This item is only available as the following downloads:
THE EFFECT OF DIVERTED DISCLOSURE ON COGNITIVE PROCESSING WITH APPLICATIONS TO SEXUAL ASSAULT BY KATHERINE HOWARD OGLESBY A Thesis Submitted to the Division of Social Sciences New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Steven Graham Sarasota, Florida April, 2012
! "" Dedication Most of all, I'd like to dedicate this to survivors of sexual assault who have the courage to keep on keeping on in the face of painful and wholly undeserved difficulties. I'd also like to dedicate it to all the awesome people I'm lucky enough to have in my life. You guys are the best friends ever. Finally, I'd like to dedicate this to my mom, dad, sister and brother, whom I love and admire a whole lot.
! """ Acknowledgements First and foremost, I would like to thank Professor Graham for all his guidance and encouragement throughout this process. I would also like to thank Professor Harley for providing surrogate guidance and encouragement while Graham was on leave. I would als o like to thank Professor Herzog for helping to develop my artistic skills throughout the years, and specifically for helping with the art show. I would also like to thank the amazing artists and survivors who contributed such beautiful and moving artwork and poetry to the show it would not have had the same impact without your voices. Last but not least, I would like to thank everyone who participated in this study many of the responses were so insightful and moving that I was truly touched and inspir ed by your words.
! "# Table Of Contents DEDICATION ii ACKNOWLEDGEMENTS iii TABLE OF CONTENTS iv LIST OF TABLES v ABSTRACT vi INTRODUCTION 1 BENEFITS OF TRAUMA DISCLOSURE 2 SOCIAL RESPONSES TO DISCLOSURE 3 DISCLOSURE AND COGNITIVE PROCESSING 6 DISCLOSURE, COGNITIVE PROCESSING, AND PTSD 9 APPLICATIONS TO DIVERTED DISCLOSURE OF SEXUAL ASSAULT 14 SEXUAL ASSAULT DISCLOSURE 17 SOCIAL REACTIONS TO SEXUAL ASSAULT DISCLOSURE 20 SEXUAL ASSAULT DISCLOSURE AND PTSD 22 DIVERTED DISCLOSURE 25 THE CURRENT STUDY 27 METHOD 29 RESULTS 36 DISCUSSION 40 REFERENCES 51 TABLES 54 APPENDICES 58
! # List of Tables Table 1: Changes in Positive and Negative Affect Before and After the Art Show 54 Table 2: Overall Descripti ve Statistics for the Linguistic Variables 55 Table 3: Descriptive Statistics for the Linguistic Variables by Condition 56 Table 4: Comparison of Cognitive Processes in Experimental Responses to 57 Base Rates in Emotional Writing
! #" THE EFFECT OF DIVERTED DISCLOSURE ON COGNITIVE PROCESSING WITH APPLICATIONS TO SEXUAL ASSAULT Katherine Oglesby New College of Florida, 2012 ABSTRACT Diverted disclosure, a common response to sexual assault disclosure, involves initially listening to the survivor confide that the assault occurred, but soon after distracting or discouraging the survivor from discussing the experience The present study investigated the impact of diverted disclosure on cognitive processing about a distressing experience. It was predicted that disclosure would enhance cognitive processing, but that a diverting response from the listener would mitigate this effect. To induce negative affect, undergraduate students (n=24) initially viewed an art show dealing with personal experiences of sexual assault. Disclosure conditions were determined by a 2 X 2 design : participants either disclosed or spoke about an unrelated topic to a confederate who either diverted the conversation or allowed full discussion. The extent of cognitive processing was determined by the overall proportion of words related to cognition in a written narrative of the art show experience S cores on a standard measure of affect revealed a significant increase in negative emotion from baseline to post show. Diverting disclosure did not impact the extent of cognitive processing. Applications to sexual assault and PTSD, and limitations of the experimental design are discussed. Steven Graham Social Sciences Division
! $ THE EFFECT OF DIVERTED DISCLOSURE ON COGNITIVE PROCESSING WITH APPLICATIONS TO SEXUAL ASSAULT "There is no greater agony than bearing an untold story inside you." Maya Angelou When people encounter an upsetting experience, they most often seek to express that experience to others. They might do so simply to "get it off their chest," to seek support or comfort, or to gain a better understanding of the experience through considering the insight of others or conversationally developing their own inferences and opinions. When their chosen confidant is supportive and r eally listens, people tend to feel a little better about the situation, a little more aware, a little more assure. Conversely, when the listener responds with an unsupportive response, it sometimes makes the individual feel a little more upset, a little mo re confused, a little less certain. However, there are some responses of a more ambiguous nature such as the recipient actively diverting the individual's attempts to express their experience that have an unclear effect on the person. When a person is prevented from discussing an experience, they might start to wonder why they received such a response. Does this person not believe me? Do they not care what I have experienced? Have I done something wrong? Especially when disclosing a particularly distre ssing or stigmatizing experience, such as sexual assault, individuals might be especially sensitive and reactive to the responses of others. Thus, even when a response to disclosure is not overtly negative, it may nonetheless negate the benefits people typ ically derive from talking to a supportive listener.
! % Benefits of trauma disclosure Disclosure, or the act of sharing personal information or experiences with others, is understood as an important means through which people construct meaning about events incorporate those events into their identity, and express themselves to others (for review see Singer & Bluck, 2001). Additionally, disclosure can improve an individual's abilities to cope with distressing or traumatic situations, and even reduce associa ted physical and psychological symptoms (Pennebaker & Graybeal, 2001). Through disclosing events, individuals invite the recipient to provide reactions and responses that influence their own understanding of the experience (Nelson, 2003). The listener play s an essential role in the disclosure experience: an engaged and responsive listener can encourage the individual to elaborate on the details of their experience and develop greater insight and understanding, but an unresponsive or unaccepting listener can cause disclosure to fall flat (Bavelas, Coates, & Johnson, 2000). Most people tend to disclose notable everyday experiences, and personally significant experiences that remain undisclosed are most likely to be about an experience considered a serious transgression of social norms or appropriate behavior (Pasupathi, McLean, & Weeks, 2009). To get a sense of day to day disclosure behaviors, participants recorded the most memorable event of their day in an open ended diary format for one week, rated the emotionality of the event, and stated whether or not the event was disclosed. The researchers later coded events by type: relationships, trauma, leisure/exploration, achievement, and transgressions. Analyses of the diaries indicated that the majority of ev ents are disclosed by the evening of the day they occur. Further
! & analyses of emotion and event type did not prove fruitful, which was attributed to the low incidence of personally significant events in the short period studied. Moving away from the munda ne experiences reported in first study, the researchers asked participants to generate memories of undisclosed and disclosed events that were important to their lives (Pasupathi et al., 2009). Participants completed a questionnaire eliciting four narrative s: two regarding important experiences, one told and one not, and two regarding unimportant experiences, one told and one not. For each narrative, participants rated their emotions related the event, how important the memory is, and stated in a free respon se format why they chose not to disclose the undisclosed events. Relative to disclosed events, undisclosed events are more likely to involve the individual perpetrating a social transgression and are more emotionally negative and less emotionally positive. Individuals most often chose not to disclose an experience if they felt disclosure might result in aversive social consequences, such as punishment or shame (Pasupathi et al., 2009). Social responses to disclosure. Of course, disclosure cannot ex ist in a vacuum, and how the recipient of the disclosure responds likely influences the derived benefits. For example, attentive listeners encourage better recall after a three week delay and encourage individuals to remember more new and elaborate inform ation, than either a distracted listener or no listener at all (Pasupathi, Stallworth, & Murdoch, 1998). Undergraduate students viewed five movie clips, and immediately after viewing listed all the events recalled from each clip. Then, they chose two clips to discuss with two different "listeners;" these listeners were actually confederates, instructed to either remain attentive, or distracted by an irrelevant counting task. After both disclosure
! experiences participants rated their emotional state and answ ered questions about the listener and the overall experience. The three movie clips not chosen for recitation became part of the "no listener/ no retelling" condition by default. Researchers recorded the number of items spoken about during the conversation and participants returned three weeks later to once again list what they recalled of the five movie clips. Scores were determined by the proportion of items recalled from a previously decided checklist of important actions, events, descriptions, and dial ogues of each clip. When speaking to distracted listeners, participants produced significantly more disfluencies (e.g., "um") per minute and tended to speak for less time; they also reported that the experience was less pleasant (Pasupathi et al., 1998). Attentive listeners elicited more information during the conversation than participants provided immediately after viewing the movie, whereas individuals speaking to distracted listeners actually provided less information during the conversation than the y initially recalled. Attentive listeners also elicited more "new" information (not included in initial recall) than distracted listeners, and encouraged significantly more elaborative information, such as opinions and inferences. Interestingly, there was no significant difference in recall for clips spoken about to a distracted listener or not spoken about at all. The better recall for stories told to an attentive listener, in contrast to a distracted listener, suggests it i s not merely increased rehearsal time providing memory benefits, but a unique consequence of the interaction with the attentive listener. The increase in new and elaborative information suggests that there may be increased cognitive processing about the movie clips as a result of speaking to an attentive rather than distracted listener. Furthermore, the finding that there was no significant difference on recall between
! ( movies that were spoken about to a distracted listener and not spoke about at all, alludes to the powerful influence on ho w the listener responds as to whether or not disclosure will effectively facilitate a more elaborate and complete memory for events. In addition to enhancing processing and elaboration about experiences, disclosure in a supportive social environment tends to reduce the amount of distress and post traumatic stress symptoms associated with a stressful event. Talking about stressors with a supportive other contributes to reductions in intrusive thoughts and perceived stress, compared to speaking with an unsup portive respondent or not speaking at all (Lepore, Ragan, & Jones, 2000). Participants viewed a relaxing nature documentary followed by footage from the Holocaust, and indicated their levels of stress during each film using both physiological and self repo rt methods. After viewing the films, participants received instructions to either not talk about their response to the films at all, to talk independently, or talk with either a validating or invalidating confederate. Two days later, participants returned to view the films again and complete the same measures as before. At that time, all participants also completed two subscales of the Impact of Events Scale to measure intrusive thoughts and avoidance relative to the Holocaust film. Participants not given the opportunity to talk at all reported significantly more intrusive thoughts than participants instructed to talk about their reactions to the Holocaust film. Not talking about the film also contributed to higher perceived stress than either talking alon e or to a validating confederate. Furthermore, the extent of intrusive memories reported by participants fully mediated the relationship between talking condition and declines in perceived stress, whereas avoidance did not. Because intrusive thoughts are c onsidered to be sign of inadequate cognitive integration, the results suggest
! ) that the opportunity to organize and understand an experience through discussion may be responsible for the impact of disclosure on reducing distress (Lepore et al., 2000). In an y given day, people tend to disclose the majority of both common and notable experiences; however, many people choose not share some experiences that might make them the target of disdain or shame (Pasupathi et al., 2009). Speaking with others and receivin g collaborative feedback enables us to better process and understand our experiences, and to more effectively cope with distressing experiences (Bavelas et al., 2000; Pennebaker & Graybeal, 2001). Indeed, cognitive processing resulting from discussing even ts may be responsible for the relationship between disclosure and decreased distress (Lepore et al., 2000). However, how the listener responds is an essential and inalienable aspect of disclosure, which appears to sometimes enhance or mitigate the benefits often broadly ascribed to disclosure (Pasupathi, et al., 1998). Therefore, it is important to establish the relationship between disclosure, listener response, and cognitive processing, in order to understand the effect a listener's response to disclosure might have on coping following traumatic events. Disclosure and cognitive processing. Pennebaker personally conducted and furthermore inspired a plethora of studies on the impact of disclosing stressful events on improving various physical and psychologi cal health outcomes. Since beginning this research in 1989, Pennebaker has explored a number of possibilities to explain the reason for the physical and psychological benefits of disclosing stressful events. He conducted studies to address whether writing about events may be inspiring behavioral changes that would influence health, or changing the way in which trauma related events were represented in memory, but neither of these explanations could be empirically supported.
