blame subscale. As this may be an artifact
¼ .83 (.72), Omega total ¼ .97 (.98). Cronbach's Alpha analyses showed of the number of items in each subscale, to test whether there were any
factor 1 (Blame) and factor 2 (Minimisation/Exoneration) displayed significant within participants difference in scores on these two subscales
excellent or close to excellent reliability (.96 and .89 respectively). in each of the outlined demographic categorisations in Table 7, Z score
versions of these subscale variables were created in SPSS. This was
3.2.3. Confirmatory factor analysis designed to place each of the subscale scores in the same unit of mea-
Confirmatory factor analyses were conducted using Robust maximum surement (using the total sample mean and SD of each subscale in the
likelihood estimation (MLM variant), which is available through the calculation of that Z score variable). These were then compared using a
Lavaan package in R. Robust maximum likelihood estimation was used, selection of Paired sample t-tests where the file was split by one of the
as this allows for some violation in the normal distribution assumption demographic grouping variables (Sex or Student/General population).
when compared to the standard version (Schmidt, 2011). Comparisons Neither of the Sex categories (Male or Female) displayed significant
will be performed in the model fit across the raw item scores, and the two within participant differences in Blame and Minimisation/Exoneration Z
sets of normalised item scores separately. The intention of using CFA is to scores. However, the Student group and the General population group
compare the fit of the general factor, two factor (Blame and Mini- did. The Student group displayed significantly higher levels of Blame
misation/Exoneration) and six factor solutions (Masculinity, Sexuality, than Minimisation/Exoneration (Blame mean Z ¼ 0.23, Minimisation/
Effect, Context, Pleasure, Perpetrator) in the Study 2 data to see how well Exoneration mean Z ¼ 0.14, t (345) ¼ 2.054, p ¼ .041), while the general
each solution fits. population group displayed significantly higher levels of Minimisation/
In each of the models the error co-variances were set to be estimated Exoneration than Blame (Blame mean Z ¼ -0.44, Minimisation/Exoner-
between the items falling within each of the six type areas (Masculinity, ation mean Z ¼ -0.26, t (181) ¼ -3.607, p < .001). It is worth noting that
Sexuality, Effect, Context, Pleasure, Perpetrator) as there is likely to be the Student group displayed higher levels of both Blame and Mini-
commonly shared item variance in the measurement of the items tapping misation/Exoneration characteristics than the General population group
into each of the six previously theory-determined areas that were used to (note the mean across all Z scores is 0, this direction is also supported by
inspire the creation of the items. Several different indices of model fit Table 7).
were estimated, including the chi square/df ratio (<3 ¼ good),
Comparative fit index (CFI, >.95 ¼ good fit, >.90 ¼ acceptable fit), 3.3. Relationships between variables
Tucker-Lewis fit index (TFI, >.95 ¼ good fit, >.90 ¼ acceptable fit), Root
mean square error of approximation (RMSEA, <.06 ¼ good fit, <.08 ¼ Additionally, in order to assess the construct validity of the MRMAS
acceptable fit) and the standardised root mean square residual (SRMR, scales, it was administered alongside a number of conceptually proximal
<.08 ¼ good fit). These guidelines of model fit are based on reviewing variables (as described above under materials; see Table 8).
