Introduction

Sexual assault is startling common on college campuses where one in four women will experience rape during their time on campus; this paper will focus on women’s experiences as women have been the participants in the majority of research on sexual assault (Gross, Winslett, Roberts, & Gohm, 2006). Most often, these are assaults committed by acquaintances, often on dates or at parties, and are associated with a host of social, emotional, and physical difficulties (Classen, Field, Koopman, Nevill-Manning, & Spiegel, 2001; Koss, 1993; Tjaden & Thoennes, 2000). Women respond to the threat of acquaintance rape by juggling concerns about the relationship, the possible impact of the threatening situation on their social circles, as well as concerns about their own safety (Macy, Nurius, & Norris, 2007). These competing concerns may create barriers to engaging in effective and protective behavioral responses to the threat of rape. Moreover, the process of behaviorally responding to the threat of rape is an extraordinarily complex one with variable outcomes, especially given the relational nature of the threat in the case of acquaintance or date rape (Nurius & Norris, 1995). Feminist self-defense is a common risk reduction strategy for women; women who participant in these interventions report positive benefits and find meaning in their participation (Hollander, 2004). Indeed, there is some evidence that women who have previously experienced sexual assault may more often opt to participate in these programs (Brecklin, 2004). This paper uses the descriptive term, women who have experienced sexual assault, rather than victim or survivor in order to avoid using a label that was not chosen by the person who experienced the event(s) herself. Interventions for sexual assault risk reduction have low efficacy (Orchowski, Gidycz, & Raffle, 2008) perhaps due to lack of understanding of the mechanisms of the intervention. Further research on behavioral responses to the threat of sexual assault has the potential to inform and improve interventions for people who have or may experience sexual assault.

The complex process of behavioral response is likely to be especially true in the case of threats from an acquaintance where the aggressor may have both social and physical power. Acquaintance rape, forced or coerced vaginal, anal, or oral intercourse, is the most common form of rape, accounting for nearly 90 % of all rapes (Tjaden & Thoennes, 2000). Although recent research has begun to identify barriers to assertive responses towards acquaintance or date rape situations, less is known about what specific styles or types of behavioral responses are elicited in these situations and the factors that facilitate different responses. A better understanding of the behavior elicited by the threat of sexual assault may inform interventions, advocacy efforts, and policies focused on responses to sexual assault. As such, the goal of the present study was to qualitatively explore the range of behavioral responses elicited by an experimental date rape threat (a specific type of acquaintance rape) using an open-ended response procedure to a brief, realistic audio stimulus.

Responses to the threat of sexual assault can be manifested in a variety of ways, from tonic immobility, (i.e., motor inhibition caused by intense fear) to kicking and screaming. For the current study, we use the term behavioral response to refer to any behavior, verbal or non-verbal, that is elicited by the threat of sexual assault. The term behavioral response is used to encompass the entire continuum of possible behaviors associated with the threat of rape including both planned, active behaviors such as kicking an attacker and involuntary, automatic responses such as tonic immobility. Additionally, behavioral response is used rather than “behavioral resistance” to indicate that some behaviors may be engaged in without conscious recognition or perception of a risk and that some of these behaviors, such as bargaining, may not be perceived as “resistance” although they are enacted with that purpose. Past research examining behavioral responses to the threat of rape has generally categorized them by two opposing dimensions based on assertive behavior as the presumed model or default, physical or non-physical (i.e., verbal) and forceful or non-forceful (Gidycz, Van Wynsberghe, & Edwards, 2008). This model cannot capture the entire range of responses which may include behavior not easily categorized in this manner. For example, turning the body away could be seen as forceful or non-forceful depending on the context.

This approach has overwhelmingly found that physically forceful behavior in response to a threat of rape is most often associated with protective or less severe outcomes (Clay-Warner, 2002; Fisher, Daigle, Cullen, & Santana, 2007). Although the dominant model, there are important limitations in this approach to categorizing behavioral responses; factors such as age, gender, culture, power dynamics, and substance consumption may influence and/or alter one’s ability to respond to threat in a forceful physical matter. These factors play a role in enabling women’s “enforced inaction” by encouraging social myths that women cannot effectively defend themselves (encouraging “freezing” or immobile responses) or might face dramatic social consequences if they do (for a primer, see Rozee, 2000). Feminist self-defense interventions therefore seek to overcome barriers to effective responses by repeated practice of skills and psycho-education on possible barriers to implementing skills (Gidycz, Orchowski, & Edwards, 2011). Although this research has been fruitful in identifying effective responses and inspiring interventions to train assertive responses, it is limited because it does not focus on changing men’s attitudes and behaviors and it has not characterized the possible scope of responses beyond physical vs non-physical and forceful vs non-forceful.

Exploring behavioral responses as a complex psychological process that can be measured, evaluated, and targeted for intervention could be critical to providing a greater number of effective interventions for people who are at risk of or have experienced sexual assault. Recent research has established that the behavioral responses women describe hypothetically in laboratory scenarios correspond well to the responses they employ in real life, making hypothetical and analog scenarios powerful tools for learning more about behavioral response to threat (Turchik, Probst, Chau, Nigoff, & Gidycz, 2007). A more comprehensive understanding of the range of possible responses that may be elicited in response to acquaintance rape threats is needed in order to better understand the phenomena of sexual assault. This can be helpful for risk reduction intervention as well as psychotherapy for survivors to help contextualize and normalize experience. Existing research is limited by overwhelmingly quantitative outcomes, which inherently limits the number and type of responses women can describe (Gidycz, McNamara, & Edwards, 2006). Thus, a less constrained, qualitative depiction of the ways people respond to the threat of acquaintance rape is necessary in order to learn more about the general tactics, broad themes, and behavior elicited in response to threat. To our knowledge, only one study has examined how women respond behaviorally to threat using a qualitative design.