! Pennebaker ( 1993) conducted post hoc analyses of narratives from three previous studies using the expressive writing paradigm. He focused on participants' cognitive word use throughout their narratives as a potential explanation for the difference in who benefits the most and who benefits the least from expressive writing. The Linguistic Inquiry and Word Count (LIWC) software was used to produce a percentage of words reflecting cognitive processing, such as insight words (realize, understood) and causal reasoning words (because, why). On average over the course of the three to five day period, both those whose health did and did not improve used approximately the same amount of cognitive words. However, individuals with improved health went from using fewer cognitive words on the first day, to using an abundance of cognitive words on the last day of writing; in contrast, those whose health did not significantly improve used a consistent amount of cognitive words across all days of the experiment. Thus, written disclosu re can influence cognitive processing about events as reflected in the percentage of cognitive words used in describing the event. Furthermore, increased cognitive processing does appear to be implicated in positive changes in physical and psychological he alth. However, this study leaves open the question of the impact of verbal disclosure of events on cognitive processing, and even more so how different listener responses inherent in verbal disclosure influence these benefits. To begin to suggest the uniq ue influences of conversational disclosure, modifying the written disclosure paradigm by adding instructions to write an imagined dialogue leads to language use suggesting greater cognitive and emotional processing than in a traditional written narrative f ormat (Burke & Bradley, 2006). Undergraduates wrote
! + about a personal experience in one of three conditions: trauma narrative, trauma dialogue, or control writing. Individuals in the trauma narrative condition were instructed to write a narrative of a trau matic experience similar to Pennebaker's procedures, whereas those in the trauma dialogue condition received further instructions to write as if "talking to someone else." The dialogue included a written account of both what the participant imagined they s aid during the conversation, and how they imagined the recipient responded. The control writing condition wrote a narrative of how they spent their time the day before the experiment. Text analyses were conducted using the Linguistic Inquiry and Word Count (LIWC) software. Participants also completed the Positive and Negative Affect Schedule (PANAS) before and after writing. Participants in the dialogue condition demonstrated greater levels of cognitive processing as opposed to those in the narrative cond ition, as evident by the increase in causal and insight words. The dialogue group also used a greater proportion of affect words than the narrative group, and a greater proportion of the words were in present tense for the dialogue as opposed to the narra tive condition. Taken together, the heightened use of affect and present tense words suggests that participants writing dialogues may have experienced recounting the traumatic memory more vividly, and with more emotional engagement than the participants wr iting narratives. Additionally, the increased use of causal and insight words suggest that the "disclosure" enhanced cognitive processing relative to the event. The increases in cognitive and emotional processing when writing an imagined dialogue disclo sing a traumatic experience, as opposed to an independent narrative
! account, support the social account making model of adapting to trauma. The social account making model proposes that adapting to trauma involves engaging in intra and inter personal verb al disclosure behavior about events, which allows the person to construct a more coherent narrative of the experience; the more coherent narrative in turn contributes a more adaptive, verbally accessible memory for the traumatic events (Burke & Bradley, 20 06). This compliments the cognitive model of PTSD, and suggests that disclosure may be one factor influencing how much cognitive processing individuals engage in about the trauma. Disclosure, cognitive processing, and PTSD. The cognitive processing theor y of Post Traumatic Stress Disorder (PTSD) development is increasingly accumulating empirical support for its contention that the amount and quality of cognitive processing individuals engage in following traumatic events is one of the most significant con tributors to who develops PTSD. Most people report at least some symptoms of PTSD immediately after experiencing a traumatic event; however, for many of these people symptoms diminish in a few weeks or months after the event. For a subgroup of individuals symptoms persist for months or years, thus meeting the criteria for a PTSD diagnosis (Ehlers & Clark, 2000). According to the cognitive model of PTSD, the disorder occurs when individuals process a traumatic event and/or its sequelae in a way that produc es a sense of serious current threat (Ehlers & Clark, 2000). Two key processes lead to the lingering sense of current threat: individual differences in cognitive appraisals of the trauma and its sequelae, and individual differences in the nature of the mem ory for the event and its link to other autobiographical memories.
! $! Individuals with PTSD have difficulties viewing the trauma as a time limited event that does not have global negative implications for their future. This creates a sense of serious current threat that is maintained by inaccurate cognitive appraisals directed both externally (e.g. "The world is dangerous) and internally (e.g., "I am incapable of handling events."). Overgeneralizing from the traumatic event often leads to beliefs that normal experiences are actually more dangerous than the y are, which in turn exacerbates and maintains symptoms of PTSD, such as hyper reactivity and avoidance of trauma cues. For example, a wom a n who was raped by an acquaintance might subsequently attribute the assault to her not being able to read people well; thus, at social settings in the future, she might feel vulnerable and afraid, or avoid them altogether. Memory abnormalities represent o ne of the defining features of PTSD (McNally, 1998). Individuals affected by PTSD demonstrate an apparent discrepancy between intentionally retrieving a complete memory and involuntarily experiencing intrusive memories of the event. Intentional attempts to retrieve trauma related memories are often difficult and fragmented, unlike the cohesive narrative structure characteristic of memories unrelated to the trauma. On the other hand, involuntarily triggered "flash backs" of the trauma are often extremely viv id, but rely more on emotional and sensory content than cognitive or linguistic information. Unlike typical autobiographical memories that are understood as occurring in the past, involuntarily retrieved memories often feel as if they are happening right n ow, with the associated emotions and sensations being almost as strong as when the event originally occurred. Additionally, trauma memories such as these cannot be intentionally retrieved, and instead occur when the individual encounters an external (e.g. a stocking like the one the perpetrator used to
! $$ bind the victim) or internal (e.g. a body position a victim was forced into) stimulus reminiscent of the trauma. The qualities of involuntarily retrieved traumatic memories in PTSD, as outlined above, point to the role of cognitive processing in the development of PTSD. To explain, there a re two routes to retrieving autobiographical memories: higher order meaning based retrieval, or stimuli based triggering. Normal processing of autobiographical memories red uces the ease at which they can be unintentionally retrieved in everyday life. Instead, events are incorporated in the autobiographical memory knowledge base and organized according to themes and personal time periods, thereby enhancing meaning based retri eval and inhibiting stimuli based triggering (for a review see Singer & Bluck, 2001). This is an important feature of autobiographical memory, because it allows us to bring up memories when we want to, and to not be bogged down by irrelevant memories of stimul i we frequently encounter: imagine, for example, how taxing it would be if every time a person saw a peanut butter sandwich, s/he suddenly felt all the emotions and sensations ever felt when eating previous peanut butter san dwiches. Ehlers and Clark (2000) explain that the abnormalities observed in PTSD are primarily the result of insufficient or inappropriate cognitive processing about the trauma: "We propose that in PTSD one of the main problems is that the trauma memory i s poorly elaborated and inadequately integrated into its context in time, place, subsequent and previous info, and other autobiographical memories. This explains problematic intentional recall (weak semantic route to retrieval), the "here and now" quality (no context in time, hence the perception of current threat), the
! $% absence of links to subsequent info (e.g. "I did not die."), and the easy triggering by physically similar cues." (Ehlers & Clark, 2000; p. 325) Clearly cognitive processing appears to play a role in which individuals develop PTSD following traumatic events. However, the cognitive model of PTSD does not address factors that might explain individual differences in cognitive processing. Understanding the relationship between disclosure, listener responses, and cognitive processing may assist in addressing this gap. Dunmore, Clark, and Ehlers (1999) generated preliminary support for the cognitive model through a retrospective survey design that indic ated negative appraisals of the trauma and its sequelae were significantly associated with the onset and maintenance of PTSD. Lending support to their proposal that cognitive processes are fundamentally implicated in the development of PTSD, and not merely correlates, the researchers devised a prospective study following survivors of assault across a nine month period (Dunmore, Clark, & Ehlers, 2001). Victims of a recent physical or sexual assault took part in a semi structured interview capturing their psy chiatric history and the characteristics of the assault, and completed a battery of questionnaires assessing trauma relevant cognitions identified by Ehlers and Clark's model (2000). Participants completed the PTSD Symptoms Scale: Self Report Version (PSS SR) and Beck Depression Inventory (BDI) in the initial session and each month after for a series of nine months. In accordance with previous studies, PTSD symptom severity tended to decrease over time, and subjective perceptions of the event (e.g. percei ved threat) affected symptom severity more than objective characteristics of the event (e.g. weapon presence; Kilpatrick, Saunders, Amick McMullan, Best, Veronen, & Resnick, 1989). The majority
! $& of cognitive variables were significantly associated with PTS D severity at the initial, six months, and/or nine months markers. Through a multiple regression analysis, the researchers found that six of the fifteen cognitive variables significantly contributed to PTSD scores above and beyond the effects of individual and assault characteristics. These relationships persisted independently even when controlling for assault characteristics and initial PTSD severity. Participants' negative appraisals of other's responses to their trauma disclosure significantly contribu ted to PTSD symptom severity at the initial measurement, six months, and nine months later. A path analysis, controlling for initial PTSD severity, revealed a direct relationship between negative appraisals of other's responses and PTSD severity nine month s later. Thus, negative reactions given to assault disclosures may significantly contribute to development of persistent post traumatic stress symptoms qualifying for PTSD diagnosis in the months following an assault. Importantly, the focus on the individual's appraisals of other's responses highlights that the perception of the response is more important than whether or not the response was intended to be negative. Therefore, even ambiguous responses to disclosure could conceivabl y contribute to the persistence of PTSD symptoms if the individual interprets the response as negative. Thus cognitive processing relative to a traumatic event particularly individual appraisals of the trauma and its sequelae significantly contribute s to individual variability in the persistent of post traumatic stress symptoms. Dunmore, Clark, and Ehlers (2001) argue that avoiding thinking or talking about a traumatic event prevents elaboration and processing of the trauma memory. Furthermore, they c ontend that successful treatment of PTSD requires exploration and correction of the maladaptive
! $' cognitive processes, and ultimately elaboration of the trauma memory (Dunmore, Clark, & Ehlers, 2001). This assertion is supported by experimental evidence that frequency, vividness, and amount of distress associated with intrusive memories (a hallmark symptom of PTSD) gradually decreased during five sessions of Cognitive Behavioral Therapy (Hackmann, Ehlers, Speckens, & Clark, 2004). In summation, disclosure to supportive others is associated with a variety of benefits that may be explained by enhanced cognitive processing. Therefore, in reference to the cognitive model of PTSD, disclosure may be implicated in who experiences the most post traumatic stress sympt oms following a traumatic event. However, how the recipient responds to disclosure is another factor that may mitigate or enhance the effect of disclosure on cognitive processing. In order to gain a more complete picture of the role of disclosure in PTSD d evelopment, the effect of various listener responses to disclosure on cognitive processing relative to events must be discussed. For the purposes of this study, these factors will be discussed in terms of sexual assault. Applications to the diverted discl osure of sexual assault The relationship between disclosure and listener response, cognitive processing, and PTSD may be especially relevant to sexual assault, because it is frequently undisclosed or responded to with unsupportive reactions, both of whi ch are associated with increased post traumatic stress symptoms (Ahrens et al., 2010; Filipas & Ullman, 2001). Additionally, victims of sexual assault exhibit higher symptoms of traumatic stress relative to physical assault victims, and the characteristics of sexual assault disclosure and associated social responses may contribute to this distinction (Elliot, Mok, and Briere, 2004).
