Research involving the population of female inmates is limited; however the research that does exist supports the notion that incarcerated women have experienced more interpersonal trauma throughout their lives than the general female population (Browne, Miller, & Maguin, 1999). Although studies have investigated the mental health correlates of interpersonal trauma (Banyard, Williams, & Siegel, 2001), less is known about the impact of these experiences on incarcerated women. Given empirical findings that female inmates report extensive trauma histories (Bradley & Davino, 2002; Cantrell & Davis, 2004), the possible relationship between these experiences and criminal behavior, specifically violent criminal behavior, is in need of further research.
Most research evaluating violent behavior in society has found a much higher perpetration rate for men than women (e.g., Bureau of Justice Statistics, 2006), however an argument exists as to whether there are true gender difference in the expression of violent behavior or if the data is not reflective of the true state of affairs (e.g., Bennett, Farrington, & Huesmann, 2005; Straus, 2006). Specifically, it is argued that actual frequency of female perpetrated violence is much greater than what is depicted in the literature, due to (a) research concentrating on measuring violence perpetration among men and victimization among women; (b) men underreporting female perpetrated violence due to stigmatization and socialization processes; and (c) a majority of prevalence studies concentrating on arrests and criminal records (Dutton & Nicholls, 2005; Straus, 2006).
Due to the limited research on perpetration of violence among samples of women, it is unclear if the motivations for violence are the same for both genders. It is theorized that men may have more instrumental motives for violence while women may be more motivated by past victimization experiences and self-defense (Swan & Snow, 2003). Regardless of the true prevalence of female perpetrated violence or of the motivations for this violence, it is necessary for all aspects to be evaluated such as: social, historical, and cultural contexts, past victimization experiences, and psychopathology including posttraumatic stress disorder (PTSD) and substance abuse (Swan & Snow, 2006).
As previously noted, prevalence rates of trauma are much higher among incarcerated women than women in the community (e.g., Bradley & Davino, 2002). Therefore, research has begun to look at the possible association between interpersonal trauma and criminal behavior, specifically violent criminal behavior. It has been theorized that for women, violence may serve as outlet for their emotional distress due to post-trauma symptomology (Giftus, 2007).
At this time, there are many unanswered questions in regards to the possible relationship between prior interpersonal trauma and violence among women. First, does such a relationship exist and if so, is this a direct pathway or are there other contributing factors influencing this relationship? In order to consider what possible factors may contribute to the relationship between violent behavior of female inmates with trauma with PTSD and substance abuse, it is necessary to evaluate variables that have an association with both trauma and violence. PTSD is one of the primary mental health effects of experiencing trauma (APA, 2000). Considering that populations of incarcerated women have higher prevalence rates of interpersonal trauma compared to women in the community, it is no surprise that the same is true for prevalence of PTSD (e.g., Battle, Zlotnick, Najavits, Gutierrez, & Winsor, 2003). Given this high prevalence rate, it is possible that PTSD may be associated with criminal behavior and violent criminal behavior.
Research has found support for an association between PTSD and violent behavior (e.g., Calhoun, Malesky Jr., Bosworth, & Beckham, 2004; Savarese, Suvak, King, & King, 2001). However, research in this area has focused on utilizing samples of men whose traumatic experiences involved combat experiences in Vietnam. Therefore, the possible relationship between PTSD symptomology and violent behavior in women and/or victims of trauma other than combat is unknown.
Substance abuse is another construct that has been associated with prior interpersonal trauma. Again, prevalence rates for substance use disorders are much higher among incarcerated women than women in the community (Battle et al., 2003). Therefore, the possible relationship between substance use and violent behavior among women remains to be empirically determined.
Another frequent association with the experience of interpersonal trauma is various forms of psychopathology (e.g., McFarlane, Schrader, Bookless, & Browne, 2006). Research also suggests a relationship between mental illness and violent behavior (e.g., Corrigan & Watson, 2005). In fact, when inpatient samples are utilized or mental illness is taken into account, gender differences in the perpetration of violence disappear (e.g., Krakowski & Czobor, 2004).
