Scripts play an important role in how individuals conceptualize and enact behavior, including sexual behavior. Scripts are a type of schema, or cognitive structure, that consists of organized knowledge about a given domain (Schank & Abelson, 1977). Specifically, a script is a schema for a particular type of event, and includes information about props, roles, and rules regarding the chronology of events within the script (Fiske & Taylor, 1991). For example, someone’s script for eating a meal in a restaurant may include props (e.g., menu, bill), roles (e.g., customer, server), and the sequence of events in the situation (e.g., the host seats the customer, the customer looks at the menu). Like all schemas, scripts operate unconsciously, are highly resistant to change (Demorest, 1995), and influence cognitive processes, such as attention, organization, interpretation, and recall of information (Baldwin, 1992; Zadney & Gerard, 1974).

Several investigations support the existence of a script for rape (Bondurant, 2001; Kahn, Mathie, & Torgler, 1994; Littleton & Axsom, 2003; Ryan, 1988). In addition, empirical evidence suggests that individuals’ rape scripts can influence a number of processes including the extent to which observers hold rape victims responsible or blame them for the rape (Burt & Albin, 1981; Frese, Moya, & Megías, 2004; Krahé, 1988) and whether a rape claim is regarded as legitimate (Krahé, 1991). Rape scripts also appear to affect whether victims label their experience of forced, unwanted sex as a rape (Bondurant, 2001; Kahn et al., 1994; Peterson & Muehlenhard, 2004). Littleton, Axsom, Radecki Breitkopf, and Berenson (2006) also found that rape scripts may affect recovery following rape, including victims’ coping strategies, whether they disclose the assault to others, and how others respond to the victim’s disclosure.

Two studies of college students’ rape scripts demonstrated that these scripts are often highly stereotypical (Littleton & Axsom, 2003; Ryan, 1988). Specifically, many scripts involved a highly violent rape between strangers or two individuals with a minimal prior relationship. Similarly, rape attribution studies, which have been conducted primarily with college students, indicate that forced sex scenarios are less likely to be labeled as rape if they occur between two individuals in a romantic relationship, involve less violence, or if the woman has previously engaged in consensual sexual activity with the man (Bridges, 1991; Frese et al., 2004; Hannon, Kuntz, Van Laar, Williams, & Hall, 1996; Monson, Langhinrichsen-Rohling, & Binderup, 2000; Proite, Dannells, & Benton, 1993; Simonson & Subich, 1999). Thus, situations that deviate from the stereotypical rape script are less likely to be conceptualized as rape.

One limitation of the research in this area is an almost exclusive focus on European American college women. Thus, little is known about the rape scripts of ethnic minorities, including Latinas, and lower-income women. However, it seems plausible to hypothesize that the rape scripts of ethnic minorities and women of lower socioeconomic status (SES) may differ in some respects from the rape scripts of European Americans and women of higher SES. For example, Latinas may have less well-elaborated rape scripts than European Americans do. One important source for the formation of rape and abuse scripts is disclosures by others (Buzi, Tortolero, Smith, Ross, & Roberts, 2002). Latinas may be less likely than European Americans to know someone who was raped because Latina victims appear to be even less comfortable than women of other ethnic groups disclosing a rape experience. Indeed, focus group research has shown that Latina women often are uncomfortable discussing the topic of rape and are less likely than European Americans to provide to others a detailed description of rape in general or of their own rape experience in particular (Bletzer & Koss, 2004; Ramos-Lira, Koss, & Russo, 1999). Thus, even if Latinas have had someone disclose a rape experience to them, the disclosure may have been vague and lacking in detail.

