|Title:||Police officers' perceptions of rape.|
|Abstract:||Examines how police officers in the United States define rape. Reformation of rape laws in the United States; How this reformation has affected police officers and their definition of rape; Analysis of definitions of rape that were given by police officers.|
|Full Text Word Count:||7959|
|Database:||Academic Search Premier|
This study examined how police officers (N = 91) define rape. In the past decade, most states have dramatically reformed their rape laws, shifting the emphasis from the behavior of the victim to that of the assailant. This study provides an exploratory picture of how such reforms have affected police by examining the degree to which there was consistency between officers' personal definitions of rape and state law. Officers were asked to define rape/sexual assault in their own words. These definitions were content analyzed and hierarchical cluster analysis was performed. Three clusters of definitions emerged Nineteen percent of the sample described many of the reformed legal factors when defining rape, such as the use or threat of force (Force Definition of Rape). Thirty-one percent focused primarily on penetration and consent (Consent Definition of Rape). Fifty-one percent of the officers provided definitions that mixed old legal definitions with some victim blaming views (Mixed Definition of Rape).
Since the mid 1970s many states have dramatically reformed their rape laws (Berger, Searles, & Newman, 1988; Fischer, 1989; Searles & Berger, 1987; Spohn & Homey, 1992). Rape victim advocates criticized early laws because the focus had been on the behavior of the victim, rather than the assailant (Fischer, 1989). Demonstrating that the victim did not consent had been the primary mandate and challenge for prosecutors, rather than proving that the assailant used force or the threat of force to sexually violate. Many activists and researchers at this time argued that these laws institutionalized victim blaming (Brownmiller, 1975; Schechter, 1982). To address these problems, most states redefined the crime of rape to place the legal focus on the behavior of the assailant--was there penetration? was there force? was there the threat of force? was there coercion? Evaluation of these reforms has concentrated at the prosecutorial level of the criminal justice system, documenting how cases are prosecuted under these revised laws (e.g., Caringella-MacDonald, 1985; Spohn & Homey, 1992).
Absent from this literature, however, is an examination of how these legislative changes affected law enforcement. Police officers are the entry level to the criminal justice system, and in some jurisdictions, they hold considerable power in determining whether a reported rape will advance to the prosecutors or be dropped (Kerstetter, 1990). Extensive legislative reform may not be as effective if it does not permeate all levels of the system it was intended to change. A first step in examining how these reforms affected police is to explore how officers understand these changes: Is there is congruence between what they consider to be rape and state law? The extent to which there is mismatch, extralegal factors, some of which may be victim blaming beliefs, could be influencing this process. This would suggest the need for further training for police officers. In the current study, officers in a state that has a revised rape law were asked to define rape/sexual assault in their own words. The consistency between their personal definitions and the state legal definition was examined. A discussion of rape law and previous research on police offers' beliefs about rape is first presented.
The legal definition of rape can be examined at two levels. First, some legal theorists have adopted a macrolevel of analysis, focusing on the culture of the legal system (e.g., Estrich, 1987). The attention is not on individual laws and cases, but on their collective meaning and their implicit and explicit messages. From this perspective, rape is defined by broader ideas and concepts rather than specific behaviors. A second approach refocuses the level of analysis at the individual state laws and delineates specific illegal behaviors.
Rape and Culture of the Legal System
Estrich (1987) argued that the legal system defines two types of rape: "real" rape and "simple" rape. "Real" rapes are committed by someone unknown to the victim (i.e., stranger rape), often involve the use of a weapon, and result in some physical injuries to the victim. "Simple" rapes are committed by someone known to the victim (e.g., acquaintance rape). In practice, however, it is primarily the "real" rapes that receive attention by the criminal justice system. Based on her reading of case law, Estrich cited four elements implicit in legal doctrine that work together to maintain a system that is suspicious of (at best) or erases (at worst) nonstranger rapes. First, prior relationship cases are viewed as "private" disputes, which removes them from the business of the public prosecution system. Second, these cases are seen as less serious and the defendants less blameworthy because they often involve a "claim of right," whereas attacks by strangers do not. The "claim of right" philosophy stipulates that if a woman has consented to sex in the past, then the man has a continuing right to sexual satisfaction from her. Third, prior relationship cases are thought to involve contributory fault by the victim, but stranger offenses do not. Simple rapes are "victim precipitated"-- she agreed to dates, drinks, rides, or failed to strongly resist sexual overtures. These acts precipitate sex and "she should have known better." Finally, attacks by nonstrangers are thought to be less terrifying to victims and therefore less serious.
Smart (1989) extended Estrich's arguments and emphasized that the "sex" in nonstranger rape adds to this culture of suspicion. Violence is not seen as violence if it is sexualized, and violence is exonerated if pleasure could be achieved. If there has been previous contact between the victim and assailant, then sexual pleasure for both parties from the assault is an even more likely conclusion and the rape is excused (Smart, 1989). Violence becomes sexy, and sexiness is not criminal. The work of Estrich (1987) and Smart suggests that the culture of the legal system defines rape as a crime between strangers. Sexual violence between nonstrangers is not enough of a crime.