literature debating fit indices published by Hu and Bentler (1998), Highest correlations were, as anticipated, present for those variables
Marsh, Hau and Wen (2004), Russell (2002), and Scmitt (2011). Hu and most proximal, and sharing greatest conceptual overlap. In order of
Bentler (1998) recommendations (“good fit”, outlined above) are the strength of relationships, a significant positive correlation was noted
standard that analyses aim to adhere to, Marsh et al., (2004) highlight between the MRMAS total and Struckman-Johnson and Struckman-
that in models with a larger number of variables/items (such as ques- Johnson Male Rape Scale (1992), r (514) ¼ .69, p < .001; The MRMAS
tionnaires) reaching these model fit criteria become challenging to meet, and the rights sub scale of the LGB-KASH inventory, r (506) ¼ .67, p <
even at the “acceptable” levels outlined above (Russell, 2002, also briefly .001; with slightly weaker relationships to the LGB-KASH knowledge, r
addresses this). The use of data that is non-normally distributed for (506) ¼ .56, p < .001; The AMMSA measure of modern myths directed
theoretically valid reasons in its raw format (as we have in this study) towards male on female rape (Gerger et al., 2007), r (508) ¼ .55, p <
further increases this challenge. .001; The LGB-KASH hate subscale, r (506) ¼ .53, p < .001; and The
Table 6 provides a summary of the fit indices for the general factor LGB-KASH internalised affirmativeness subscale, r (506) ¼ .50, p < .001,
and two factor solutions. Unfortunately, the six factor solution did not were also all significantly, positively correlated. The MRMAS subscales
converge. Table 6 also highlights how many of the fit indices have met at both displayed trends in a similar direction, although there was some
least an acceptable level of fit in a particular model based on the divergence in the correlations to LGB-KASH rights, with Blame displaying
guidelines above (last column). Note that due to the sample size collected a much stronger correlation, r (506) ¼ .71, p < .001, than Mini-
the chi square/df ratio and SRMR may be displaying a bias towards better misation/Exoneration, r (506) ¼ .46, p < .001. The same pattern was
fit. This is why greater credence is being placed on a combination of two displayed with correlations to the Struckman-Johnson and
indices out of CFI/TLI and RMSEA meeting (at least) acceptable fit con- Struckman-Johnson Male Rape Scale, with Blame displaying a much
ditions. The two factor models consistently met this criteria, regardless of stronger correlation, r (514) ¼ .75, p < .001, than

15
B.A. Hine et al. Heliyon 7 (2021) e07421

Minimisation/Exoneration, r (514) ¼ .43, p < .001. The opposing di- research utilizing this scale, and thus greater insight into the experiences
rection was found for the correlations to the AMMSA with Blame dis- and responses to male victimization.
playing a slightly weaker correlation, r (508) ¼ .47, p < .001, than In determining where to begin, the literature exploring attitudes to-
Minimisation/Exoneration, r (508) ¼ .59, p < .001. wards incidences involving female victims provides a useful framework.
Specifically, examining the beliefs, judgments and actions of those in
4. Discussion specialist populations (e.g., those within the criminal justice system,
counsellors etc.) provides an important and highly impactful route to
This study introduces a new measure of male rape myth acceptance, improving the experiences of male victims. For example, the existence of
appropriately named the Male Rape Myth Acceptance Scale (MRMAS). robust ‘traditional’ rape myth acceptance scales has allowed for the
As such, this study provides the first highly reliable, peer reviewed, measurement of rape myth acceptance within police officers (Murphy
comprehensive measure of male rapes myths specifically. It is argued that and Hine, 2019), the assessment of their impact on police judgments
development of such a scale will allow for future measurement of male (Hine & Murphy, 2017, 2019), and the relationship between rape myths
rape myths, across a variety of populations, institutions and settings, as and case processing (Hine et al., 2021; Hohl and Stanko, 2015; Murphy
well as allowing for the correlation of such attitudes with judgments et al., 2021). However at present, whilst it has been argued that male
towards male victims and reporting. As with ‘traditional’ rape myths, the rape myths are held by police officers, and are influential in their case
opportunity to conduct research of this nature on male victims should decision making and approach (Javaid, 2016; Rumney, 2008), these
allow for the development of training, educational programs and in- observations are anecdotal. This new measure thus provides a previously
terventions targeting negative beliefs about the crime of male-on-male unavailable avenue to more rigorous, empirical assessment in this and
rape, in attempts to improve the experiences of victim-survivors upon other areas.