! $( Sexual assault requires attention from researchers, because, unfortunately, so many women and men experience this type of vio lence during their lifetime. Elliot, Mok, and Briere (2004) investigated the prevalence, associated symptoms, and sex differences in the general United States population regarding adult experiences of sexual assault. All participants completed two surveys: the Traumatic Events Survey, which assess personal experiences of sexual and physical trauma during both adulthood and childhood, and the Trauma Symptom Inventory, which is a standardized clinical measure of trauma related symptoms. In this study, 22% of women and 3.8% of men in the general U.S. population reported an experience of sexual assault at some point during their adult years; furthermore, about half of sexual assault victims reported experiencing assault more than once. Victims of adult sexual a ssault (ASA) are most likely to be female and have a lower income; however, race, amount of education, and employment status do not appear to be associated with ASA status. A lover or a spouse assaulted the majority of participants reporting a sexual assau lt: women were especially likely to be assaulted by a lover or a spouse compared to sexually assaulted men. This statistic suggests that "non stereotypical" assaults that is, sexual assaults not committed by a stranger with a weapon or extreme physical force actually make up the majority of assaults. Whereas similar studies have been critiqued for lacking economic and racial dive rsity, this sample was comparable to 1990 U.S. census data on most demographic variables and therefore is likely to be a good representation of the actual rates of sexual assault in the United States. None the less, the researchers suggest that these num bers might reflect lower rates than are actually present in the population, because sexual
! $) assault tends to be underreported; however, since these results categorized participants according to their responses on the Traumatic Events Survey, they are lik ely more accurate than studies that rely on participants to self identify as sexually assaulted. These results also indicate that, while women are more likely to be victims of sexual assault, men are also affected. This is an important finding because much of the research on sexual assault has implicitly or explicitly focused on women as the victims and men as the perpetrators. Sexual assault is not only important to examine because of its prevalence: it also tends to cause a significant amount of distress in victims. In the study by Elliot, Mok, and Briere (2004), participants with a history of ASA reported more symptoms than their non assaulted peers on all 10 subscales of the Trauma Symptom Inventory, even when controlling for the confounding experience of Childhood Sexual Abuse (CSA) and other demographic variables. Perhaps most shockingly, an average of 14 years had elapsed since participants' last experience of sexual assault, yet the majority of victims still reported significant levels of distress in the present day. Male victims of sexual assault tended to report more symptoms of post traumatic stress than female victims. Unlike experiences of sexual assault, adult experiences of physical assault were not independently related to higher scores on the TSI in this sample. Thus, sexual assault strongly impacts trauma related symptoms and self reported feelings of distress, and perhaps slightly more so for male victims; it is also important to note that the trauma related symptoms and distress persisted for many years following the actual experience of sexual assault. This illustrates that the old adage that "time heals all wounds," may not be enough in cases of sexual assault, emphasizing the need for specific
! $* interventions that could lessen trauma relat ed symptoms. Additionally, the finding that physical assault was not independently related to trauma symptoms reflects how sexual assault, even compared to other types of interpersonal violence, appears to be particularly and persistently distressing for i ndividuals. Finally, it is interesting to recognize that sexually assaulted males tended to display more trauma related symptoms than sexually assaulted females. One potential explanation for this finding may be that there is even less of a culturally endorsed script or narrative for how males should respond to sexual assault compared to females, and there is considerable pressure on men to conform to the male sex role of being strong and aggressive, particularly in the sexual domain. Therefore, men may be less able to put their experience of victimization into words, and especially may be less willing to share their experience with others. With that in mind, it is important to understand the patterns of disclosure and listener responses to sexual assault, and what e ffect these might have on PTSD development. Sexual assault disclosure. Recall th at people generally disclose the majority of day to day events by the evening of the day on which they occurred (Pasupathi et al., 2009). People are even more apt to disclose notable, personally significant events whether they are positive or negative. However, it is also apparent that there are some events that individuals choose not to disclose, and that these events are mo st often "transgressions," or violations of social norms or accepted morals perpetrated by the individual (Pasupathi et al., 2009). This finding is interesting in terms of disclosure of sexual assault, because the social stigma that still surrounds sexu al assault due to lingering rape myths make it an experience that is not commonly or comfortably
! $+ disclosed, and at worst may even be subject to unfair character or moral judgments. Thus, we turn now to data on the rates and characteristics of sexual assaul t disclosures. In the past, disclosure has been primarily thought of as a dichotomous, yes or no variable: a personal event is either spoken about, or it is not. However, some studies looking at the specifics of disclosure behaviors indicate that th is is not always the case. The experience of disclosure, therefore, may be subject to a variety of differences based on individual and social factors, which could conceivably influence the impact disclosing an event has on the person. Therefore, identifyin g different patterns of sexual assault disclosure and the characteristics and outcomes associated with each one, may be important in determining whether or not talking about sexual assault will produce the benefits typically associated with full disclosure. In fact, there are four distinct patterns of sexual assault disclosure (Ahrens, Stansell, & Jennings, 2010). Trained Undergraduate and Masters students interviewed survivors regarding their sexual a ssault experience(s) and their disclosure history for the assault, or the most recent assault if more than one had occurred. Responses regarding the assault experience were coded for the total number of past assaults, for non stereotypical assault characte ristics ("stereotypical" assaults were defined as being executed by a stranger using extreme physical force, such as a weapon), and whether or not the participant considered the experience rape at the time that it happened. A disclosure timeline was creat ed for each participant, starting form the most recent experience of sexual assault and encapsulating how many people were told, how many of those people were police or medical officials, and how long after the experience each disclosure occurred. Particip ants also completed a battery of survey materials
! $, relating to social reactions to disclosure, depression, post traumatic symptoms, and physical health symptoms. Four distinct patterns of disclosure following sexual assault emerged from analyses of the dis closure timelines: non disclosers, slow starters, crisis disclosers, and on going disclosers. The non disclosers category represents individuals who had never disclosed the sexual assault prior to coming in for the interview. Slow starters waited at leas t two weeks before disclosing, but continued to talk about the event following initial disclosure; these individuals waited, on average, four years before first disclosing the experience and told four people. Crisis disclosers first disclosed within two da ys of the assault, but their last disclosure occurred within the first week of the assault. Crisis disclosers tended to confide their experience in fewer people than the other three groups. Finally, the on going disclosers initially disclosed their experie nce within the first week of the assault and continued to talk about their experience throughout their lives. On going disclosers on average told more people about their sexual assault than any other group. Survivors of sexual assault were almost equally a s likely to be categorized as a n on discloser, delayed discloser, crisis discloser, or on going discloser. The sexual assault experience of those in the non disclosers group tended to have more features of non stereotypical assaults (e.g. assaulted by a loved one), and they were more likely than other groups to not consider the experience rape at the time of the assault. This alludes to the powerful impact cultural assumptions about sexual asault have on survivors: when the experience does not fit the cultural narrative of what constitutes rape, people may be less likely to talk about their experience or seek support. There was no significant difference between the groups in positive responses they received upon
! %! disclosure. However, crisis survivors received significantly more negative responses compared to slow starters. This may indicate that crisis survivors stopped disclosing because of the negative responses they received initially, which highl ights the importance of social reactions to disclosure of sexual assault (Ahrens et al., 2010). Even when choosing to disclose the assault, survivors spoke with very few people (two to five on average), which emphasizes the potential for each individual r esponse to be more impactful. The types of listener responses received to disclosure may influence some of the differences in how often people disclose, how many people they disclose to, and, most dishearteningly, whether they stop disclosing all together (Ahrens et al., 2010). Thus, it is important to examine what types of responses individuals receive to sexual assault disclosure, and whether these are or are not perceived as supportive. Social reactions to sexual assault disclosure. Survivors of sexual assault receive a variety of helpful and unhelpful reactions upon disclosure to informal and formal support providers (Filipas & Ullman, 2001). Female survivors volunteered to complete a survey packet that included measures related to the characteristics o f the assault (Sexual Experiences Survey), when and to what extent the participant first disclosed (two Likert scale items), social reactions to disclosure, and psychological adjustment. Social reactions were assessed in two question formats: closed ended and open ended questions. The Social Reactions Questionnaire (SRQ) provided the closed ended data, asking participants to rate the frequency with which they received a variety of positive and negative responses to their disclosure of sexual assault. Open e nded questions prompted participants to describe the four responses that were most helpful, most unhelpful, most "uniquely helpful" (particularly helpful in a way other responses were not), and one
! %$ response that they wished they received but did not. After each question, participants indicated who made the response they mentioned (i.e., partner, friend, family member). The average age of participants at the time of the assault was 20 years and an acquaintance or romantic partner most often committed the a ssault, in keeping with the findings reported for the general population in the United States by Elliot, Mok, and Briere (2004). Most of the participants in this sample disclosed their experience of sexual assault to three people on average, but many parti cipants delayed disclosure for months or years and/or reported that they did not discuss the assault in detail. Participants most often reported disclosing to a friend or relative, followed by mental health professionals, and romantic partners. When cons idering formal and informal support providers together, participants reported a greater frequency of negative social responses to their disclosure of sexual assault compared to positive responses. However, friends and romantic partners tended to provide mo re helpful responses, especially emotional support and listening to the victim talk about the experience and their emotions. Reactions from support providers commonly endorsed as helpful include listening to the survivor, not blaming her, and encouraging h er to talk about her experience or not distracting her with other things. The most common unhelpful responses included stigmatization responses (e.g., "My mom disowned me and ignored the situation.") and statements perpetuating rape myths (e.g., "You had no business gong out dressed like that!"). The most common rape myths espoused by responders included blaming the victim for "inappropriate" clothing choices or allowing herself to be alone in the perpetrator's home, and the impossibility of calling a nonc onsensual sexual experience rape if committed by a boyfriend or spouse.
