**2. CSA research and the plague of null findings**

Despite robust and consistent findings throughout the CSA literature, research has also been plagued with null findings. It seems for any measured variable found to show an association with negative outcomes, another study will show the opposite. Age at the beginning of abuse, a closely related offender, longer duration of abuse, the use of force to elicit abuse acts, and a lack of support following disclosure, have all been shown to be related to greater levels of symptomology at time of measurement (O'Leary et al., 2010; Reyes, 2008; Zinzow et al., 2010). However, studies can also be found that show no such associations (see Paolucci et al., 2001). The one finding that maintains consistency however is that of intrusion: more intrusive, invasive, penetrative abuse acts have repeatedly been associated with worse outcomes for individuals (Nelson et al., 2002; Ullman, 2007).

Although consistency in research findings far outweighs the discrepancies, the fact that null findings continue to arise has led some to believe CSA is not as harmful as first thought (Rind et al., 1998). It can be argued, however, that the reason so many inconsistencies are found within the research base is not due to the possibility that CSA is not severe or damaging, but instead the ways in which research into this area is conducted is in and of itself problematic. For example, the choices of populations to study, a lack of standardisation in the definition of CSA, and the vast catch-all approach to including any form of sexual exposure of an individual during childhood in research samples, are all areas that contribute to the inconsistencies identified in the literature.

#### **2.1 Populations**

Population estimates of the prevalence of CSA among the general community in Western societies indicate an estimated 9% - 35% of women and 4% - 19% of men have experienced some form of sexual abuse in childhood (Pereda et al., 2009; Putnam, 2003). These figures themselves, though alarmingly high, are cautioned to be conservative estimates due to a combination of the sensitive and personal nature of CSA, the relatively high rates of nondisclosure (McGregor et al., 2010), and the likelihood that the data contains a healthy percentage of false-negatives (Nelson et al., 2002). These estimates are testament to the fact that a substantial percentage of functioning, non-clinical men and women in the general community have been affected by CSA. However, beginning research into CSA was often conducted on clinical female populations, with many participants being in-patients within psychiatric facilities (Gold et al., 1999). In general psychological clinical samples the rate of CSA is remarkably high and the type of abuse reported is often intrusive, enduring, and severe (Calam et al., 1998; Gold et al., 1999). The outcome of these investigations clearly

between CSA and risk for adverse outcomes (Molar et al., 2001; Nelson et al., 2002; Wiffen & MacIntosh, 2005). With this being said, due to the heterogeneity of CSA experiences, individual differences in coping strategies during and after abuse, the interplay of the various mediating and moderating variables, as well as a host of other factors, a simple cause and effect model cannot be created to account for the myriad adverse outcomes that co-occur with the often traumatic experience of CSA (Putnam, 2003). This disparity not only exists among the types of experiences or environmental conditions that make up the population of those exposed to CSA; discrepancy and contradictions exist within the CSA

Despite robust and consistent findings throughout the CSA literature, research has also been plagued with null findings. It seems for any measured variable found to show an association with negative outcomes, another study will show the opposite. Age at the beginning of abuse, a closely related offender, longer duration of abuse, the use of force to elicit abuse acts, and a lack of support following disclosure, have all been shown to be related to greater levels of symptomology at time of measurement (O'Leary et al., 2010; Reyes, 2008; Zinzow et al., 2010). However, studies can also be found that show no such associations (see Paolucci et al., 2001). The one finding that maintains consistency however is that of intrusion: more intrusive, invasive, penetrative abuse acts have repeatedly been

associated with worse outcomes for individuals (Nelson et al., 2002; Ullman, 2007).

