Abstract
Intimate partner sexual violence (IPSV) is a significant aspect of intimate partner violence (IPV). While intimate partners commit one third of sexual assaults, IPSV is often overlooked in studies about IPV and in research on sexual violence. There are difficulties identifying, defining, and measuring IPSV, and research lacks consistency in terminology and measurement. The purpose of this article is to review the terms, definitions, and measurements associated with IPSV. Academic journals and nonscholarly documents from the United States were searched for articles and reports associated with the study of sexual violence and IPV. Forty-nine documents met the criteria for inclusion. A four-part taxonomy defining IPSV was developed, which included IPSV, intimate partner sexual coercion, intimate partner sexual abuse, and intimate partner forced sexual activity. The average weighted prevalence rates of these various forms of IPSV were calculated across included research studies. However, the measurements generally used to assess IPV do not adequately measure IPSV. Future research should consist terms to ensure consistent conceptualization and measurement of IPSV and to inform practice with survivors.
Keywords
Key Findings
Researchers do not use common terminology when studying intimate partner sexual violence.
Researchers use different definitions for the same terms when studying intimate partner sexual violence.
Inconsistencies in terminology and definitions lead to inconsistent studies and difficulties estimating the effects and consequences of intimate partner sexual violence.
These language differences exist because (1) sexual violence is a sensitive issue and terms such as rape may prevent women from reporting sexual violence; (2) people may conceive of rape differently than sexual assault or violence; and (3) the interdisciplinary nature of studying sexual violence has led to variance across disciplines.
It is important the terminology be reconciled so that practitioners, researchers, and clients are using the same language with the same meaning.
Consistent use of terminology may lead to consistent construction of incidences.
Some of the most commonly used measurement tools have been critiqued for their inability to measure intimate partner sexual violence.
In the police files of a Domestic Violence Unit in a Northern California city, an incident of violence is described by the survivor: “Today … at about 3:30 am, I was at home in bed sleeping with my boyfriend. [My boyfriend] and I have been living together for about seven months and dating for about a year. [My boyfriend] had been drinking since about two in the afternoon. [He] woke up and got on top of me. I was still asleep. When he got on top of me, he pulled off my underwear and forced me to have sex with him. I told him stop and get off me. I was begging [him] to stop, but he wouldn’t. He held my arms down to the bed by my biceps. [He] came all over my body. I was crying. I was lying on the bed facing the window with my back to [him]. He kept trying to turn me towards him and hug and hold me but I wouldn’t ….” “…I wanted to clarify about the part where [my boyfriend] and I had sex. I do not feel I was raped and I do not want to press rape charges. I also do not want to take any type of sexual assault exams. I only want to press charges for him pushing and biting me. This is a true statement.”
Recent political discourse also highlights the controversial nature of identifying and defining rape in the United States. The recent use of the terms forcible rape (2010, H.R. 3) and legitimate rape (Jaco, 2012) by politicians brought definitional and conceptual issues to the forefront of public consciousness and generated an outcry regarding how sexual assault is defined and how status as a victim of sexual violence is ascribed. These terms also exemplify the continued propagation of rape myths and misinformation; physical trauma is prioritized over mental trauma, and issues of consent are inconsistently theorized. This problem becomes even more complex when broadening the definition to include multiple forms of sexual violence and, in particular, when the perpetrator and victim have previously engaged in consensual sexual activity. These challenges are reflected in the academic literature and complicate estimates of the prevalence of intimate partner sexual violence (IPSV). As such, there is a need for more complete and standardized definitions of sexual violence that occurs in the context of intimate relationships (Jordan, 2007).
IPSV
Intimate partner violence (IPV) impacts a significant proportion of women in the United States, with more than one in three women reporting that they have experienced physical or sexual violence or stalking by an intimate partner in their lifetime (Black et al., 2011). Annual prevalence of IPV is estimated to be somewhere within the range of 2–12%, with a more recent population-based sample indicating that 5.9% of women experienced IPV within the past year (Black et al., 2011; Tjaden & Thoennes, 2000; Walton-Moss, Manganello, Frye, & Campbell, 2005). IPSV is one component of IPV (Saltzman, Fanslow, McMahon, & Shelly, 2002). Intimate partners commit approximately one quarter (26%) of sexual assaults (Bachman & Saltzman, 1995), and lifetime rates of sexual assault by an intimate partner in national random samples have ranged from 7.7% to 13% (Basile, 2002; Tjaden & Thoennes, 2000). The most recent national prevalence rates rest in the center of this range, with 1 in 10 women reporting that they have experienced sexual assault by an intimate partner (Black et al., 2011). When examining IPSV among women whose partners have physically assaulted them, the rates are much higher and range from 28% (Eby, Campbell, Sullivan, & Davidson, 1995) to 68% (McFarlane et al., 2005).
Despite its widespread prevalence, IPSV is often overlooked, both in studies of IPV and in studies of sexual violence. Researchers studying IPV often subsume IPSV under a comprehensive definition of IPV without isolating and examining the unique correlates and consequences of IPSV apart from other types of violence in intimate relationships (Campbell, Dworkin, & Giannina, 2009). Likewise, research on the correlates and consequences of sexual violence does not consistently specify whether participant reports of sexual violence occurred within or outside of an intimate relationship. Because IPSV lies at the intersection of IPV and sexual violence, it is often overlooked, and research on the specific consequences of sexual violence committed by an intimate partner lags behind research on other forms of violence against women.
When IPSV is identified and examined as a separate construct, it is defined and operationalized inconsistently across research studies. For example, some researchers suggest that co-occurring physical and sexual violence in an intimate relationship is a distinct form of victimization, differing from either type of IPV alone (Katz, Moore, & May, 2008; White, McMullin, Swartout, Sechrist, & Gollehon, 2008). Others use differential terminology—such as rape (Testa, VanZile-Tamsen, & Livingston, 2007), forced sex (Campbell & Soeken, 1999), and sexual coercion (Katz, Carino, & Hilton, 2002)—to describe either similar or distinct forms of victimization. To add complexity, there are additional difficulties with measurement across studies. In some cases, different measurement instruments have been used to measure the same IPSV construct; in other cases, the same measurement instruments have been used to measure different IPSV constructs.