! %% Police officers and family members were most likely to respond to sexual assault disclosure with disbelief and rape myths. This research is valuable because the combination of closed and open ended questions produced a breadth of material about how recipients frequently respond to sexual assault disclosure, and which of the responses survivors perceive as helpful or unhelpful. The data suggest that survivors' receive a variety of posi tive and negative responses, with a somewhat higher frequency of negative responses; thus, negative social responses to sexual assault, such as stigmatizing or blaming, remain a potential exasperator of the associated distress. But does sexual assault disc losure and associated social responses actually affect post traumatic stress symptoms and associated disorders (PTSD)? Is it enough to just say that the assault happened, to "get it off your chest," or does how the listener responds and whether or not they allow the survivor to discuss the event influence the effect disclosure has on post traumatic stress? Sexual assault disclosure, listener response, and PTSD. Different physical and mental health outcomes are associated with the four distinct patterns of sexual assault disclosure, even when controlling for confounding variables such as negative social reactions to disclosure, the number of past assaults, and the time since assault (Ahrens et al., 2010). Not disclosing the assault predicted health outcomes on the posttraumatic stress and depression inventories above and beyond the e ffects of the control variables (Ahrens et al., 2010). Among the three patterns of disclosure, there was no significant difference in amount of posttraumatic stress, depre ssion, or physical health symptoms. However, individuals reporting more negative social reactions to disclosure also reported more symptoms on all three measures. Thus, it is evident that negative social reactions
! %& may negate the benefits of disclosure to s ome degree, and powerfully influence coping following sexual assault. One major limitation to this study, as with most studies of sexual assault, is that individuals self identifying as sexually assaulted may differ, perhaps significantly, from those who do not or do not choose to do so openly. Furthermore, the interview process represents an additional, or in some cases initial, incident of disclosure, which fundamentally changes the label ascribed to non disclosers and crisis disclosers. However, the res ults of the depression and posttraumatic stress scales reveal that participants have levels of symptoms exceeding the levels observed in the general population, and which are more comparable to clinical populations in regards to depression and posttraumati c stress. Thus, it is important to generate more research about factors that hinder or promote recovery from post traumatic stress following sexual assault. Studies have examined a variety of variables attempting to understand why some survivors of sexual assault develop PTSD while others do not. Earlier studies tended to focus on pre assault individual factors, such as trauma history, or factors related to the characteristics of the assault itself, such as victim offender relationship and assault severity However, a substantial relationship between individual or assault factors and PTSD symptoms has not been consistently demonstrated, and recent research has shifted focus to post assault experiences, such as attributions of blame and social reactions, a s potential contributors to PTSD development. In an attempt to sort out variables that do and do not contribute to PTSD, Ullman, Fillipas, Townsend, and Starzynski (2007) devised a regression equation capturing
! %' demographic, pre assault, assault, and post assault characteristics. Victims completed a survey packet capturing a broad range of variables, including demographics (age, race, income, and education), assault characteristics (Sexual Experiences Survey), non sexual traumatic experiences (Stressful Lif e Events Screening Questionnaire), and the quality of their social support network (Social Activities Questionnaire). Additionally, post assault experiences measured include disclosure behavior, social reactions to assault disclosure (Social Reactions Ques tionnaire), avoidance coping strategies (Brief COPE), and PTSD symptoms (Post Traumatic Stress Diagnostic Scale). Post assault experiences were located last in the regression equation to provide a conservative estimate of their impact on PTSD symptoms. Re sults indicated that the regression model was significant, and explained 45% of the variance in PTSD scores. Demographic variables did not significantly predict symptom severity. Assault characteristics, such as victim offender relationship and assault sev erity, were not related to PTSD symptoms, with the exception of perceived life threat. In accordance with previous research, the number of other traumatic events and a history of childhood sexual abuse were both significantly associated with PTSD symptoms. However, the post assault variables accounted for the largest amount of variation in PTSD symptoms compared to all other variable groups: demographics, pre assault and assault variables together explained 20% of the variation in PTSD score, whereas post a ssault variables alone explained 25%. In particular, delayed disclosure, self blame, and negative social reactions significantly contributed to heightened PTSD symptoms. Thus, the social sequelae following experiences of sexual assault may actually
! %( be more responsible for who develops PTSD than the characteristics of either the individual or the assault. It is evident that not disclosing or disclosing after a significant delay contributes to experiencing symptoms of post traumatic stress followings sexual assault (Ahrens et al., 2010; Ullman et al., 2007). Additionally, negative social responses to disclosure, such as blaming or disbelief, also contribute to increases in symptoms of PTSD (Ullman et al., 2007). Most studies examining social responses to sexu al assault disclosure have focused on responses that can be discretely categorized as positive or negative. However, it is likely that conceptualizing listener responses in such a way fails to capture the complexity of how disclosure recipients react to a disquieting topic such as sexual assault. Diverted Disclosure. Distracting the victim from talking about the assault represents one response that is sometimes construed as helpful, and at other times as unhelpful (Filipas & Ullman, 2001). Whether distract ion is perceived as helpful or not likely depends on the individual's interpretation of the responder's intention, and the timing of the response. For example, distraction may be positive when in the form of an invitation to do something to "get your mind off it" or a compassionate suggestion not to dwell on it. However, victims often report distraction as exceptionally unhelpful if they are attempting to express the experience to another but are interrupted or discouraged from continuing. One survivor desc ribed her unhelpful experience of diverted disclosure in the following words: "The person I was talking with became uncomfortable and tried to say a couple of words like Oh, that's awful,' or What a jerk,' and quickly changed the subject."
! %) Many particip ants indicated that being distracted from discussing the assault was an unhelpful response, or one that they wished they had not received (Filipas & Ullman, 2001). In fact, one of the least desired responses to sexual assault disclosure was to "be distract ed or discouraged from talking about the experience." Conversely, 66.7% of participants reported that a friend encouraging them to talk about the experience was the most "uniquely helpful" response to disclosure. Furthermore, a large percentage of particip ants reporting an experience of diverted disclosure chose being encouraged to talk about the experience as their most desired, but previously not received, social response. Interestingly, this study revealed a common social response that has not received much attention in the literature on disclosure of sexual assault. It seems that recipients of sexual assault disclosure frequently respond in a way that is neither overtly de grading nor supportive to survivors, by simply not responding at all. This behavior, which I refer to as diverted disclosure involves initially listening to the survivor confide that someone sexually assaulted him or her but providing minimal response and soon after distracting or discouraging the survivor from expressing or discussing the experience. In regards to diverted disclosure of sexual assault, Clark and Ehlers (2000) raise an interesting point of how a trauma victim might perceive unresponsive recipients of disclosure. Others are often unsure of how to respond to trauma victims particularly in regards to sexual trauma and may avoid talking about the event out of personal discomfort or in an attempt to not d istress the victim. Victims may interpret this lack of response as a sign that no one cares, or that others are even blaming them for what happened. As Clark and Ehlers put it,
! %* "Such interpretations are likely to directly produce some symptoms of PTSD (es trangement from others and social withdrawal) and are also likely to prevent victims from discussing the trauma with others, hence reducing the opportunity for therapeutic reliving and for feedback from others that might help correct excessively negative v iews about the meaning of the event." (Ehlers & Clark, 2000; p. 323). This quote alludes to the potential for diverted disclosure to discourage thorough cognitive processing about the traumatic assault experience While not necessarily intended as a negative response, diverted disclosure could have a similar effect on cognitive processing and associated post traumatic stress as non disclosure or an overtly negative response, such as victim blaming. However, this particular response has not been studi ed in the literature connecting listener responses to cognitive processing. The Current Study Disclosing experiences to attentive and supportive listeners is associated with a variety of benefits, including enhancements in memory, increases in elaborati ve information such as inferences and insights, and reduced distress; however, the same benefits are not derived from disclosure to an inattentive or unsupportive individual (Pasupathi et al., 1998; Lepore et al., 2000). One potential explanation of these benefits might lie in attentive listeners facilitating cognitive processing about events: cognitive processing appears to mediate the relationship between disclosure and lowered distress, as well as improved physical health (Lepore et al., 2000; Pennebaker 1993). Furthermore, maladaptive cognitive appraisals regarding the trauma and its sequelae,
! %+ including negative appraisals of social responses, contribute to PTSD development above and beyond characteristics of the individual or the trauma (Dunmore et al. 2001). The relationship between disclosure, cognitive processing, and PTSD may be particularly relevant to sexual assault, because sexual assault is frequently not disclosed or responded to negatively, and is more likely to result in PTSD than non sexua l assault (Ahrens et al., 2010; Elliot et al., 2004; Filipas & Ullamn, 2001). The influence of non disclosure, delayed disclosure, and negative social reactions on PTSD development supports this assumption (Ahrens et al., 2010; Ullman et al., 2007). Resear ch on social responses to sexual assault disclosure tend to focus on the dichotomous influences of positive responses on enhancing coping versus the influences of negative responses on hindering coping. The impact of ambiguous responses on PTSD development such as actively deterring individuals from fully speaking about the details of the event and their reaction to it, has not been addressed. Likewise, the impact of this particular response on cognitive processing is not present in the literature on the b roader benefits of disclosure. Thus, the question remains as to the effect of diverted disclosure on the extent of cognitive processing individuals engage in regarding a negative emotional experience. An art show speaking out against sexual assault was created to be the stimulus, because most individuals report that hearing about personal experiences of sexual assault is moderately upsetting (Banyard, Moynihan, Walsh, Cohn, & Ward, 2010). Participants completed the Positive and Negative Affect Schedule ( PANAS) scales before and after viewing the art show to confirm that it increased negative affect. Then, participants either disclosed their experience of the art show or talked about an unrelated topic with an experimental confederate instructed to allow full disclosure or divert the discussion.