The Specifics of Rape Law
In the last 10 to 15 years, most states have dramatically reformed their rape laws. The traditional legal definition of rape--unlawful sexual intercourse with a female without her consent--focuses the attention and burden of the trial on the victim (Estrich, 1987; Fischer, 1989; Spohn & Homey, 1992). The rapist is innocent until proven guilty, but many activists were concerned that with this focus on the victim in the law, the behaviors of the assailant were taking a very distant back seat. Consequently, it was the victim's state of mind and behavior, rather than the defendant's, that was subject to intense scrutiny by the police, the prosecutors, the defense lawyer, the jury, and the community as a whole (Fischer, 1989). Under common law, courts required that the prosecution show, in addition to the statutory elements of the crime, that the victim resisted "to the utmost" or failed to resist because of extreme fear, that the victim made a prompt complaint, and that the victim was "chaste" (DuBois, 1988; Fischer, 1989). Old rape laws also included evidentiary requirements not imposed in other criminal offenses. Until recently, the state could only prove rape where some piece of independent evidence corroborated the victim's story (Estrich, 1987; Fischer, 1989; Spohn & Homey, 1992). The implied elements of the crime of rape, together with the special evidentiary requirements, made effective rape prosecution extremely difficult.
In response to growing criticisms, many state legislatures from the late 1970s to mid 1980s reformed rape statutes by redefining the crime of rape (Berger et al., 1988; Fischer, 1989; Spohn & Homey, 1992). Many states replaced the offense of rape with the offense "criminal sexual assault" or "criminal sexual conduct." Whereas each state law varies in specific language, three changes were consistent across most laws (Fischer, 1989). First, the focus on consent (the state of mind of the victim) was shifted to an emphasis on the use of force or the threat of force (the behavior of the assailant). For example, Michigan law served as model statute for other states and defined criminal sexual conduct as an act of sexual penetration by the use of force or the threat of force (Michigan Criminal Sexual Conduct Law, 1974). Second, the meaning of consent and its role in the crime was also modified. In some jurisdictions, consent can only be raised as an affirmative defense, otherwise it is not an issue at trial (Fischer, 1989). Many evidentiary requirements of consent were also changed. Consent cannot be established by showing that the defendant believed that the victim consented. Lack of verbal or physical resistance by the victim also is not consent. This evidentiary shift was intended to change the treatment of victims in the courtroom by eliminating questions about her behavior and state of mind. Third, most states passed victim-shield laws that limit or prevent the defense from cross-examining victims about their sexual history (Berger et al., 1988; Fischer, 1989; Searles & Berger, 1987; Spohn & Homey, 1992). Overall, these reforms were intended to focus attention on the behavior of the assailant rather than on the victim.
Estrich's proposition that rape is defined as a crime between strangers was not ostensibly codified in these reformed statutes. Digging deeper into the evidentiary requirements, however, indicates that stranger rape and nonstranger rape have slightly different rules of procedure. Whereas almost all states (42) make no statutory distinctions between stranger and nonstranger rape, all 42 of these states add evidentiary guidelines that limit victim-shield laws and allow testimony on previous sexual contact between the victim and assailant (Berger et al., 1988; Searles & Berger, 1987; Spohn & Homey, 1992).(n1) For example, the case of Michigan v. Lucas (1991) set a standard for other states. Michigan has one of the most progressive sexual assault laws, but the court decided that evidence of the victim's sexual contact with the assailant is almost always an exception to the shield laws. The admissibility of this evidence is crucial. These guidelines can create difficulties for nonstranger rape victims because if a woman previously consented to any connection with a man, a presumption arises that she subsequently consented to sexual contact during the incident in question (Balos & Fellows, 1991; Estrich, 1987; MacKinnon, 1987; Smart, 1989). Feminist legal theorists have observed that when indices of closeness (e.g., dating) are taken as evidence of the victim's behavior during the offense, many common types of coercive sexual encounters go unregulated by the law (Estrich, 1987; MacKinnon, 1983, 1987; Smart, 1989). Consequently, MacKinnon (1987) argued that rapes are not reported, prosecuted, or sanctioned based on the force that was used, nor is the criterion how coercive the rape was or how violated the woman feels. Instead, rape prosecution is often based on how intimate the woman is with the assailant. Despite the promise of rape legislative reform to place the act itself on trial, the behavior of the victim can remain an issue, particularly in cases of nonstranger rape.