The problem of identifying, defining, and measuring IPSV is not new. In a seminal work on rape in marriage, Finkelhor and Yllo (1985) ask the following question: Is it rape when a woman has sex to ‘keep the peace in the house’? On the one hand, calling it rape highlights how oppressive and coercive sex is under such circumstances. On the other hand, calling it rape means substantially expanding—and at the same time diluting—the meaning of the word ‘rape.’ (p. 85).
Therefore, the aim of this article is to review and synthesize the U.S. literature on sexual violence in intimate relationships. We build upon the review conducted by Bennice in 2003, focusing on the terms, definitions, and measurements used in research on IPSV. Specifically, we will (1) review the terms and definitions used in research on sexual violence in intimate relationships, (2) review the measurement tools used to examine the various forms of IPSV, and (3) report the weighted mean prevalence of various forms of IPSV reported across previous research studies. Based on this review of the literature, we conclude by providing suggestions for a consistent linguistic and definitional taxonomy for IPSV; suggestions for measurement based on this taxonomy; and discuss research, practice, and policy implications.
Method
Literature Search
The first author searched academic, computer-based journal databases, including Cumulative Index to Nursing and Allied Health Literature, PsychINFO, PubMED, and EBSCO, through December 2013.Article titles, abstracts, and subject lines were searched using the terms rape, sexual assault, sexual coercion, or sexual violence, paired with domestic violence, intimate partner violence, relationship, or intimate partner. Each of the sexual assault terms were matched with each of the IPV terms in the search. In an ancestral approach, reference lists of articles that met inclusion criteria were surveyed for articles that had not been identified through the electronic search. After screening academic literature, the first author searched the gray literature—unpublished academic sources including dissertations, books, and nonacademic documents and reports—to identify additional references. These additional resources were identified through web-based searches, reference lists of selected articles, and in-person visits at our University library. The second author helped to identify missing articles and seminal works. In total, the search yielded 590 articles, 13 books, and 4 nonacademic publications for consideration.
Following the search process, the first author screened all documents for inclusion (Figure 1). In the first round of screening, where possible, articles and documents were excluded based on a review of titles and abstracts. In the second round of screening, the full text of remaining articles, documents, and seminal works was reviewed to determine which would be included. All questionable inclusion or exclusion decisions were debriefed with the second author.
Inclusion/exclusion of resources.
Inclusion and Exclusion Criteria
Peer-reviewed articles and nonacademic publications were included in this review if they reported on any form of adult male-to-female perpetrated sexual violence or sexual abuse within the context of an intimate relationship; concurrent physical violence was not a requirement for inclusion. An intimate relationship was defined as an ongoing romantic relationship with a current or former partner or spouse. Same-sex couples were excluded from this review because there may be additional social or cultural factors influencing IPSV among lesbian, gay, bisexual, and transgender couples that go beyond the scope of this review. Both qualitative and quantitative research studies were included. Manuscripts and publications that did not specify the perpetrator–victim relationship or that examined sexual assault by a nonintimate partner (including sexual assault by strangers, acquaintances, and date rape) were excluded. Those focusing on perpetrator behaviors (motivations, etc.) were also excluded (n = 2; Monson, Langhinrichsen_Rohling, & Taft, 2009; Meyer, 1998), as this perspective was beyond the scope of the present review. Likewise, international articles (n = 58) were omitted, as our focus was on research conducted within the United States due to the ongoing political, legal, and social debates within the United States and the social and cultural constructs that influence these debates concerning sexual violence in intimate relationships.
After screening, 43 peer-reviewed articles met the criteria for inclusion (Table 1). Twenty-four of these articles focused specifically on the prevalence or conceptualization of IPSV. An additional 19 articles addressed the correlates and health consequences of IPSV; however, only 12 were included in this article, as they also addressed prevalence rates and/or definitions. As such, 36 peer-reviewed articles are included in this review. After screening available books (n = 11), two seminal works were selected, given the impact that they have had on the literature (Finkelhor & Yllo, 1985; Russell, 1982). Nonacademic research reports and practice guidelines were also identified and included (n = 4) as were terms and definitions provided on the Center for Disease Control and Prevention website (n = 1). Thus, a total of 43 resources informed this review.
Summary of included articles, books, and reports.
*When an article discussed the conceptualization of IPSV but neither a definition was given nor a prevalence rate provided, the article was included in the text of the paper where indicated.
Average Weighted Prevalence Rates
The average prevalence rate weighted by sample size for each of the various forms of IPSV was calculated across included studies. Studies were included in the average weighted prevalence if they reported the prevalence of any form of IPSV in their sample, regardless of whether assessing prevalence was an aim of the research study. Research studies were first separated by sample type: studies reporting the prevalence of IPSV among IPV survivors (n = 14) and those reporting the prevalence of IPSV among the general population (n = 8). A single article was excluded from those reporting on IPSV among IPV survivors, as the sample was selected to include 50% women who had experienced sexual assault (Logan, Cole, & Shannon, 2007). Once separated by the sample type, the type of IPSV measured was assessed. Each prevalence rate reported was then weighted by study sample size, and the frequency weighted prevalence rates were averaged across studies within a particular category.
Defining IPSV
Across the included literature, 14 different terms defined in 29 ways have been used to describe IPSV. There is a great degree of overlap between terms and definitions, with similar definitions labeled by different terms, different definitions used for the same terms, and some terms housed within the definitions of other terms (Table 2). This level of discord and overlap leads to confusion and uncertainty about what is meant when particular terms are used.
Terms and Definitions From Previous Research on Intimate Partner Sexual Violence.
In order to organize the terms used in previous research on IPSV, we utilize a framework developed by the Centers for Disease Control and Prevention (CDC). The CDC outlines four defining characteristics of sexual violence (not specific to intimate partners): (1) lack of consent, (2) whether the act was completed or attempted, (3) type of force (i.e., physical or nonphysical), and (d) the type of sexual activity, ranging from noncontact sexual harassment to penetration (Basile, Hertz, & Back, 2007; Black et al., 2011; Saltzman et al., 2002).
Lack of Consent
Nonconsent is used with consistency across definitions of IPSV in the literature. Twelve (of 29; 41.4%) definitions explicitly specify that the sex act is unwanted (Basile, 2008; Black et al., 2011; Campbell & Soeken, 1999; Katz et al., 2008, 2002), against the victim’s will (CDC, 2010, n.d.; McFarlane, 2007; Saltzman et al., 2002), or without the victim’s consent (Basile et al., 2007; Black et al., 2011; Campbell & Soeken, 1999; McFarlane, 2007). In two additional definitions, nonconsent was not explicit, but other terms embedded within the definition (e.g., sexual assault and sexual aggression) indicated nonconsent (Phiri-Alleman & Allman, 2008; White et al., 2008).