! %, Finally, participants wrote a free response narrative of their reaction to the art show; these narratives were subsequently analyzed for word use indicating cognitive processing using the Linguistic Inquiry and Wor d Count (LIWC; Pennebaker, Chung, Ireland, Gonzales, & Booth, 2007 ). It was hypothesized that disclosing the art show experience would increase cognitive processing about the event. Furthermore, it was hypothesized that a diverting response from the list ener would decrease cognitive processing, but only in the disclosure conditions. Thus, Non Diverted Disclosure would produce higher levels of cognitive processing than Diverted Disclosure, which would not differ from either Non Disclosure condition. If the se hypotheses are confirmed, then diversion may be one of many listener responses that can actually mitigate the benefits generally associated with disclosure. This finding would have particular implications for disclosure of sexual assault, to which dive rted disclosure is a common response. Additionally, if diverted disclosure influences cognitive processing it could represent yet another facet of the social environment that could contribute to the high incidence of PTSD development following sexual assau lt. Method Participants Students at a small liberal arts college in Southwest Florida participated in this study (n = 24). Participants responded to a student forum email requesting people willing to participate in a study in which they would view and subsequently discuss an art show speaking out against sexual assault. A five dollar gift card to a local coffee shop was offered as compensation. The average age of participants was 20. 58 years (SD = 1.61);
! &! 20 participants identified as female, two participants identified as male, one participant identified as female bodied, and one participant identified as transgendered male. Materials Stimulus. All participants viewed an art sho w raising awareness about the frequency of sexual assault, the challenges the survivors face, and the silence and even hostility that often permeates discussions of sexual assault in our society. Members of the student body and the researcher created the a rtwork about personal experiences of sexual assault. The art show was expected to invoke a mild amount of stress in participants, in keeping with literature that suggests talking to a victim of sexual assault is mildly stressful for the confidant (Banyard et al., 2010). The art show took place in a community space on campus previously used for student art shows. Four students displayed artwork in the show, and there was a total of seven pieces of visual art and three poems. Four posters displaying stat istic s related the frequency of sexual assault, assault disclosure, PTSD development, and offender prosecution were also included as part of the show. Refer to Appendix A through Appendix F for images of the artwork included in the stimulus. Affect. Participants completed the Positive and Negative Affect Schedule (PANAS; Waston, Clark, & Tellegen, 1988) to assess their emotional state at baseline and after viewing the show. The PANAS is composed of two 10 item mood scales addressing overall positive and negative affect. Participants rated a list of 20 intermixed positive and negative affect words according to how much they felt each emotion "right now, that is, at this very moment." Each item was rated on five point scale from 1 (very slightly or not at all) to 5 (extremely). Watson and colleagues demonstrated the factorial
! &$ independence, highly internal reliability, and validity of the scales. See Appendix G for the PANAS Manipulation check. Participants completed an edited form of the Social Reactions Questionnaire (SRQ) in order to determine whether or not the diversion manipulation was effective (Ullman, 2000). The SRQ assesses social reactions commonly experience d in response to sexual assault, and is comprised of seven subscales: Emotional Support/Belief, Treat Differently, Distraction, Take Control, Tangible Aid/Informational Support, Victim Blame, and Egocentric. The Emotional Support/Belief and Distraction subscales were used for the purposes of this study. The items on the SRQ incorporate some social reactions that are specific to sexual assault and would not be applicable to the art show experience (e.g., "Told you that you were not blame.") as well as some responses that cou ld be generalized to any kind of disclosure (e.g., "Tried to discourage you from talking about the experience."). Therefore, in order to create a questionnaire capturing how much participants felt the confederate either supported or distracted them from d iscussing their reactions to the art show, the sexual assault specific items were removed from the Emotional Support/Belief and Distraction subscales leaving only items that could be generalized to the art show experience. Thus, four of the original 15 ite ms constituted the revised Emotional Support/Belief subscale, and four of the original six items composed the revised Distraction subscale. The shortened Emotional Support/Belief subscale demonstrated adequate internal reliability (Chronbach's = 0.65). T he four item Distraction subscale produced an unacceptable level of internal reliability (Chronbach's = 0.19); items seven and eight were dropped from the survey,
! &% producing an improved internal reliability for the two item scale (Chronbach's = 0.62). S ee Appendix H for the revised SRQ. Cognitive processing. Participants wrote a narrative of their personal experience of the art show on a sheet of paper bearing the following instructions, which were also verbally provided by the researcher: "Please ta ke a moment to write down your personal experience of the art show. You may write about whatever you wish to include, such as descriptions of what you saw, thoughts you had, or emotions you experienced. There is no minimum or maximum writing time, and you are free to write as little or as much as you feel you need to say." The researcher typed participants' hand written narratives verbatim into digital text files. The extent of cognitive processing present in the narrative recollections was assessed using t he computer based microanalytic linguistic analysis system Linguistic Inquiry and Word Count (LIWC2007; Pennabker, Chung, Ireland, Gonzales, & Booth, 2007). The LIWC2007 refers to an extensive dictionary to identify the proportions of particular target w ords used in a text sample. The LIWC2007 is used to identify word use related to 22 standard linguistic dimensions (e.g. pronouns, verbs tense), 32 word categories capturing psychological constructs (e.g. affective, cognitive, and perceptual processes), 7 personal concern categories (e.g. work, home, and leisure activities), 3 paralinguistic dimensions (assents, fillers, nonfluencies), and 12 categories related to punctuation usage (periods, commas, etc.). The influence of diversion and disclosure condit ions were only examined in terms of the total word count, linguistic dimensions, and psychological constructs.
! && Procedure Upon arriving at the art show location, the researcher reminded participants that they would view art created by members of the student community regarding personal experiences and reactions to sexual assault. The researcher once again emphasized that if a t any point the participant felt that viewing or talking about artwork describing the experience of sexual assault had become too upsetting, they were welcome and encouraged to stop participating in this study. Once any questions were answered, participant s signed the informed consent indicating they understood the study. Before viewing the art show, participants completed the PANAS with paper and pencil in a separate room close to the art show location. Then, participants were given seven minutes to view the art show independently while the researcher waited outside. Immediately after viewing the show, the researcher led the participant back to the original room and seated him or her at a table to complete the PANAS for a second time. In order to maintain anonymity, participants placed all experimental measures inside of a manila envelope marked only with a number and condition immediately after completion. Experimental conditions were randomly assigned according to the order in which participants volunteer ed for the study, because there was no reason to believe the order would not be random: for example, the researcher assigned the first volunteer to condition one, the second volunteer to condition two, and so on repeating that pattern. Four conditions w ere determined by a 2 X 2 design in which the variables included whether or not participants disclosed the art show experience, and whether or not the confederate diverted the discussion.
! &' The researcher told participants that the next section was intended to examine how students communicate about experiences with art. In all four conditions, participants were introduced to a confederate described as a student who recently viewed the art show and were given two minutes in which to talk. All participants int eracted with the same female confederate. In order to create a situation in which only the participant disclosed, the researcher instructed both participant and confederate to reach into a box and select slips of paper that would randomly assign them to be either the "speaker" or the "listener;" however, both slips indicated "speaker" and the confederate claimed hers said "listener" in order to ensure the participant was the one disclosing. In both disclosure conditions, participants were verbally instruct ed to speak about their experience and reactions to the art show they just saw, whereas the confederate was instructed to listen and respond "as you would in any other conversation." In the Non diverted Disclosure condition, the confederate responded b y listening attentively for the entire two minutes maintaining eye contact and providing simple, naturalistic responses designed to encourage participants to continue talking, but not to influence the content of their disclosure. In the Diverted Disclosur e condition, the confederate responded for the first minute with eye contact and simple responses; after one minute was over, the confederate abruptly changed the subject by interjecting the question, Let's not talk about this what're you taking this sem ester?" The confederate responded attentively to the new topic, and if the participant tried to resume talk about the art show, the confederate directed his/her attention to the new topic again. In both non disclosure conditions, participants were not give n the opportunity to disclose their reactions to the art show, and instead were instructed to describe their
! &( "experience with art, dance or music, including any shows you have attended, classes you have taken, or things you have created." In the Non Diver ted Non Disclosure condition, the confederate responded to the unrelated discussion as described in the Non Diverted Disclosure condition. In the Diverted Non Disclosure condition, the confederate responded as described in the Diverted Disclosure condition Following the manipulation, the researcher thanked the confederate for their "participation" and removed them from the room. For the first 11 participants, the researcher administered the SRQ immediately after the confederate interaction. Upon finding t hat the SRQ appeared to be revealing the other student to be an experimental confederate, the experimenter administered the SRQ after participants completed their response to the show for the remaining 13 participants. Either immediately before or after co mpleting the SRQ, participants were asked to write down their personal response to the art show: the instructions dictated that they may write about whatever they want, and may write as little or as much as they would like. Participants were given as much time as they need ed to write the narrative (time range = 1.57 to 24.11 minutes; M = 8.03, SD = 5.12). Finally, participants answered two open ended questions regarding their gender and age, and placed them in the anonymous envelope containing all other mea sures. After completing the study, participants were asked if they had any questions or comments about the artwork or the study. They were then informed of the true purpose of the study. Additionally, they were told that the individual they spoke with was a confederate, and was instructed to behave as she did in the conversation. Finally, participants were thanked and given the gift card. Attached to the gift card was a slip of paper providing them with a website to visit for more information about sexual assault
! &) (RAINN.org) and the telephone number and instructions to call the college's Counseling and Wellness Center if the study had upset them in any way. Results Changes in Affect from Baseline to Post Show Measures of affect were collected before and after viewing the art show, and compared using a paired t test, to determine whether the stimulus effectively increased negative emotions. Participants reported higher levels of negative affect after viewing the show (M = 17.83, SD = 5.01), compared to before viewing the show (M = 13.5, SD = 3.71). Thus, the art show produced a statistically significant increase in overall negative affect, t(23) = 4.21, p < .001. Specifically, as can be seen in Table 1, there was a significant increase in the extent to which participants reported feeling distressed (t(23)= 5.72, p <.0001), upset (t(23) = 7.37, p <.0001), guilty (t(23) = 2.07, p = .045), scared (t(23) = 2.07, p = .05), ashamed (t(23) = 2.29, p = .032), and afraid (t(23) = 2.84, p=0.009). The mean score on overall positive affect was virtually unchanged before (M = 27.96, SD = 6.05) and after viewing the show (M = 27.92, SD = 8.24). Thus, the art show did not produce a statistically significant change in po sitive affect, t(23) = 0.03, p = .976. However, feelings of enthusiasm significantly decreased (t(23) = 4.45, p<.001), whereas reported inspiration significantly increased (t (23) = 4.75, p<.0001). In summation, while the change in positive affect may h ave been slight or inconsistent, the stimulus effectively induced negative affect in a variety of domains.
