Sexual assault continues to be a pervasive crime in the USA as approximately one in four adult women are raped in their lifetime (Basile et al., 2022; Campbell & Wasco, 2016). Sexual assault is most commonly characterized as nonconsensual sexual contact or penetration in which a perpetrator uses force, coercion, or other means to acquire sexual contact from another person (Cantor et al., 2015). Sexual assault, depending on the definition, can range from nonconsensual sexual behaviors such as groping or fondling, to forcible rape (Muehlenhard et al., 2016). How a victimFootnote 1 labels an experience might constitute a crucial step in understanding the likelihood of reporting and, subsequently, affect the measured prevalence of sexual assault. As such, the purpose of this study is to qualitatively examine how victims label their experiences as well as examine two related concepts that contribute to these descriptions: severity of violence and in-the-moment victim reactions.

Labeling Sexual Assault

Rape Myth’s Role

People whose experiences legally constitute sexual assault or rape yet do not label the experience as such are termed “unacknowledged victims” (Kahn et al., 2003; Koss et al., 1988; Peterson & Muehlenhard, 2004). The prevalence of unacknowledged victims ranges from approximately 43% to 73% (Bondurant, 2001; Kahn et al., 1994; Koss et al., 1988). Peterson and Muehlenhard (2004) suggest that the high prevalence of unacknowledged victims may be a result of rape myths—“prejudicial, stereotyped, or false beliefs about rape, rape victims, and rapists” (Burt, 1980, p. 217). Lonsway and Fitzgerald (1994) suggest that men use such myths to justify or deny sexual violence and women use them to deny personal vulnerability to rape (Lonsway & Fitzgerald, 1994). This process of justifying rape or sexual assault not only blames rape victims for their own victimization but also limits what may actually get counted as rape (Lonsway & Fitzgerald, 1994; Peterson & Muehlenhard, 2004). Peterson and Muehlenhard (2004) conclude that rape myths contribute to a narrow definition of rape, which includes specifying the severity of what constitutes a rape and that victims should physically resist. In turn, this narrow definition may limit how people end up labeling their own experiences—as sexual assault or rape or something else.

Severity of Violence’s Role

The severity of violence during a sexual assault can affect how people label their experiences as well. For example, Payne et al.’s (1999) widely used Illinois Rape Myth Acceptance scale includes several items which assess people’s perceptions of the severity of the experience. Indeed, the scale’s construct “It wasn’t really rape” contains items like “If a woman claims to have been raped but has no bruises or scrapes, she probably shouldn’t be taken seriously” (p. 50). Also, in items comprising the scale’s construct “Rape is trivial,” rape is compared with other violence in the item “being raped isn’t as bad as being mugged or beaten” (p. 49). These items exemplify how sexual assault’s severity of violence is an underlying consideration of rape myths and thus influences how people may conceptualize rape and their own experiences of rape. Similarly, researchers have found that if the offender used a weapon, a victim is more likely to report sexual assault (Koss & Gidycz, 1985; Starzynski et al., 2005) and the more physically violent and injurious the sexual assault is, the more likely a victim will label it as a sexual assault and report it to authorities (Bondurant, 2001; Fisher et al., 2003; Kahn et al., 2003). Conversely, if the perpetrator engages in coercive tactics instead of physically violent tactics, victims are more likely to harbor harmful coping strategies, including self-blame (Ullman et al., 2007).

The severity of the event not only affects the victim’s perceptions of the experience and subsequent labeling reactions, it also can affect whether others perceive the event or the victim as deserving of support or justice. Victim self-blame and blame from others are closely connected to rape myths (Davies et al., 2012) with some researchers arguing that “Rape myth acceptance is… a form of victim blaming” (Hayes et al., 2013, p. 206).

Victim Blaming as a Result of In-the-Moment Victim Behavior

Victim blaming is often predicated on the idea that victims should have done something different prior to or during the assault to prevent its occurrence (Janoff-Bulman, 1979). Further, victim blaming holds that because the victim did not do these things “correctly,” they are somewhat or entirely to blame for their own assault. Victim behavior is important to study because assumptions of “common” or “correct” victim responses can and do affect prosecutors’ likelihood of pursuing a sexual assault case as well as likelihood of jury conviction (Beichner & Spohn, 2012; Campbell et al., 2015).

Most research assessing victim response focuses on resistance strategies. Sexual assault resistance strategies can be conceptualized as engagement in protective actions (e.g., verbal, behavioral) by the victim with the intention of avoiding or evading sexual assault (Ullman et al., 2007).Footnote 2 But the resistance strategy research has two overarching limitations. One limitation is the use of police reports to document victim behavior (e.g., Balemba et al., 2012; O’Neal & Kaiser, 2015). Behaviors reported to police could differ from behaviors that typically occur because victims could be concerned that disclosing certain behaviors (e.g., dancing with perpetrator earlier in the night) may lead police to be less likely to pursue the case. These concerns are legitimate; when victims report certain behaviors (e.g., they drank alcohol), police are more likely to consider the report fraudulent (Jordan, 2004). Additionally, written police records are found to draw on rape myths to deny or justify a sexual assault (Shaw et al., 2017). Therefore, using police reports to examine victim behavior during the sexual assault may lead to erroneous conclusions.

The second limitation relates to the scope of resistance strategy literature, because it largely focuses on strategies that deescalate sexual assault (Edwards et al., 2014). Focusing primarily on actions that deescalate sexual assault misses other potentially common behaviors. For example, disassociation is a common behavior during sexual assault (Campbell, 2012; Möller et al., 2017; Van der Kolk, 2015). This behavior serves to protect the victim as an emotional coping mechanism but does not deescalate the assault (Ullman et al., 2007). Therefore, if we are to combat victim blaming by broadening assumptions regarding common victim reactions, the current resistance literature falls short of achieving this.