Attempted Versus Completed Sex
Some authors clarify that their definitions of IPSV include sexually violent acts that are attempted even if not completed, a practice less common in the literature. Attempted acts were included in seven definitions; on four of these occasions, this specification occurs in the academic literature (Katz et al., 2002; Testa et al., 2007; White et al., 2008). The remaining three references to sexually violent acts not completed occur in documents sponsored by the CDC utilizing the CDC criteria outlined previously (Basile et al., 2007; Black et al., 2011; CDC, 2010; Saltzman, 2002). Thus, compared to the other definitional criteria, the distinction between attempted and completed sexually violent acts is not generally made within the literature on IPSV.
Type of Force
In defining IPSV, there is a lengthy history of controversy and debate as researchers have attempted to establish what is meant by force. After delineating four types of coercion—social coercion, interpersonal coercion, threatened physical coercion, and actual physical coercion—Finkelhor and Yllo (1985) chose to limit the term rape to include “situations of actual or threatened physical force,” although they “readily acknowledge that other kinds of force can be frightening and traumatic” (pp. 89–90). Similarly, Russell (1982) distinguished between use of physical force and nonphysical coercion, even to the point of determining which actions constitute physical force. For example, Russell’s team “determined a minimal level of physical force, which included such acts as pushing, pinning, and being held down by a husband’s weight so that the woman couldn’t move” (p. 48).
In more recent literature, the term force (including the terms forced or forcible) is used in over half (16/29) of the definitions. Terms indicating physical force (i.e., physical violence, physical aggression, and beating) were used in an additional three definitions. The term force is often accompanied by a distinction between physical force and nonphysical force. In eight definitions, used in describing the terms IPSV/sexual violence, rape, sexual coercion, and sexual aggression, authors specified the use of physical force or physical violence (Black et al., 2011; Campbell & Soeken, 1999; CDC, 2010; McFarlane, 2007; Meyer, Vivian, & O’Leary, 1998; Katz et al., 2008; White et al., 2008). In three cases, nonphysical force, pressure, or controlling tactics were specified in reference to sexual coercion (Black et al., 2011; DeGue & DiLillo, 2005, Logan et al., 2007). In one instance, coercion was implied in reference to sexually abusive and controlling acts (Campbell & Soeken, 1999). In five cases, the definition of sexual coercion was broader, including both physically forced and nonphysically forced sexual violence in the same category (Broach & Petetric, 2006; Finkelhor & Yllo, 1985; Katz et al., 2002; Meyer et al., 1998; Starrat, Goetz., Shackelford, McKibbon, & Stewart-Williams, 2008). Finally, threats of physical violence or threats of physical force were included in nine of the definitions describing the terms forced sex, rape, sexual aggression, sexual coercion, sexually abusive and controlling acts, and being made to penetrate someone else (Black et al., 2011; Campbell & Soeken, 1999; Finkelhor & Yllo, 1985; McFarlane, 2007; Meyer et al., 1998; Phiri-Alleman & Alleman, 2008; Russell, 1982); an additional two definitions for the terms rape and forced sex included threats to end the relationship or being abusive or threatening without specifying physical violence (Black et al., 2011; Campbell & Soeken, 1999).
Type of Sexually Abusive Act
The other dimension on which many of the definitions rest is the type of sexually abusive act that the perpetrator forces/coerces. These can be broken down into penetrative sex acts, which include unwanted intercourse or penetration (Black et al., 2011; McFarlane, 2007; Phiri-Alleman & Alleman, 2008); vaginal, anal, or oral sexual assault (Black et al., 2011); and sexual assault with objects (Campbell & Soeken, 1999). Each of these acts meets the legal criteria for sexual assault. Nonpenetrative sexually abusive acts included within the definitions are touch without penetration, forced kissing, and fondling (Black et al., 2011). Other sexually abusive acts not involving penetration or force include “exposing their sexual body parts, flashing, masturbating in front of the victim, someone making a victim show his or her body parts, someone making a victim look at or participate in sexual photos or movies, or someone harassing the victim” (Black et al., 2011, p. 17); sexual degradation and emotional abuse (Campbell & Soeken, 1999; Logan et al., 2007); refusal to use condoms or use contraceptives (Campbell & Soeken, 1999); and sex outside of the relationship (Campbell & Soeken, 1999). Naming and studying each of these sexually abusive acts as IPSV acknowledges them as sexual violence with notable consequences.
When quantifying the distinctions between types of sexually abusive and violent acts in the definitions of terms, penetration was used in eight definitions to specifically identify that the following acts were penetrative: rape, sexual coercion, and IPSV (Basile et al., 2007; Black et al., 2011; Campbell & Soeken, 1999; McFarlane, 2007; Phiri-Alleman & Alleman, 2008; Russell, 1982). In eight other definitions of the terms IPSV, sexual aggression, sexual coercion, and sexual victimization, the word rape is embedded within the definition and, therefore, penetration is implied (Broach & Petetric, 2006; Meyer et al., 1998; Phiri-Alleman & Alleman, 2008; Russell, 1982; Starrat et al., 2008; Testa et al., 2007; White et al., 2008); and in an additional definition of rape, the term intercourse implied penetration (Testa et al., 2007). In the other definitions, the type of sexually abusive act is left unclarified, with the use of generic terms such as sexual activity (Katz et al., 2002), have sex (Logan et al., 2007), and sexual contact (DeGue & DiLillo, 2005). It is not clear whether these definitions are referring to penetrative acts.
A Proposed Taxonomy of IPSV Terms and Definitions
The terms and definitions utilized by researchers in the bulk of existing IPSV research tend to vary by only two of the definitional characteristics described previously: the type of force used by the perpetrator to obtain the sexually abusive activity (i.e., type of force) and the type of sexually abusive activity. We propose a new taxonomy of IPSV based on these two definitional criteria (Figure 2). Under this framework, a quadrant is formed by the level of forcefulness on the horizontal axis and level of invasiveness on the vertical axis. The level of forcefulness refers to the degree of physical force used, ranging from physical violence at the high end of physical force to nonphysical manipulation at the low end of physical force. The level of invasiveness refers to how invasive the type of sexually abusive act is, ranging from vaginal, oral, or anal penetration at the high end of invasive acts to unwanted touching at the low end of noninvasive acts.