In conclusion, the exact pathway from interpersonal trauma to future violent behavior among women is unknown. It remains unclear if the comorbidity of posttrauma symptomology, substance use, and psychopathology or specific combination of these factors may influence this relationship. The purpose of this study will be to evaluate more closely the relationship between violent behavior of female inmates with trauma with PTSD and substance abuse. Specifically, possible mediating variables in this relationship will be assessed including: post-trauma symptomology, substance abuse, and psychopathology.
Review of Literature
According to the 2007 report Crime in the United States issued by the Uniform
Crime Reporting Program (UCR), approximately 82% of arrests for violent crime were men compared to 18% women. Controversy exists in the literature as to whether this wide gap between male perpetrated violence and female perpetrated violence is accurate. Specifically, the argument is whether there are true gender differences in the perpetration of violent crime or whether the current research methodology used does not provide an accurate representation of violence perpetrated by women (e.g., Bennett, Farrington, & Huesman, 2005; Dutton & Nicholls, 2005).
The 2006 UCR data is based on arrest data from law enforcement agencies nationwide. When victim self-report is examined, the gap between male perpetrated and female perpetrated violence remains consistent. For example, the Bureau of Justice Statistics' National Crime Victimization Survey (2006) which gathers self-report data from a nationally representative sample of 77,200 households, found that 79% of violent crimes were reportedly perpetrated by men compared to 19% perpetrated by women and 2% reported they did not know the gender of their perpetrator. However, some researchers argue against this notion and purport that the prevalence rates of violent behavior perpetrated by females are inaccurate, especially in regards to intimate partner violence. Dutton and Nicholls (2005) reviewed the literature on intimate partner violence perpetrated by men and women and concluded that women perpetrate violence much more frequently (e.g. Brown, 2004) than what is portrayed throughout the literature.
For example, Brown (2004) found that women were four times more likely to report intimate partner violence to the police than men. It is unknown if men would also be unlikely to report victimization by female perpetrators to sources other than the police (e.g., researchers utilizing self-report). Dutton and Nicholls (2005) hypothesized that men underreport violence perpetrated by women because men are less likely to view female violence as a crime. They further postulate that feminist theory, focused on the need to protect women in domestic violent relationships, disregards the reality of the situation. However, research in and of itself is meant to be objective and unbiased. Policy should be an outcome of research but not a dictation of research. In Straus' (2006) review of the literature on gender symmetry in intimate partner violence, it was further noted that police were more likely to be called to a crime scene where injury was involved and male perpetrated violence is more likely to cause injury, therefore inflating the statistics of male perpetrated violence that comes from crime report data. Straus (2006) further comments on his personal experiences regarding the pressures within the field to avoid this type of research due to extreme criticisms and isolation experienced when publishing these results.
Factors Contributing to Gender Differences in the Perpetration of Violence
While some researchers argue that inaccuracies in research methodology account for gender differences in violent behavior, others argue in favor of true gender differences in the expression of violent behavior between men and women. In Bennett and colleagues' (2005) review of the literature, they found biological and environmental factors that account for true gender differences in the perpetration of violence. Specifically, they hypothesized that differences in social cognition between the sexes may be one possible explanation for the unequal prevalence rates of violence. They define social cognition as the following:
Social cognition refers specifically to interpersonal knowledge, perceptions, attitudes, and behavior, and social information processing refers to mechanisms such as coding, rehearsing, storage, and retrieval that are employed in interpreting social data and that guide behavior. An individual faced with a social event appraises and interprets situational cues, searches memory for guides to behavior, assesses and decides on the optimal behavior, and then enacts that response. Individuals use their social information processing capabilities to decide what behavior is suitable for a particular situation (p. 265).
Overall, women seem to acquire these social cognitive skills earlier and have superior skills in this area which may be due to both biological and socialization processes. Earlier brain maturation in females is given as an example of the biological process underlying gender differences, specifically, the earlier the brain matures, the more practice with various cognitive abilities an individual will have, leading to greater competency (Bennett et al., 2005).