Although consistency in research findings far outweighs the discrepancies, the fact that null findings continue to arise has led some to believe CSA is not as harmful as first thought (Rind et al., 1998). It can be argued, however, that the reason so many inconsistencies are found within the research base is not due to the possibility that CSA is not severe or damaging, but instead the ways in which research into this area is conducted is in and of itself problematic. For example, the choices of populations to study, a lack of standardisation in the definition of CSA, and the vast catch-all approach to including any form of sexual exposure of an individual during childhood in research samples, are all areas

Population estimates of the prevalence of CSA among the general community in Western societies indicate an estimated 9% - 35% of women and 4% - 19% of men have experienced some form of sexual abuse in childhood (Pereda et al., 2009; Putnam, 2003). These figures themselves, though alarmingly high, are cautioned to be conservative estimates due to a combination of the sensitive and personal nature of CSA, the relatively high rates of nondisclosure (McGregor et al., 2010), and the likelihood that the data contains a healthy percentage of false-negatives (Nelson et al., 2002). These estimates are testament to the fact that a substantial percentage of functioning, non-clinical men and women in the general community have been affected by CSA. However, beginning research into CSA was often conducted on clinical female populations, with many participants being in-patients within psychiatric facilities (Gold et al., 1999). In general psychological clinical samples the rate of CSA is remarkably high and the type of abuse reported is often intrusive, enduring, and severe (Calam et al., 1998; Gold et al., 1999). The outcome of these investigations clearly

literature itself.

**2.1 Populations** 

**2. CSA research and the plague of null findings** 

that contribute to the inconsistencies identified in the literature.

showed the severe and lasting effects of the impact of CSA, with high impairments found across a number of domains within these women. However, clinical populations do not provide a representative picture of the majority of those who have experienced CSA, as evidenced by the large number of people within the general community who report experiencing abuse.

The rise of university student samples has been a popular choice for researchers for some time now (see Finkelhor & Browne, 1986). The benefit of using university students as research populations is that it provides researchers with access to large numbers of easily accessible participants. Although ease of gathering information with the use of this population is enticing, there are also inherent problems with the use of university student samples. First, within university samples, CSA prevalence rates have been shown to be lower than what is found within the general community, and the types of offences reported show a disproportionately higher percentage of non-contact abuse (exposure, pornography), and single-episode, non-penetrative abuse (Ullman, 2007; Zinzow et al., 2010) than those found within the general public (McGregor et al., 2010). A further limitation in utilising university students is that they often consist of relatively young people, the majority of whom are in their late teens to early thirties, who most often are not married and have no children (Harding et al., 2010; Zinzow et al., 2010). What this means in practice is that the sample consists largely of individuals who would be considered high-functioning, as evidenced by the fact they have met requirements to enter tertiary education, and who are also yet to go through significant life stages of partnering, marriage, and child rearing; stages that often cause one to think about the impacts of their own childhood on the way in which one conducts themselves as a spouse, mother or father (O'Dougherty et al., 2007). Not surprisingly then, university populations have been found to show less impairment than that found when studying individuals within the general population (Molner et al., 2001; Rojas & Kinder, 2009).

In reality, neither of these sample bases can be considered truly representative to the scope of individuals affected by CSA. With estimates of one in four women and one in seven men experiencing some form of CSA in childhood (Finkelhor, 1994) we only need to look at our workplaces, the local community, and indeed our own families, to see that if we have a mother, sister, aunty, and grandmother, the likelihood exists that one of these women have been affected (not withstanding familial correlates of CSA incidents). Similarly, a company board meeting consisting of 25 men could mean three or more of these men have also experienced CSA. What this means for research is that in order for our results to be truly meaningful, and to encapsulate the spectrum of individuals affected by this experience, broadening the use of general population samples would provide a more representative view of CSA and arguably, a reduction of null findings would also follow.