Most revised rape laws defined sexual assault as an act of penetration by the use of force or the threat of force. There may be other factors, however, that influence how police define rape. Previous research on officers' perceptions of rape suggests that their beliefs are indeed shaped by many factors, not all of which are legally relevant. In one study, police stated that they thought real rapes included the presence of injuries, complete consistency in the victim's story, willingness of the victim to take a lie detector test, and chaste behavior of the victim (i.e., she did not engage in premarital or extra-marital affairs, and she did not have any previous social contact with the assailant; Feldman-Summers & Palmer, 1980). By contrast, LeDoux and Hazelwood (1985) reported that police were more sympathetic to rape victims than previous research and popular belief suggested. Specifically, they reported that police were not "insensitive to the plight of rape victims" because, as a group, the officers tended to disagree with statements that suggested the raped woman is not truly a victim (p. 215). Officers were, however, suspicious of victims who had had prior sex with a man or who "provoked" rape through appearance and behavior. This aspect of their findings is quite consistent with Feldman-Summers and Palmer's (1980) results that indicated that police may not construe nonstranger assaults as rape. Another study by Krahe (1991) found similar results. She asked East German police officers to define a "typical" rape and a "dubious" or hard-to-believe rape. The definition of the "typical" rape was an assault by a stranger (often "crazy") that occurred at night, outdoors, and resulted in some minor physical injuries to the woman. The "dubious" rape occurred between two people who knew each other (the man was "normal"), happened at either the man's or woman's house when the woman was drunk, and resulted in no physical injuries to the woman. These findings together suggest that characteristics of the victim help define what officers consider to be rape.
The literature to date suggests that police officers may use a variety of "guidelines" to determine what is rape, some of which are legally relevant, others are not. Exploring how police who work in a state with a reformed rape law understand rape may provide some insight into how these reforms have trickled down to the entry level of the criminal justice system. A combination of qualitative and quantitative methods were utilized in this study. Officers were asked to define rape/sexual assault in their own words, and these narratives were content analyzed for the themes that were mentioned in the definitions. In keeping with the exploratory nature of this study, cluster analysis was then used to create typologies of definitions. These typologies were compared to the legal definition of sexual assault in the state where this research was conducted (see the appendix for the state law). The validity of this cluster solution was evaluated by comparing the police in each cluster on demographic and work history variables as well as their attitudes toward women and their attitudes toward interpersonal violence. These variables were selected because they could provide useful information for training programs for police officers on rape law. Demographic information and work history could help identify who could be targeted for training, and the attitude variables may suggest other underlying beliefs that should be discussed in such interventions.
Two police departments agreed to participate in this study (N = 91). The first group of officers was from a medium-sized city in the Midwest (n = 53) and the second was from a police department for a large, Midwestern university in a different city (n = 38). The officers from the two departments did not differ significantly on the variables studied, so they were combined and treated as one sample.
Relevant demographic information regarding the research participants is presented in Table 1. The majority of the participants were male (88%) and 91% of the sample was White. Ages of the police officers ranged from 24 to 57, with a mean age of 39 years. Many of the officers were married (60%), and most had children (79%). Almost three quarters (74%) had obtained either an associate's or bachelor's degree.
The police had been officers for an average of 15.95 years, with a range of 1.8 years to 30 years. They had been with their current department for an average of 14.17 years, with a range of 7 months to 29.6 years. More than half (56%) of the participating officers were street patrol officers, 30% were command officers, and 11% were investigators. More than half (60%) of the officers had had special training on the topic of rape.
The investigators attended briefing meetings at the police departments to describe the project to the officers and then distribute the questionnaires. Every available sworn officer at both departments was given a questionnaire to be completed anonymously. The officers were given time on duty, in private offices, to complete the questionnaires. They spent approximately 1 hour completing the survey. The investigators returned one week later to pick up returned questionnaires. The response rates in this study were quite high as 87% of the officers from the first department and 86% of the officers from the second department completed the questionnaire. These rates were higher than those obtained in previous research (e.g., Feldman-Summers & Palmer, 1980, obtained a 24% response rate).
The police officers were asked to write, in their own words, how they define rape/sexual assault. The specific wording of this item was "As you know, it's the legislators that make the laws and decide how to define crimes and what punishments will be. But, you are actually in the community, dealing with victims and criminals. Based on your work as a police officer, how do you define rape/sexual assault?" It is therefore likely that these definitions reflect both personal beliefs and statutory guidelines. A content analysis was conducted on their definitions coding the themes expressed in the participants' answers. An initial list of themes was developed a priori, modified from previous research (Krahe, 1991). The narratives were then examined by a different research assistant to see if the answers could be accurately coded by such themes. Based on this exploration, two other themes were added (see Table 2 for a list of the final seven themes coded).
A coding manual was developed that defined each of these seven themes. Two other raters then used this manual to code the officers' definitions of rape. Both raters coded all (N = 91) definitions. Due to the detailed and specific nature of the coding scheme, the raters were likely somewhat aware of the purpose of the study. The themes were dichotomously coded "yes" (theme was mentioned in the definition) or "no" (theme was not mentioned in the definition). Ten questionnaires were randomly selected and coded. Initial intercoder agreement was 75%. This percentage agreement index was corrected for chance agreement with the K statistic (Cohen, 1960). Initial K was .50. The coding manual was revised and another 10 questionnaires were then randomly selected and coded. Intercoder agreement improved to 90% (K = .80). Once acceptable intercoder reliability had been established, all of the text was coded and a final assessment of intercoder agreement was calculated. Intercoder reliability was assessed after every 30 questionnaires coded (i.e., reliability was assessed four times). The average intercoder agreement was 96% with a range of 90% to 98%. Average K was .89.