Another difference that could exist between the rape scripts of European American and Latina women is a greater emphasis on shame and blame of the victim among Latinas due to internalization of gender stereotypes that focus on a woman’s sexual purity as a major contributor to her dignity and that of her family (Niemann, 2004; Sorenson, 1996). Indeed, Ramos-Lira and colleagues (1999) in their qualitative study of Latina women’s ideas about rape found that many women discussed issues of shame and blame including the notion that a rape victim, particularly a victim of date rape, had violated the confidence of her parents and that her future husband would be disappointed by her loss of virginity. Similarly, Lefley, Scott, Llabre, and Hicks (1993) found that Latina rape victims engaged in more self-blame and perceived their community as more victim blaming than did African or European Americans. Finally, there is some evidence that Latinos/as are more likely than European Americans or African Americans to adhere to traditional, patriarchal gender roles (Kane, 2000). Adherence to patriarchal gender roles could be associated with greater belief in the notion of sexual precedence (Caron & Carter, 1997; Livingston, Buddie, Testa, & VanZile-Tamsen, 2004), that is the idea that if a woman has had previous consensual activity with a man or is in a committed relationship with him, then he has a legitimate claim to her future sexuality (Shotland & Goodstein, 1992). Thus, greater belief in sexual precedence would be associated with less belief that rape can occur within the context of an established romantic relationship.

SES differences in individuals’ ideas about rape are clearly an understudied area. However, it seems plausible to hypothesize that women of lower SES may be less likely than women of higher SES to have been exposed to educational programming about rape. For example, a likely frequent source of rape education is programming provided on college campuses that often focuses on dispelling rape myths and helping individuals to adopt a broader conceptualization of rape (Breitenbecher, 2000). Given that lower SES women are less likely to have attended college, they may similarly be less likely to have been exposed to this programming and, as a result, they may adopt a more stereotypical conceptualization of rape than do women of higher SES. Indeed, a meta-analytic review indicated that lower SES was associated with greater acceptance of rape myths (i.e., stereotypical and false beliefs about rape; Anderson, Cooper, & Okamura, 1997). Individuals of lower SES may also tend to be more accepting of patriarchal gender roles than are individuals of higher SES (Rank, 2000).

However, no researchers have examined the rape scripts of ethnic minorities or individuals of lower SES through use of a quantitative methodology. Thus, the primary goal of the current study was to evaluate the rape scripts of women of lower SES by drawing on a sample of women attending family planning clinics that primarily serve low-income women. In addition, given the potential differences in the rape scripts of Latinas and European Americans and the implications of these differences for individuals’ conceptualization of rape incidents and help seeking, we sought to compare the rape scripts of European American and Latina women. Finally, given the possibility that factors such as immigration stress, educational opportunities, and language use could affect women’s rape scripts, the scripts of monolingual Spanish speaking Latina women and Latina women who were recent U.S. immigrants were compared to the rape scripts of other Latinas.

Method

Participants

Data were collected (as part of a larger study of sexual victimization) from women who visited one of two university-administered family planning clinics in southeast Texas between July and October of 2005. The refusal rate was approximately 27%. In the current study, analyses were restricted to self-identified European American and Latina women (n = 393) because there were insufficient numbers of women who self-identified as other ethnicities to make meaningful comparisons (n = 71). There was no compensation for participation, and the study was approved by the university’s Institutional Review Board.

The average age of women in the final sample was 28 years (SD = 8.2; range 18–56). Forty-six percent (n = 180) identified themselves as European American and 54% (n = 213) identified themselves as Latina, primarily Mexican or Mexican American. As summarized in Table 1, 31% of participants reported that they had not completed high school nor received a GED, whereas 28% reported at least some college education or other form of postsecondary education. Sixty percent of participants reported that their household income was $15,000 per year or less, and only 5% reported a yearly household income over $35,000. Fifty-two percent of Latinas reported that they were born outside of the United States. In addition, 46% of Latinas reported that their household was bilingual, and another 35% reported that they spoke Spanish exclusively at home.

Table 1 Demographics of participants.
Full size table

Procedure

Female outpatients were recruited from the clinic waiting room. Potential participants were approached by a research coordinator and asked to complete a brief, anonymous questionnaire, available in Spanish or English. Participants were informed that the purpose of the research was to learn about women’s ideas regarding, and experiences with, unwanted sex. If necessary, a bilingual clinic staff member explained the purpose of the study. Women who consented to participate by signing a written consent form in their language of choice were informed that they could complete the questionnaire in either the clinic waiting room or the examination room. Participants returned the questionnaire to the research coordinator in an envelope provided with the survey. Completing the questionnaire took approximately 30 min. Women who were under 18 years of age, over the age of 65, currently pregnant, or could not read or write in English or Spanish were not eligible to participate.