#### **2.2 Definitions and measurement**

Other problematic issues within research lays both within the way CSA is defined and measured. As it stands, no universal definition exists as to what actually constitutes CSA. This alone is contentious, as an obvious question is then of how is it possible for contradictory findings *not* to emerge if there is yet to be an agreement on what CSA actually is. Definition and measurement discrepancies lay within the types of experiences classified as abuse, the age of the victim at time of the assault, and a lack of enquiry into the subjective

A Salutogenic Approach to Healing Following Child Sexual Assault 39

coupled with the inclusion of unwanted sexual experiences of teenagers well into their sexual experimentation years, has the potential to confound results with the inclusion of acts that are not what would be considered to be within the realm of sexual abuse of a child, nor incorporates the relevant factors of grooming, complicity, shame, power-over, and secrecy

Gaining an accurate account of subjective distress, rather than assuming inferred harm, is perhaps one avenue that could contribute to reducing null findings and inconsistencies within the literature. All too often, the phenomenological experience of the individual is overlooked within research of any description. Without an understanding of the subjective distress at the time of the abuse, there is a higher likelihood that researchers are comparing "apples with oranges". A point has been made that just because an act is considered morally wrong, that in and of itself is not enough to assume that harm has been done (Rind et al., 1998). Obviously, individual differences in coping abilities and factors associated with notions of dispositional resilience play their part, for it is not possible to know all the underlying factors that cause someone to be more or less resilient than someone else who has experienced the same type of encounter. Nor is it possible to measure in detail how any particular risk or resilience factor, let alone all of the ones currently known, may contribute to a person's perception and subsequent reaction to sexual abuse. However, it must be considered that not all incidents are experienced or interpreted equally. For instance, in relation to non-contact abuse, a pre-teen girl who was 'flashed' by a stranger in a park may show no long-term ill-effects, while a similar-aged girl who's older brother intentionally invades her privacy by leering at her whilst showering, may show more distress. Similarly, in relation to contact abuse, someone who experienced being touched on their breast over the top of their clothes by the brother of an older playmate may be less affected than someone who experienced being fondled on their genitals underneath their clothes by their

Given the above examples of areas of potential confounds within CSA research, it could be argued that research into CSA would do well to increase its rigour around these particular areas. The use of community samples as the choice of populations to study would provide a more useful participant base from which to extrapolate findings that hold more meaning across the scope of individuals who have experienced CSA. Finding an agreed-upon operational definition of what constitutes CSA, one that includes pertinent aspects of CSA such as grooming and the use of trust and/or power to gain compliance and secrecy, as well as one that is more descriptive of CSA rather than simply sexual experiences, could assist in decreasing discrepancies, by reducing confounds of false-positives through methodological wording. Finally, gaining an accurate understanding of the subjective impact CSA has had on an individual can be an avenue that reduces assumptions in research, and allows for a

Yet, even with the potential areas for problems, research has provided us with a wealth of information into the serious, damaging, and pervasive negative effects that a history of sexual abuse in childhood can have on an individual. This information arms researchers and clinicians alike with a valuable knowledge base into not only the workings of CSA, but also the likely effects that may follow, and in light of this, efforts can be made to alleviate distress within individuals who have suffered this experience. However, the pathological paradigm has its limitations, and perhaps as an antidote to this, a relatively new wave of research is

more authentic view of the real impact of different degrees of CSA exposure.

that are important factors in the initiation and continuation of CSA.

uncle.

distress of the individual at the time of the experience. These factors are expanded on in the following sections.

Definitions of what constitutes CSA vary widely between studies. Some studies use questionaries from previous research (Rojas & Kinder, 2009; Ullman, 2007), others choose to use a more legal definition (see Paolucci et al., 2001) and some researchers use no definition at all, choosing to simply ask if participants had experienced sexual abuse in childhood (O'Leary et al., 2010; Phanichrat & Townshend, 2010). The latter approach in directly asking participants if they have experienced sexual abuse in childhood appears to be in the minority, with researchers often framing questions to read more on the lines of enquiry into sexual experiences (Ullman, 2007), sexual contact (Harding et al., 2010), or sexual activities (Zinzow et al., 2010) in childhood. The problem with such varied definitions of what constitutes CSA, is the likelihood of samples being inflated with the inclusion of falsepositives, because not all experiences classified as CSA by the researcher may be equally classified as abuse by the participant. This point becomes pertinent when looking at responses from men who have been designated into a 'sexually abused' group on the definition of 'sexual experiences in childhood'. By this definition, a surprising majority of these men state their experience as being 'positive' (Schultz & Jones, 1983). In their qualitative study on men who had endured sexual abuse as boys by members of the clergy, Isley et al., (2008) did not find one man who rated their experience as being 'positive'. However they did report that all the men themselves spoke of experiencing pervasive feelings of inadequacy, shame, isolation, and a belief they were 'damaged' by the abuse. In contrast, for a man who, at 15 had a sexual encounter with a woman in her 20's, the likelihood of that man rating his experience as 'positive' is substantially greater. However, all too often both of these men would likely be grouped together as being 'abused'. When the operational definitions of abuse vary so greatly, it is hardly surprising that such discrepancies exist within research findings.