Assessing the Validity of the Cluster Solution
The validity of the cluster solution was checked by contrasting the officers in each cluster on their demographic characteristics, their work history, and their general attitudes toward women and interpersonal violence. First, the clusters were compared on four demographic variables: gender, race, education, and age.
Second, the officers' work histories were contrasted across each cluster: job title, whether they had ever had special training on rape, and their experience working with rape cases, which was measured with the Experience With Rape Scale. This scale was developed by the primary researcher, and contained four items that asked if they have ever had professional contact with a stranger/nonstranger rape victim, and if they had ever professionally interviewed a stranger/nonstranger rape victim. The officers' responses were dichotomously coded and averaged to create a scale score so that higher scores reflect more experience with rape cases (M = .59, SD = .30, variance = .09). Corrected item-total correlations ranged from .37 to .42, and alpha for this scale was .60.
Third, the officers in each cluster were compared on their general attitudes toward women and their general attitudes toward interpersonal violence. A modified version of Spence, Helmreich, and Stapp's (1973) Revised Attitudes Toward Women Scale (ATW) was used to assess sex role stereotyping. The Revised ATW consisted of 25 items that assessed opinions toward the rights and roles of women in society (e.g., "Women should not expect the same freedom as men"). Respondents rated their agreement with each item on a five-point scale (1 = strongly disagree; 5 = strongly agree). Five items were dropped a priori due to ambiguous wording, and an additional 10 were dropped because of low factor loading values reported by Spence and colleagues. The remaining 10 items were administered to the participants in this study. Participants' ratings for each item were averaged to create the final scale: Some items were reverse scored so that higher scores reflect more favorable attitudes toward women (M = 4. 18, SD = .67, variance = .45). Five items were dropped due to low corrected item-total correlations. These five items did not form a separate scale. The final five-item scale had corrected item-total correlations ranging from .56 to .67, with alpha = .82.
Attitudes toward interpersonal violence were measured with Burt's (1980) scale of Adversarial Sexual Beliefs (ASB) and her scale of Acceptance of Interpersonal Violence (AIV). The ABS consisted of nine statements about the adversarialness of male-female intimate relationships (e.g., "A man must show who is boss"). Participants rated their agreement with each item on a five-point scale (1 = strongly disagree; 5 = strongly agree). Based on the item-total correlations reported by Burt, four items were dropped a priori and a five-item version was used in this study. The AIV has an identical format and consisted of six attitude statements about violence in intimate relationships (e.g., "Being roughed up is sexually stimulating to women."). Again, based on the published psychometric properties, two items were dropped a priori and a four-item scale was administered to the participants. The modified ASB and AIV were highly correlated (r = .80, p < .01), so the items were combined into one scale. Reliability analysis was conducted on the combined nine-item scale. Participants' ratings for each item were averaged for this combined scale and higher scores indicated more progressive attitudes toward the acceptability of violence in intimate relationships (i.e., ideas that dating is not fundamentally exploitive and that violence is not acceptable in dating situations) (M = 1.74, SD = .54, variance = .29). One item from the ASB was dropped due to lack of variability. Items on this scale had corrected item-total correlations ranging from .59 to .79, with an a of .89.
The officers' definitions of rape/sexual assault were content analyzed as described previously. To explore similarities and differences within the officers' definitions, cluster analytic procedures were used. Rapkin and Luke (1993) noted that cluster analysis forms homogeneous groups of entities based on similarities and differences among the entities with respect to measured characteristics. The goal of these analyses was to form groups that maximize intergroup dissimilarities and intragroup similarities. In the current study, the seven themes coded in the police definitions were used to create these groups (clusters). Thus a cluster represents a group of police officers who defined rape/sexual assault in a similar manner.
A binary measure of proximity (Phi) was used to estimate similarities in these data (in the content analysis, themes were coded "yes"--theme was mentioned in definition--or "no"--theme was not mentioned in the definition). This proximity matrix was then used with Ward's (1963) method of hierarchical cluster analysis, which maximizes cluster homogeneity by minimizing the within-cluster error sum of squares (Rapkin & Luke, 1993). Whereas there is no generally accepted rule for choosing the correct number of clusters, a combination of empirical procedures and overall interpretability is used (Anderberg, 1973; Luke, Rappaport, & Seidman, 1991; Rapkin & Luke, 1993; Romesburg, 1984). A three-cluster solution was chosen for several reasons. First, the sizes of the clusters were fairly similar, which aids interpretation. Second, the partitions containing the four- and five-cluster solution were created by forming a cluster based on a single location that split-off from one of the three larger clusters. These single split-offs appeared after the three-cluster solution. Finally, these clusters suggested a conceptually meaningful organization of the officers' definitions. The stability of this three-cluster solution was examined using a split-half test (Luke et al.,1991). Half of the officers' data were randomly selected, and the cluster procedures described above were employed. The same three-cluster solution emerged from this split-half test, suggesting that the clusters represented a stable organization of the data. Table 3 presents the percentage of times officers in each cluster mentioned the seven themes.