More current research suggests that hormones may play a role in the gender differences found in perpetration of violence. Ramirez (2003) postulates that these gender-based differences in violence occur from differences in the euroendocrinological systems during prenatal development. Specifically, a link has been found between testosterone and aggressive behavior in samples of animals and humans (e.g., Easterbrook, Kaplan, Glass, Pletnikov, & Klein, 2007; Ramirez, 2003). Popma, Vermeiren, Geluk, Rinne, van den Brink, Knol and colleagues (2007) found that not only is testosterone involved in aggressive behaviors, but this relationship may be moderated by levels of Cortisol. Specifically, individuals with higher levels of testosterone and lower levels of Cortisol are more likely to engage in violence.
Motivation of Self-Defense
Swan and Snow (2006) proposed a theory of women's violent behavior in the context of intimate relationships with women having different motivations for perpetrating violence than men. Specifically, they argued that women are more likely to report self-defense, fear, defense of children, and retribution as motivators for their perpetration of violence. Swan and Snow (2002) sampled 108 women from the community who reported using some form of physical violence in the previous 6 months against their intimate partner. The Revised Conflict Tactics Scale (CTS2; Straus, Hamby, Boney-McCoy, & Sugarman, 1996) and the Psychological Maltreatment of Women Inventory (PMWI; Tolman, 1989) were utilized to assess perpetration and victimization experiences within their intimate partner relationships. Participants also completed a item measure of social desirability created for this particular study. Results indicated that only six women in the entire sample reported not experiencing physical abuse from their partner within the previous six months. The women reported perpetrating significantly more moderate acts of physical violence (e.g., threw things) against their male partners than their partners perpetrated against them; however, they also reported significantly more severe physical violence (e.g., choking) used against them by their male partners than severe physical violence perpetrated by them. They were also more likely to report that their male partners used methods of control and utilized sexual coercion more frequently than they did against their partners. In regards to motivations, women who were more likely to be victims of intimate partner violence reported that they utilized violence in self-defense. Results from this study suggested that intimate-partner violence perpetrated by women is better understood within the context of the current victimization experiences of the women in the relationship. Although the women reported perpetrating more moderate acts of physical violence towards their male partners, they also reported that their male partners perpetrated more severe physical acts of violence which caused injury, and utilized more methods of control and sexual coercion. Furthermore, it is important to note that the women who were aggressors towards their intimate partners were still experiencing a high frequency of physical abuse from their partners. Specifically, on average, the women aggressors reported experiencing approximately abusive behaviors within the last six months. It remains unclear if this reveals empirical support for the self-defense theory of violence, since no information was collected on who initiated the violence or how often violence was perpetrated in response to the partner's violence.
Motivation of Prior Abuse
Utilizing 95 women from the Swan and Snow (2002) sample above (10 were excluded due to missing data), Swan and Snow (2003) further analyzed the data utilizing the following variables: childhood abuse (Early Trauma Inventory; Bremmer, Vermetten, & Mazure, 2000) and motivations. The participants were categorized into four groups based on their relationship information gathered from the CTS2 and PMWI as reported by the women: (1) women as victims; their male partner perpetrates more severe physical violence and coercion against the female partner than she commits against him; (2) women as aggressors; the woman commits more severe physical violence and coercion against her male partner than he commits against her; (3) mixed-male coercive; the male partner uses more coercive control than the woman but the rate of severe physical violence perpetrated by the woman is equivalent or greater than what is perpetrated by the man; and (4) mixed-female coercive; the woman's use of coercion is equivalent to or greater than the man's but the man's use of severe physical violence is equivalent to or greater than the woman.
Results indicated for the entire sample that 35% of the women reported experiencing childhood physical abuse, 37% childhood emotional abuse, and 52% childhood sexual abuse. The women as the aggressor group reported experiencing significantly more childhood abuse than the women as victim group or the mixed-male coercive group. No significant differences were found in regards to childhood abuse between the women as aggressors group and the mixed-female coercive group (Swan & Snow, 2003). However, there were some limitations within this study. First, the categorization process utilized has not been done before; therefore, more research is needed to establish these categorizations as empirically relevant and useful. Further, traumatic events other than childhood abuse and intimate partner violence were not assessed, therefore, it remains unclear what role these variable may play in this context if any. Overall, this study provides support for the complexities in the theoretical formulations of female perpetrated violence.