Taxonomy of intimate partner sexual violence.
Intimate Partner Sexual Assault (High Force and High Invasiveness)
Eight terms have been used to describe physically forced penetrative sexually abusive acts in the context of an intimate relationship: IPSV, forced sex, rape, sexual assault, sexual coercion, sexual aggression, sexual victimization, and being made to penetrate someone else (refer to Table 2). For example, the term forced sex has been used to refer to physically forced sex acts, sex acts that include violent victimization, and sex acts obtained through threats of physical violence (Campbell & Soeken, 1999). However, it has been recommended that language generally used to describe consensual sex should not be used to describe sexual assault (End Violence Against Women International, 2013), making the use of the term sex in forced sex potentially problematic.
Other authors use the terms rape or sexual assault when describing forced sexually assaultive acts that include penetration and actual or threatened physical violence (Bergen & Bukovec, 2006; Bennice & Resick, 2003; McFarlane, 2007; Phiri-Alleman & Alleman, 2008; Testa et al., 2007). Campbell and Soeken (1999) make a persuasive argument for using the term sexual assault, as opposed to the term rape, to refer to physically forced vaginal or anal penetration in the context of an intimate relationship; these authors state that the term rape commonly gives the impression that the perpetrator is a stranger or acquaintance as opposed to an intimate partner. While the terms marital rape and spousal rape delineate an intimate relationship, these terms narrowly define the relationship as marital. The term intimate partner rape may resolve this particular issue, although survivors may have a difficult time referring to sexual assault by an intimate partner as rape. As such, we recommend the term intimate partner sexual assault to describe the use of physical violence or the threat of physical violence to obtain, or attempt to obtain, unwanted oral, vaginal, or anal intercourse, including forced penetration and sexual assault with objects. This term also applies to unwanted penetration when a victim/survivor is unable to consent or is unaware, that is, asleep or under the influence of drugs and alcohol.
Intimate Partner Sexual Coercion (Low Force and High Invasiveness)
This category of IPSV, intimate partner sexual coercion, is similar to intimate partner sexual assault, in that the sexual violence is highly invasive, meaning sexual penetration occurs (e.g., oral, anal, or digital penetration or penetration with objects; Black et al., 2011). It is different than intimate partner sexual assault in that the unwanted sex act is obtained through manipulative tactics and control rather than physical force (Black et al., 2011; Broach & Petetric, 2006; DeGue & DiLillo, 2005; Logan et al., 2007). Nonphysical tactics include verbal manipulation, withholding of resources, and threats to end the relationship or seek sexual fulfillment outside of the relationship (Logan et al., 2007). For example, a partner might threaten to humiliate a victim in public if she does not concede to sexually assaultive activity at home. A challenge to understanding and defining intimate partner sexual coercion is that consent outside the context of physical force may be more difficult to identify. Women may submit to coercive sexual tactics and consent to unwanted intercourse to avoid the negative outcomes of refusing it (Livingston, Buddie, Testa, & Vanzile-Tamsen, 2004) or out of perceived obligation to a spouse or partner (Basile, 2002). Despite this challenge, research is beginning to demonstrate the deleterious impact of intimate partner sexual coercion on the psychological and physical well-being of victims (Broach & Petetric, 2006; Miller et al., 2010). Thus, although perpetrators of intimate partner sexual coercion are less likely to be held accountable by the legal system, many researchers advocate for a broader understanding of sexual assault that includes coercive acts (Broach & Petetric, 2006; Muehlenhard, Sympson, Phelps, & Higby, 1994).
Various terms have been used to describe nonphysically forced IPSV, including the terms coercion, intimidation, and pressure (Black et al., 2011; CDC, 2010; Katz et al., 2002; McFarlane, 2007; Phiri-Alleman & Alleman, 2008; Starrat et al., 2008; Testa et al., 2007). We have selected the term sexual coercion for our taxonomy because it is the most commonly used. It is important to point out, however, that the term sexual coercion has not always been used to singularly refer to nonphysical force in the domain of forcefulness. In some instances, authors have used the term sexual coercion to encompass sexual assault, that is, physically forced IPSV (Katz et al., 2002; Starrat et al., 2008); others have used it to refer to general IPSV without making a distinction between physical and nonphysical force (Basile, 2008). In our proposed taxonomy, we reject the inclusion of physically forced IPSV as a type of sexual coercion. Consistent with previous research, this allows a distinction between physically forced and nonphysically forced sexual violence (see, e.g.,, Messing, Bagwell & Thaller, 2014).
In addition to the level of forcefulness, our taxonomy adds distinction based on a second dimension, differentiating between penetrative and nonpenetrative acts by including a domain of the level of invasiveness. In the literature, the term intimate partner sexual coercion has been used to refer to unwanted sexual penetration in six definitions (Basile, 2008; Black et al., 2011; Broach & Petetric, 2006; Finkelhor & Yllo, 1985; Logan et al., 2007). We adopt this definitional component and narrow our definition based on this criterion, using the term intimate partner sexual coercion to describe IPSV that is both nonphysically forced and penetrative. Thus, intimate partner sexual coercion is the use of nonphysical, controlling, degrading, and manipulative tactics to obtain, or attempt to obtain, unwanted oral, vaginal, or anal intercourse, including other penetrative acts such as sexual assault with objects. To describe less invasive sexual activity, we use the terms intimate partner sexual abuse and intimate partner forced sexual activity, as discussed subsequently.
Intimate Partner Sexual Abuse (Low Force, Low Invasiveness)
In abusive relationships, perpetrators often use manipulative, psychologically abusive tactics, or keep their partners in submissive positions of power through sexual degradation (Campbell, 2002; Campbell & Soeken, 1999). When looking at the terms and definitions previously utilized, four sources specify these forms of sexual abuse in describing IPSV (Black et al., 2011; Campbell, 2002; Campbell & Soeken, 1999). Two of the articles specifically refer to sexually abusive and controlling acts (Campbell, 2002; Campbell & Soeken, 1999), while one nonacademic source supported by the CDC referred to the control of reproductive and sexual health by a partner (Black et al., 2011). In addition, one qualitative study refers to degrading tactics within the context of sexual activity (Logan et al., 2007) among women who do not experience sexual assault by their abusive partners.