The Current Study

If victim blaming is predicated on what people think a victim should have done differently, then it is important to conduct a study which examines victim behaviors. Currently, only the resistance strategy literature—which has notable limitations—describes what victims, in fact, actually do during a sexual assault. Therefore, a study on victim behaviors is especially needed which does not repeat the limitations from previous resistance strategy studies. Edwards et al. (2014) state that the resistance strategy literature has mostly been quantitative, has excluded coercive sexual assault experiences, and focuses on deescalation strategies. They then conducted a qualitative analysis of 51 college students that included coercive sexual assault and framed their findings in comparison to the perpetration tactics used in each event. Using similar definitions of sexual assault, we expand on Edwards et al.’s (2014) work by collecting a larger sample from the general US population and heeded their call for qualitative work with “…larger samples that might allow for the development of more nuanced categories and understanding of women’s resistance strategies” (p. 2541). Additionally, we aim to overcome one of Edwards et al. (2014) limitations by assessing participants’ “most memorable” experience instead of asking them to report on their “most severe” experience. More severe experiences may illicit different reactions from the victim.

In this study, we explored the research question “how do women recount their experiences of sexual assault and rape?” Due to the broad research question, several themes inductively emerged. We look only at three main themes from our qualitative data in the current study: Labeling, Judging Severity, and Victim Reactions. We connect these three concepts because (1) labeling sexual assault can be affected by the severity of the experience and (2) both severity of experience and victim reactions are sources of victim blaming. For Labeling and Judging Severity, we assessed how victims categorize their experience beyond a dichotomy of sexual assault/not sexual assault and severe/not severe. In Victim Reactions, we assessed common victim reactions to sexual assault disregarding whether or not it deescalates the assault. We do this to further inform sexual violence prevention programs, norming campaigns, and criminal justice services and proceedings.

Methods

Participants

Women (N = 906) were drawn from a larger study (N = 1382) of sexual assault among lesbian, bisexual, and heterosexual identified women in the USA. Women who did not experience sexual assault were not retained in the sample for the current study. Participants were included in the current study if they responded in the affirmative to at least one of the items on a modified version of the Sexual Experience Survey-Short Form Revised (SES-SFV; Koss et al., 2007).

The sample consisted of lesbian (32%, N = 285), bisexual (41%, N = 367), and heterosexual (23%, N = 208) identified women with about 5% identifying as another sexual orientation (e.g., asexual, queer, or unsure/questioning). The average age of participants was 39.8 years old (SD = 16.8 years). Many participants identified as White (78%) with most other participants identifying as Black (9%), Latino/a (4%), and bi- or multi-racial (5%). Many reported a household income between $30,001 and $75,000 (43%), while 34% made less than $30,000 and 23% made more than $75,001. A slight majority reported having no college degree (53%), while the remaining reported having an Associate’s degree (13%), Bachelor’s degree (22%), Master’s degree (10%), or a Doctoral degree (2%). Demographic information is provided in Table 1.

Table 1 Participant demographics
Full size table

Procedures

Data were collected via a cross-sectional survey disseminated to a national panel sample via Qualtrics Online Survey Company in 2016. Qualtrics is a sample aggregator who identified individuals from their national participant pool of over one million people who met the study’s eligibility requirements (i.e., women over the age of 18 living in the USA) and sent out email invitations to participate in the study. To aid in recruitment, Qualtrics offered incentives to participants for completing the survey (Qualtrics, personal communication, August 15, 2014). Study protocol was approved by the host institution prior to data collection.

Measures

Modified Sexual Experiences Survey-Short Form Victimization (SES-SFV).

Sexual assault victimization was assessed using a modified version of the SES-SFV (see Canan et al., 2020 for modified measure). Participants who reported experiencing any perpetration tactics greater than “0” times on either of the two periods (birth to 13 years of age, 14 years of age to present) for any of the nonconsensual sexual behaviors (sexual touching, oral penetration, vaginal penetration, anal penetration) and/or indicated being raped were guided into a specific branch of the survey that included additional questions about their experiences. All other participants who selected “0” times for all tactics over both time periods for all behaviors and answered “No” to being raped were guided into a different alternative branch of the survey that comprised a slightly different set of questions. Participants completing this alternative branch of the survey are not included in the present study.

Open-Ended Sexual Assault Response

Participants in the present study were instructed “If you selected that you had more than one experience, please respond to the following questions about the one experience that you remember in the most detail.” Following several close-ended questions about the participant’s experience with sexual assault, an open-ended question allowed participants to broadly detail their experience. The prompt specifically indicated:

In as much detail as you can remember and are able to share, please tell us about the context (e.g., location, time of day, actions before the event, and other things about the experience) in which this event / these events happened. Any and all information you are willing to provide would be appreciated.

Data Cleaning

Some open-ended responses included responses such as “N/A,” “nothing,” and “does not apply.” Others included open-ended responses that were detailed enough that it seemed clear that the participant was describing a consensual behavior. For example, one participant wrote, “Moved to an unfamiliar city for school. Met a girl, hung out, partied together and then started dating. We both consented to the experience. It was my first. We stayed together for 7 years.” Because of this, we explored the option that some participants may have mistakenly answered affirmatively on the SES-SFV earlier in the survey.