Age of the victim at time of abuse is another area that varies widely between studies, and has the potential to contribute to inconsistent findings. Common cut-off ranges for age at beginning of abuse or abusive episodes are 14 years (Ullman, 2007; Ullman et al., 2009), 16 years (Rojas & Kinder, 2009) and 18 years of age (Harding et al., 2010; Zinzow et al., 2010). Some studies chose to include both an upper and lower age range (Palesh et al., 2007), while others provide no age descriptors at all (Del Castillo & Dougherty, 2009; McGregor et al., 2010; O'Leary et al., 2010). The age of an individual is a potentially important factor due to the relevance of emerging sexuality and sexual experimentation that is integral to the adolescent life stage. A five-year age gap between a 16 and 21 year old is not unusual for a consenting relationship; a sexual relationship between a 17 and 22 year old is also not uncommon (Darroch et al., 1999). And within these relationships, or within other contexts that adolescents find themselves in, intrusive and unwanted sexual experiences do most certainly occur. However, these types of unwanted sexual experiences are not representative of the dynamics of what CSA is known for. The sexual abuse of a child is just that, it is an *abuse* of power, an *abuse* of trust, and an *abuse* of authority over a minor, through the *abnormal use* of a child for an adult's, or significantly older teenager's, own sexual gratification. The inclusion of age ranges that extend to 18 years of age, well above the legal age of consent of 16 years of age in many places, creates the potential for individuals being included in the cohort of those experiencing CSA, when in fact this may not be the case. Therefore, the compounding effects of the lack of an agreed-upon definition of CSA,

distress of the individual at the time of the experience. These factors are expanded on in the

Definitions of what constitutes CSA vary widely between studies. Some studies use questionaries from previous research (Rojas & Kinder, 2009; Ullman, 2007), others choose to use a more legal definition (see Paolucci et al., 2001) and some researchers use no definition at all, choosing to simply ask if participants had experienced sexual abuse in childhood (O'Leary et al., 2010; Phanichrat & Townshend, 2010). The latter approach in directly asking participants if they have experienced sexual abuse in childhood appears to be in the minority, with researchers often framing questions to read more on the lines of enquiry into sexual experiences (Ullman, 2007), sexual contact (Harding et al., 2010), or sexual activities (Zinzow et al., 2010) in childhood. The problem with such varied definitions of what constitutes CSA, is the likelihood of samples being inflated with the inclusion of falsepositives, because not all experiences classified as CSA by the researcher may be equally classified as abuse by the participant. This point becomes pertinent when looking at responses from men who have been designated into a 'sexually abused' group on the definition of 'sexual experiences in childhood'. By this definition, a surprising majority of these men state their experience as being 'positive' (Schultz & Jones, 1983). In their qualitative study on men who had endured sexual abuse as boys by members of the clergy, Isley et al., (2008) did not find one man who rated their experience as being 'positive'. However they did report that all the men themselves spoke of experiencing pervasive feelings of inadequacy, shame, isolation, and a belief they were 'damaged' by the abuse. In contrast, for a man who, at 15 had a sexual encounter with a woman in her 20's, the likelihood of that man rating his experience as 'positive' is substantially greater. However, all too often both of these men would likely be grouped together as being 'abused'. When the operational definitions of abuse vary so greatly, it is hardly surprising that such