In the case of intimate partner sexual abuse, the degree of physical force is low as is the degree of invasiveness. This type of sexual violence is not penetrative and not physically forced; these acts are not considered sexual assault and cannot be prosecuted legally. Nevertheless, evidence overwhelmingly indicates that psychological abuse, even in the absence of physical violence, can have long-lasting consequences (Coker, Smith, Bethea, King, & McKeown, 2000; Pico-Alfonso et al., 2006). Emotional and psychological abuse can take on many forms, and intimate partner sexual abuse is a form of emotional or psychological abuse used as a tactic for controlling an intimate partner (for more on coercive control and intimate terrorism, see Johnson, 2008; Pence & Paymar, 1993, 1996; Stark, 2009). An example of intimate partner sexual abuse is described in one qualitative study, where a woman recounts being made to bark like a dog during sex (Logan et al., 2007). This participant’s story demonstrates how dominating and humiliating tactics were used to keep her in a subservient position. Intimate partner sexual abuse is distinct from intimate partner sexual assault and coercion, in that it is not necessarily used to obtain a specific sexual act. In some cases, the sexual act is already occurring with mutual consent when the degrading behavior occurs. In other cases, sexual humiliation and degradation occur without physical sexual activity. Slinging sexual insults (i.e. “you’re lousy in bed”) and forcing a partner to watch pornographic material exemplify intimate partner sexual abuse in the absence of coerced or physically forced sexual activity.
In our taxonomy, intimate partner sexual abuse includes refusal to use condoms, sex outside the primary relationship, control over sexual decision making, and birth control sabotage (Campbell & Soeken, 1999). By this definition, reproductive coercion, that is, when a male partner attempts to get his female partner pregnant when she does not want to be, is a form of intimate partner sexual abuse. The inclusion of reproductive coercion as a type intimate partner sexual abuse contrasts slightly with the definitions put forth by the CDC. According to the CDC, sexual and reproductive control is categorically different than IPSV and, thus, not presented as a subtype. However, we believe the literature on the health consequences of IPSV supports including sexual and reproductive control as one component of intimate partner sexual abuse, as this type of sexual violence is associated with a number of negative health outcomes and often occurs in the context of IPV (Miller et al., 2010; Moore, Frohworth, & Miller, 2010).
Intimate Partner Forced Sexual Activity (High Force, Low Invasiveness)
The fourth and final category of IPSV, characterized by high force and low invasiveness, is intimate partner forced sexual activity. In previous literature, only one term and definition covers this concept (and only partially so): the CDC’s intimate partner and sexual violence survey report (Black et al., 2011). In this document, the term unwanted sexual contact is utilized and defined as “unwanted sexual experiences involving touch but not sexual penetration, such as being kissed in a sexual way, or having sexual body parts fondled or grabbed” (Black et al., 2011, p. 17). In our taxonomy, for an act of IPSV to meet the criteria of physically forced sexual activity, some degree of physical force is required. In the CDC’s definition, the type of sexually abusive act is specified as nonpenetrative but the degree of force is not specified. Given that our classification system relies on force as one of the dimensions, we make a distinction between nonpenetrative sexually abusive acts that are coerced (intimate partner sexual abuse, see above) and nonpenetrative sexually abusive acts that are physically forced (intimate partner forced sexual activity). No previous research has documented intimate partner forced sexual activity; while this limits knowledge of the phenomenon and related sequelae, there may be significant negative consequences for survivors. For example, being held down and masturbated on or experiencing unwanted physical violence during intercourse, such as forcible hitting, biting, or strangulation, could be extremely traumatic and result in physical injury. As such, intimate partner forced sexual activity should not be minimized. Physical violence, targeted toward a sex organ or occurring during a sexual act, should be named and classified as a type of IPSV, as it may have unique effects on survivors compared to physical violence outside the sexual domain, such as more shame and difficulty with disclosure, given the sexual nature of the violence.
Measuring IPSV
In the reviewed literature, seven separate measurement tools were used alone or in conjunction to measure IPSV (Table 3). The most commonly used measures are the Revised Conflicts Tactics Scale (CTS-2; Straus, Hamby, Boney-McCoy, & Sugarman, 1996) and the Sexual Experiences Survey (SES; Koss & Oros, 1982). The CTS-2 is commonly used to measure physical IPV and includes a includes a 7-item subscale of sexual coercion, with 5 items addressing intimate partner sexual assault and 2 items addressing intimate partner sexual coercion. One critique of the CTS-2 sexual coercion subscale is that the test–retest reliability for it has been weaker than other portions of the CTS-2 (r = .30; Vega & O’Leary, 2007).
Measurement Instruments: Description, Reliability, and Validity.
The modified SES is an11-item instrument used to measure degrees of sexual victimization on a continuum. Although it is not specifically intended to measure sexual victimization by an intimate partner, language could be modified from “someone” to “intimate partner.” The SES assesses types of violence and victimization resulting from coercion, threats, drugs, authority, or use of force. One item is specific to nonphysically forced unwanted sexual contact (i.e., other forced sexual activity), 2 items are specific to sexual coercion, 6 items are specific to completed or attempted sexual assault, and two are specific to sexual abuse (Testa, VanZile-Tamsen, Livingston, & Koss, 2004). The SES has shown strong reliability and convergent validity in a sample of African American adolescent females, although not specifically in reference to intimate partners (Cecil & Matson, 2006). Among IPV survivors, it demonstrated high construct validity for measuring experiences of sexual assault and sexual coercion, yet low validity for measuring forced sexual activity and attempted sexual assault incidents (Testa et al., 2004).
Five other scales have been used to measure some form of IPSV alone or in conjunction with other constructs: the Severity of Violence Against Women Survey (SVAWS; Marshall, 1992), the Index of Spousal Abuse (ISA; Hudson & McIntosh, 1981), the Danger Assessment (DA; Campbell et al., 2003), the Sexual Coercion in Intimate Relationships Scale (SCIRS; Shackelford & Goetz, 2004), and the Partner Abuse Scale (PAS; Hudson, 1992). A description of each of these scales is made available in Table 3. In three of these scales, (DA, SVAWS, and ISA), it was unclear which category of IPSV one or more items measured. The words, “make you” or “force you” to “have sex,” for example, specify intercourse, but it is not clear that it would be apparent to a reader whether the force used was physical force (as in intimate partner sexual assault) or nonphysical force (as in intimate partner sexual coercion).