After a review of all the open-ended responses, 63 stories were highlighted as possible false positives. While triangulating the qualitative open-ended responses with quantitative measures in the survey, three issues came up that allowed us to confirm a false positive. First, some participants’ response patterns to the SES-SFV indicated insufficient effort responding potentially due to the survey layout (e.g., selecting zero for all tactics in all behaviors except selecting all tactics one time in the very last row of the last behavior (attempted anal, N = 4). Second, some of the highlighted participants tried to further express that they were in the wrong branch of the survey by using the “other” option of closed-ended questions (e.g., writing in “this does not apply” for several questions, N = 3). Third and finally, some participants gave impossible answers (e.g., they were assaulted at zero years old, N = 9) to closed-ended questions. If participants had any one of these types of responses, we deemed them a likely false-positive and removed them from the victim group. After this systematic cross-checking, we found that of the 63 cases investigated, 16 participants (1.2% of the entire sample, 1.7% of participants categorized as victims) were likely a false-positive whereas 47 participants of those highlighted were likely correctly categorized as a victim.

Data Analysis

Qualitative Analysis

The open-ended sexual assault responses were analyzed using a multi-step thematic analysis procedure laid out by Braun and Clarke (2006) to retain the flexibility to systematically analyze very diverse response types. Our sample not only had diverse response types due to the broad prompt, but it also had a diversity in response lengths. Some participants wrote multiple paragraph responses while others wrote a single sentence. Given that a more robust, interpretive analysis would be inappropriate for our shorter participant responses, we elected to perform a content-driven, thematic categorization of the data. We used an inductive or ‘bottom-up’ approach to analyzing the data with a goal of creating themes. These methods use the “small q” approach to qualitative research as described by Braun and Clarke (2013) in that we make “use of specific qualitative data collection and techniques, not (necessarily) within a qualitative paradigm” (p. 4).

In step 1, the entire dataset of responses was read twice by the first author while taking analytical notes of patterns and meaningful text. In step 2, the research team worked through the entire data set systematically, reviewing each potential data item and developing the coding scheme to identify emerging patterns (themes) across the data (Braun & Clarke, 2006). Some codes from step 2 eventually formed main themes, sub-themes, and/or were discarded to develop a first draft of a data codebook. To decide this, the open-ended responses were randomly divided into three sections that were assigned to three external sexual health researchers; each of them independently read and noted patterns within the open-ended responses. Then, patterns identified by these three researchers were merged and organized with our original list of potential patterns. These identified patterns represented codes that were then organized into themes in step 3.

Step 4 involved theme and codebook refinement. Operational definitions were created to ensure each theme was mutually exclusive. If themes were not mutually exclusive, they were merged with other conceptually similar themes. This was done by first reviewing all the coded data items and then rereading the entire data set using ATLAS.ti 8 qualitative software to (1) ensure themes were reflected in the data and to (2) analyze additional data missed in Step 2 that fit the established themes. In step 5, the research team revisited and reorganized codes to ensure the accompanying narrative matched the definition of themes. Through this process, we assigned sub-codes to main themes; sub-codes are more specific descriptors of relevant concepts that fit under a main theme because they are conceptually tied to that theme. In Step 6, a team of seven research assistants used the codebook to systematically apply finalized sub-codes directly to the open-ended responses while maintaining a 100% overlap between coders to assess inter-rater reliability using Cohen’s kappa (Hallgren, 2012). Intercoder reliability for the sample was 0.80 indicating high agreement (McHugh, 2012). Due to the variety in length and narrative direction of each participant response, our analysis produced 19 main themes. However, only three of these main themes are discussed in the current manuscript; other themes are not as directly related to the current research question.

Results

Three main themes of interest inductively emerged from the data set: Labeling (4 sub-codes); Judging Severity (4 sub-codes), and Victim Reactions (15 sub-codes). Labeling refers to how the victim describes their experience—as sexual assault or some other descriptor. Judging Severity captures if victims state that the experience was severe, mild, or compares the severity to something else (e.g., not being believed was worse than the sexual assault itself). Victim Reactions refers to how the victim responded (behaviorally, verbally, or emotionally) to the behavior during the assault. Refer to Table 2 for main theme operational definitions, sub-codes, and sub-code example text.

Table 2 Open-ended narrative general themes, sub-codes, and examples of participant responses
Full size table

Labeling

Many women used the words “rape,” “sexual assault,” or “sexual abuse” when describing their experiences (Sexual Assault sub-code). This contrasted with a few women who explicitly said that they do not view their experience as rape, sexual assault, or sexual abuse (Not Sexual Assault sub-code).Footnote 3 Although how a participant labels their experience is neither correct nor incorrect, it is important to note that of those few women who described their experience as not sexual assault, some described experiences which would meet the definition of rape (e.g., “I have never been raped, however sometimes during sex men will put their penis into the wrong hole on purpose with no warning” and “I guess it’s not rape but I didn’t particularly like waking up with someone doing me when I was obviously passed out”). A few women indicated that they did not describe it as rape because behaviors were attempted but not completed (e.g., “Me and my boyfriend got into a fight and he held me down and tried to have sex with me but I started crying and he stopped. I do not consider it rape but he was really forcing it”). Lastly, some of these women who described their experience as Not Sexual Assault described experiences that were coercive:

“I have not been raped, but I have had people…put their hands on me without my permission and tried to get them to stop at which point I was either met with anger, or pleading to let them engage in the sexual act with me.”