! &* The Effect of Diversion and Disclosure on the SRQ In order to determine whether the confederate successfully distracted participants in the diver sion conditions, participants completed the revised Emotional Support/Belief and Distraction subscales of the Social Reactions Questionnaire (SRQ) in reference to their interaction with the "other participant." Participants in the Diverted Disclosure group rated the confederate as the most distracting (M = 1.25, SD = 0.61), followed by Diverted Non Disclosure (M = 0.33, SD = 0.41), Non Diverted Disclosure (M=0.25, SD=0.42), and finally Non Diverted Non Disclosure (M= 0.17, SD = 0.26). Even when indicating t hat the confederate was distracting, respondents tended to, at most, indicate that they were "rarely" or "sometimes" distracting; no participants indicated that the confederate was "frequently" or "always" distracting. A 2 X 2 analysis of variance was con ducted to determine the impact of disclosure and diversion on perceiving the confederate as distracting. Participants met with a diverting confederate did report that the "listener" distracted them from discussing their experience more so than non diverted groups, F (3,20) = 10.43, p = .004, R 2 = .240. However, participants who disclosed about their experience of the art show also experienced the listener as more distracting, F (3,20) = 7.66, p = .012, R 2 = .176. Additionally, there was a statistically sign ificant interaction between disclosure and diversion such that individuals experienced the confederate as more distracting when attempting to disclose their experience of the art show, than when talking about the unrelated topic, F (3,20) = 5.32, p = 0.032 R 2 = .539.
! &+ Additionally, a 2 X 2 analysis of variance was conducted to determine the impact of disclosure and diversion on perceiving the confederate as supportive. Across conditions, participants did not report a difference in the amount of emotional s upport provided by the confederate. Participants found the listener to be equally supportive whether they were disclosing about their reactions to the art show or talking about their experience with art, F (3,20) = 0.11, p = .743, R 2 = .005. Additionally, participants described the listener as equally supportive regardless of whether the confederate was instructed to divert the topic of conversation or not, F (3,20) = 0.44, p = .514, R 2 = .019. Thus, while diverting the discussion did produce an increase i n perceiving the confederate as distracti ng it did not produce a complimentary decrease in the perception of overall emotional support. Descriptive Statistics for All Linguistic Variables The overall means and standard deviations for each dependent variable measured by the LIWC are reported in Table 2. Additionally, the means and standard deviation s for each dependent variable are reported by condition in Table 3. The Effect of Diversion and Disclosure on the Linguistic Variables A 2 X 2 analysis of variance was conducted to determine the effect of whether or not an event is disclosed and whether or not the discussion is diverted on the linguistic, social, perceptual, and cognitive features of the narrative describing participant s reacti ons to the art show. There were no significant differences in the majority of linguistic, social, and perceptual processes related to the art show experience. However,
! &, being diverted from the discussion, regardless of whether disclosing the experience or n ot, produced a greater use of past tense verbs, F (3, 20) = 6.52, p = .019, R 2 = .234. Additionally, diverted participants in both disclosure and non disclosure conditions used a greater proportion of words related to visual processes than non diverted par ticipants, F (3, 20) = 4.79, p = .041, R 2 = .187. The first hypothesis stated that Non Diverted Disclosure would produce higher levels of cognitive processing relative to the other three conditions. This hypothesis was not confirmed. Instead, Non Diverted Disclosure (M = 19.32, SD = 4.62) showed levels of overall cognitive processing comparable to Diverted Disclosure (M = 19.38, SD = 4.23), and exhibited less cognitive processing than either Non Diverted Non Disclosure (M = 22.7, SD = 2.24) or Diverted No n Disclosure (M = 23.58, SD = 3.99). A 2 X 2 analysis of variance indicated that participants who did not disclose about the art show demonstrated significantly more cognitive processing tha n participants who did, F (3, 20) = 5.72, p = .027, R 2 = .221. Ad ditionally, those who had the opportunity to disclose used more words indicating cognitive inhibition regarding their experience of the art show, F (3, 20) = 6.34, p = .021, R 2 = .217. Thus, non disclosure actually enhanced cognitive processing about the art show. The second hypothesis stated that Diverted Disclosure would produce levels of cognitive processing similar to both Non Disclosure groups. As previously stated, this w as not the case because those in the N on D isclosure conditions actually showed higher levels of cognitive processing relative to both disclosure groups. A 2 X 2 analysis of variance revealed that there was no main effect of diversion on cognitive process ing, F (3,20) = 0.09, p = .769, R 2 = .003. Furthermore, no changes in cognitive processing could
! '! be explained by an interaction between the disclosure and diversion conditions, F (3,20) = 0.07, p = .798, R 2 = .227. Thus, diverting disclosure did not signif icantly influence cognitive processing. Comparison of the Linguistic Variables to Base Rates in Emotional Writing Because individuals in this study volunteered to speak about sexual assault, they may have exhibited higher cognitive processing across the board if they were already motivated to discuss sexual assault upon arriving at the show. Therefore, a t test was computed testing the proportion of cognitive words observed in this sample against the base rates reported for emotional writing in the LIWC manual (Pennebaker, Chung, Ireland, Gonzales, & Booth, 2007). The descriptive statistics and t tests for each cognitive variable can be viewed in Table 4. Participants in this sample reported higher overall cognitive processing (M = 21.25, SD = 4.11), compared to the base rate (M = 19.66, 2.85). This difference was marginally significant, t (23) = 1.89, p = .072. Participants also exhibited greater insight ( M = 5.6, SD =2.12) relative to the base rates (M =3.25, SD = 1.05), which was highly significant, t (23) = 5.44, p < .0001. However, participants in this sample were also a little more hesitant to make firm assertions about their experience: they demonstra ted a higher proportion of words indicating discrepancy (e.g., words like "maybe, guess;" M = 1.54, SD = .9) and a lower proportion of words indicating certainty (M = 1 .00 SD = .79), compared to base rates (M = 2.13, SD = .79, and M = 1.73, SD = .64, r espectively). Both the difference in discrepancy, t (23) = 3.23, p = .004, and certainty, t (23) = 4.55, p = .0001, were found to be significant.
! '$ Discussion The Emotional Effect of the Art Show That the sexual assault art show so effectively inspi red emotional engagement represents an interesting subsidiary finding of this study. The PANAS scores indicate that a variety of negative emotions were effectively induced, including distress, upset, and fear. However, it is also interesting to note that w hile viewing art on the topic of sexual assault was surely upsetting for people, it also left many of them feeling inspired. Therefore, in addition to increasing negative emotions, the art show may have some benefits and significance for participants that could protect against feeling excessively distressed by the topic of sexual assault. Sample responses provided by participants themselves best illustrate the powerful emotional impact of the art show: "I found myself getting lost in the artwork, noticing emotions on faces and seeing pointy edges and uncomfortable twists. Some of the artwork was beautiful, some was tragic and scary, but it was all unsettling." "I cried when I read the poems, especially because they were right after the sta tistics. They made the situation seem more hopeless." "The art show made me feel sad and worked up as I was viewing the pieces but after leaving the room I noticed that I felt inspired." The art show may be a particularly effective stimulus for inducing n egative affect, and it would be interesting to compare it to methods commonly used for inducing distress, such as film clips (e.g., Lepore, Ragan, and Jones, 2000). The powerful emotional impact of the art show could be explained by viewers engaging and
! '% sympathizing with the strong emotions portrayed by the survivors of sexual assault. Furthermore, moving through the art show is an active experience, and therefore might be more emotionally engaging and arousing than a passive film v iewing. Additionally, some researchers have suggested that movie clips attempting to induce negative affect by exposing participants to topics such as the Holocaust may be a less effective means of creating emotional change because, while undoubtedly horri fic, such an experience has a limited amount of personal significance for the viewers (Lepore, Ragan, and Jones, 2000). In contrast, the topic of sexual assault may have a high personal significance for participants, perhaps especially females, because of the high er risk of personally experiencing the event. In addition to being an effective stimulus for inducing affect, the art show might even represent a powerful means for inspiring awareness, understanding and compassion for sexual assault survivors. Of the 24 participants, nearly every single narrative expressed feeling moved or making new realizations as a result of the art show. In the words of some of the participants: "I thought the art show was short and simple yet I felt such strong emotions fr om it. As a survivor, the artwork and poetry really hit me. ... The show pretty much described my entire coping and victim experience. From feeling vulnerable, scared, and silent to being strong and feeling empowered enough to speak out about and against it." "I am glad I had an opportunity to read [poems] like those because they really put the idea of sexual assault into perspective. I felt like I could connect with the authors and even feel their pain."
! '& "As an observer it made me think about what they we nt through and what they must have felt. It made me feel bad for them and even a little nervous because you don't think it could happen to you but it can." "I think the statistics on the wall also show that there is a need for more expression and educati on about this topic. This art show inspired me; I hope more sexual assault survivors find the strength and courage to express themselves." "Overall, the experience made me feel motivated to keep situations of sexual assault out of the world however I can possibly do that. No one should have that memory in their beautiful mind." Clearly, the art show had a variety of impacts on the viewers. In the future, it would be interesting and advantageous to examine whether an art show about sexual assault might be a mechanism for inspiring insight and compassion towards survivors of sexual assault. The Effect of Diverted Disclosure on Cognitive Processing Diverted Disclosure did not appear to effect individuals' cognitive processing about the sexual assault art show experience. Specifically, the hypothesis that disclosure would enhance cognitive processing was not supported. Likewise, the hypothesis that diversion would decrease cognitive processing was not supported. Unexpectedly, individuals not disclosing their art show experience actually exhibited gr e ater cognitive processing than those who did disclose. This result is both counter intuitive, and contrary to the literature. Therefore, it is more likely that a methodological error or unique circumstance of the study produced this result.