Masters, Norris, Stoner, and George (2006) recruited women to read a vignette describing an acquaintance rape attempt in progress. Women were then asked to write the ending of the story and describe anticipated behavioral responses. Examining women’s sequential responses to the aggressive action (i.e., aggressor action 1, defensive response 1, aggressor action 2, defense response 2, etc.), Masters et al. found that women tended to increase their use of physically assertive behaviors while decreasing their use of verbally assertive behaviors from the first to second action (Masters et al., 2006). This is critical in elaborating that a forceful, physical response which is likely the most effective response may not be the first naturally occurring behavior. Additionally, they found that descriptions of assertive behavior were common, but a small minority of women described non-assertive responses such as making excuses for why they did not want to engage in sex; this kind of response is not well characterized by the physical/non-physical, forceful/non-forceful dichotomy. That women described non-assertive responses suggests that the range of behavioral responses may include behavior that has not typically been studied or well characterized by researchers. Therefore, qualitative research may have a particularly important role to play in the development of tools to measure and assess behavioral responses because it is well situated to elicit the broadest possible range of responses. The study by Masters et al. was limited, however, in utilizing only one severe stimulus for women to describe their behavioral response to, thereby potentially limiting the range of behavioral responses that might be elicited. Therefore, research using a variety of stimuli (including less severe stimuli) is needed as women would benefit from learning to respond to threats as early and quickly as possible.

Given the need to better understand behavioral responses to the threat of acquaintance rape and the limitations of current quantitative assessment of behavior, the purpose of the current study was to explore qualitative themes in college women’s hypothetical behavioral responses to a date rape stimulus. Because little is known about the possible range of behavioral responses to threats of date rape, the current analysis will explore themes in the broadest manner possible by recruiting college women of any background and any assault history using an open-ended response format. Furthermore, a hypothetical response paradigm with varying degrees of threat within the same basic scenario was used to elicit a broad range of responses without introducing responses specific to the environmental characteristics of the stimulus rather than the threat level of the stimulus.

Method

Participants

A total of 143 college women were recruited in Fall 2010 and Spring 2011 semesters through psychology courses offering credit for participation. Inclusion criteria were that women needed to be 18 years of age or older. The data of two participants who identified as exclusively lesbian were excluded from data analyses, but not participation, as it was theorized that they may have difficulty imagining themselves in a scenario that portrayed a heteronormative date (two cisgender people who chose to go on a stereotypical date to the movies) and this subsample was too small to analyze separately. Two participants’ responses included information that would permit identification of their assigned condition possibly introducing bias in the coders (a person may feel certain stimuli should be associated with certain responses), thus making blind coding of the response impossible. Following, these two responses were also excluded leaving a final sample of 139 participants and corresponding transcripts.

Participants were 139 undergraduate women, ages 18–39 years (M = 21.8, SD = 4.1, mode = 19), enrolled at a medium-sized Midwestern university. Participants were predominantly Caucasian (77 %); 10 % reported being African American, 4 % reported Asian or Pacific Islander, 6 % selected their race as “other,” 7 % as Hispanic or Latino, and 1 % bi- or multi-racial. Four participants (3 %) identified their sexual orientation as bisexual and the remainder identified as heterosexual.

Procedure

The current study was a qualitative analysis of an experimental study that investigated the utility of an audio vignette—analog threat paradigm to quantitatively evaluate behavioral responses to the threat of acquaintance rape (Anderson & Cahill, 2014). To assess the relationship between the intensity of responses and the intensity of threat stimuli, participants were randomized to four different conditions representing different levels of intensity of the same coercive stimulus. The variable of interest for the experimental study was the quantitative intensity or clarity of women’s hypothetical responses to the stimuli; for further details, see Anderson and Cahill (2014). Participants were also asked to respond to the stimuli in an open-ended format. For the current study, we are qualitatively analyzing the open-ended responses.

The Vignette Stimulus

The audio recording used as the threat response stimulus was created by trained actors. The validity of the scenario depicted in the audio vignette has been rated as realistic by college student participants and used extensively to study factors related to threat perception and sexual assault (Marx & Gross, 1995). The recording depicts a couple on a date, Jenny and Dan, who have recently returned to the man’s apartment after a movie. Limited background information was provided on the context of the date; instructions noted that the couple portrayed in the recording had been on two dates before but never had sexual intercourse.

The scenario begins with casual conversation followed by mutual kissing. Coercive sexual behavior is later enacted by the man and escalates as the scenario continues. In the phases of escalating coercive sexual behavior, the woman politely refuses his advances but the man persists. The woman continues to verbally refuse the man who then apologizes. The couple continues to kiss and the man begins to verbally and physically pressure the woman into escalating their sexual intimacy (e.g., touching her buttocks, etc.) in which the woman refuses verbally, angrily shouting at the man.