Some women indicated that their label for the situation had changed over time. More specifically, they previously did not call it sexual assault/rape, but now they do label it as such (Previously Not but Now Is Sexual Assault sub-code). This process unfolded clearly in the following story:

“I started crying in the car, and I told them that I had just made a bad decision. Over the weekend, my roommate gently suggested that I was assaulted and that I should at least see a mental health counselor. The next Monday (3 days post assault), I did. The counselor convinced me to go to the hospital. I was reluctant because I was ashamed and afraid to tell my family - I thought they would be ashamed of me. They were supportive. The hospital called police even though I asked them not to - I think a language barrier resulted in miscommunication. I decided that if my family knew, I had no reason not to speak with law enforcement/try to fight back against what happened. I started to accept what happened as rape, and I wanted to press charges.”

Lastly, there was a small minority of women (n = 4) who indicated that they had Grey Area Experiences of consent. Of these, two described experiences which would qualify as sexual assault according to the SES-SFV (e.g., “…once forced anal sex on me but stopped after one thrust when I cried out… like most women I’ve had ‘grey area’ rapey things happen to me”). One instance seemed to result from an established pattern of feigned resistance between romantic partners, “I never know what to call it… He started doing it and I said ‘no,’ but I would always say ‘no’ usually, but as a joke. But this time meant it, but I let him do it. I don’t know why.” While the final instance was described by a 64 year old woman about an instance when she was 13 year old—“I don’t remember all the details…the whole situation was very uncomfortable and sort of halfway consensual.”

Judging Severity

Some women commented in their stories about how severe they judged their experience to be. This main theme contained four sub-codes. Two contrasting sub-codes—Not a Big Deal and Worst Experience of My Life—represented two ends of a spectrum of severity. On one end of the spectrum, women expressed things like “it was a minor event” and “I don’t think of it as a bad experience.” On the other end of the spectrum, women said things like “The pain doesn’t stop when the horrific act is over, it continues to interfere with the rest of your life!”.

Additionally, instead of judging their experience as one of two extremes, women sometimes took the opportunity to compare their experience with other victims. In the sub-code Not as Bad as Other’s Experiences, they reported being different from other victims in that their experience was less violent and/or less traumatic than others’ experiences. For example, one woman wrote, “In comparison to what other people go through as far as sexual crimes, this is tame and this survey helped me realize my privilege as someone who hasn’t experienced many of these incidents.” One final sub-code in this main theme was Normal Experience where participants reported this experience as something that was common among women. Examples of this included statements like “my story is basic” and “It is what I would call a normal teen-age sex/dating behavior.”

Victim Reactions

Participants described a variety of behavioral, verbal, or emotional responses that they engaged in while the sexual assault was happening. This resulted in 15 sub-codes. Due to the large number of sub-codes in this main theme, we organized the sub-codes below into four meaningful clusters: overt resistance, covert resistance, passive actions, and negotiation. Additionally, one sub-code, Exhibited Confusion, was meaningfully different than these clusters and is presented by itself. The final two of the 15 sub-codes are “other” type of responses and are only presented in Table 2.

Overt Resistance

The overt resistance cluster represents when women clearly and actively withdrew consent via words or defensive actions. This included the three sub-codes: Said No; Yelled or Screamed; and Fought Back. These are all three exemplified in the following exert:

“I spun around angrily and yelled ‘What the hell do you think you are doing? Don’t touch me!’ which caused the other young men to burst out laughing and cheering him on. He smiled and laughed himself, then attempted to grab my breasts and I struck him on the side of his head. The other young men then started to run over to defend their friend.”

Additionally, some participants attempted to gain control in the situation as exemplified in the sub-code Confronted Perpetrator. Here they said things like “I smacked his hand away and told him if he ever tried touching me again I would tell my parents” and “I…demanded he take me to the train – the first available one.”

Covert Resistance

The covert resistance cluster represents when women used more subtle and covert, yet still active resistance strategies. This included two sub-codes: Pushed or Pulled Away; and Left. One participant reported “I pulled his hands away from my vagina an estimated 10 times or more…,” while another stated “I said I had to go to the bathroom (I was going to climb out the window) but he followed me so I couldn’t.” Some participants also indicated that they attempted to Get Help, another sub-code. These attempts were sometimes subtle (e.g., “I walked up to my mom… and asked her a random question just something to let the man know that this woman was my mom without scaring her”) and other times more obvious (e.g., “He had his hand across my mouth, so I couldn’t scream… I then heard someone walking nearby and I started kicking in the brush and causing dirt to move around”).

Passive Actions

The passive actions cluster represents when women used inactive, passive methods instead of overt or covert forms of resistance. This included two sub-codes involving involuntary freezing or voluntary freezing. Involuntary freezing occurred in the sub-code Frozen / Shock. An example of the Frozen / Shock sub-code was seen in a recollection of a rape that occurred when the woman was 10 years old, “…I just went to another place in my mind and totally shut down for the ordeal.” Voluntary freezing occurred in the sub-code Played Asleep or Stayed Still. One woman said that, “I felt powerless, so I pretended to sleep.” Another described an experience at a club where “…he started touching my ass in this uncomfortable desperate way and I was literally just there with my eyes closed like ‘lol when is this going to end.”’

Negotiation

The negotiation cluster represents times when participants negotiated with their perpetrator in some way. This included three sub-codes: Gave Excuses; Negotiated “Lesser Sexual Acts”; Complied Out of Fear, Confusion, or Fatigue. When women Gave Excuses they mentioned having a boyfriend/girlfriend, stated they were on their period, said they were a lesbian, or described that they needed to leave for some other reason. When Negotiating Lesser Sexual Acts, one woman stated,

“He then pulled his penis out and put my hand on it. I told him I would touch him if he please didn’t have sex with me because I had never (and still have never) had sex with a boy. He said fine and made me touch him until one of his friends walked in and I ran for the door.”