Measures

In addition to demographic questions, women were asked to complete a questionnaire regarding their ideas about rape. Specifically, women were provided with a list of 37 potential rape script elements and were asked to rate on a 5-point Likert scale bounded by 1 (not at all) and 5 (completely) the extent to which they thought that each was true of a typical rape. Twenty-one of the potential rape script elements were drawn from a prior study of the rape scripts of college students (Littleton & Axsom, 2003) that were generated based on open-ended rape script descriptions provided by students. The items drawn from this prior study assessed a number of potential aspects of individuals’ rape scripts including the location of the rape (e.g., the rape happens in the man’s or woman’s home), the psychological effects of the rape on the woman (e.g., the woman feels depressed after the rape), and the behavior of the man and woman prior to the rape (e.g., the woman dresses in sexy clothing, the man has been drinking alcohol). Because these potential aspects did not include a number of potentially relevant components of individuals’ rape scripts, a number of other items were added. These items consisted of specific types of force used by the man (e.g., holding the woman down), specific resistance tactics used by the woman (e.g., screaming), specific types of relationships between the man and woman in the rape (e.g., the man and woman are in a romantic relationship), items consistent with a highly violent rape (e.g., the woman is seriously injured), and other potential rape locations (e.g., the rape happens outdoors). All items were worded so that they reflected a man raping a woman. The entire questionnaire was translated and back-translated by bilingual research coordinators with experience translating health surveys.

Scoring and analysis plan

To reduce the number of analyses conducted, a principal components analysis with a varimax rotation was conducted. An orthogonal rotation was chosen because it was expected that components would not be highly correlated (Floyd & Widaman, 1995). Items with loadings of less than .30 on all components were excluded. Participants’ mean score on the items completed within each component was then calculated. This scoring approach was chosen to minimize missing data. However, women who had extensive missing data on a component, defined as 50% or more items missing on a component with at least five items or one or more items missing on a component with four or fewer items, were excluded from analyses of that component. Bonferroni-adjusted t-tests were conducted to compare the mean component scores of European Americans and Latinas; the modified Bonferroni procedure recommended by Jaccard and Becker (1997) was used.

The same procedure was utilized to compare the scripts of Latina women who were recent immigrants (moved to the U.S. in adolescence or adulthood) to those who were born in the U.S. or moved to the U.S. in childhood, as well as to compare the scripts of monolingual Spanish speaking Latinas to bilingual or English speaking Latinas. Language use and age at immigration were utilized to compare subgroups of Latina women as opposed to an acculturation measure because of the numerous weaknesses inherent in such measures. These include poorly defined concepts of “mainstream” and “ethnic” cultures, downplaying of the historical and geographic links between so-called separate cultures (e.g., Mexican and U.S.), and downplaying of the role of other factors that could affect attitudes and behaviors, such as socioeconomic status, immigration stress, and facility with English (Hunt, Schneider, & Comer, 2004).

Results

Principal components analysis yielded seven interpretable components that comprised 34 of the 37 items. These components were: negative psychological effects of the rape on the woman, a “hook-up” or date rape, severe violence in the rape, less severe violence in the rape, a non-violent rape, the woman feels stigmatized, and rape within an established relationship. The items that make up each component and the loadings of each item can be found in Table 2. Together these components accounted for 52% of the variance in scores on the questionnaire. The three items that did not load on any components (all loadings less than .30) were: the man and woman are strangers, the rape happens outdoors, and the rape happens in the man’s or woman’s home.