Threat conditions were created by increasing the length of the audio recording participants heard before being asked to respond; an additional 18 s was added in each condition thus introducing small amounts of additional coercive behavior. The scenario participants listened to for each condition was, respectively: the woman politely refusing the man’s advances (low threat, Condition 1), the man apologizing for touching her breasts a second time (medium threat, Condition 2), and the woman angrily refusing the man for touching her buttocks (high threat, Condition 3). Participants in Condition 4, the control condition, chose the threat level themselves by pushing a button to indicate the man had “gone too far.” This allowed for comparison of how participants viewed threat when the threat was identified by themselves versus when it was designated by the experimenter as in conditions 1–3 as presumably participants may only be able to generate a behavioral response after they themselves have identified the threat. This control condition is in accordance with how the stimulus has been utilized in past research (Marx & Gross, 1995). Approximately three quarters of participants selected stimuli in the same range as the other three conditions indicating control condition participants in general selected and responded to the same stimuli (Anderson & Cahill, 2014).

Participants completed the study in individual appointments in a private room with the assistance of a female experimenter to complete informed consent, explain how to complete the study procedures independently, and illustrate how to contact experimenters for further help or questions. Participants were instructed to imagine themselves in the place of Jenny in the scenario. When the audio recording automatically paused, participants alerted the experimenter, who entered the room and provided further instruction. The following instruction was given: “What would you do now if you were Jenny [the woman in the audio recording]? Please say and/or show what your response would be in this situation. There are no right or wrong answers and please be as honest as possible.” After participants responded, experiments gave a final cue in order to capture all possible responses saying in a neutral tone, “anything else you would do or say?” Experimenters audio recorded all responses and took notes as to whether any physical demonstrations or cues were utilized by the participant. No participants made physical gestures without also giving a verbal explanation. Experimenters were allowed to ask follow-up questions to ensure clarity. During this time, experimenters also responded to participant questions. The majority of participants did not have questions but several asked for clarification of the instructions or confirmation regarding details of the vignette. All participants were able to generate a response that clearly indicated what they would do next; no participants demonstrated a physical response without also verbally describing it. At the end of the appointment, participants completed debriefing where they were provided information about the nature of the study, local resources, for survivors, and the opportunity to ask questions and provide feedback about their participation in the study. Participants rated their participation on average as mildly upsetting M = 3.1 (SD = 2.1) on a scale of 1–10.

Responses were audio recorded and transcribed for analysis. Ten percent of transcripts were randomly selected and examined for errors by listening to the original audio while reading the transcript. Transcripts averaged less than one error in spelling or wording per respondent. No transcripts were identified in which an error of content (e.g., part of the response was missing) was identified.

Data Analysis

Transcriptions of participants’ verbal responses were the source of data for this study. Data were analyzed using the Consensual Qualitative Research (CQR) methodology (Hill et al., 2005). CQR methodology is an approach to qualitative data analysis wherein multiple researchers come to a consensus on themes generated from data review (Hill, Thompson, & Williams, 1997). CQR is a replicable process that maintains scientific rigor and the validity of the data via a three-step procedure (Hill et al., 2005). In the first step, researchers independently develop general themes, also called domains, by reviewing the data line by line. The same piece of data may be evaluated for more than one theme in this approach allowing for the fullness and richness of the data to come forth. Although data may be coded for more than one theme, parsimonious coding is encouraged. The independently generated themes are then discussed by the team of coders and together they come to a consensus on the core themes; this process facilitates consensus as well as parsimony in the themes. Once a consensus has been reached, core ideas are created by analyzing the raw data (i.e., individual participants’ statements) from each theme. In the final phase, called cross analysis, themes are compared across all participants and, where necessary, sub-themes are established or themes combined.

To reduce bias, strengthen the validity of the data, and determine accuracy of coding, an auditor reviews decisions made by the coders at each stage in the data analysis. Any coded sections of transcripts that the auditor disagrees with or any definitions of developed codes that the auditor finds unclear are identified and then given to the coding team for review. The coding team then discusses whether to accept or reject the auditor’s concerns. When no new themes emerge, saturation, or the stability of the findings, is said to be achieved (Williams & Hill, 2012). In CQR, typicality is established by indicating how frequently themes emerge in the study. Themes and sub-themes are then labeled as an experience that is general (i.e., all participants experienced it), typical (i.e., half or more of the participants experienced it), variant (i.e., less than half of the participants experienced it), or rare (i.e., only one or two participants experienced it) (Hill et al., 2005).

The core team of CQR coders consisted of three undergraduate women who were experienced research assistants. One of the research assistants had no previous experience working with topic of sexual assault, whereas the other two research assistants assisted in data collection in the original study. Another experienced male undergraduate research assistant, who was unfamiliar with the study or sexual assault research, served as the auditor. The auditor identifies as a masculine demisexual genderqueer person with homoromantic tendencies and was not familiar with sexual assault research. The auditing process strengthens the validity of results by introducing a different perspective than that of the study team and one that is diverse in gender and sexual orientation. In the current study, the auditor agreed with the majority of the developed codes and coded transcripts. His suggested changes or disagreements were primarily related to clarifying theme definitions and the coding team accepted these minor revisions. To train the team of coders, a seminar was held wherein the coders and first author were instructed in qualitative methods by the second and third authors and practiced CQR methodology using examples from previously published work. In the following coding sessions, the coders first met independently to compare coding and discuss the coding process. After this initial part of the meeting, the first author was then invited to the coding session to help discuss any difficulties that may have arisen in coding, i.e., disagreements in coding, questions about transcripts, et cetera.