disclosure. Engagement in such areas of research would allow for the develop-
ment of informed and targeted training and educational programs on
male rape myths, and their impact on responses to victimization (both by
4.1. The Male Rape Myth Acceptance Scale (MRMAS)
the victim themselves, and those around them). Indeed, traditional rape
myth acceptance scales are routinely used to evaluate the success of
As mentioned above, the MRMAS achieved excellent levels of reli-
police training programs on rape investigation (see (Lonsway, Welch, &
ability across various methods as assessment. Favorable reliability scores
Fitgerald, 2001) for example). As such, this new measure not only pro-
were complimented by high levels of correlation with proximal measures
vides an empirical framework upon which to construct training programs
(in support of previous findings by Walfield, 2018), including measures
specific to male rape, but to evaluate them also. It can further be argued
of traditional rape myths (i.e., the AMMSA), previous measures of male
that, alongside these specific research avenues, simply raising awareness
rape myths (i.e., the Struckman-Johnson and Struckman-Johnson scale),
of the existence and nature of male rape myths is important in and of
and other attitudes (i.e., homophobia), thus suggesting excellent
itself. Particularly important is to incorporate relationships between at-
construct validity. In this sense, it can be confidently claimed that the
titudes into training programs; in the case of male rape myths, this would
MRMAS provides a robust measure of myths relating to male-on-male
mean acknowledging how homophobic attitudes towards sexual minor-
rape, and that this measure it improves upon previous scales which
ity men underpin these beliefs and constructing programs accordingly. In
have not reported strong reliability values (such as Struckman-Johnson
short, by successfully chronicling and measuring both broad male rape
and Struckman-Johnson, 1992 scale). It is also important to note that the
myth themes, and specific male rape myth examples, the validation of
MRMAS assesses agreement with a greater variety of myths than previous
this measure provides a clear catalogue of extant male rape myths for
measures (including Melanson’s 1999 scale), with items assessing mas-
specialist populations and the public at large.
culinity, sexuality, pleasure, perpetrators, context and effect. This, along
with a greater number of items overall, also allowed for a more
4.3. Limitations
comprehensive assessment of latent structure. A two-factor structure was
revealed, suggesting a grouping of items which blame the victim or
There are three principal limitations with the current study. First, as
suggest that only certain groups of men are raped, and those which
with any process of validation, some items were excluded during the
exonerate the perpetrator or minimise the incident. Indeed, Omega totals
process of analyses (for a variety of reasons: poor item-total correlations,
demonstrate that this multiple factor was most appropriate (though
item means, poor factor loadings, feedback from researchers regarding
reliability for the scale overall was still excellent, as measured by Cron-
participant understanding etc). As a consequence, it should be noted that
bach's Alpha and Hierarchial omega). This provides valuable information
some themes lost more items than others. Specifically, whilst most
as to the possible function of said myths, and how participants may be
‘themes’ have between seven and nine items, pleasure myths are only
using different subsets of myth to support their reasoning and responses
represented by three items. However, the scale appears to be structured
to victimisation (similarly to processes identified in research using
around a two-factor solution, and the number of items in each of these
traditional rape myths; see Bohner et al., 2009). Importantly, this moves
factors remains robust.
discussions beyond the thematically-driven six factor structure proposed
Second, as observed in the methods section of both studies, many of
earlier in this piece; suggesting instead that myths are grouped by the
the items, and the scale overall, appears to suffer from floor effects. This
broader latent structures around blame versus minimisation/exoneration
is common in scales which measure emotive or sensitive topics (such as
for the victim and perpetrator respectively.
rape), and in scales which are among the first to measure a specific
concept (similar issues were found for earlier traditional rape myth
4.2. Implications scales, such as the IRMAS; (Lonsway et al., 1999). Part of the issue may be
the wording of questions, as, for this first scale, the wording was not
As detailed above, the MRMAS appears to provide an extensive, veiled. Specifically, the word rape was used in several places, as were
robust measure of male rape myth acceptance. In the introduction of this other words such as erection, which may make some participants un-
paper, it was argued that producing such a scale would allow for the comfortable, and lead them to answer in socially desirable ways. Whilst
proliferation of literature on male rape; an endeavor that is, arguably, this was partially compensated for through the use of square root and
long overdue (M. Davies and Rogers, 2006). In this sense, this study LOG10 transformations, future research should perhaps look to develop
represents a ‘call to arms’ to those interested in both attitudes and re- more obviously valenced items, as occurred with the development of the
sponses towards incidences of rape and those involved, to develop AMMSA (Gerger et al., 2007).