Table 2 Items that comprise rape script components and item loadings.
Full size table

Women with extensive missing data on a component were eliminated from analyses of that particular component. The percentage of women eliminated from analyses of each component score ranged from 12% (severe violence in the rape) to 20% (non-violent rape). Latinas were more likely than European Americans to have extensive missing data on one or more components (41% vs. 13%), χ 2(1) = 37.5, p < .005, and women who completed the questionnaire in Spanish were also more likely to have extensive missing data on at least one component than were women who completed the questionnaire in English (66% vs. 17%), χ 2(1) = 82.3, p < .005. In addition, women who reported less than a high school education/GED were significantly more likely to have extensive missing data on one or more components than were women who reported having a high school education/GED or higher (51% vs. 19%), χ 2 (1) = 38.8, p < .005. However, it should be noted that only a minority of women in each of these groups were eliminated from all analyses. Seventy-one percent of Latinas had no missing data or had missing data on only one component, and 17% had extensive missing data on all components. Similarly, 56% of women with less than a high school education/GED had no missing data or had missing data on only one component, and 29% had extensive missing data on all components. Finally, 48% of women who completed the questionnaire in Spanish had no missing data or had missing data on only one component, and 37% had extensive missing data on all components.

Means, standard deviations, and effect sizes of European Americans’ and Latinas’ factor scores are summarized in Table 3. Overall, participants rated negative psychological outcomes for the woman, severe violence (e.g., the woman being seriously injured), and less severe violence (e.g., the man holding the woman down or using his superior body weight) in the rape as highly typical. In contrast, rape within an established relationship, hook-up or date rapes, and non-violent rapes were rated as only slightly typical of rape. The woman feeling stigmatized was rated as somewhat typical of rape.

Table 3 Means, standard deviations, and effect sizes of ratings of rape script components by European American and Latina women.
Full size table

Three of the Bonferroni-adjusted t-tests used to compare the rape scripts of European American and Latina women were significant, with small to moderate effect sizes. Specifically, European Americans rated rape that occurs in a hook-up or date context as significantly more typical than did Latinas, t (287) = 2.9, p < .05. European American women also rated the woman feeling stigmatized as significantly more typical than Latina women did, t (322) = 4.3, p < .05. Finally, European Americans rated rape that occurs within an established relationship as significantly more typical than Latinas did, t (319) = 4.1, p < .05. There were no differences in European American and Latina women’s typicality ratings of the three items that did not load on any of the factors; all of these items received low mean typicality ratings (data not shown).

The means, standard deviations, and effect sizes of Latina women’s responses by language use (English or bilingual versus monolingual Spanish) are summarized in Table 4. Two of the Bonferroni-adjusted t-tests used to compare these two groups of women were significant. Specifically, monolingual Spanish speaking women rated rapes that occurred within an established relationship as significantly less typical of rape than did English speaking and bilingual women, t (149) = 3.7, p < .05. In addition, monolingual Spanish speaking women rated the woman feeling stigmatized as significantly less typical of rape than did English speaking and bilingual women, t (152) = 3.3, p < .05. Similarly, the means, standard deviations, and effect sizes of Latina women’s responses by age at immigration (born in U.S. or immigrated before age 15 versus immigrated after age 15) are summarized in Table 5. Two of the Bonferroni-adjusted t-tests used to compare these two groups of women were significant. Specifically, Latina women who had immigrated to the U.S. at an older age rated less severe violence by the man as significantly less typical of rape than did women who were born in the U.S. or immigrated at a younger age, t (150) = 4.0, p < .05. In addition, later immigrants rated the woman feeling stigmatized as less typical than did Latina women who were born in the U.S. or had immigrated at a younger age, t (152) = 3.5, p < .05.

Table 4 Means, standard deviations, and effect sizes of ratings of rape script components by monolingual Spanish speaking and English speaking or bilingual Latina women.
Full size table
Table 5 Means, standard deviations, and effect sizes of ratings of rape script components by Latina women who immigrated to the U.S. during adolescence or adulthood and women who were born in the U.S. or immigrated during childhood.
Full size table