When women Complied Out of Fear, Confusion, or Fatigue, they said things like “I responded unwillingly to the oral stimulation and tolerated vaginal penetration after fearing he would slap or hit me” and “I would usually let him have his way for fear that if I didn’t I might lose him to someone else who would.” Another woman described,

“I felt like if I refused he would become verbally aggressive again and leave. I feel like he purposefully used the vulnerable situation I was in to get his way. I just laid there, never actually giving my consent and let him do what he wanted.”

Exhibited Confusion

Many participants described the emotions they experienced during the sexual assault (e.g., scared, disgusted, tormented, horrified, embarrassed, annoyed, angry, upset, or ashamed). However, one emotion repeatedly stood out: Exhibited Confusion. It was common for women to report not understanding what was going on or not knowing how to react to what was happening. Sometimes this confusion happened because the sexual assault occurred at a young age (e.g., “Some guy came up behind me [in a crowd at the circus] and rubbed himself against my backside. I wasn’t sure what was happening”). While other times confusion happened because the assault occurred so quickly (e.g., “Before I knew it, he had his finger inside me, he grabbed my hand and shoved it down his pants. I was in shock and didn’t know what to do.”). Women were also confused due to conflicting guidance (e.g., “I was always taught to respect my elders so I didn’t know what to do”) or lack of guidance (e.g., “I would tell him to stop and that I was uncomfortable, but didn’t do much fighting past that because I didn’t know what [I] was supposed to do. They don’t tell you how to deal with that, you know?”).

Discussion

In the present study, we examined three main themes that inductively emerged from victims’ qualitative reports of their sexual assault. We identified how victims labeled their experiences, which yielded four subcodes: Sexual Assault, Not Sexual Assault, Previously Not but Now Is Sexual Assault, and Grey Area Experiences. Another main theme pertained to how participants judged the severity of their experience, which yielded four subcodes: Not a Big Deal, Worst Experience of My Life, Normal Experience, and Not as Bad as Other’s Experiences. Lastly, victims identified behavioral, verbal, and emotional reactions they had during the experience. This yielded 15 sub-codes which we then grouped together into four clusters: overt resistance, covert resistance, passive actions, and negotiation.

Labeling

Many of the participants who labeled their experience identified it as sexual assault or a similar descriptor (i.e., rape, sexual violence, sexual abuse). However, there were some participants who took the time to explain that their experience was specifically not sexual assault despite their endorsements on our quantitative prevalence measure indicating that they had experienced at least one sexually assaultive behavior in their lifetime according to a modified version of the SES-SFV (Canan et al., 2020; Koss et al., 2007). This phenomenon of not acknowledging the experience as sexual assault or rape has been widely documented in the literature (e.g., Blayney et al., 2021; Newins et al., 2018; Wilson & Miller, 2016; Wilson & Newins, 2019). Peterson and Muehlenhard (2004) find one of the reasons that victims do not acknowledge their experience as sexual assault or rape is the absence of penile penetration. Our findings support this notion given that some victims reasoned their experience was not sexual assault because it was attempted but not completed; therefore, there was no penile penetration. However, the Rape, Abuse, and Incest National Network (RAINN) includes attempted rape in their definition of “what is sexual assault?” (RAINN, 2021). Conversely, many laws do hold attempted rape as a separate crime from completed rape, and some states (e.g., North Carolina) do not define attempted rape or attempted sexual assault as crimes at all (North Carolina General Assembly, 2018). Given this previous research and these mixed societal messages regarding whether attempts “count” as sexual assault, it is unsurprising that some participants may not label their attempted sexual assault as sexual assault.

Our findings also align with previous literature in that some women who labeled their experience as not sexual assault described coercive experiences. Fisher et al. (2003) found that victims were more likely to acknowledge the event as sexual assault when perpetrators used force or threats of force compared with when perpetrators did not. Perpetrators also treat coercion differently than other tactics in that they likely to underreport their illegal sexual assault tactics on self-report measures but are willing to admit their coercive tactics (Strang & Peterson, 2020). It is possible that our participants describing coercive experiences as not sexual assault may be drawing on rape scripts that sexual assault is physically forceful. Or they could be describing the experience as not sexual assault because in coercive experiences victims acquiesce. But because acquiescence under pressure of coercion does not meet the established “freely given” standard of consent (Hickman & Muehlenhard, 1999), these experiences would not be considered consensual and thus could count as sexual assault nonetheless.

Relatedly, a small minority of women said that they had Grey Area Experiences of consent. This is particularly interesting due to feminist criticisms of both media messages and legal arguments that assert that there are ambiguous lines between sexual assault and consent. It is important to note that of the 172 labeling codes in our dataset, only four women mentioned Grey Area Experiences of consent. Two of these four described a scenario that clearly met operational definitions of sexual assault, but, for whatever reason, the participants still called it a grey area. Another participant reflected on an instance more than 50 years prior, in which she did not provide enough detail for us to fully understand what happened during her experience. As such, only one story out of 906 stories was labeled in this way and seemed to represent a “grey area.” Although this one valid experience in our sample represents the concept of Grey Area Experiences of consent, this was not a common experience among our participants.