16
B.A. Hine et al. Heliyon 7 (2021) e07421

Third and finally, the decision was taken in this study as part of scale References
development, based on the differential content and affect of male-female
versus male-female myths, to only include items pertaining to male-on- Ahmad, S., Bhugra, D., 2010. Homophobia: an updated review of the literature. Sex.
Relatsh. Ther. 25, 447–455.
male rape. This decision will allow for targeted and appropriate admin- Anderson, E., 2009. Inclusive Masculinity: the Changing Nature of Masculinities.
istration and application of this scale to cases involving men assaulted by Routledge, New York.
other men (e.g., when examining the relationship of such attitudes to Anderson, I., Quinn, A., 2009. Gender differences in medical students’ attitudes towards
male and female rape victims. Psychol. Health Med. 14, 105–110.
criminal cases), particularly in the UK (where the legislative definition Bartlett, M.S., 1954. A note on the multiplying factors for various chi square
defines rape as perpetrated by only men towards either men or women). approximations. J. Roy. Stat. Soc. 16, 296–298.
This decision is also supported by the fact that the sample in this study Bohner, G., Eyssel, F., Pina, A., Siebler, F., Tendayi Viki, G., 2009. Rape myth acceptance:
cognitive, affective and behavioural effects of beliefs that blame the victim and
was from the UK, and it may have been confusing for participants to exonerate the perpetrator. In: Horvath, M., Brown, J. (Eds.), Rape: Challenging
answer items about ‘rape’ perpetrated by females. However, it is Contemporary Thinking. Routledge, New York, NY, pp. 17–45.
important to recognize that the measurement of distinct attitudes per- Burt, M.R., 1980. Cultural myths and supports for rape. J. Pers. Soc. Psychol. 38,
217–230.
taining to the sexual assault of men by women is an important future
Campbell, A., 2005. Keeping the ‘lady’ safe: the regulation of femininity through crime
research direction. Ergo, the generation of a more general ‘male sexual prevention literature. Crit. Criminol. 13, 119–140.
assault scale’, or the addition of a sub-scale to the scale generated in this Chapleau, K.M., Oswald, D.L., Russell, B.L., 2008. Male rape myths: the role of gender,
study, can both be considered worthy and much needed endeavors. violence, and sexism. J. Interpers Violence 23, 600–615.
Clements-Schreiber, M.E., Rempel, J.K., 1995. Women’s acceptance of stereotypes about
male sexuality: correlations with strategies to influence reluctant partners. Can. J.
5. Conclusion Hum. Sex. 4, 223–231.
Clifton, J.D.W., 2019. Managing validity versus reliability trade-offs in scale-building
decisions. Psychol. Methods 25, 259–270.
It has been argued that research on male rape myths is around 20 Connell, R., 2002. Gender (Short Introductions). Polity Press, Cambridge, UK.
years behind that on female rape myths (Davies and Rogers, 2006). One Costello, A.B., Osborne, J., 2005. Best practices in exploratory factor analysis: four
of the principal barriers to progress in this area is the absence of a reliable recommendations for getting the most from your analysis. Practical Assess. Res. Eval.