Discussion

The rape scripts of the low-income women in this investigation fit that of a highly stereotypical rape (i.e., a violent, stranger assault). Women rated factors associated with a violent rape, such as the use of various forms of strong physical force by the man and serious injury to the woman, as highly typical. In contrast, factors associated with rape within a dating context or an established relationship were rated as much less typical, which supports the notion that sexual precedence influenced women’s perceptions of whether incidents of forced sex are regarded as rape. Non-violent rape items also were not rated as highly typical. It was somewhat surprising that the woman experiencing feelings of stigma was rated as only moderately typical overall. Thus, women appeared to conceptualize the typical rape victim as one who is emotionally damaged by the rape (e.g., suffers depression, loss of self-esteem), but not one who experiences strong feelings of stigma as a result of her status as a victim. This could reflect the fact that most women viewed the typical rape victim as one who would likely be perceived as legitimate by others, that is, a woman who suffered a violent attack by a man with whom she has no romantic history and who did not engage in any behaviors that suggested that she was interested in sex or sexually available. The women in the current sample were less likely than the college women in Littleton and Axsom’s (2003) study to regard a typical rape as occurring in a date or hook-up setting. This difference could reflect somewhat less elaborated rape scripts among the women in the current sample, as compared to samples of university students. However, both groups reported similar ratings of the extent to which a typical rape has negative psychological effects on the victim, leads to stigma, and occurs without overt violence.

There also were some ethnic differences in this sample of low-income women’s rape scripts. Latinas tended to rate factors associated with a rape that occurs within a hook-up or date context as less typical than did European Americans. In addition, they rated a rape that occurs between a man and a woman with a prior relationship history as less typical than European Americans did. Thus, even more so than European American women, Latinas appear to regard rape as highly stereotypical, that is, a crime that involves a man assaulting a woman with whom he does not have a prior relationship. Latinas also rated the victim feeling stigmatized as significantly less typical of rape than did European Americans. One possible explanation of this final finding is that, as Latinas regarded the typical rape victim as one who would be seen by most people as a legitimate victim, they would not expect her to experience strong feelings of stigma.

The pattern of comparative findings presents the possibility that Latina women may have a less elaborated rape script than European Americans do, such that their scripts may be less likely to contain information about the subtle and long-term effects of victimization (e.g., stigma) and multiple types of rapes (e.g., rapes that occur on dates, hook-up rapes at a bar or party). The results also suggest potentially greater belief in the notion of sexual precedence among Latinas than among European Americans. In addition, the findings regarding differences among Latina women based on language use and age at immigration suggest that women who are experiencing more immigration stress or who are at a greater socioeconomic/educational disadvantage (e.g., lack of access to educational resources available in the U.S., inability to speak English) have a less elaborated rape script than other Latinas do. For example, these women tended to rate all items as less typical of rape than did other Latina women. In addition, these women were especially likely to rate factors that could be considered representative of a more well-elaborated or comprehensive rape script as less typical (e.g., the victim experiences stigma, the rape occurs in an established relationship) than did other Latina women. It is also possible that this subgroup of women held more traditional and patriarchal gender role beliefs than other Latinas, and thus, may be more likely to believe in the notion of sexual precedence.

Limitations of the study should be noted. First, women in the current study were asked to respond to a list of items, and, thus, certain aspects of their rape scripts may not have been tapped by these items. Similarly, the potential rape script items presumed a man raping a woman, and we did not evaluate whether women held scripts for other types of rape (e.g., rape between individuals of the same gender). In addition, less educated, Spanish speaking Latina women were most likely to have extensive missing data, which limits generalizability to these women. The more extensive missing data among these women is likely to reflect the confluence of cultural, educational, and language barriers. In addition, variables that may affect script formation and content were not evaluated (e.g., media exposure, rape education exposure, being the recipient of a victim’s disclosure). Finally, the current sample consisted of low-income European American and Latina women in southeast Texas, and, therefore, the findings may not be generalizable to other groups of women.

Future researchers should continue to examine the sexual scripts, both consensual and non-consensual, of individuals of different cultural, ethnic, and socioeconomic backgrounds. Similarly, future researchers should examine how cultural (e.g., gender role beliefs), social (e.g., how society views rape victims, media depictions of rape), legal (e.g., whether men who commit rape are prosecuted), and individual (e.g., knowing a rape victim) factors influence rape scripts. There is a clear need to investigate how rape scripts influence normative sexual interactions and interpersonal relationships. There also is a clear need to explore the recovery experience of rape victims of different cultural and ethnic backgrounds, including how women label experiences of unwanted sex, whether they tell others about their experiences, how others respond, and the psychological and physical effects of such experiences. Research in these areas will increase our understanding of the social, cultural, and personal influences on the development and maintenance of sexual scripts, including rape scripts, as well as how these scripts affect victims of rape and women in general.