Policy Implications for Labeling

One of our subcodes in the Labeling main theme represented experiences where victims previously did not label the experience as sexual assault but then did label it as such at a later time. It is important to highlight that someone changing their initial reaction to an experience is valid. In fact, this shift is sometimes a step in the process of understanding sexual assault due, in part, to the ways victim blaming is deeply rooted in our understanding of sexual violence (Filipas & Ullman, 2001). Such information could inform policy regarding statute of limitations for sexual assault. Given that it may take a while for victims, especially child victims (McElvaney, 2015), to even recognize and label what happened to them as assaultive, extending or removing statutes of limitations on sexual assault may be warranted. This is especially important because it can take victims years before they seek mental health services after a sexual assault (Monroe et al., 2005). The push for extending or eliminating statute of limitations has strengthened since the #MeToo campaign brought attention to many cases, both new and old. Of course, policy officials have to balance this reality with the research on memory and eye-witness testimony finding that adults can recall certain details about childhood memories (e.g., activities, locations, who was present; Wells et al., 2014), especially if those memories are related to childhood sexual assault (Alexander et al., 2005), but that other overly specific memories may be false (Wells et al., 2014).

Judging Severity

The sub-codes in the Judging Severity main theme indicate that sexual assault can be a severe and traumatic event for some women and a mundane event for others. Of course, much research has rightfully focused on the trauma of sexual assault; a cross-database search for the terms sexual assault, rape, and trauma yielded 59,077 journal articles published in the last twenty years. But when women sometimes describe their experiences as less traumatic, clinical sexual assault professionals and researchers sometimes call this minimization. For example, Edwards et al. (2014) notes that, “Indeed, 26% of women [spontaneously] specifically mentioned that the sexual assault experience was ‘not a big deal’…which seemed to indicate a minimization and normalization process (p. 2537).”

The concept of minimization assumes that a person is curtailing the impact that sexual assault has on them. In other words, it assumes that sexual assault has a bigger impact than is being stated. Although this may be the case for some victims, as minimization can be a coping mechanism for trauma (Dörfel et al., 2008), there is also a possibility that some victims are authentically and appropriately describing the impact of their experience. This can present a difficult situation for victim advocates and clinicians as it is their job to support victims by both believing victims and helping victims sort through trauma. For the former, they must take victims at their word even if victims describe their experience as Not a Big Deal; for the latter, they must not take victims at their word to help victims process trauma if the advocate suspects the victim is experiencing minimization. Ultimately, we are limited in what we can say about what makes sexual assault severe for some and not for others. But acknowledging that both experiences exist—that sexual assault can be Not a Big Deal for some and the Worst Experience of My Life for others—is important because rhetoric which ignores either reality can either invalidate some people’s trauma or pressure others to situate their experience in an inauthentic way.

Nevertheless, minimization is a well-documented behavior. It is a problem because researchers speculate that minimization and normalization of sexual assault can increase revictimization risk (Edwards et al., 2014). Indeed, minimization seemed to have emerged in other parts of participants’ open-ended responses in our sample. Specifically with the sub-code Not as Bad as Other’s Experiences, it is interesting to note that the comparison only occurred in one direction—saying that their experiences was not as bad as others. No participants indicated that their experience was worse than other women’s experiences. The reason we only see the comparison in one direction could be due to minimization.

Policy Implications for Judging Severity

Because one of the sub-codes in this main theme was Normal Experience and, in general, sexual assault is common (e.g., Canan et al., 2019; Muehlenhard et al., 2017), rhetoric which frames sexual assault as a unique experience or a special interest topic is clearly incongruent with our data. Therefore, it is inappropriate to treat sexual assault prevention programming as a special interest issue. Further, because our participants indicate that their experience is normal, a wide breath of policy is needed to address this issue (e.g., expand and reinstate the federal Violence Against Women Act, ratify the Equal Rights Amendment, limit the ability for convicted perpetrators to serve in governmental positions of power, classify online rape threats as crimes, allow immunity for undocumented immigrants and sex workers to report sexual assault). We base these suggestions in the assumption that sexual assault should not be nor feel normal.

Victim Reactions

Because our last main theme had 15 subcodes, we grouped them together in four meaningful clusters. Our first cluster of sub-codes, overt resistance, are actions that are arguably the most socially accepted and expected reactions for a victim to have during sexual assault. This cluster contains the sub-codes Said No; Yelled or Screamed; Fought Back; or Confronted Perpetrator. Our findings are consistent with some of the qualitative themes Edwards et al. (2014) reported in their college sample. These actions are often portrayed in media as common actions that people use to refuse sex (Jozkowski et al., 2019) and therefore are perhaps the most socially accepted reactions because they demonstrate, in the most obvious way, that the event was not consensual. In fact, several rape myths are predicated on the idea that “real” rape only happens when victims overtly refuse in these ways. For example, McMahon and Farmer (2011) found that college students agreed or were neutral about the statements “If a girl doesn’t physically resist sex—even if protesting verbally—it can’t be considered rape” or “If a girl doesn’t say ‘no’ she can’t claim rape” (26.8% and 60.3% of the time, respectively). However, according to a review of resistance strategies, victims engage in physical resistance or fighting less often, with only 20–25% of victims using these strategies during a rape (Ullman, 2007). Nevertheless, many sexual assault prevention programs focus on strengthening the self-efficacy of performing such actions (Hollander, 2018; Orchowski, et al., 2020a, b) despite critiques that this focus may perpetuate these rape myths (Basile, 2015). Although our findings suggest that these behaviors are clearly the way some women react during their assault, it is important to remember that this cluster is one of four common response types.