In order to fully evaluate the range of responses to the threats presented, we analyzed all responses blinded to the original condition or threat level. Sub-dividing samples prior to thematic analysis is not recommended as this could lead to the creation of different themes due to the artificial separation that might render later comparison impossible (Ladany, Thompson, & Hill, 2012). Therefore, we analyzed themes in a way that included all participants together and then compared themes across conditions post-coding in order to minimize possible bias and remain consistent with the CQR approach and prior research (Paul et al., 2010).

Results

In analyzing women’s responses to a date rape stimulus threat, six major themes emerged (sub-themes are listed in parentheses): assertion (physical, verbal), compliance/acceptance, conditional decision making (on Dan, on self), avoidance (de-identification with victim role, deflection), expressions of discomfort, and allusion to future contact. Following CQR methodology, definitions for each theme were derived from the data itself rather than from prior scholarly work and are subsequently presented. The number of times each theme appeared is presented in Table 1. Sub-themes were identified within all themes except for compliance/acceptance, expressions of discomfort, and allusion to future contact. Sub-themes will be discussed for each respective code.

Table 1 Results of cross analysis: Frequency of theme by condition, N (% of total codes within theme)
Full size table

Assertion

Assertion was a typical experience among participants as a response to a perceived attempted sexual assault in the current study. Assertion is defined as responses where a confident declaration expressing or enacting behavioral change was made. This declaration was directed at the threat and the person’s desires were clearly articulated. Assertive responses were viewed as a continuum from relatively less assertive responses such as physically moving away from the threat (i.e., a specific behavioral change to escape the proximity of the threat) to relatively more assertive responses such as slapping and yelling at Dan (i.e., specific behavioral changes made to directly counteract or react to the threat). Previous work has similarly defined assertive behavior in a broad fashion that spans a range of possible behaviors (Macy, Nurius, & Norris, 2006; Masters et al., 2006; Parrot, 1996). Women in the current study described two forms of assertion and consequently two sub-themes were created: verbal and physical assertion.

Verbal Assertion

Many participants described their response to threat by refusing advances verbally or by verbally communicating behavioral boundaries to Dan’s advances. Verbally assertive responses frequently included justification or explanation for why participants would respond in the way that they did. Two categories within verbal assertion were created: verbal refusal and expressing and enforcing boundaries.

Verbal Refusal

Verbal refusal responses included those that directly asked Dan to stop or told Dan “no.” For instance, women provided a variety of assertive statements often including explanation or reprimands such as “I would just say no if I was not comfortable with that situation,” “I probably would tell him not to do that,” and “I would tell him that ‘no’ means no.” Some verbal responses came with further explanations of why women in the study were rejecting Dan’s advances. For example, one woman said, “…I would just be like ‘no’…and explain to him why I didn’t want that [physical advances].” Another woman said, “I would tell him to slow down…[then] say, ‘you came on a little too strong’.” Yet another participant described her response like this, “I would probably say that I’m not ready for that yet, that type of what they were doing I guess. So I would tell him I wasn’t ready.”

Expressing and Enforcing Boundaries

In addition to outright refusal, participants also provided confident verbal declarations of their physical boundaries. For instance, one woman explained how she would explicitly state her boundaries to Dan: “[I would] tell him exactly what [was] and was not acceptable.” Another woman was more specific about her description of boundary lines, responding, “I would tell him very specifically that I am only ok with kissing and nothing else.” Another woman indicated, “I would just tell Dan like that I’m not comfortable with going to be that sexual [sic], I don’t want him to touch my breasts, like we can kiss but that’s a little too fast for me.” Others were more vague with exact behavioral boundaries, but clearly expressed their desire to maintain a boundary. For example, one woman said, “I would tell him I don’t want to go any further.” The enforcement of such boundaries often included limits that the women ascribed to. For example, one woman shared, “you shouldn’t have to ask more than one time like to be respected.” Women also gave less directive accounts of how they would enforce boundaries. For instance, one woman explained, “If I was Jenny, I would…not really continue into having sex or let Dan touch my breasts.”

When boundaries were perceived as violated, women indicated that they would verbally state their boundaries to Dan and confidently declare the need for them to be respected. As one woman put it, she would “lay down the law.” Another woman stated, “I would just say [to Dan]…he’s not respecting what I’m asking him not to do.” Other women’s responses were similar, with one noting that she would tell him, “I don’t appreciate the way that [you are] not respecting my wishes.” One woman described how her delineation of boundaries would even lead to a decision point: “[I would] say look, ‘you know, we can… continue to hang out; I’m having a good time with you but I need you to respect my boundaries.’”

Physical Assertion

A second sub-theme was physically assertive responses wherein the women in the current study identified physical behaviors that they would engage in as a response to the threatening situation. Physically assertive responses were described as refusing Dan with some type of active physical response, distancing themselves from Dan, physically removing themselves from the situation, and simply stopping the behaviors. In refusing Dan’s advances, many women provided explicit examples of behaviors that they would engage into stop him. For example, women indicated that they would “push him off,” “slap him,” “stop kissing him,” “not let him touch me,” or “make him stop doing what he was doing.” Others described how they would distance themselves from Dan and from the situation. For example, one woman said, “I would get up and like sit somewhere else, maybe turn on the lights, kinda put some space between us.” Another described she would “kinda back away and be more forceful in my answer.” For others, they responded by physically removing themselves from the situation by leaving. For instance, women stated, “I would just leave,” or “I would get up off the couch and leave the apartment.” Finally, some women responded by stopping all behavior, saying “I would just stop.”