Age of the victim at time of abuse is another area that varies widely between studies, and has the potential to contribute to inconsistent findings. Common cut-off ranges for age at beginning of abuse or abusive episodes are 14 years (Ullman, 2007; Ullman et al., 2009), 16 years (Rojas & Kinder, 2009) and 18 years of age (Harding et al., 2010; Zinzow et al., 2010). Some studies chose to include both an upper and lower age range (Palesh et al., 2007), while others provide no age descriptors at all (Del Castillo & Dougherty, 2009; McGregor et al., 2010; O'Leary et al., 2010). The age of an individual is a potentially important factor due to the relevance of emerging sexuality and sexual experimentation that is integral to the adolescent life stage. A five-year age gap between a 16 and 21 year old is not unusual for a consenting relationship; a sexual relationship between a 17 and 22 year old is also not uncommon (Darroch et al., 1999). And within these relationships, or within other contexts that adolescents find themselves in, intrusive and unwanted sexual experiences do most certainly occur. However, these types of unwanted sexual experiences are not representative of the dynamics of what CSA is known for. The sexual abuse of a child is just that, it is an *abuse* of power, an *abuse* of trust, and an *abuse* of authority over a minor, through the *abnormal use* of a child for an adult's, or significantly older teenager's, own sexual gratification. The inclusion of age ranges that extend to 18 years of age, well above the legal age of consent of 16 years of age in many places, creates the potential for individuals being included in the cohort of those experiencing CSA, when in fact this may not be the case. Therefore, the compounding effects of the lack of an agreed-upon definition of CSA,

following sections.

discrepancies exist within research findings.

coupled with the inclusion of unwanted sexual experiences of teenagers well into their sexual experimentation years, has the potential to confound results with the inclusion of acts that are not what would be considered to be within the realm of sexual abuse of a child, nor incorporates the relevant factors of grooming, complicity, shame, power-over, and secrecy that are important factors in the initiation and continuation of CSA.

Gaining an accurate account of subjective distress, rather than assuming inferred harm, is perhaps one avenue that could contribute to reducing null findings and inconsistencies within the literature. All too often, the phenomenological experience of the individual is overlooked within research of any description. Without an understanding of the subjective distress at the time of the abuse, there is a higher likelihood that researchers are comparing "apples with oranges". A point has been made that just because an act is considered morally wrong, that in and of itself is not enough to assume that harm has been done (Rind et al., 1998). Obviously, individual differences in coping abilities and factors associated with notions of dispositional resilience play their part, for it is not possible to know all the underlying factors that cause someone to be more or less resilient than someone else who has experienced the same type of encounter. Nor is it possible to measure in detail how any particular risk or resilience factor, let alone all of the ones currently known, may contribute to a person's perception and subsequent reaction to sexual abuse. However, it must be considered that not all incidents are experienced or interpreted equally. For instance, in relation to non-contact abuse, a pre-teen girl who was 'flashed' by a stranger in a park may show no long-term ill-effects, while a similar-aged girl who's older brother intentionally invades her privacy by leering at her whilst showering, may show more distress. Similarly, in relation to contact abuse, someone who experienced being touched on their breast over the top of their clothes by the brother of an older playmate may be less affected than someone who experienced being fondled on their genitals underneath their clothes by their uncle.

Given the above examples of areas of potential confounds within CSA research, it could be argued that research into CSA would do well to increase its rigour around these particular areas. The use of community samples as the choice of populations to study would provide a more useful participant base from which to extrapolate findings that hold more meaning across the scope of individuals who have experienced CSA. Finding an agreed-upon operational definition of what constitutes CSA, one that includes pertinent aspects of CSA such as grooming and the use of trust and/or power to gain compliance and secrecy, as well as one that is more descriptive of CSA rather than simply sexual experiences, could assist in decreasing discrepancies, by reducing confounds of false-positives through methodological wording. Finally, gaining an accurate understanding of the subjective impact CSA has had on an individual can be an avenue that reduces assumptions in research, and allows for a more authentic view of the real impact of different degrees of CSA exposure.