Another response type was covert resistance, which included subcodes of Pushed or Pulled Away, Left, or Got Help, the first two of which were also reported by Edwards et al. (2014). These actions are not as directly combative as the overt resistance tactics, but still actively attempt to stop the sexual assault and are often successful in doing so (Ullman, 2007). The emergence of this covert cluster can be partially explained by women’s socialization as relationship builders and peacekeepers (Ullman, 2007), which also applies in sexual encounters. For example, previous research finds women feel a responsibility to caretake men’s egos in college hookup situations by acquiescing to sex the women do not want (Jozkowski et al., 2017). So, some victims will want to maintain as much social harmony as possible while still stopping the attempted assault (otherwise known as walking the “cognitive tightrope”; Norris, 2011). For example, pushing or pulling away allows a victim to nonverbally refuse the behavior. By using nonverbal cues, the victim can subtlety refuse, ideally increasing the likelihood that the perpetrator can avoid potential social humiliation of being rejected. Additionally, victims may use covert resistance with regard to leaving because it is one of the best chances of stopping the sexual assault while avoiding direct confrontation and possible escalation; the assault cannot happen if the two people are not in the same room. Getting help is also an effective option because a non-enabling bystander’s presence will thwart most sexual assault attempts—a fact that many bystander intervention programs rely on (Koelsch et al., 2012). When a bystander enters the interaction, any potential social disruption is now no longer the “fault” of the victim.

Our next cluster of sub-codes, passive actions, involved involuntary freezing (Frozen / Shock sub-code) and voluntary freezing (Played Asleep or Stayed Still sub-code). Again, these findings are similar to Edward et al. (2014) findings. It is well documented that traumatic responses, such as sexual assault, can create a flight response in the autonomic nervous system causing a person to freeze and dissociate (Campbell, 2012; Möller et al., 2017; Van der Kolk, 2015). As such, the emergence of the sub-code Frozen / Shock, is not in fact shocking. Similarly, voluntary freezing is also unsurprising for two reasons. First, hiding and “playing dead” can help to avert danger for humans and animals alike because it can disincentivize predators, although research finds this generally does not thwart sexual assault attempts (Ullman, 2007). We see this thought process unfold in the quote in Table 2 where the victim says “I thought if I acted like I was asleep [a]nd didn’t respond he would leave me alone.” Second, voluntary freezing may further be relied on due to societal messages that people, and women in particular, should avoid direct sexual communication. Pretending to be asleep and staying still are both ways of avoiding direct sexual communication. Given that many people, even trained clinical professionals (Lavie-Ajayi, 2016), experience anxiousness or embarrassment around discussing sex, victims may engage in voluntary freezing as a means of avoiding having to refuse, which would require them to communicate about sex.

Negotiation was our next cluster of subcodes where some victims seemed to bargain with their perpetrators to avoid the assault or make it less severe. To avoid the assault entirely, our participants Gave Excuses like having a boyfriend/girlfriend, they were on their period, they were a lesbian, or they needed to leave for some other reason. But, perhaps if the assault felt inevitable, victims sometimes reported using one of two forms of harm reduction. In one form, they Negotiated “Lesser Sexual Acts” (e.g., agreeing to perform manual or oral sex instead of vaginal/penile sex). The second form happened when victims Complied Out of Fear, Confusion, or Fatigue. For example, they allowed the sexual assault in lieu of being beaten or they allowed sexual assault because they did not know what else to do. To be clear, both of these forms of harm reduction are no less sexual assault than other experiences given that sexual assault hinges on “freely given” consent (Hickman & Muehlenhard, 1999, p. 259), which is not present in these seemingly forced-choice circumstances. Edwards et al. (2014) also noted this theme in their narratives adding that students regularly quantified the extent (e.g., number of minutes, number of refusal cues) that they tried other resistance strategies before conceding.

Lastly, in addition to the various behaviors that victims engaged in, they also commented on the emotions they experienced during the assault itself. A range of emotions were present: from annoyance to horror. But one noteworthy emotion that was repeatedly mentioned was confusion. Participants were either too young to understand, the actions happened too suddenly to understand, or they were given too little guidance to understand. Most sources—police departments (e.g., Amherst MA Police Department, 2021), universities (e.g., Saint John’s University, 2021; Villanova University, 2021), and empirical articles (e.g., Dunmore et al., 2001)—discussing the association between confusion and sexual assault only acknowledge the confusion that occurs after the assault has taken place, often as a byproduct of Post-traumatic Stress Disorder. Confusion as an emotion experienced during the assault is less frequently discussed, which may mean that researchers assume victims are aware of what is happening to them in-the-moment. Assumptions like this will have implications for what we expect victims to be able to tell researchers, police officers, and sexual assault support services about the experience of sexual assault.

Policy Implications for Victim Reactions

The current federal definition of sexual assault delegates the power of defining the term to local state and tribal laws (Department of Justice, 2019). As a result, what can be prosecuted as sexual assault varies across the country. For example, North Carolina state law (North Carolina General Statutes Sect. 14–27.21, et seq.; Bill NC S199) indicates that:

A person commits a sexual offense by engaging in a sexual act with another person by force and against the will of the other person; or with another person who lacks capacity, either mentally or physically, and the defendant knows or reasonably should know of that incapacity. (para 7)

In definitions such as this, a case would have to establish both force by the perpetrator and the “will” of the victim. Our findings suggest that a range of behaviors can demonstrate unwillingness. This is especially important because, in the case of North Carolina, what is “against the will of another person” is judged by whether it is “in a manner that would cause a reasonable person to believe consent is revoked” (§ 4–27.20.1a.b). If we typically get messages that only overt resistance tactics meet the threshold of a “reasonable person” believing consent is revoked, like Orchowski et al., (2020a, b) finds in their sample of college men, then we disqualify many instances of sexual assault from criminal prosecution because the victim did not indicate their unwillingness in a “reasonable” manner. It would be important for district attorneys and other stakeholders in the criminal justice system to know that victims will engage in a range of behaviors (e.g., negotiate lesser sexual behaviors to avoid other sexual behaviors) and that these behaviors are trauma responses and harm-reduction strategies, not consent. Because victim blaming largely asserts that the victim should have done something different to avoid being assaulted, the open-ended responses in this study serve to normalize a range of behaviors that victims engage in. We would recommend that norming campaigns focus efforts on illuminating the wide range of typical responses to assault found in this study.