Compliance/Acceptance

Compliance and acceptance to a hypothetical date rape scenario was a less frequent theme (i.e., the theme was variant), but it did occur on 35 occasions. Compliance and acceptance were identified in the transcripts as very similar behaviors and therefore were coded together; some participants described complying with the perceived threat without specifically mentioning acceptance, whereas other participants specifically mentioned acceptance and implied compliance. In both compliance and acceptance, participants responded to the threatening situation by opting to allow the situation to continue in accordance with Dan’s behavior. For example, in response to the situation one woman indicated that she would comply, “just keep going, whatever he [Dan] wants.” Another woman said, “I probably wouldn’t do anything [to stop Dan].” Other women expressed acceptance of Dan’s behavior on account of stereotypes associated with men and sex. One woman said, “I mean guys try things all the time; that’s just what they do.” Another woman expressed how aggressive tendencies are natural for men and should be expected and accepted; “men have that type of, you know, natural aggressive attitude, where they just want to go in and go for it.”

Conditional Decision Making

The third theme was conditional decision making, that is, hypothetical strategies that rely on gathering additional information. Conditional decision making was a variant response described by the women in our sample. Fifty-four of the 139 women in the sample indicated that they would wait to make a behavioral response to the threat because they needed additional information that would likely influence their response to the threat of acquaintance rape. This information could come in the form of waiting to see what Dan’s next actions were or wanting more information to take into consideration of personal emotions or Jenny’s emotions. Thus, two sub-themes were identified: conditional on Dan and conditional on Jenny/self. Notably, many of the women who described conditional decision making also described assertive responses.

Conditional on Dan/Relational

Many participants indicated that they would wait for further action from Dan before making a decision or doing anything. Thus, they would often provide conditional statements prior to indicating how they would behave in this scenario. In particular, women stated that their response would depend on whether he stopped his advances. A characteristic response in this domain included an element such as, “If he didn’t stop or try to calm down…” alluding to behavioral actions occurring at the point at which Dan did not stop his advances. One woman bluntly described the role conditionality plays in her response, “It’d all depend on if he stopped or not.” In this quotation, “It” refers to her behavioral response which was unspecified, suggesting that whatever she would do was based on Dan’s response.

Conditional on Jenny/Self

Fewer participants indicated that their behavioral response would depend on perception of the female’s emotions (i.e., Jenny’s) in the encounter. Some women responded as if they identified as the woman in the scenario, while others referred to a consideration of Jenny’s emotions. A characteristic response in this domain included an element such as, “if I felt really uneasy by the fact, you know if I’m her…and he keeps…” A consideration of attraction or liking for the perpetrator was a condition that women considered in the scenario. For example, one woman put it simply, “It also depends on how much I like him.” Another woman described how consideration of liking and a desire for a relationship could have influenced her decisions along the sequence of sexual advances, stating, “If I didn’t want or had no thoughts of pursuing anything with him, I probably wouldn’t have gotten that far.”

Avoidance

Avoidance, defined in this analysis as an implicit or explicit resistance to responding to the threat directly, was a variant response in the present study. Thirty-three of the 139 women in the sample gave responses that spoke to an avoidance of fully engaging the scenario or avoiding the escalating sexual advances in the scenario by attempting to redirect the perpetrator to another activity. Others noted that they would have avoided the threatening situation altogether by responding differently than what the women in the vignette would have done. Accordingly, three sub-themes were identified: de-identification with the victim role and deflection.

De-identification with a Victim Role

Twenty-eight participants gave responses that reflected a separation of oneself from Jenny’s behaviors and the hypothetical acquaintance rape scenario. The difficulty in relating to Jenny was seen with respect to identification with the emotional or physical aspects of the scenario and the woman’s experience. In some cases, participants framed their responses with what Jenny, and not themselves, should do in the scenario. For example, one participant stated, “I think he tried to feel her up three times already, so then she should for sure go home.” Another participant answered the question by explaining why Jenny might respond as she does rather than directly providing her own response, “She might feel, you know, embarrassed or a little ashamed of her body or something. She just [might] not be in a comfortable position.” In other cases, women instead referenced the differences between themselves and Jenny. For instance, one woman, in responding to escalation in physical intimacy when Jenny refused to allow Dan to touch her buttocks after repeatedly refusing him to touch her breasts, responded, “Well, first of all if I’m on the date with a guy and I’m making out with him I’m going to assume that he’s going to want to do that and it’s not going to bother me. So I don’t think I’d have that initial reaction.” In some cases, the women adopted a critical stance toward Jenny in their de-identification. For example, one woman explained,

So I felt, as disrespectful as he was being she was being very misleading…if I would have entered your apartment and asked you to kiss me I would already decided [I] want to have sex with you. I guess I don’t see if I’m going to be intimate to a certain extent, then all of sudden expect you to know that I want you to stop when everything I’m doing and saying is exhibiting behavior otherwise. So I never would have been there had I not you know wanted to do it with him, I never would have gone in. You know, ‘kiss me more but don’t touch me there,’ that just seems real unclear on her part, from my perspective.