When using our taxonomy to organize measures of IPSV (Table 4), each measurement scale, except the PAS, has at least 1 item measuring intimate partner sexual assault. The SES is the only measurement tool that assesses other forced sexual activity. Five of the scales measure intimate partner sexual coercion: the SES (with a minor adaptation to focus on an intimate partner), the CTS-2, the SCIRS, the SVAWS, and the PAS. The SCIRS has the most items assessing different types of intimate partner sexual coercion, with a total of 28 items. Two scales measure intimate partner sexual abuse: the SES and the CTS-2. The SES is the only scale to measure all of the constructs identified; yet it still has been critiqued for not accurately measuring IPSV (Rinehart & Yeater, 2011). Many researchers have combined one of the other six measurement tools with the CTS-2 in order to measure both IPSV and intimate partner physical violence. This may be a common strategy, particularly given the widespread use of and the inclusion of physical violence on the CTS-2; however, the CTS-2 has been widely critiqued (see, e.g., DeKeseredy & Schwartz, 1998; Dobash, Dobash, Wilson, & Daly, 1992; Kimmel, 2002) and the sexual violence items on the CTS-2 have shown inconsistent reliability (Vega & O’Leary, 2007). Given that the SES addresses all types of IPSV, this is the most comprehensive measure available, once adapted to specify that questions refer to sexual violence from an intimate partner. The SES, combined with the SVAWS, would measure both physical and sexual violence in an intimate relationship. Researchers may create their own IPSV scale, choosing items from various scales, but this raises additional questions about the reliability and validity of adapted scales.
Measurement Instruments: Number of Items Addressing the Four Subtypes of IPSV.
Note. IPSV = Intimate partner sexual violence.
Prevalence Rates
We expected that the prevalence rates of IPSV reported in the literature would vary by the type of IPSV measured—intimate partner sexual coercion, intimate partner sexual assault, intimate partner sexual abuse, or intimate partner forced sexual activity (Table 5). The average prevalence of intimate partner sexual assault among IPV survivors weighted by sample size across 11 studies (n = 3,178) is 36.1% (95% confidence interval [CI]: [35.5%, 36.8%]), and the average prevalence of intimate partner sexual coercion weighted by sample size across five studies (n = 1,408) is 24.9% (95% CI: [24.6%, 25.3%]). There is a wide range in prevalence rates for both types of IPSV, with a difference of 82% for intimate partner sexual assault (9% minimum and 91% maximum) and 73.6% for intimate partner sexual coercion (17.4% minimum and 91% maximum). Too few studies assessed rates of intimate partner sexual abuse to create summary statistics; the prevalence rate for intimate partner sexual abuse among IPV survivors, provided by one study (n = 68) is 72% (El Bassel, Gilbert, Rajah, Foleno, & Frye, 2000). Two studies (n = 214) provided information about IPSV without specifying the type of IPSV being measured. For these two, the weighted average prevalence rate is 55.0% (95% CI: [53.6%, –56.4%]). No research, to date, has examined the prevalence of intimate partner forced sexual activity among IPV survivors.
Prevalence Rates of IPSV in Previous Research.
Note. SES = Sexual Experiences Survey. SVAWS = Severity of Violence Against Women Survey; CTS-2 = Conflicts Tactics Scale; ISA = Index of Spousal Abuse; DA = Danger Assessment; PASPH = Partner Abuse Scale: Physical. aExcluded from summary statistics because sample was selected to have 50% experienced sexual assault. bExcluded from summary statistics because men and women included in sample.
For the studies that examined prevalence rates in a broader population, not limiting their sample to IPV survivors, prevalence rates of intimate partner sexual assault are lower, ranging from 7.7% to 21%. The weighted mean prevalence across five samples (n = 19,310) was 9.1% (95% CI: [9.06%, 9.12%]) and the median rate was 10%. Two studies (n = 9,386) assessed the prevalence of sexual coercion, finding a weighted mean prevalence of 9.8% (95% CI: [9.8%, 9.8%]). Only one study (n = 9,086) examined sexual abuse, finding a prevalence of 6.4%. Another study (n = 927) measured the prevalence of IPSV among the general population without specifying the type of IPSV being measured, finding a prevalence of 7.7%
Overall, these findings show a similar mean and median for each construct of IPSV, indicating little skew or kurtosis. Also, there is a wide prevalence range in all measures, which is a pattern that may be explained by the inconsistencies of terms and definitions as outlined previously. When examining prevalence rates, problems with the self-report of IPSV must also be kept in mind; it has long been suggested that IPSV prevalence rates are underestimated because women in relationships are hesitant to call the sexual violence that they experience rape or forced sex.
Discussion
IPSV is a serious social problem and health disparity and is particularly common for women experiencing physical IPV. The high prevalence rates and devastating health consequences of IPSV, such as injury, sexually transmitted infections, depression, and post-traumatic stress disorder (Campbell, 2002), indicate a need for effective treatment and prevention strategies. Findings from this review demonstrate a significant problem in consistency of terminology, with 14 different terms and 29 corresponding definitions in the literature assessed. In some instances, the same terms had conflicting definitions; in other instances, different terms were used to refer to the same definitions. This creates significant problems in operationalizing constructs and leads to confusion in definition and measurement.
Further complications arise, given the multiple tools used to measure the various forms of IPSV. Seven different measurement scales are currently used to assess sexual violence in intimate relationships, and only one of these tools independently covers the four major types of IPSV presented in the literature and in the above analysis. Further, some existing scales ask whether sex was forced without distinguishing between physical force and nonphysical force, making it impossible to distinguish between sexual assault and sexual coercion in these instances. Survivors taking these surveys may be unclear as to what force means, leading to either under or overreporting. Moreover, some of the most widely used measurement tools, such as the CTS-2 and the SES, have been critiqued for their inability to accurately measure IPSV (Rinehart & Yeater, 2011). Without more specific and nuanced measures of IPSV, it is not possible to effectively determine the prevalence and consequences of IPSV and develop the best strategies for intervention.