Yet, even with the potential areas for problems, research has provided us with a wealth of information into the serious, damaging, and pervasive negative effects that a history of sexual abuse in childhood can have on an individual. This information arms researchers and clinicians alike with a valuable knowledge base into not only the workings of CSA, but also the likely effects that may follow, and in light of this, efforts can be made to alleviate distress within individuals who have suffered this experience. However, the pathological paradigm has its limitations, and perhaps as an antidote to this, a relatively new wave of research is

A Salutogenic Approach to Healing Following Child Sexual Assault 41

ago, in the words of Buddha, "*Life is suffering".* From this perspective, the focus is on coping resources that contribute to movement towards the healthy end of the wellbeing continuum, or at the very least, which assists in the maintenance of one's position. In this way, salutogenesis is an investigation of the total story of a person, discovering how one successfully resolves tension in their lives and maintains or enhances their position on the

At the core of the salutogenic paradigm is the theory behind the perceptual differences Antonovsky discovered in individuals who were able to maintain wellness despite their previous or current circumstances, what Antonovsky coined the sense of coherence (SOC). The SOC comprises three main components across cognitive, behavioural, and emotional domains, being comprehensibility, manageability, and meaningfulness. Comprehensibility is the extent to which an individual views stressors as understandable, clear and ordered. Manageability is the extent that resources are perceived to be available to the individual, and that these resources are adequate to meet the challenges a person might face. Meaningfulness is the extent to which an individual believes their emotional life makes sense, and the emotional demands they face are worth investing energy into. Therefore, healthy coping, according to Antonovsky, is when one is able to make sense of their situation, believe they have the abilities to cope with what is in front of them, and believe that the emotional struggle to deal with their difficulties is of importance and worthy of

The overall outlook of the salutogenic orientation is to offer a theoretical perspective of successful coping. Inherent in this theory is the notion that human wellness is more than an absence of pathology, and takes into consideration the human ability to flourish and experience positive change after the experience of a major disruption or trauma. Positive changes that arise from the struggle to cope with a traumatic event have been termed *posttraumatic growth* (PTG) (Calhoun & Tedeschi, 2006; Tedeschi & Calhoun, 1996). The notion of PTG is a burgeoning field of research that takes the notion of resilience and resistance to stress to a new level. Consequently, the past 15 years of inquiry has provided research demonstrating positive post-trauma changes in individuals who have struggled

One theoretical process model of the pathways to PTG asserts that the experience of trauma is so powerful that it shakes the very foundations on which we view ourselves, our world, and our place in it (Janoff-Bulman, 2006). Fundamental cognitive assumptions held about the self and one's environments, assumptions that serve to make life predictable, logical, and intelligible, and which are based on concepts such as safety, benevolence, and good things happening to good people, and bad things happening to someone else, are believed to be in dissonance with reality after a perceived traumatic event, due to the profound loss of safety, protection and certainty the experience of trauma often brings (Janoff-Bulman, 2006; Calhoun & Tedeschi, 2006). The internal disruption of one's inner world through the experience of loss of safety and security is believed to overwhelm ordinary defences, leaving one with a sense of threat, helplessness, and vulnerability. Coping, then, becomes a process of reworking internal fundamental assumptions, moving from overgeneralisation of danger and helplessness, to a place that incorporates both the new reality of the uncertainty of the

with many types of trauma (Janoff-Bulman, 2006; Tedeschi & Calhoun, 1995).

well-being continuum despite, or perhaps because of, their difficulties.

investment (Antonovsky, 1987).

**3.1 Post-traumatic growth** 

beginning to emerge; one that seeks to add a new body of information on CSA; a complimentary, innovative way in which we both understand and work with survivors of this particular trauma.