In other cases, women had identified being able to avoid the hypothetical situation in the first place. For instance, one woman commented, “You mean they were dating for two days?…and she’s there in his apartment. If I were her…I won’t go with him to his private place.”

Deflection

Six participants indicated that their behavioral response would include redirection to another activity or subject of conversation, in an attempt to stop the escalation of sexual threat. In most cases, deflection followed a verbal response, either a verbal refusal or the perpetrator’s response of “no.” In all cases of deflection, the participant indicated that she would not necessarily physically remove herself from the situation but would change the activity. The purpose of deflection was to divert attention, as one woman described, “I’d try to be diplomatic. I wouldn’t get upset or mad. Just divert attention.” The attention diversion tactics suggested by women were that they spend some more time talking, watch another movie, or play a game.

Eight women deflected or diverted attention thru the use of lying or making up excuses in an attempt to avoid possible social and/or other perceived consequences. These instances varied from other examples of deflection in that they appeared to be undertaken to escape the situation. Women gave examples of general intent to make an excuse if they were in that situation. For instance, one woman indicated, “[I would] probably try to make up some excuse to leave.” In some cases, the excuses that would be given were specific. In these cases, excuses often referenced the late time of night. For instance, one woman indicated that she would tell Dan, “My parents texted me to come back home.” In all cases of excuse-making as specific kind of deflection, the response was an avoidant one and did not directly address the unwelcome sexual advances.

Expressions of Discomfort

The fifth theme was expressions of discomfort. In this theme, participants referred to feeling uncomfortable or experiencing the emotion of discomfort in their responses when taking on the role of Jenny or referencing the scenario. Expression of discomfort was a variant response endorsed by only 13 of the 139 women in our sample. Women discussed that they would feel uncomfortable if they were Jenny in this scenario due to perceived risky verbal and physical advances. For instance, one woman indicated, “I would probably feel uncomfortable because he was touching me in places I didn’t want him to.” Another woman responded, “When he started saying stuff about touching her—that just makes me uncomfortable.” Yet another woman indicated that Dan’s ignoring of boundaries made her uncomfortable:

You know if I’m her and I felt really uneasy by the fact that he keeps ignoring what I’m saying and it hasn’t progressed to that point where I feel comfortable to start letting him touch me in those places.

Other women referenced Jenny’s likely global discomfort in the situation, as one woman described, “It was a little too fast, she seemed uncomfortable with it.” One woman referenced the short duration of their relationship, stating that since they had only been on two dates “[the sexual action] would probably be a bit uncomfortable.” In many cases, women explained how their level of comfort would influence their behavioral responses in the scenario. For example, one woman indicated, “I wouldn’t do anything that I wasn’t comfortable with.” Another woman indicated that she would use verbal refusal if she felt uncomfortable in the situation, stating, “I would just say no if I was not comfortable with that situation.”

Allusion to Future Contact

The sixth, and final, theme was allusion to future contact, that is, responses that made a reference to future contact with Dan. This acknowledgement of future contact could include statements that they would see Dan again, would not see Dan again, or were unsure that they would see Dan again. Allusion to future contact was a variant theme, endorsed by 15 of the 139 participants. References were framed as what they would tell Dan and what they told the experimenter. This reference could include an indication of uncertainty as whether they would see Dan again, as stated by one woman, “I don’t know I would continue seeing him again.” Respondents also varied in their certainty about the prospect of a future relationship. For example, one woman responded, “I probably won’t go on another date with him,” while another woman was more ambiguous, “I would…hesitate to be around him more. And maybe not go on a third date, or something. Or fourth one.” One woman indicated that she might see him again in a different, potentially safer setting, commenting, “[I would] Maybe give him another chance on a date in public. Others referred to future contact, but were more directive with Dan about their wishes: “I would probably say let’s do this again just let’s not take it as fast.” Another woman referred to the need for more time or familiarity before going further in their relationship, stating, “I need to get to know you better or go on more dates.” Future contact could hypothetically occur soon after the scenario, as one woman described regarding her dialog with Dan: “It’s only been two dates. How about you call me this weekend maybe?” Only one participant indicated that she would tell him their dating relationship had no future. She stated, “[I would] tell him I didn’t want to see him anymore.”

Analysis of Themes by Threat Level Condition and Co-Occurrence

We also analyzed the frequency of each theme by each threat level condition and the degree to which participants described multiple themes in their responses (see Tables 1, 2, respectively). Notably, each theme is present in each condition, indicating some universality to the ideas expressed by those themes. However, there was some variation in the frequency of some themes by condition. Some themes, such as compliance/acceptance, and to a lesser extent conditional responding, decreased in frequency as threat became greater (i.e., in higher conditions). Other themes, including assertion and allusion to future contact, increased in frequency as threat became greater. Other themes such as conditional, avoidance, and expressions of discomfort do not seem to vary considerably by condition.

Table 2 Frequencies of co-occurring themes from each participant response
Full size table

The following response illustrates the co-occurrence of themes and the complex nature of responses described by participants. Underlines were used to label the different themes coded in this response.

figure a

As illustrated above, multiple themes and theme co-occurrence were common; most participants’ responses included more than one theme. Table 2 was created to examine patterns of co-occurrence within each participant’s response and how specific themes may have co-occurred. Given that most participants had at least one assertive statement, all other themes co-occurred with assertion. Only the avoidant theme was found to co-occur with just one other theme. One participant had responses with avoidance and compliance themes and another participant had co-occurring themes of avoidance and conditional. Several participants’ responses included multiple themes (e.g., 3+ themes). The most frequent occurrence (n = 9) was with the co-occurrence among assertion, compliance, and conditional themes. The next most frequent form of co-occurrence among more than two themes was among assertion, conditional, and expressions of discomfort. Other forms of multiple co-occurrences ranged in frequency from one participant to six participants describing responses that were coded with more than two themes.