When moving outside of the research community, the language, measures, and constructs that are used by advocates, clinical practitioners, and survivors demonstrate an even greater diversity. Rape advocates, for example, may believe in calling sexual assault rape, whether or not it was committed by an intimate partner, acquaintance, or stranger with the rationale that using terms such as sexual violence or sexual assault dilutes the impact of the experience. However, researchers have a similarly strong rationale for rejecting this language: using the term rape may prevent potential participants from identifying their experience or discussing the sexual violence in their relationship. Among survivors, qualitative work with women has demonstrated that they often struggle with how to categorize and label negative sexual experiences committed by an intimate partner (Logan et al., 2007). In more than one case, for example, a woman indicated she had never experienced sexual assault, yet concurrently cited an incident of waking up to her partner having sex with her (Logan et al., 2007).
These differences may reflect wider divisions among how people conceive of rape, sexual violence, or sexual abuse. Within the IPSV literature, this dilemma dates back to Russell’s (1982) qualitative examination of sexual violence in marriage; researchers struggle with, on one hand, allowing women to define their own experiences and, on the other hand, labeling women’s experiences using operationally defined terms. These examples demonstrate the difficulties of naming and defining IPSV not only for researchers but also for survivors, advocates, and other stakeholders who are impacted by research.
Differences in conceptualization are associated with differences in measurement. Assessment tools used in practice settings to screen for sexual violence are rarely the same as those used in research settings to measure and study IPSV. This leads to difficulty in transdisciplinary communication and in translating research into practice. For example, the Basile, Hertz, and Back (2007) compiled a list of the preferred sexual violence screening tools for health care settings, and the DA (Campbell et al., 2003) was the only recommended measurement tool that coincided with the measurements utilized in the research literature. Common screening tools used and evaluated in medical settings include the Hurt, Insult, Threaten, and Scream (Sherin, Sinacore, Li, Zitter, & Shakil, 1998), the Woman Abuse Screening Tool (Brown, Lent, Schmidt, & Sas, 2000), the Partner Violence Screen (Feldhaus et al., 1997), and the Abuse Assessment Screen (McFarlane, Parker, Soeken, & Bullock, 1992; Rabin, Jennings, Campbell, & Bair-Merrit, 2009). Combined with the differential use of terminology, the use of different tools across research and practice settings may lead to a disjuncture between estimates of the problem when examined by researchers and the practical application of these estimates for practitioners.
This collective lack of clarity may be rooted in the relatively recent identification of IPSV as a social and health problem. As traced in a historical account of marital rape by Bennice and Resick (2003), marital rape exemptions—by which a man could not be convicted of rape if the victim was his wife—existed in all states throughout most the of the 20th century. These laws were not eradicated until the 1970s and 1980s, and marital rape was not criminalized in the penal code of all 50 states until the 1993 (Bennice & Resick, 2003). Changes in these laws occurred only after much debate. For example, in 1979, a democratic California senator asked, “If you can’t rape your wife, who can you rape?” (Freeman, 1981, p. 1). This type of thinking is reflected over 30 years later in the more recent political debates around the terms legitimate rape and forcible rape. Until relatively recently, social and cultural norms have prevented women from finding legal protection against intimate partners who committed sexual violence against them. As presented in this history, the more recent understanding and naming of IPSV could explain why researchers are still grappling with definitions.
Another reason for the different terms, definitions, and measures could be the cross-disciplinary nature of IPV and sexual assault research. Scholars who study violence against women come from a variety of disciplines, each of which may have different conceptualizations of constructs and measures. There are also parallel but separate research agendas of researchers studying IPV and those studying sexual assault; IPSV is the intersection of these two research arenas. Yet, the separate treatment of these two types of violence is incongruous, given their common co-occurrence. Researchers and advocates in the domain of public health have attempted to create unified terms and definitions previously (see, e.g., Black et al., 2011; Basile & Saltzman, 2002; Saltzman, Fanslow, McMahon, & Shelley, 1999), but, despite these attempts, this review makes it clear that no consistency in terms or definitions has emerged. We believe that the inductive process taken in this review—taking other’s specific examples, and coming up with generalizations—may be more likely to be adopted, as it was generated based on previous research and conceptualizations.
Implications and Recommendations
As other researchers have observed (Jordan, 2007), there is a need for employing common terminology across research and practice settings to encourage consistent evaluation of prevention and treatment interventions. Based upon this review of the literature, it appears that IPSV can be used as an umbrella term to encompass all forms of sexual violence by an intimate partner. Although this term has the same limitations as the term “intimate partner violence” in that it does not specify the gendered nature of the problem, the benefits of using this umbrella term outweigh the limitations. Namely, it specifies that sexual violence is occurring within the context of an intimate relationship but does not limit that relationship to the parameters of marriage. In this respect, it is broad enough to include sexual violence that occurs in committed dating relationships, among cohabiting but unmarried couples, and in more complex relationships (i.e., married but separated and divorced). Furthermore, the term IPSV encompasses a range of types of sexual acts including the continuums of invasiveness and force identified based on previous research. Although all sexually violent and abusive acts subsumed under IPSV are not recognized as prosecutable crimes, it is important to include coerced and nonpenetrative sexually abusive acts in the definition of IPSV, as they are physically and psychologically harmful to women.
Our proposed taxonomy reduces the challenges of multiple terms and definitions by taking 29 definitions (Table 2) and collapsing them into four types of IPSV. This simplified taxonomy could streamline communication between researchers and service providers across multiple disciplines, while simultaneously acknowledging the full range of sexually violent acts that may be perpetrated by an intimate partner. All forms of IPSV fall within the four quadrants characterized as a spectrum of forcefulness by invasiveness. As such, any experience of IPSV that a survivor describes—even when that particular experience that falls outside of the stereotypical perception of sexual assault—can be recognized and named. For a survivor who may believe that her experience is unique, shameful, or who may not otherwise characterize her experience as abuse, the process of naming may be legitimizing, may connect her with other women who have experienced this form of abuse, and may assist in the healing process.