! '' One explanation might be that, by nature of volunteering for the study, participants were already motivated to think about and attempt to interpret the art show regardless of how the confederate responded. This assertion is supported by finding that participants in this sample showed a somewhat high er level of cognitive processing relative to base rates typically observed, particularly in regards to insights. However, if this were the case we would likely expect to see high cognitive processing across the board, and no significant difference between groups. Alternatively, it could be that the confederate's behavior did not accurately capture the experience of diverted disclosure. Particularly, the confederate was instructed to behave the same as in the Non Diverted conditions for the first half of D iverted conditions: they nodded, responded affirmatively, and acted engaged in what the person was saying, before changing the topic. As the high ratings on the Emotional Support subscale of the SRQ regardless of condition indicates, participants in the Di verted conditions experienced the confederate as no less supportive than in the Non Diverted conditions. In an actual experience of diverted disclosure, the recipient would most likely provide minimal response or appear uncomfortable before initiating the subject change, and thus would conceivably be viewed as less supportive. Furthermore, the relatively low responses (only ones and two on a four point scale) on the Distraction subscale for individuals in the Diverted conditions further alludes to the lack of strength of the diversion manipulation. Another factor that might be at play is the limitations of using the LIWC to measure cognitive processing about the sexual assault art show. As Tau sczik and Pennebaker (2010) explain, the LIWC is not able to dete rmine the context or content of
! '( what the person means to say; rather, the LIWC simply compares each word to an established dictionary, and computes a proportion of words used form each predetermined category of linguistic, social, cognitive, and perceptual features. Thus, in this study, it was only possible to determine the extent of cognitive processing about anything not only cognitive processing specific to the sexual assault content of the art show. Thus, some of participants may have shown a lot of c ognitive processing about the intended topic of sexual assault, while others exhibited processing about periphery, surface level details, such as physical descriptions of the art work. This distinction can be viewed in the actual responses provided by pa rticipants. For example, these individuals appear to be thinking about and attempting to understand the message of the art show as it relates to sexual assault: "I noticed that a lot of the poems dealt with the idea of isolation and feeling silenced so it seems important to display them publicly so that the stories aren't ignored." "The strongest thing I felt upon leaving the show was what a gift it is to be artistically inclined since it's so nearly impossible to express such an unthinkable experience. Clearly expression is crucial to processing. Trauma is such an incredible thing! Like an emotional time capsule." "I feel like sexual oppression the statistics helped is everywhere and has a stronger manifestation in rape and assault but its just g od damn everywhere. I feel as though I was thinking when I walked through most people assume it has not happened to anyone in our circle but it's everywhere. I thought about some aggressive tendencies of my partner and why I am so deeply turned off b y
! ') them I don't want to be dominated. Anywhere. Anytime. Thanks this helped I haven't written in a long time. In contrast, individuals such as the following appeared to be doing more thinking and responding to the actual physical details of the artwork or their own movements through the art show, rather than the art's content or relationship to sexual assault: "One poem had certain words or syllables separated by parenthesis, which made me think those had special significance to the author, but I could n't discern a pattern." "I read the signs on the wall first and then looked at all of the artworks in a clockwise order. Then, when I had more time, I went back and looked at some of them again." Thus, being able to sort out what respondents were talking a bout may have elucidated some interesting findings. Since participants in the diverted conditions were significantly more likely to use words related to vision processes, it is plausible that diversion responses made individuals less likely to speak about their reactions to the emotionally provocative subject of sexual assault, causing them instead to focus on perceptual features of the art show. While all of these methodological considerations likely influenced the results, none of them are suff icient to explain why participants not disclosing the experience actually exhibited greater cognitive processing. Oddly enough, the answer to this unusual finding might indirectly be attributable to diverted disclosure. It may be that the manner in which t he confederate interaction was set up caused all disclosing participants to perceive that their disclosure was diverted not by the confederate, but actually by the
! '* experimenter. To explain, the experimenter ended the very brief interaction between the co nfederate and participant by walking back into the room and instructing them to stop the discussion. In many cases, the participant was cut off mid sentence, and in some cases even attempted to continue speaking, only to inevitably be interrupted again. Thus, as a product for the experimental design, participants may have felt that their discussion was prematurely curtailed. Considering that participants believed the study was about how students respond to an art show on sexual assault, they may have fel t upset that they were not provided with more of an opportunity for discussion, and subsequently been discouraged from fully processing the experience in their narrative response. Being interrupted in such a way likely would not have affected participan t s in the non disclosure condition as much, since they were speaking about their experience with art, which is a much less emotional topic. This proposal is supported by the finding that not only diverted participants felt that the "listener" was more dist racting disclosing participants felt so too. Thus, the short discussion period and manner in which the experimenter ended the discussion may have caused participants in the disclosure group, who were talking about a personally significant and affective e xperience, to perceive the experience as more distracting than participants talking about an unrelated and not highly affective experience. Alternatively, it may be the confederate herself was actually more uncomfortable talking to people about sexual assa ult, and reacted to these participants in such a way that made them feel more distracted than participants in the non disclosure conditions. Either way, despite diversion not having a significant impact on cognitive processing directly, it may
! '+ indirectly a ccount for the unexpected increase in cognitive processing in the non disclosure conditions. Limitations and Suggestions for Future Research First and foremost, the small sample size (N = 24) and disproportionately high number of female participants (N = 20) limit the results and generalizability of this study. Considering there were only six participants in each group, it could be that participants more inclined to process the art show were distributed into the non disclosure group by h appenstance. Thus, future studies with a larger and more diverse sample should be conducted to more conclusively suggest the influences of diverted disclosure. As previously mentioned, the diversion manipulation did not appear to be very strong. Th ere are two issues likely contributing to this finding that future studies should address. First, quite simply, the confederate was not extensively trained in how to respond, and may have not been very consistent or forceful in her distraction attempts. Ad ditionally, the design of the diversion manipulation may not capture the qualities of diverted disclosure in real life. Specifically, so as not to conflate the diverted response with a negative response, the confederate was told to act supportively before diverting the disclosure. However, in actual diverted disclosure, individuals tend to provide minimal or no support before changing the subject (Filipas & Ullman, 2001). Thus, future studies should find better was of conceptualiz ing and operationalizing diverted disclosure while avoiding conflating it with a negative response. Additionally, the LIWC may not have been the best tool for measuring cognitive processing because of its inability to discern the context and content of t he responses.
! ', Future studies may benefit from conducting and coding a more qualitative interview, a survey instrument measuring cognitive processing, or measuring traumatic stress related symptoms as opposed to directly measuring cognitive processing. Fi nally, to determine the effect of diverted disclosure on coping with sexual assault, future studies should be conducted within the actual population. Specifically, studies looking at the effect of social, post assault variables on subsequent PTSD development should include diverted disclosure as one of the many potential exasperators of post traumatic stress In this way, not focusing exclusively on positive and negative responses would contribute to a more robust understanding of how the social context surrounding sexual assault affect s PTSD development Summary This study examined the effect of diverted disclosure on cognitive processing in an effort to generate preliminary support for the pos sibility that diversion is one of many responses to sexual assault disclosure that may contribute to PTSD development. The results indicated that diversion did not effect cognitive processing; however, counter intuitively and in contrast t o prior research, disclosure a ffected cognitive processing such that not disclosing actually result ed in enhancements. While a number of methodological considerations may have contributed to this finding, none of them are sufficient to explain the significant difference between disclosure groups. One explanation may be that some aspect of the experimenter's or the confederate's interaction with the participants caused a ll disclosing individuals to feel diverted from their disclosure. Thus, it may be plausible that diverted disclosure does influence cognitive processing, but both a more refined study and a larger more diverse sample is necessary to actually suggest this possibility.
! (! In conclusion, cognitive processing appears to be malleable by the influence of disclosure and listener responses. While it is plausible that diverte d disclosure represents on e fac e t in the spectrum of social responses to disclosure tha t may influence cognitive processing, its effect remains unclear. Understanding th e nuances of disclosure benefits and how they vary by listener response is essential, since cognitive processing about a traumatic event is causally implicated in PTSD development (Dunmore et al., 2001). Through i ncreasing understanding of how a variety of social responses might contribute to experiencing post traumatic stress following sexual assault, improved educational interventions could be devised to not only target blaming or degrading responses, but indifferent or distracting responses as well While coping with the experience of sexual assault will always be diff icult and painful, coping with the social reactions to sexual assault disclosure should never have to be.
! ($ References Ahrens, C.E., Stansell, J., & Jennings, A. (2010). To tell or not to tell: The impact of disclosure on sexual assault survivors' recovery. Violence & Victims, 25 (5), 631 648. Banyard, V.L., Moynihan, M.M., Walsh, W.A., Cohn, E. S., & Ward, S. (2010). Friends of survivors: The community impact of unwanted sexual experiences. Journal Of Interpersonal Violence 25 (2), 242 256. doi:10.1177/0886260509334407 Bavelas, J. B., Coates, L., & Johnson, T. (2000). Listeners as co narrators. Journal of Personality and Social Psychology, 79 (6) 941 952. Burke, P.A., & Bradley, R.G. (2006). Language use in imagined dialogue and narrative disclosures of trauma. Journal of Traumatic Stress, 19 (1), 141 146. Dunmore, E.C., Clark, D.M., & Ehlers, A. (1999). Cognitive factors involved in the onset a nd maintenance of posttraumatic stress disorder (PTSD) after physical or sexual assault. Behaviour Research and Therapy, 37 809 829 Dunmore, E.C., Clark, D.M., & Ehlers, A. (2001). A prospective investigation of the role of cognitive factors in persistent Posttraumatic Stress Disorder (PTSD) after physical or sexual assault. Behaviour Research and Therapy, 39 : 1063 1084. Ehlers, A. & Clark, D.M. (2000). A cognitive model of posttraumatic stress disorder. Behavior Research & Therapy, 38 (4), 319 345. Elliott, D.M., Mok, D.S., & Briere, J. (2004). Adult sexual assault: Prevalence, symptomatology, and sex differences in the general population. Journal of Traumatic Stress, 17 (3), 203 211.
! (% Filipas, H.H., & Ullman, S. E. (2001). Social reactions to sexual as sault victims from various support sources. Violence and Victims 16 (6), 673 692. Hackmann, A., Ehlers, A., Speckens, A., & Clark, D. M. (2004). Characteristics and content of intrusive memories in PTSD and their changes with treatment. Journal Of Traumat ic Stress 17 (3), 231 240. doi:10.1023/B:JOTS.0000029266.88369.fd Kilpatrick, D.G., Saunders, B. E ., Amick McMullan, A., Best, C. L., Veronen L. J., & Resnick, H. S. (1989). Victim and crime factors associated with the development of crime related post traum atic stress disorder. Behavior Therapy, 20, 199 214 Lepore, S.J., Ragan, J.D., & Jones, S. (2000). Talking facilitates cognitive emotional processes of adaption to an acute stressor. Journal of Personality and Social Psychology 78 (3), 499 508. McNally, R J. (1998). Experimental approaches to cognitive abnormality in posttraumatic stress disorder. Clinical Psychology Review 18 (8), 971 982. doi:10.1016/S0272 7358(98)00036 1 Nelson, K. (2003). Self and social functions: Individual autobiographical memory and collective narrative. Memory 11 (2), 125 136. doi:10.1080/741938203 Pasupathi, M., McLean, K. C., & Weeks, T. (2009). To tell or not to tell: Disclosure and the narrative self. Journal of Personality, 77 (1), 89 124. Pasupathi, M., Stallworth, L.M., & M urdoch, K. (1998). How what we tell becomes what we know: Listener effects on speaker's long term memory for events. Discourse Processes, 26( 1), 1 26.