Discussion

The aim of this study was to better understand the range of behavioral responses to the threat of acquaintance rape through a qualitative analysis. A variety of themes and behaviors were elicited by similar threat conditions including assertion, conditional responses, avoidance, compliance/acceptance, expressions of discomfort, and allusion to future contact. Certain themes, such as assertion, did reflect traditionally recognized behavioral responses to threat. Many participants responded assertively, describing both physical and verbal ways to escape the escalating threat. Verbal assertion took forms of both refusal and expressing clear boundaries for acceptable behavior if the interaction were to continue. Although assertion was the most commonly expressed response, other themes, notably conditional and compliance/acceptance, reflected behavioral responses rarely, if ever, assessed in current research on sexual assault, behavioral response, or self-defense.

A smaller but significant number of participants expressed the desire to let the perceived threat continue, while others noted the need for more information about what would happen next before knowing how they might respond, thereby creating a response that was conditional on the instigator of threat. Some participants also chose to communicate their responses through expressions of discomfort. It is of note that expressions of discomfort was a relatively less common theme; some research has indicated that emotional reactions can be indicators of threat processing (Bart & O’Brien, 1984). Many described some form of avoidance by physically or emotionally distancing themselves from the instigator of the threat. This sometimes took the form of enacting individual agency by stating a reason to leave or do another activity. Others expressed avoidance by redirecting attention to another activity as a means of attempting to de-escalate the threat. This range of themes, some of which seem to contradict one another, highlights the complicated nature of the task—participants may or may not have viewed the stimuli as threats depending on their own personal experiences, beliefs (including potential internalization of rape culture), et cetera. Overall, the diversity of themes suggests that in this sample, participants shaped their responses to threat of date rape in more than just opposing dimensions of physical or non-physical and forceful or non-forceful responses.

A number of responses included non-assertive themes or behavior. These themes exemplify the variety in behavioral responses that is not currently well characterized in the literature or typically measured in research on the experience of rape or self-defense. In the case of compliance/acceptance, some women felt that the kind of aggressive behavior displayed in the vignette was to be expected. This theme may reflect social expectations that the male sex drive is unrelenting and uncontrollable or that men have an implicit right to access women’s bodies as they wish (Flood, 2003; Vitellone, 2000). It may also be reflective of an internalization of rape culture; recent research has demonstrated a link between rape myth acceptance and tolerance for greater risk of sexual assault in a vignette task (Yeater, Treat, Viken, & McFall, 2010).

Participants who described avoidant responses may also represent a unique group. An important, though subtle, distinction between avoidant responses and compliance/acceptance is that women who described avoidant responses used their individual agency to indirectly respond to the hypothetical threat condition directly (e.g., they diverted attention or suggested an alternate activity). They did, however, construct a response, but utilized less direct strategies, as illustrated by the avoidant sub-themes of de-identification from victim role, deflection, and diversion. There is, however, a potential value in some of the avoidant behaviors found in the current study. Excuse-making, although avoiding the threat directly, demonstrates a form of coping that might (or may not) be protective. Indirect but agentic responses, then, may have been perceived by participants as protective or as providing a way out of the threatening situation that they thought could protect the individual for such negative consequences, particularly social consequences, thereby providing a form of creative agency in responding to threat. Yet, some avoidant responses had a similar character to the diplomatic response characterized in prior literature that is associated with a history of sexual assault and consistent with gender differences in socialization (Macy et al., 2007). Avoidance was a relatively less common theme as threat increased indicating that this theme may be less common when threat cues are more easily detected or more severe. However, utilizing behavior consistent with the avoidant theme may increase risk; in epidemiological research, non-assertive behavioral responses have been associated with increased risk for rape (Clay-Warner, 2002). More research is needed to identify how these kinds of response behaviors may be related to risk.

For women who described conditional responses, many also described assertive responses. At face value, these two themes may appear to be in conflict, but their overlap suggests that the conditional theme is frequently followed by the assertive theme suggesting other factors that may mediate the process between these two behaviors. Many of the conditional responses indicated that some women would wait to assess whether the man in the scenario continued to act aggressively and then only if the man made further advances would they take assertive action. This type of “wait and see” responding has not been recognized well in prior research, although research on the process of appraisals indicates that this likely occurs with some frequency (Norris, Nurius, & Graham, 1999). This could also reflect the relational nature of date rape threats—these threats are inherently couched in a social context—participants who described conditional responses may have been seeking further information, particularly relational information, to shape their response. Alternatively, for participants who described conditional responses, the process of appraisal may be different as they may be sensitive to specific types of cues (e.g., expressions of sexual or romantic interest; see examples such as Byers, Giles, & Price, 1987) that disrupt threat processing. Or they may have greater difficulty estimating their own risk, due to the common positivity bias, tending to underestimate risk (Norris et al., 1999). More work clearly needs to be done to examine potential risk reduction strategies on date rape, including work with men to reduce sexual violence and date rape.