The first cluster of officers (N = 27, 31% of the sample) defined rape almost exclusively by penetration and a lack of consent. Whereas penetration is a key element that legally defines rape, lack of consent is not an explicit legal element in this state. Lack of consent is implicit in the crime of rape, so it is not surprising many of the officers attuned to this aspect. Only 30% of the time did these officers mention the use of force as a defining feature, which is a main statutory focus in this state. Hence, this cluster was named "Consent Definition of Rape"-a partially reformed legal definition of rape. For example, some officers in this cluster wrote:
"If she did not give consent to sex, it was rape."
"Rape is unwanted sex."
"Rape is an act of penetration without a woman's consent and is done through the use of force."
"It all comes down to her word against his on consent--if she gave it, OK, if she didn't, it's rape. It's hard to figure out sometimes if she gave consent."
"Rape is anything that happens after a woman says `no.'"
"Rape is to make someone have sex or touch another against their own will."
"One sided sex against the will of the victim."
"I guess guys now need written consent to have sex with a girl. Rape is sex without consent, but it's not always clear what is and isn't consent."
By contrast, officers in the second cluster (N = 17, 19% of the sample) included many of the elements that currently legally define rape. Most listed penetration (84%), almost all (95%) mentioned the use of force, and more than half (58%) discussed the threat of force. They down-played the role of consent considerably (only 21% listed this theme) compared to the other clusters (100% and 51%). More than one third (37%) of these officers noted that assailants often coerce their victims, which is mentioned in this state's law. A few officers (11%) also mentioned that they thought rape is an act of power, either against the victim individually or against women in general. This cluster was named "Force Definition of Rape"--a reformed legal definition of rape. For example, some officers in this cluster wrote:
"Rape is violent intrusion into an individual's physical and emotional existence. One of the most degrading, insultive, and hurtful actions upon another individual."
"Rape is a crime of violence. It is the forcible penetration of the sexual organs of one person by another. In my opinion, this is to achieve dominance and violation of the victim, and not sexual gratification."
"Rape is forced or coerced sex activities whether under influence or begged until he/she gives in. Or when one person promises something. Overall, rape starts when `not' is said."
"Rape is a psychological statement of power/control by the rapist over his/her victim, by violating the most intimate and private part of the victim's being."
"Rape is any forced, coerced, or unconsented sexual contact. Sex contact is defined as an oral, anal, or vaginal contact with the sex organs of either the same or different gender."
"Rape is forced or coerced sex acts whether under the influence of alcohol or begged until he/she gives in. You don't have to clobber someone to make it rape."
"Rape is being forcibly violated either without consent or proper consciousness- the person may seem willing, but may be unable to pronounce him or herself carefully."
"Rape is a violent sexual act often committed to obtain a sense of superiority or control."
The third cluster of officers (N = 43, 50% of the sample) focused on the issues of penetration (42%), rape as an act of sexual gratification (40%), and consent (51%). Of these commonly cited elements, only penetration is legally relevant in this state. Eleven percent mentioned the use of force, and 2% listed threat of force. Whereas some officers did write that rape is an act of power (4%), more believed that rape is committed to satisfy the sexual drives of the assailant (40%). The officers who did mention that rape is an act of power also heavily emphasized the role of consent, which differentiated them from the police in the second cluster. Such definitions in this third cluster stated that rape is committed for power, but consent and behavior of the victims are important factors to consider as well. Officers in the second cluster did not add such qualifiers to their explanations of the role of power (see previous example definitions). With this focus on old and reformed legal elements, as well as rape myths, this cluster was named "Mixed Definition of Rape." For example:
"From a male's point of view, all a female has to do is say she was raped and the male has to defend himself. It does not matter that the female consented and changed her mind afterward."
"I do not believe that consensual sex that is later regretted and reported as nonconsensual sex is rape. The `man in the bushes' is a much bigger threat to the community than the date that got carried away."
"What if someone has had sex 20 or 30 times over a 3-month period. Then one night they say "no." Should this be rape? I don't think so."
"If a young lady dates a man and they go home together--she allows the young man to get into bed with her and just before he has an orgasm she shouts "NO." How then could you prosecute a young man for rape? I believe the definition of rape should apply to any circumstances as long as the evidence and testimony can prove it beyond a reasonable doubt."
"Sometimes a guy can't stop himself He gets egged on by the girl. Rape must involve force--and that's really rare."
"I can tell you what rape isn't: an individual going on a date and getting into foreplay and then finally consenting after being verbally persuaded."
"Rape is just rough sex that a girl changed her mind about later on. Technically, rape is a sex act done by the use of force, but so many girls are into being forced, that you can't tell the difference and you wouldn't want to convict an innocent guy"
"Men taking what women really want at that moment but decide they didn't the next morning when they sober up."