10, 1–9.
measure of male rape myth acceptance. The present study provides such Cotton, P., 1992. Attacks on homosexual persons may be increasing, but many 'bashings'
a tool, in an attempt to facilitate increasing research in this area. It is still aren't reported to police. J. Am. Med. Assoc. 267, 2999–3000.
hoped that the generation of scholarly work in this area, facilitated Coxell, A.W., King, M.B., 2010. Adult male rape and sexual assault: prevalance, re-
victimsation and the tonic immobility response. Sex. Relatsh. Ther. 25, 372–379.
through the use of this scale, will result in improved understanding of Crown Prosecution Service CPS, 2015. Joint CPS and Police Rape Action Plan. Retrieved
male victimization, and subsequent reporting and coping behaviour. from. [Link]
This, in turn, should enable more effective support of male victims, and Davies, M., Gilston, J., Rogers, P., 2012. Examining the relationship betwen male rape
myth acceptance, female rape myth acceptance, victim blame, homophobia, gender
improved victim experience, particularly within the criminal justice roles, and ambivalent sexism. J. Interpers Violence 27, 2807–2823.
system. Davies, M., Pollard, P., Archer, J., 2001. The influence of victim gender and sexual
orientation on judgments of the victim in a depicted stranger rape. Violence Vict. 16,
607–619.
Declarations Davies, M., Rogers, P., 2006. Perceptions of male victims in depicted sexual assaults: a
review of the literature. Aggress. Violent Behav. 11, 367–377.
Author contribution statement Dinos, S., Burrowes, N., Hammond, K., Cunliffe, C., 2015. A systematic review of juries'
assessment of rape victims: do rape myths impact on juror decision-making? Int. J.
Law Crime Just. 43, 36–49.
Benjamin A. Hine, Anthony D. Murphy: Conceived and designed the Estrich, S., 1987. Real Rape. Harvard University Press, Cambridge, MA.
experiments; Performed the experiments; Analyzed and interpreted the Fabrigar, L.R., Wegener, D.T., MacCallum, R.C., Strahan, E.J., 1999. Evaluating the use
data; Contributed reagents, materials, analysis tools or data; Wrote the of exploratory factor analysis in psychological research. Psychol. Methods 4,
272–299.
paper. Feild, H., 1978. Attitudes toward rape - a comparative analysis of police, rapists, crisis
Jamie S. Churchyard: Analyzed and interpreted the data; Wrote the counselors, and citizens. J. Pers. Soc. Psychol. 36, 156–179.
paper. Field, A., 2009. Discovering Statistics Using SPSS. Sage, London.
Garnets, L., Herek, G., Levy, B., 1990. Violence and victimisation of lesbians and gay men:
mental health consequences. J. Interpers Violence 5, 366–383.
Funding statement Gerger, H., Kley, H., Bohner, G., Siebler, F., 2007. The acceptance of modern rape myths
about sexual aggression (AMMSA) scale: development and validation in German and
English. Aggress. Behav. 33, 422–440.
This research did not receive any specific grant from funding agencies Gonsiorek, J., 1994. A critique of current models in sexual abuse. In: Gonsiorek, J.C.,
in the public, commercial, or not-for-profit sectors. Bera, W.H., LeTourneau, D. (Eds.), Male Sexual Abuse. Sage, Thousand Oaks, CA,
pp. 21–36.
Goyer, P.F., Eddleman, H., 1984. Same-sex rape of nonincarcerated men. Am. J. Psychiatr.
Data availability statement 141, 576–579.
Groth, N., Burgess, A., 1980. Male rape: offenders and victims. Am. J. Psychiatr. 137,
806–810.
The authors do not have permission to share data. Grubb, A., Harrower, J., 2008. Attribution of blame in cases of rape: an analysis of
participant gender, type of rape and perceived similarity to the victim. Aggress.
Violent Behav. 13, 396–405.
Declaration of interests statement Grubb, A., Harrower, J., 2009. Understanding attribution of blame in cases of rape: an
analysis of participant gender, type of rape and percieved similarity to the victim. J.
The authors declare no conflict of interest. Sex. Aggress. 15, 63–81.
