Definition of Child Sexual Abuse

Definition of Child Sexual Abuse Researchers have not yet reached a consensus on how inclusive the definition of child sexual abuse should be. "One of the major difficulties in assessing the extent of abuse and its effects is that the definition of sexual abuse is not consistent in the literature" (Craine, Henson, Colliver, & MacLean, 1988, p. 303), Across studies, variability is found in (a) the type of abuse allowed (i.e., contact versus noncontact abuse); (b) the age differential between victim and offender; (c) the age cutoff for adolescent victims; and (d) the type of victim/offender relationship. Why does so much variability in the definition of child sexual abuse across studies exist? Why has no consensus been reached? A portion of the problem is in defining an event that has so much situational variability. As many writers are quick to point out (see, for example, Kelly, 1988), a too conservative definition of child sexual abuse invariably omits legitimate cases of child sexual abuse. On the other hand, a broad definition may allow for incidents of sexual contact that are difficult to classify as child sexual abuse. The difficulty is in creating an operationalized definition of child sexual abuse that is neither too broad nor too narrow. Perhaps the best method is that adopted by Russell (1983) and others in which respondents were asked to discuss all incidents of unwanted sexual contact. A research panel then made the decision regarding whether the incident was abusive. This latter method, although costly and time-consuming, appears to be an ideal method for assessing whether an incident is truly abusive. By doing so, the definition of child sexual abuse can also be modified later to be more or less inclusive (Wyatt & Peters, 1986). Eventually the researcher must determine the study's criteria for child sexual abuse. Various factors enter into the researcher’s decision to use a more narrow or broad definition. Sometimes the researcher's bias about what constitutes child sexual abuse may be the deciding factor. Alternately, the legal definition of child sexual abuse for that region may be used, or legal guidelines might predicate the criteria (Violato & Genuis, 1993). Another important concern may be the credibility of the study if a broad definition of child sexual abuse is used. Finally, researchers may choose to use a more narrow or broad definition based upon the needs for their study. Researchers attempting to ascertain the effect of child sexual abuse upon later psychopathology, for example, may choose to use only those victims with more extreme abuse. This use of a more restrictive population might obviate the problem of floor effects in which many victims may not exhibit clinical ranges of psychopathology (Berliner & Saunders, 1996). While good reasons for making the definition of child sexual abuse more or less narrow exist, the problem is when comparing studies to each other. In all types of studies, differences in criteria for inclusion as a child sexual abuse incident have a significant effect upon rates of abuse. In Russell's (1983) community survey, for example, when noncontact sexual abuse was included, the prevalence of women sexually abused as children (i.e., under the age of 18) was 54%, but when noncontact abuse was excluded, the prevalence was 38%. It is therefore critical to assess how differences in the definition of child sexual abuse across studies might affect the outcome of the research. These next sections discuss the criteria for a definition of sexual abuse and their potential impact on the outcome of research. Type of abuse allowed: The first issue is the type of abuse allowed (i.e., whether both contact and noncontact abuse should be included). With this issue, researchers may be concerned that including noncontact abuse may diminish the credibility of the study. For example, Haugaard and Emery (1989) found that victims of child sexual abuse could be divided into three subgroups. The first subgroup was primarily composed of those victims who experienced only noncontact abuse. This small subgroup included approximately 20% of all female victims and 38% of all male victims. Victims in this group were noteworthy for the limited effects of the abuse. The other extreme was a small clinical group (18% of all victims) in which victims had been penetrated vaginally, anally, or orally and had suffered the greatest respondent to provide sufficient detail about the incident, after which a research
panel decides whether the incident is abusive based upon criteria for inclusion.
The issues of the age differential between victim and offender and the upper
age limit for adolescents are of great concern because of the potential, when setting
arbitrary criteria, for missing valid cases of child sexual abuse. In a secondary
analysis (Bolen, 2000a) of Russell’s (1983) community prevalence study, peer abuse
and abuse by dates and others in a romantic/sexual relationship with the victim were
some of the most severe of all abuse incidents. Perpetrators in a romantic/sexual
relationship were almost as likely as strangers to physically assault their victims.
Penetration and the use of force or threat in both abuse by friends and those in a romantic/sexual relationship occurred more fkequently than for any other extrafamilial
abuse perpetrator type. Furthermore, multiple attacks occurred well above average
for both of these groups. Criteria that exclude these cases therefore minimize the
serious issue of sexual assault by same-age or near-age peers.
Victim/offender relationship: While child sexual abuse may be committed by a person of any filial or nonfilial relation to the victim, many clinical studies often restrict their samples to victims of intrafamilial abuse. This practice stems from a
critical problem in the current assessment and treatment of child sexual abuse
victims—that priority is given to victims of intrafamilial abuse, and especially
victims of father-daughter incest, The available pool of child victims is therefore
highly skewed toward populations of victims of incest.
Because much of the current literature derives primarily from samples of
intrafamilial abuse, the definition of incest is an important consideration. The reader must know, for instance, whether the definition includes only father-daughter incest,
incest by the nuclear family, or abuse by any relative. Even within father-daughter
incest it is important to know what constitutes the relationship. For example, does the study include all father figures or just legally recognized fathers? Does the study
include stepfathers? It is equally important that custody be specified. There is some indication that a significant amount of father-daughter incest may be committed by biological fathers who do not have custody of or who are not living with their children
(Faller, 1990). Therefore, the reader must know exactly what type of relationship the
sample constitutes. It is especially important not to generalize the findings beyond
the limits of the sample. For example, it is inappropriate to draw conclusions for all victims of child sexual abuse from a study using only intrafamilial abuse victims.