Additionally, our findings can be used to further support the growing popularity of “Yes Means Yes” and similar affirmative consent policies in defining sexual consent. Relying on victims’ overt refusals to stop sexually assaultive behavior is a flawed method because overt refusals represent only a portion of the range of common victim actions. Further, because some victims reported experiencing confusion due, in part, to how quickly their assault progressed, it is important to move away from “no means no” policies as some perpetrators subvert the victim’s ability to refuse by surprising them with the behavior (Canan et al., 2020) or not giving them the opportunity to say no (Jozkowski & Peterson, 2013).

In addition to the behaviors, the presence of the Exhibited Confusion sub-code further exemplifies the need for sex education more broadly. Because policies of sex education deliverance are inconsistent in the U.S. (Guttmacher Institute, 2021) and existing sex education seems to infrequently address consent (Willis et al., 2019), it is unsurprising that some victims report being confused when consent is ignored or rejected. One of our participants articulated this lack of experience and knowledge when she reported feeling confused in response to her perpetrator continuing to pursue sex post-refusal: “…They don’t tell you how to deal with that, you know?” It seems that for women like her, sex education policy needs to better address expectations and skill building, particularly to prepare them for what to do if their boundaries are not respected in a sexual encounter.

Limitations and Future Directions

Due to concerns about the lifetime recall period and potential recall bias, we instructed participants with more than one experience to write about their “most memorable” experience. We did this instead of using the “most recent” experience because not all recent experiences are easily remembered due to inebriation, dissociation during the event, etc. Given the request for “most memorable” experiences, the responses could be subject to the von Restorff effect which finds that when people have to select from a group of similar things, the item that is the most different from other items will be the one most easily noticed and remembered (Parker et al., 1998). Therefore, if participants are selecting from several sexual assault experiences, they may be more inclined to recall an experience that differs from other experiences in some way. As such, the accounts reported in our study may represent an atypical sexual assault experience. However, with the sheer number of participants in the current study, we feel confident in the number of endorsements per theme throughout these open-ended responses to say that these are representative of our sample’s experiences of sexual assault.

One more methodological limitation is that our qualitative prompt did not ask about any one particular detail of the sexual assault experience; instead, we asked participants to provide any information about the context of the experience. Therefore, we cannot treat themes in our study as an accurate measure of frequencies. We cannot meaningfully compare which themes occurred more or less than others. For example, there could be substantially more unacknowledged sexual assault victims in our study, but because we did not specifically ask participants how they labeled their experience, they might not have commented on that aspect. Also, this means that we can only discuss the literature that our findings are similar too, not the research that it is divergent from. Because we did not ask participants to specifically comment on certain aspects of their experience, we can only describe what participants said, not what things they did not say. Researchers may consider quantifying our qualitative findings in future studies to compare frequencies and plot trends.

Lastly, the current study was delimited to cisgender women over the age of 18 living in the USA. To extend our findings, we recommend researchers study these experiences in other populations (e.g., transgender and nonbinary populations). Notably, according to research on the sexual assault experienced by men, there is a wide range of prevalence rates depending on the definitions used (Peterson et al., 2011). Considering that some states legally define sexual assault in terms of whether the person communicated non-consent “in a manner that would cause a reasonable person to believe consent is revoked” (North Carolina General Statutes Sect. 14–27.21, et seq.; Bill NC S199), it is important to examine what a reasonable person would expect a man’s non-consent to look like. Further, it is then important to assess if these expectations are in line with how men actually respond during their sexual assaults. Similarly, given that research examining sexual assault experiences within transgender and nonbinary communities is even less developed than research of cisgender men, we also recommend researchers expand this topic to include these groups.

Conclusion

Victims of sexual assault react in several different ways during their assault experience. Some reactions that are commonly portrayed in media or focused on by the criminal justice system and law—such as overt resistance—tend to evoke less victim blaming. But victims may respond in a variety of other ways, such as via covert resistance, passive actions, and negotiation. Normalizing these responses is important especially given that some state sexual assault statutes define sexual assault in terms of what a “reasonable person” would recognize as a refusal cue. Additionally, victims experience a wide range of emotions during their assault—confusion being a notable one, which has implications for sex education.

When victims give a label to their assaults, we find that “grey area” experiences are exceedingly rare and that some labels change over time, the latter of which may have implications for statute of limitations policies. When judging the severity of their experiences, women mentioned two ends of a continuum (i.e., worst experience of my life vs. not a big deal) in addition to judging the experience as normal or judging it as less severe than other victims. Although we cannot speak to why certain participants judged the severity in the manner they did, we assert that these experiences are valid. We also caution researchers and clinicians to think about if and when they should describe any of these severity statements as minimization. Ultimately, the sub-code Normal Experience under the Judging Severity main theme, is one of the most concerning. If the policy changes suggested in this paper are ever fully realized, we argue that those would have a deep impact on combatting the idea that sexual assault should feel normal.