Although an integrated and consensual trauma-focused violence theory is yet to be created, research continues to evaluate the possible relationship between trauma history and perpetration of violence and thus far has found equivocal results for women. Further, a clear understanding of this possible association is lacking. Not all people who experience abuse perpetrate violent criminal acts. Therefore, what are the contributing factors within this paradigm? It is also unclear if the same predictors of violent behavior for women outside of prison also predict violence while they are incarcerated. In a review of the "cycle of violence" research, Widom (1989a) noted that there is no straight pathway from victimization to perpetration firmly established in the literature. This is due in part to ambiguities in the definitions of abuse, broad definitions of violence, and methodological limitations in research.
High prevalence rates of the experience of interpersonal trauma have been found among the population of female inmates. Further, high prevalence rates of PTSD, substance abuse/dependence, and other psychopathology (e.g., bipolar disorder, major depression, personality disorders) have also been found within this population. Each of these factors (interpersonal trauma, PTSD, substance use, and other psychopathology) has been associated with the engagement in violent behavior. However, much of this research has focused on samples of men and/or victims of combat. It remains unknown how or if these factors influence violent behavior among women. Furthermore, the complexities in the relationship between violent behavior of female inmates with trauma with PTSD and substance abuse are yet to be discovered.
The purpose of the present study is to evaluate more closely the relationship between violent behavior of female inmates with trauma with PTSD and substance abuse. Specifically, possible mediating variables in this relationship were assessed including: post-trauma symptomology, psychopathology, and substance abuse or dependence. The hypothesis for this study includes:
Frequent experiences of childhood physical abuse of women engages in more recurrent violent behavior towards their intimate partners and persons with whom there is no intimate-partner relationship in the year before they are arrested compared to women with less frequent incident of abuse during their childhood.
Physical abuse during childhood years will forecast more of the difference in violent behavior commited against intimate and non-intimate partners compared to other forms of abuse which includes emotional abuse, sexual violence, and abandonment when adult trauma is controlled for.
The relationship between Physical abuse during childhood years and commitment of violent behavior will be arbitrated by post-trauma, Axis I psychopathology, and substance abuse/dependence.
Inclusion criteria for the study will be women who have been incarcerated within the last seven years. This time frame is chosen in an attempt to gather more accurate information that is less confounded by memory mistakes. For example, an inmate who has been incarcerated for 30 years may have more difficulty remembering events and feelings experienced in the year prior to arrest than women incarcerated for seven years or less. The only exclusion criteria will be the appearance of active psychoticism through behavioral observation (e.g., tangential thought process, word salad, bizarre behavior) and/or active suicidality (inmates on suicide precautions were not able to participate). There will be no restrictions based on ethnicity or race. If an inmate reported to the primary investigator that she was unable to read, the primary investigator will schedule another time to complete the survey where the survey would be administered verbally by the primary investigator in a private setting.
Flyers advertising the study along with a sign-up sheet will be displayed in each housing unit of the minimum, medium and maximum-security portions of the prison.
Inmates who will choose to participate in the study placed their name and Department of Correction's identification number on the sign-up sheet in their housing units.
A power analysis will be conducted utilizing G*Power 3 (Faul, Erdfelder, Lang, & Buchner, 2007). The largest regression equation that will be utilized is the equation examining at post-trauma symptomology as a possible mediator and potentially includes the following 11 predictors: childhood physical abuse, PTSD diagnosis, the frequency score for each of the three clusters of PTSD symptoms: intrusive experiences, avoidance/numbing, hyperarousal; and the following post-trauma symptomology: anxious arousal, anger/irritability, defensive avoidance, dissociation, impaired self-reference, and tension reduction behavior. Therefore, 11 predictors will be entered into the G*power 3 analysis along with a moderate effect size and an alpha level set at 0.05.