Alternatively, conditional responding may reflect varying levels of wanting and consent that are dimensional but may be in opposition to one another. In other words, sexual wanting and sexual consent are separate dimensional facets that may conflict. For example, a woman may experience sexual wanting but for various reasons, such as feeling social pressure to abstain, not consent (Peterson & Muehlenhard, 2007). These dimensions may conflict in ways that may impact coping with sexual assault. Peterson and Muehlenhard found that women who rated nonconsensual experiences higher in wanting were less likely to label these experiences as sexual assault, in spite of the lack of consent. Other research has found that participants who did not acknowledge their assault experience had slower risk recognition (Marx & Soler-Baillo, 2005). The conditional theme may represent a conflict between wanting and consenting, a conflict which could impact risk recognition and behavioral response to threat. This theme represents a type of behavioral response that is not currently recognized in the literature and worthy of future study to better characterize this response style.

Examination of themes by condition demonstrated that all themes were present in all conditions—indicating the strength of the qualitative coding process and the broad applicability of the themes identified. Variation of the themes by condition appeared to be appropriately contextual to the strength of the threat. Analysis of co-occurrence among themes indicates that participants frequently described responses that included multiple behavioral responses to threat. Furthermore, participants frequently described assertive and non-assertive behaviors within the same response. At face value, this would appear contradictory but likely reflects the complex psychological process of responding to the threat of date rape wherein multiple concerns are weighed and juggled against another.

Responding in ways that are less stereotypically assertive is rarely assessed in research on threat response or self-defense, which often assumes that women will find the situation threatening enough to act in a physically assertive manner. Future research should continue to investigate the possible range of behavioral responses in order to better characterize the variety of possible responses to perceived date rape and to empirically research whether or not these responses lead to decreased or increased risk for sexual assault. This information would be important for providing data to women; it would also help normalize sexual assault survivors’ experiences in psychotherapy and for law enforcement and the judicial system to better understand the phenomena of sexual assault and create more sensitive practices and policies. Future research should also investigate issues related to specific behavioral responses within the cognitive-ecological model as it is likely that specific background (prior abuse), intrapersonal (personal beliefs), and interpersonal (type/length of relationship) factors may shape behavioral response styles (Nurius & Norris, 1995). Future research can also bolster gender-transformative work with men to reshape gendered power relations and shift the specific inequitable gendered attitudes and behaviors that foster all forms of sexual violence (Dworkin, Treves-Kagan, & Lippman, 2013; Pulerwitz & Barker, 2008). Future research should also investigate how men perceive behavioral responses and how men can be intervened with to be understood and respect cues of consent. Research examining this area from the perspective of men who may aggress is extremely limited but has great potential.

Limitations

The results of this study are limited by the use of a contextually restricted, brief stimulus to a hypothetical situation that likely does not elicit all possible response behaviors. Indeed, it is reasonable that some participants (particularly in condition 1) may not have perceived the stimulus as threatening, depending on their own values and beliefs. This study is also limited by the use of a hypothetical stimulus. While research has shown good correspondence between hypothetical response behaviors and responses in real life, a laboratory stimulus cannot fully capture the relational nature of the threat of date rape (Turchik et al., 2007). To wit, as seen in some of the quote above, some participant’s responses were laden with narratives of victim blaming and criticism of the woman portrayed in the vignette. Given the analog scenario, it may be difficult for participants themselves to know what their response might be as they juggle competing internal demands, including demands to position themselves as correct against the hypothetical woman who is “incorrect” by being in the threatening situation itself.

Additionally, the woman depicted in the experimental story did model assertive behaviors that may have influenced participants to also describe these types of behaviors. The scenario portrayed, an acquaintance date rape, corresponds only generally to the sexual assault threats many college women face. Given the specific scenario to which participants hypothetically responded, the results of this study are most relevant to the threat of date and perhaps acquaintance rape for high school or college-aged women; however, this type of threat is exceedingly common in this high-risk group. While the auditing process introduced a new perspective designed to strengthen coding, any one perspective is inherently a partial one. Not all possible perspectives based on gender or sexual orientation identity were able to be represented in the coding process. The process of deciding upon and executing a behavioral response is complex, yet this study was only able to explore participants’ initial responses. Although this research is important for learning about the experience of sexual assault in order to help survivors and reduce risk, research on those who engage in sexual aggression is critical to reducing rape.

Conclusions

Assertive behavioral responses to the threat of acquaintance rape were easily described by most of the sample. Themes such as expressions of discomfort and allusion to future contact indicate that even when being asked to focus on a specific, contextually limited threat, women internally juggle multiple concerns, especially concerns about social relationships. Given that nearly all participants described some type of assertion, this response style has intuitive appeal but likely comes into conflict with social pressures to privilege male sexual desire. Future research should examine the psychological factors that influence behavioral response styles and how assertive behavior is perceived by men who enact sexual aggression.

However, some of the themes identified in this study, such as compliant or conditional behavioral responses, were fairly frequent and seemed practical and effective to participants. Yet, these response styles are rarely assessed in studies examining behavioral response or self-defense behavior. The exclusive focus on overt assertive behavior may inadvertently encourage those who were unable to act in this manner to blame themselves. As seen in this study, it is likely that a great variety of responses are common among survivors; greater information is needed about this to inform psychotherapy with survivors as well as law enforcement professionals, intervention researchers, and policy advocates.