Validity of the Cluster Solution
To test the validity of this three-cluster solution, a series of two-way Chi[sup 2] analyses and one-way ANOVAs were performed (see Table 4). No significant relationship existed between gender, race, age, education, and cluster membership, suggesting that these differences were not due to demographic characteristics. Yet, whether the participant was a street patrol officer, command officer, or investigator was related to cluster membership. The participants in the second cluster (Force Definition of Rape), who focused on the reformed legal elements, were more likely to be command officers and investigators, rather than patrol officers: Chi[sup 2] (4, N = 91) = 21.44, p < .01. The officers in the second cluster were also more likely to have had special training on rape: Chi[sup 2] (2, N = 91) = 6.43, p < .05. Additionally, these officers had more professional experience working with rape victims: F (2, 88) = 6.27, p < .01; Scheffe post hoc (p <.05): Consent Cluster M = .65, SD = .34; Force Clusters M = .81, SD = . 11; Mixed Cluster M = .63, SD = .53.
The officers in each cluster also differed on their attitudes toward women and their attitudes toward interpersonal violence. Officers in the second cluster (Force Definition of Rape) had the most favorable attitudes toward women: F(2, 88) = 4.63,p < .05; Scheffe post hoc (p < .05): Consent Cluster M = 3.09, SD = .83; Force Clusters M = 4.57, SD = .64; Mixed Cluster M = 2.98, SD = 1.21. They were also the least accepting of interpersonal violence: F (2, 88) = 7.23, p < .01; Scheffe post hoc (p < .05): Consent Cluster M = 3.10, SD = .37; Force Clusters M = 4.81, SD = .22; Mixed Cluster M = 3.01, SD = .98.
Previous research has substantiated that police consider many factors when deciding what is rape, and this study explicitly examined how officers in one state with a reformed rape law view sexual assault in light of such legislative changes. In this research, the consistency between personal definitions of rape and state law was examined. The officers did not hold uniform perceptions as three typologies of definitions emerged, each focusing on different legal and extralegal factors. In only one of these three clusters was there strong consistency between officers' beliefs and statutory elements.
Although this sample was rather homogeneous demographically, the officers held widely differing beliefs as to what constitutes rape. Some of the officers defined rape very consistently with state law, even describing some of the more unique features of this law, such as the role of coercion (Force Definition of Rape). Yet this cluster of officers was the smallest--19% of the sample attuned to these reformed legal characteristics. Another 31% of the sample gave definitions that concentrated on penetration and consent (Consent Definition of Rape). In the state where this study was conducted, the role of consent has been down-played for more focus on the use of force. Interestingly, a distinction between stranger and nonstranger rape was absent from most of the definitions in these two clusters. Most officers in these two groups did not state or imply that nonstranger assaults are not rape. The other half of the sample wrote definitions that focused on penetration, consent, and the need for sex (Mixed Definition of Rape). Some officers did mention force and coercion, but most did not. The mismatch between state law and personal beliefs was most salient here as not only did these officers often miss key legal elements of rape, but they also ascribed to stereotypical ideas about rape (e.g., "sometimes a guy can't stop himself; he gets egged on by the girl"). Some of these officers differentiated between stranger and nonstranger assaults, expressing doubts of the veracity of acquaintance rape victims' reports.
The validity of the cluster solution was supported with additional analyses. Police in the Force Definition of Rape cluster, who concentrated on reformed legal elements, were more likely to be command officers and had had special training on the topic of rape. There are several potential explanations for these results. It could be the case that these officers had training on rape because they were command officers. Promotions are based, in part, on departmental involvement and extracareer training. Irrespective of whether the critical factor is training or being a command officer (or both), officers who did have training on rape were more likely to provide definitions consistent with state law. Not surprisingly, these officers also had more professional experience working with stranger rape and nonstranger rape victims. This may be an interactive process: With experience, officers gain more knowledge about the law, and with more knowledge and experience, they may be asked to respond to these cases more often. Taken together, these findings suggest that training and experience are factors that orient officers' attention on the legal elements of the crime of rape rather than on other beliefs. The police in this cluster (Force Definition of Rape) also expressed less traditional attitudes toward women and less acceptance of interpersonal violence. Understanding rape law, therefore, may be intertwined with both work experience and general beliefs about women and violence.
This research did not link these definitions and beliefs with actual behavioral responses in rape cases. For instance, it is not known if the officers in the Mixed Definition of Rape, who ascribed to more victim-blaming views and held less egalitarian views about women, were more likely to "unfound" cases. Conversely, it is not clear if the officers in the Force Definition of Rape, who demonstrated a more thorough understanding of rape and sexual assault law, responded to victims in a more sensitive manner. Future research should attempt to link beliefs and knowledge with job behaviors.