Women’s own definitions and conceptualizations of IPSV must also inform the terms and definitions, and a delicate balance is needed between researcher-defined and survivor-defined IPSV. More qualitative research is necessary to understand how survivors define and name various forms of IPSV. Meanwhile, differential language and constructs of IPSV between researchers, practitioners, and clients lead to a series of thought-provoking questions: Is it necessary to label IPSV in particular ways that fall outside of a woman’s conceptualization of her experience (e.g., saying “you were raped” if she does not label it as such)? What are the implications of this type of labeling in research and practice settings, given the feminist values of individual agency, self-determination, and subjugated knowledge? When intervening with clients, it is important to initially use the language of survivors of IPSV. Should practitioners introduce terms other than the ones that survivors use to classify and name their experiences and, if so, when and how should this be done? On the other hand, what are the implications of failing to assist women in naming their experiences? Using this taxonomy of IPSV to describe women’s experiences—even if the terms do not correspond with the language survivors use—may be meaningful. This naming process may assist clinicians in connecting survivors to resources for healing their trauma and services for their physical, mental, emotional, and spiritual health. This ultimately means that greater emphasis must be placed on a tripartite collaboration between clinicians, researchers, and survivors, so that all three groups use the same language and associated meanings.
Future research should note the differential effects of IPSV based on the type of violence and tactic—intimate partner sexual coercion, intimate partner sexual assault, intimate partner sexual abuse, and intimate partner forced sexual activity. It may be that, in much of the present research, distinctions between types of sexual violence are not made in an attempt to ensure that all types of sexual violence toward women are acknowledged as significant and traumatic, especially given the history of dismissing sexual assault within intimate relationships as not criminal. However, this lack of differentiation may inadvertently ignore forms of IPSV that impact women. For example, no research, to date, exists exploring intimate partner forced sexual activity. As such, the prevalence and consequences of this form of sexual violence in intimate relationships is unknown. Making clear distinctions between types of IPSV will ultimately lead to a better description of the prevalence and consequences of IPSV, allow for aggregation and comparison across research studies, and assist in the development of tailored interventions for women who have experienced the various forms of IPSV.
In addition to naming the various form of sexual violence in intimate relationships, the taxonomy presented here may assist survivors to receive targeted care related to IPSV, and perpetrators to receive criminal justice and/or social service intervention, based on the four quadrants. The quadrant model has been used to specify treatment approaches in other areas of social work practice, such as for co-occurring mental health and substance abuse issues (Van Wormer & Davis, 2013). While there are many differences between IPSV and co-occurring mental health/substance abuse disorders—primarily, that IPSV has a clear perpetrator and victim—the model has utility as an example. Specifically, according to the mental health and substance abuse treatment model, each quadrant corresponds with a set of treatment recommendations. Women whose partners coerce them into sexual activity may have different treatment needs than women whose partners physically force sexual activity, for example, feelings of guilt or self-blame may be more prevalent among women whose partners use coercive tactics. Similarly, women whose partners force or coerce penetrative sexual activity may have health needs, such as testing for sexually transmitted infections, that women whose partners force or coerce nonpenetrative sexual activity do not have. The same quadrant model could be utilized in treatment of men who coerce or force penetrative or nonpenetrative sexual activity, as treatment needs may not be the same across all forms of sexual violence perpetration. Because this model allows for distinguishing for diverse consequences and experiences, research could also investigate the effectiveness of treatment outcomes by quadrant to inform more targeted treatment recommendations.
In addition to assisting with treatment, this taxonomy may be useful in the criminal and civil justice context. Once researchers better understand the implications of all forms of IPSV, it may open the pathway to include other forms of sexual violence in criminal or civil codes. Educational campaigns presented in accessible formats (television, social media, and radio) can inform the public that, for example, sexual coercion is a type of IPSV. Policy agendas may also work to redefine consent in a way that takes into consideration the dynamics of sexual coercion in intimate relationships. As seen in the examples of marital rape exemptions and the criminalization of rape, social change in the nation’s critical consciousness may lead to changes in legal code.
In addition to the adoption of consistent terms and definitions, researchers should adopt consistently agreed upon, reliable, and valid measures that examine IPSV using the taxonomy created here. A new measure, for example, could ask a few questions to assess a “range of forcefulness” and a “range of invasiveness” of sexual violence. This would reduce participant burden by decreasing the number of questions that participants must respond to, as well as limiting questions about specific sexually violent acts. At the same time, this sort of measurement tool would provide specific information about the type of IPSV a survivor experienced. A shortened measure could advance research and clinical practice by simplifying screening and reducing the burden of retelling traumatic experiences. If the quadrants were additionally used to inform treatment decisions, a simple screening could provide a starting place for intervention as well. In regard to currently utilized instruments, researchers should clarify terms like “make” or “force” when assessing IPSV. This would lead to an improvement in communication and collaboration between participants and researchers, and among researchers who are conducting complimentary research. To facilitate communication between researchers and practitioners, the measurements used in research should be as consistent in both language and constructs as possible with the measurements utilized in health care screenings. Consistency in this respect would ensure that researchers are measuring the same issues practitioners are treating, thus clarifying patient needs and the extent to which they are being met.
In practice settings, we recommend engaging women in conversations about their sexual health and sexual experiences with their intimate partner. Given that sexuality may be an uncomfortable topic to address, strong and effective screening and assessment tools can assist in creating a mechanism to discuss these issues. Because women may identify IPSV in many different ways, it is important to meet survivors where they are, educate them concerning safe and consensual sex practices, and let them know that feelings of hurt, shame, and depression associated with IPSV are normal. This strategy will lead to opportunities to assist women in safety planning around sexual health. The negative sexual health consequences of IPSV present additional possibilities for collaboration between health care and social services, for example, by incorporating trained medical personnel into the staff at shelters or developing partnerships between shelters and hospitals. Consistency in shelter policies concerning referrals for reproductive health needs could help link women to the sources of support they need for sexual safety planning in light of their IPSV experiences.
Recommendations for Research, Policy, and Practice
Research
Use consistent language across studies.
Construct consistent definitions of terms to ensure congruent meanings across studies.
Clarify terms such as “make” and “force” (i.e., physical force vs. nonphysical force) when asking participants about IPSV.
Create new instruments and short forms of instruments to use in both research and practice.
Practice
Use the language of victims/survivors when intervening with clients.
Engage women in discussions about their sexual health.
Normalize the language used in practice.
Utilize more effective assessment tools when helping survivors of intimate partner sexual violence.
Policy
Include coercion in the definition of sexual assault.
Use consistent terms and definitions in laws in order to improve prosecutions.
Footnotes
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.