Grubb, A., Turner, E., 2012. Attribution of blame in rape cases: a review of the impact of
rape myth acceptance, gender role conformity and substance use on victim blaming.
Additional information Aggress. Violent Behav. 17, 443–452.
Hickson, F., Davies, P., Hunt, A., Weatherburn, P., McManus, T., Coxon, A., 1994. Gay
man as victims of nonconsensual sex. Arch. Sex. Behav. 23, 281–294.
No additional information is available for this paper. Hillman, R., O’Mara, N., Taylor-Robinson, D., Harris, J.R.W., 1990. Medical and social
aspects of sexual assault of males: a survey of 100 victims. Br. J. Gen. Pract. 40,
502–504.
Acknowledgements Hillman, R., O’Mara, N., Tomlinson, D., Harris, J.R.W., 1991. Adult male victims of sexual
assault: an underdiagnosed condition. Int. J. STD AIDS 2, 22–24.
The authors would like to acknowledge the dedication and hard work Hinck, S., Thomas, R., 1999. Rape myth acceptance in college students: how far have we
come? Sex Roles 40, 815–832.
of the following student research assistants in the conduct and comple-
Hine, B.A., Murphy, A.D., 2017. The impact of victim-perpetrator relationship, reputation
tion of this project; Bridget Scanlon, Gamze Er, Elara Calahorro Bello, and initial point of resistance on officers' responsibility and authenticity ratings
Ellie Woodward, Bethan Griffiths, Maryam Amin, and Nivek Jones. towards hypothetical rape cases. J. Crim. Justice 49, 1–13.

17
B.A. Hine et al. Heliyon 7 (2021) e07421

Hine, B.A., Murphy, A.D., 2019. The influence of ‘High’ vs. ‘Low’ rape myth acceptance Myers, M., 1989. Men sexually assaulted as adults and sexually abused as boys. Arch. Sex.
on police officers' judgements of victim and perpetrator responsibility, and rape Behav. 18, 203–215.
authenticity. J. Crim. Justice 60, 100–107. Nalavany, B.A., Abell, N., 2004. An initial validation of a measure of personal and social
Hine, B.A., Murphy, A.D., Yesberg, J.A., Wunsch, D., Charleton, B., 2021. Mapping the perceptions of the sexual abuse of males. Res. Soc. Work. Pract. 14, 368–378.
landscape of male-on-male rape in London: an analysis of cases involving male Office for National Statistics, 2018. Sexual Offences in England and Wales: Year Ending
victims reported between 2005 and 2012. Police Pract. Res. 22, 109–126. March 2017. Retrieved from, London.
Hohl, K., Stanko, E., 2015. Complaints of rape and the criminal justice system: fresh Parratt, K.A., Pina, A., 2017. From “real rape” to real justice: a systematic review of police
evidence on the attrition problem in England and Wales. Eur. J. Criminol. 12, 324–341. officers' rape myth beliefs. Aggress. Violent Behav. 34, 68–83.
Horn, J.L., 1965. A rationale and test for the number of factors in factor analysis. Payne, D.L., Lonsway, K.A., Fitzgerald, L.F., 1999. Rape myth acceptance: exploration of
Psychometrika 30, 179–185. its structure and its measurement using the Illinois rape myth acceptance scale. J. Res.
Horvath, M., Brown, J., 2013. Rape: Challenging Contemporary Thinking, third ed. Pers. 33 (1), 27–68.
Routledge, Abingdon, Oxon. Peterson, Z.D., Muehlenhard, C.L., 2004. Was it rape? The function of women’s rape myth
Howard, J., 1984a. The “normal” victim: the effects of gender stereotypes on reactions to acceptance and definitions of sex in labeling their own experiences. Sex Roles 51,
victims. Soc. Psychol. Q. 47, 270–281. 129–144.
Howard, J., 1984b. Societal influences on attribution: blaming some victims more than Revelle, W., 2017. Procedures for psychological, psychometric, and personality. R
others. J. Pers. Soc. Psychol. 47, 494–505. package Bpsych^., Version 1 (5).