There are, however, methodological problems with this research that limit its usefulness in explaining how police officers perceive rape. Rape is a controversial and sensitive topic, and officers have reason to bias their answers to appear more aware and sensitive to the problem of rape. Because this possibility of social desirability bias was not explicitly examined, it remains a potential threat to the interpretation of these findings. Common method variance is also an issue as all of the data collected in this study were self-report. Furthermore, the generalizability of these findings must be examined closely. Both samples of officers were selected from police departments that agreed to participate in this study. The demographic characteristics reported suggested that these officers' education levels were higher than national averages: 74% of the officers in this sample were college graduates compared to 37% on the national level (Carter & Sapp, 1991; U.S. Department of Justice, Bureau of Justice Statistics, 1992). But, due to ambiguity in the question, it is not clear if the officers in this study had degrees from a two-year or four-year college. The national data reported here are the percentages for four-year college graduates. These results may be generalizable only to other officers with similar education levels. Other factors that may affect generalizability include the gender and racial composition of this sample. Although the gender breakdown in this study is consistent with national data, there was limited racial diversity as the sample was almost exclusively White. The fact that the demographic variables did not differentiate the clusters should be interpreted in light of this restricted range. Additionally, these samples were drawn from midwestern cities, and one was a college police department. The officers in these departments do not handle the number of crimes and the number of serious crimes as do police in departments in larger cities.
Implications and Future Directions
The results of this study suggest that police officers could benefit from further education on rape. For many of the police in this sample, their definitions of rape/sexual assault were inconsistent with state law. Although this project did not link these differential beliefs with actual behavioral responses, other research has suggested that officers' attitudes and beliefs--both positive and negative--are sometimes communicated to the victim (Kerstetter & Van Winkle, 1990). Moreover, a wider-scale effort should follow this exploratory approach to see if officers in other states with reformed rape laws hold such differing beliefs. In designing training programs for police, these results suggest that addressing general beliefs about women and interpersonal violence may also be beneficial. In fact, it could be argued that many of the changes in rape law reflect changing social beliefs about women and violence against women. Addressing these implicit issues in training is a logical extension. As our laws and culture continue to examine and reevaluate violence against women, it is important to ensure that the enforcers of these changes are adequately trained so they can be of most help to victims.
(n1.) The remaining eight states have marital rape and/or "cohabitant" rape exemptions (Berger et al, 1988) In these states, a wife or woman living with her partner cannot charge her husband/partner with rape.
Medium-Sized University Total City Police Police Sample size 91 53 38 Gender Percentage Male 88 98 74 Percentage Female 12 1 26 Race Percentage White 91 90 91 Percentage African American 9 10 9 Age in years (mean) 39 40 39 Family Percentage Married 60 62 58 Percentage Children 79 87 68 Education Percentage College Graduate 74 57 91 Work history Mean years police officer 15.95 15.76 16.27 Mean years at department 14.17 14.47 13.66 Percentage Patrol 56 58 52 Percentage Command 30 27 33 Percentage Investigations 11 15 5 Rape training Percentage who had 60 53 67 training on rape
1. Penetration (e.g., have sex, have intercourse, penetration) 2. Threat of force (e.g., done through the threat of force, threatened victim with force) 3. Use of force (e.g., done through the use of force, force the victim, held her down) 4. Use of psychological force (e.g., coerced her into having sex, pressured her a lot to have sex, psychological manipulated her to have sex) 5. Lack of consent (e.g., done without her consent, unwanted) 6. Rape is committed to satisfy sexual needs (e.g., done for sexual reasons, done for sexual gratification) 7. Rape is committed as an act of power (e.g., done to make the rapist feel powerful, done to feel control over the victim, done to dominate the victim)
Cluster 1 (N = 27) Cluster 2 (N = 17) "Consent Definition "Force Definition of Rape" of Rape" Penetration 100 84 Threat of force 0 58 Use of force 30 95 Use of psychological force (coercion) 0 37 Lack of consent 100 21 Rape is committed to satisfy sexual needs 0 0 Rape is committed as an act of power 0 11 Cluster 3 (N = 43) "Mixed Definition of Rape" Penetration 42 Threat of force 2 Use of force 11 Use of psychological force (coercion) 2 Lack of consent 51 Rape is committed to satisfy sexual needs 40 Rape is committed as an act of power 4
Variable Crossed With Cluster Membership Chi[sup 2]/F df p Demographics Gender 4.21 2 .11 Race 4.59 2 .10 Education 7.38 2, 88 .50 Age 2.77 2, 88 .10 Work history Job title (patrol, command, investigations) 21.44 4 .01 Had training on rape 6.43 2 .05 Experience With Rape Scale 6.27 2, 88 .01 General attitudes Attitudes Toward Women Scale 4.63 2, 88 .05 Attitudes Toward Interpersonal Violence 7.23 2, 88 .01
Authors' Note: The authors thank Doug Luke for his assistance with the cluster analysis and Jody Venable for her help with data collection. This research was supported by a grant from the George W. Fairweather Fund awarded to the first author. All correspondence should be addressed to Rebecca Campbell, Department of Psychology, University of Illinois at Chicago, 1007 West Harrison, Chicago, IL 60607-7137.
Anderberg, M. R. (1973). Cluster analysis for applications. New York: Academic.
Balos, B., & Fellows, M. L. (1991). Guilty of the crime of trust: Nonstranger rape. Minnesota Law Review, 75, 599-618.
Berger, R. J., Searles, P., & Newman, W. L. (1988). The dimensions of rape reform legislation. Law and Society Review, 22, 329-357.
Brownmiller, S. (1975). Against our will: Men, women, and rape. New York: Bantam.