! (& Pennebaker, J. W. (1993). Putting stress into words: Health, linguistic, and therapeutic implications. Be haviour Research And Therapy, 31 (6), 539 548. doi:10.1016/0005 7967(93)90105 4 Pennebaker, J.W., Chung, C.K., Ireland, M., Gonzales, A., & Booth, R.J. (2007). The development and psychometric properties of LIWC2007. [Software manual]. Austin, TX ( www.liwc.net ) Pennebaker, J. W., & Graybeal, A. (2001). Patterns of natural language use: Disclosure, personality, and social integration. Current Directions In Psychological Science 10 (3), 90 93. doi:10.1111/1467 8721.00123 Singer, J. A., Bluck, S. (2001). New perspectives on autobiographical memory: The integration of narrative processing and autobiographical reasoning. Review of General Psychology, 5 (2), 91 99. Tausczik, Y. R., & Pennebaker, J. W. (2010). The psychologic al meaning of words: LIWC and computerized text analysis methods. Journal Of Language And Social Psychology 29 (1), 24 54. doi:10.1177/0261927X09351676 Ullman, S.E., Fillipas, H.H., Townsend, S.M. and Starzynski L.L. (2007). Psychosocial correlates of PTSD symptom severity in sexual assault survivors. Journal of Traumatic Stress, 20 (5), 821 831. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negati ve affect: The PANAS scales. Journal Of Personality And Social Psychology 54 (6), 1063 1070. doi:10.1037/0022 3518.104.22.1683
! (' Table 1 Changes in Positive and Negative Affect Before and After the Art Show Before Show After Show Variable M SD M SD t(23) p Positive Affect 27.96 6.05 27.92 8.24 0.03 .976 Interested 3.67 0.64 3.83 0.87 0.94 .357 Excited 2.53 0.98 2.04 1.27 2.01 .056 Strong 2.71 0.95 2.42 1.18 1.66 .12 0 Enthusiastic 3.04 0.86 2.21 1.14 4.45 .0002 Proud 2.17 0.96 2.04 1.16 0.59 .56 0 Alert 3.13 1.15 3.46 1.22 1.5 .15 0 Inspired 2.17 1.01 3.33 1.24 4.75 <.0001 Determined 2.38 1.13 2.71 1.43 1.28 .21 0 Attentive 3.5 0 1.02 3.5 0 1.25 0 .00 1 .00 0 Active 2.67 1.01 2.38 1.17 1.43 .17 0 Negative Affect 13.5 0 3.71 17.83 5.01 4.21 .0003 Distressed 1.38 0.65 2.46 0.72 5.72 <.0001 Upset 1.17 0.38 2.46 0.78 7.37 <.0001 Guilty 1.17 0.48 1.46 0.59 2.07 .05 0 Scared 1.33 0.48 1.63 0.71 2.07 .05 0 Hostile 1 .00 0 .00 1.21 0.59 1.74 .096 Irritable 1.46 0.83 1.38 0.71 0.49 .63 0 Ashamed 1.04 0.2 0 1.33 0.56 2.29 .032 Nervous 1.88 0.85 2.04 1 .00 0.72 .477 Jittery 1.88 1.03 2.29 1.37 1.68 .11 0 Afraid 1.21 0.51 1.58 0.72 2.84 .009 ________________________________________________________________________
! (( Table 2 Overall Descripti ve Statistics for the Linguistic Variables Variable M SD Social Processes 6.14 3.66 Affective Processes 7.45 2.71 Positive emotion 3.68 2.5 0 Negative emotion 3.52 2.1 0 Anxiety 0.93 0.69 Anger 1.36 1.32 Sadness 0.64 0.85 Cognitive processes 21.25 4.11 Insight 5.6 0 2.12 Causation 2.32 1.45 Discrepancy 1.54 0.9 0 Tentative 3.49 2 .00 Certainty 1 .00 0.79 Inhibition 0.66 1.04 Inclusive 5.11 1.59 Exclusive 2.58 1.79 Perceptual Processes 4.9 0 2.07 See 1.75 1.47 Hear 0.5 0 0.61 Feel 1.97 1.46 Writing T ime 8.03 5.12 Word Count 157.04 90.34 __________________________________________________
! () Table 3 Descriptive Statistics for the Linguistic Variables by Condition Disclosure Non Dsiclsoure Diverted Non Diverted Diverted Non Diverted Variable M SD M SD M SD M SD Writing time 8.11 3.65 7.48 4.66 9.01 8.14 7.51 4.19 Word Count 161.17 55.45 146.5 80.79 180 154.17 140.5 55.29 Pronouns 13.94 2.72 13.29 3.09 14.75 4.2 0 16.71 4.91 Personal Pronouns 8.24 1.6 0 8.93 3.21 10.58 2.43 9.97 1.68 1st Person Singular 6.44 1.94 6.4 0 2.87 8.13 2.7 0 6.75 2.34 1st Person Plural 0 .00 0 .00 0 .00 0 .00 0.04 0.09 0.4 0 0.78 2nd Person Singular 0.62 0.97 0.55 0.85 0.11 0.26 0.37 0.9 0 3rd Person Singular 0.36 0.68 1.19 1.22 0.86 1.85 0.22 0.54 3rd Person Plural 0.83 0.67 0.79 0.79 1.45 1.33 2.25 2.03 Impersonal Pronouns 5.7 0 2.38 4.36 0.68 4.17 2.46 6.74 3.9 0 Past Tense 8.41 2.34 4.98 2.31 8.42 2.34 6.67 2.89 Present Tense 5.04 1.07 5.53 2.23 4.32 2.53 6.33 3.53 Future Tense 0.45 0.69 0.23 0.36 0.46 0.65 0.51 0.6 0 Social Processes 4.3 0 1.24 7.23 2.96 6.16 5.76 6.89 3.46 Family 0 .00 0 .00 0 .00 0 .00 0.8 1.32 0.11 0.27 Friends 0.07 0.17 0 .00 0 .00 0 .00 0 .00 0.26 0.41 Humans 0.88 1.1 1.5 1.45 1.39 0.79 1.63 1.34 Affective Processes 7.77 1.88 6.88 3.67 7.02 3.46 8.12 1.85 Positive Emotion 4.8 0 2.62 2.36 2.09 3.46 3.34 4.13 1.56 Negative Emotion 2.76 2.47 4.16 2.77 3.27 2.05 3.88 0.92 Anxiety 0.64 0.85 1.05 0.71 1.27 0.69 0.77 0.46 Anger 1.59 1.65 1.53 1.77 0.85 1 .00 1.48 0.83 Sadness 0.43 0.61 0.81 0.8 0 0.87 1.36 0.44 0.54 Cognitive Processes 19.38 4.23 19.32 4.62 23.58 3.99 22.7 2.24 Insight 5.22 1.97 4.71 2.34 6.24 2.32 6.22 1.95 Causation 2.09 1.39 2.12 1.45 2.66 1.05 2.44 2.06 Discrepancy 1.08 0.78 1.59 1.09 1.86 1.02 1.64 0.68 Tentativeness 2.7 0 1.42 3.45 2.25 4.7 0 1.85 3.11 2.26 Certainty 0.59 0.55 0.96 0.83 0.94 0.68 1.49 0.96 Inhibition 0.69 0.57 1.57 1.71 0.07 0.17 0.3 0 0.34 Inclusion 5.21 1.62 4.12 1.56 5.71 1.31 5.39 1.78 Exclusion 2.84 1.84 2.1 1.99 2.27 1.38 3.12 2.18 Perceptual Processes 5.33 1.56 5.42 1.24 5.22 3.15 3.61 1.77 See 2.74 1.3 0 1.08 0.99 2 .00 1.93 1.17 1.17 Hear 0.38 0.32 0.77 0.94 0.57 0.5 0 0.3 0 0.54 Feel 1.83 1.46 2.18 1.21 2.31 1.83 1.58 1.57
! (* Table 4 Comparison of Cognitive Processes in Experimental Responses to Base Rates in Emotional Writing Experimental Responses Base Rates Variable M SD M SD t(23) p Cognitive Processes 21.25 4.11 19.66 2.85 1.89 .072 Insight 5.6 0 2.12 3.25 1.05 5.44 <.0001 Causation 2.32 1.45 1.85 0.84 1.6 .122 Discrepancy 1.54 0.9 0 2.13 0.79 3.23 .004 Tentativeness 3.49 2 .00 2.93 1.05 1.37 .183 Certainty 1 .00 0.79 1.73 0.64 4.55 .0001 Inhibition 0.66 1.04 0.46 0.39 0.93 .382 Inclusion 5.11 1.59 5.09 1.54 0.05 .96 0 Exclusion 2.58 1.79 3.49 1.06 2.47 .021 *Base rates taken from the LIWC Manual by Pennebaker, Chung, Ireland, Gonzales, & Booth (2007)
! (+ Appendix A Image #1 of the Sexual Assault Art Show Stimulus
! (, Appendix B Image #2 of the Sexual Assault Art Show Stimulus
! )! Appendix C Image #3 of the Sexual Assault Art Show Stimulus
! )$ Appendix D Image #4 of the Sexual Assault Art Show Stimulus
! )% Appendix E Image #5 of the Sexual Assault Art Show Stimulus
! )& Appendix F Image #6 of the Sexual Assault Art Show Stimulus
! )' Appendix G The PANAS Scales This scale consists of a number of words that describe different feeling and emotions. Read each item and then mark the appropriate answer in the space next to that word. Indicate to what extent you feel this way right now, that is, in the present moment. Use the following scale to record your answers. 1 2 3 4 5 very slightly a little moderately quite a bit extremely or not at all _______ interested _______ irritable _______ distressed _______ alert _______excited _______ ashamed _______ upset _______ inspired _______ strong _______ nervous _______ guilty _______ determined _______ scared _______ attentive _______ hostile _______ jittery _______ enthusiastic _______ active _______ proud _______ afraid
! )( Appendix H The Revised Social Reactions Questionnaire Please indicate how often you received each kind of response from your partner during the discussion. 1. Listened to your feelings 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4=Always 2. Distracted you with other things 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4=Always 3. Showed understanding of your experience 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4=Always 4. Tried to discourage you from talking about the experience 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4= Always 5. Saw your side of things and did not make judgments 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4=Always 6. Seemed to understand how you were feeling 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4=Always 7. Told you to stop talking about i t 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4=Always 8. Told you to stop thinking about it 0 = Never 1=Rarely 2=Sometimes 3=Frequently 4=Always