Hu, L.T., Bentler, P.M., 1998. Fit indices in covariance structure modeling: Sensitivity to Revelle, W., Rocklin, T., 1979. Very simple structure: an alternative procedure for
underparameterized model misspecification. Psychol. Methods 3 (4), 424. estimating the optimal number of interpretable factors. Multivariate Behav. Res. 14,
Isely, P., 1991. Adult male sexual assault in the community: a literature review and group 403–414.
treatment model. In: Burgess, A.W. (Ed.), Rape and Sexual Assault III: A Research Rumney, P., 2008. Policing male rape and sexual assault. J. Crim. Law 72, 67–86.
Handbook. Garland, New York, pp. 161–178. Russell, D.W., 2002. In search of underlying dimensions: the use (and abuse) of factor
Isley, P., Gehrenbeck-Shim, D., 1997. Sexual assault of men in the community. J. analysis in Personality and Social Psychology Bulletin. Pers. Soc. Psychol. Bull. 28,
Community Psychol. 2, 159–166. 1629–1646.
Jamel, J., Bull, R., Sheridan, L., 2008. An investigation of the specialist police service Sarrell, P.M., Masters, W.H., 1982. Sexual molestation of men by women. Arch. Sex.
provided to male rape survivors. Int. J. Police Sci. Manag. 10, 486–508. Behav. 11, 117–130.
Javaid, A., 2016. Giving a voice to the voiceless: police responses to male rape. Policing: Scarce, M., 1997. Same-sex rape of male college students. J. Am. Coll. Health 45,
J. Pol. Pract. 11, 146–156. 171–173.
Johnson, B.D., Kuck, D.L., Schander, P.R., 1997. Rape myth acceptance and Schmidt, T.A., 2011. Current methodological considerations in exploratory and
sociodemographic characteristics: a multidimensional analysis. Sex Roles 36, 693–707. confirmatory factor analysis. J. Psychoeduc. Assess. 29, 304–321.
Kaiser, H., 1970. A second generation little jiffy. Psychometrika 35, 401–415. Sleath, E., Bull, R., 2010. Male rape victim and perpetrator blaming. J. Interpers Violence
Kaiser, H., 1974. An index of factoral simplicity. Psychometrika 39, 31–36. 25, 969–988.
Kassing, L., Beesley, D., Frey, L., 2005. Gender role conflict, homophobia, age, and Sleath, E., Bull, R., 2017. Police perceptions of rape victims and the impact on case
education as predictors of male rape myth acceptance. J. Ment. Health Counsel. 27, decisionmaking: a systematic review. Aggress. Violent Behav. 34, 102–112.
311–328. Stemple, L., Meyer, I., 2014. The sexual victimization of men in America: new data
Kaufman, A., 1984. Rape of men in the community. In: Stewart, I.R., Greer, G.J. (Eds.), challenge old assumptions. Am. J. Publ. Health 104, 19–26.
Victims of Sexual Aggression. Van Nostrand Reinhold, New York, NY, pp. 57–77. Stermac, L., Bove, G.d., Addison, M., 2004. Stranger and acquaintance sexual assault of
King, M.B., 1992. Male sexual assault in the community. In: Mezey, G., King, M. (Eds.), adult males. J. Interpers Violence 19, 901–915.
Male Victims of Sexual Assault. Oxford University Press, Oxford, pp. 1–12. Struckman-Johnson, C., 1991. Male victims of acquaintance rape. In: Parrot, A.,
Lacey, H.B., Roberts, R., 1991. Sexual assault on men. Int. J. STD AIDS 2, 258–260. Bechhofer, L. (Eds.), Acquaintance Rape: the Hidden Crime. Wiley, New York,
Laurent, C., 1993. Male rape. Nursing Times 89, 18–19. pp. 192–213.