Burt, M. R. (1980). Cultural myths and supports for rape. Journal of Personality and Social Psychology, 38, 217-230.
Caringella-MacDonald, S. (1985). The comparability in sexual and nonsexual assault case treatment: Did statute change meet the objective? Crime & Delinquency, 31, 206-222.
Carter, D. L., & Sapp, A. D. (1991). Police education and minority recruitment: The impact of a college requirement Washington, DC: Police Executive Research Forum.
Cohen, J. (1960). A coefficient of agreement for nominal scales. Educational and Psychological Measurement, 20, 37-46.
DuBois, D. M. (1988). A matter of time: Evidence of a victim's prompt complain tin New York. Brooklyn Law Review, 53, 1087- 1115.
Estrich, S. (1987). Real rape: How the legal system victimizes women who say no. Cambridge, MA: Harvard University Press.
Feldman-Summers, S., & Palmer, G. P. (1980). Rape as viewed by judges, prosecutors, and police officers. Criminal Justice and Behavior, 7, 19-40.
Fischer, K. (1989). Defining the boundaries of admissible expert testimony on rape trauma syndrome. University of Illinois Law Review, 1989, 691-734.
Kerstetter, W. A. (1990). Gateway to justice: Police end prosecutorial response to sexual assaults against women. Journal of Criminal Law and Criminology, 81, 267-313.
Kerstetter, W. A., & Van Winkle, B. (1990). Who decides? A study of the complainant's decision to prosecute in rape cases. Criminal Justice and Behavior, 17, 268-283.
Krahe, B. (1991). Police officers' definitions of rape: A prototype study. Journal of Community and Applied Social Psychology, 1, 223-244.
LeDoux, J. C., & Hazelwood, R. R. (1985). Police attitudes and beliefs towards rape. Journal of Police Science and Administration, 13, 211-220.
Luke, D. A., Rappaport, J., & Seidman, E. (1991). Setting phenotypes in a mutual help organization: Expanding behavior setting theory. American Journal of Community Psychology, 19, 147-167.
MacKinnon, C. A. (1983). Feminism, marxism, method and the state: Toward feminist jurisprudence. Signs: Journal of Women in Culture and Society, 8, 635-653.
MacKinnon, C. A. (1987). Feminism unmodified: Discourses on life and law. Cambridge, MA: Harvard University Press.
Michigan Criminal Sexual Conduct Law (1974). Public Act 266, Michigan Senate Bill No. 1207.
Rapkin, B. D., & Luke, D. A. (1993). Cluster analysis in community research: Epistemology and practice. American Journal of Community Psychology, 21, 247-277.
Romesburg, H. C. (1984). Cluster analysis for researchers. Belmont, CA: Lifetime Learning.
Searles, P., & Berger, R. J. (1987). The current status of rape reform legislation: An examination of state statutes. Women's Rights Law Reporter, 10, 25-40.
Schechter, S. (1982). Women and male violence: The visions and struggles of the battered women's movement. Boston: South End.
Smart, C. (1989). Feminism and the power of law. London: Routledge.
Spence, J. T., Helmreich, R., & Stapp, J. (1973). A short version of the Attitudes Toward Women Scale. Bulletin of the Psychonomic Society, 2, 219-220.
Spohn, C., & Homey, J. (1992). Rape law reform: A grassroots revolution and its impact. New York: Plenum.
U.S. Department of Justice, Bureau of Justice Statistics. (1992). State and local police departments, 1990. Washington, DC: Department of Justice.
Ward, J. H. (1963). Hierarchical grouping to optimize an objective function. Journal of the American Statistical Association, 58, 236-244.
In the state where this research was conducted, sexual assault is defined as follows:
Sexual penetration [means sexual intercourse, cunnilingus, fellatio, anal intercourse, or any other intrusion, however slight, of any part of a person's body or of any object into the genital or anal openings of another person's body] (where) the actor uses force or coercion to accomplish the sexual penetration. Force or coercion includes but is not limited to any of the following circumstances:
1. When the actor overcomes the victim through the actual application of physical force or physical violence
2. When the actor coerces the victim to submit by threatening to use force or violence on the victim and the victim believes that the actor has the present ability to execute these threats
3. When the actor coerces the victim to submit by threatening to retaliate in the future against the victim
4. When the actor engages in the medical treatment or examination of the victim in a manner or for purposes which are medically recognized as unethical or unacceptable
5. When the actor through concealment or by the element of surprise is able to overcome the victim
6. The actor causes personal injury to the victim and the actor knows or has reason to know that the victim is mentally defective, mentally incapacitated, or physically helpless
By REBECCA CAMPBELL, University of Illinois at Chicago and CAMILLE R. JOHNSON, Michigan State University
Rebecca Campbell is an assistant professor of psychology at the University of Illinois at Chicago. She received her Ph.D. in ecological/community psychology from Michigan State University. Her research focuses on how the legal, medical, and mental health systems respond to victims of sexual assault and domestic violence
Camille R. Johnson received her bachelor's degree in communications from Michigan State University. Her work focuses on health education, women's health, and violence against women.