Lonsway, K.A., Welch, S., Fitzgerald, L.F., 2001. Police training in sexual assault response: Struckman-Johnson, C., Struckman-Johnson, D., 1992. Acceptance of male rape myths
Process, outcomes, and elements of change. Crim. Justice Behav. 28 (6), 695–730. among college men and women. Sex Roles 27, 85–100.
Marsh, H.W., Hau, K.T., Wen, Z., 2004. In search of golden rules: Comment on hypothesis- Struckman-Johnson, C., Struckman-Johnson, D., 1994. Men pressured and forced into
testing approaches to setting cutoff values for fit indexes and dangers in sexual experience. Arch. Sex. Behav. 23, 93–114.
overgeneralizing Hu and Bentler’s (1999) findings. Struct. Equ. Model. 11 (3), Suarez, E., Gadalla, T.M., 2010. Stop blaming the victim: a meta-analysis on rape myths.
320–341. J. Interpers Violence 25, 2010–2035.
Marshall, W.L., Hambley, L.S., 1996. Intimacy and loneliness, and their relationship to Suschinsky, K., Lalumiere, M., 2010. Prepared for anything?: an investigation of female
rape myth acceptance and hostility towards women among rapists. J. Interpers genital arousal in repsonse to rape cues. Psychol. Sci. 22, 159–165.
Violence 11, 586–592. Sussenbach, P., Bohner, G., 2011. Acceptance of sexual aggression myths in a
McGee, H., O’Higgins, M., Garavan, R., Conroy, R., 2011. Rape and child sexual abuse: representative sample of German residents. Aggress. Behav. 37, 374–385.
what beliefs persist about motives, perpetrators, and survivors? J. Interpers Violence Temkin, J., 2000. Prosecuting and defending rape: perspectives from the bar. J. Law Soc.
26, 3580–3593. 27, 219–248.
McMullen, R.J., 1990. Male Rape: Breaking the Silence on the Last Taboo. GMP, London. Temkin, J., Krah
e, B., 2008. Sexual Assault and the justice gap: A Question of Attitude.
Melanson, P.K., 1999. Belief in male rape myths: a test of two competing theories Hart Publishing, Portland, OR.
(Doctoral dissertation, Queen’s University, 1999). Diss. Abstr. Int. 59, 5620. Turchik, J., Edwards, K., 2012. Myths about male rape: a literature review. Psychol. Men
Mezey, G., King, M., 1987. Male victims of sexual assault. Med. Sci. Law 27, 122–124. Mascul. 13, 211–226.
Mezey, G., King, M., 1989. The effects of sexual assault on men: a survey of 22 victims. Velicer, W.F., 1976. Determining the number of components from the matrix of partial
Psychol. Med. 19, 205–209. correlations. Psychometrika 41 (3), 321–327.
Mitchell, D., Hirschman, R., Hall, G.N., 1999. Attributions of victim responsibility, Walfield, S., 2018. “Men cannot be raped”: correlates of male rape myth acceptance. J.
pleasure, and trauma in male rape. J. Sex. Res. 36, 369–373. Interpers Violence 1–27.
Murphy, A.D., Hine, B.A., 2019. Investigating the demographic and attitudinal predictors White, B., Kurpius, S., 2002. Effects of victim sex and sexual orientation on perceptions of
of rape myth acceptance in U.K. Police officers: developing an evidence-base for rape. Sex Roles 46, 191–200.
training and professional development. Psychol. Crime Law 25, 69–89. Worthington, R., Dillon, F., Becker-Schutte, A., 2005. Development, reliability, and
Murphy, A.D., Hine, B.A., Yesberg, J.A., Wunsch, D., Charleton, B., 2021. Lessons from validity of the Lesbian, Gay, and Bisexual knowledge and attitudes scale for
London: a contemporary examination of the factors affecting attrition among rape Heterosexuals (LGB-KASH). J. Counsel. Psychol. 52, 104–118.
complaints. Psychol. Crime Law.

18