[Page 42]
Commonly expressed opinions, both lay and professional, have implied that CSA possesses four basic properties: causality (it causes harm), pervasiveness (most SA persons are affected), intensity (harm is typically severe), and gender equivalence (boys and girls are affected equally). Qualitative and quantitative literature reviews of CSA have offered mixed conclusions regarding these properties but have suffered from various shortcomings. Problems in qualitative reviews have generally included sampling bias (i.e., overreliance on clinical and legal samples), subjectivity, and imprecision. Quantitative reviews have included larger proportions of nonclinical and nonlegal samples, reduced subjectivity, and increased precision and indicate that the intensity of CSA effects or correlates is of low magnitude in the general population. These reviews, however, have offered less clarification regarding issues of causality, pervasiveness, and gender equivalence. To address the shortcomings of the qualitative and quantitative reviews, we reviewed the CSA literature based on college samples. The advantages of this literature were (a) it contains the largest set of studies conducted on nonclinical and nonlegal populations; (b) it offers the most extensive database on moderating influences (e.g., family environment), useful for examining the issue of causality; (c) it provides a large number of male samples, facilitating gender comparisons; and (d) it provides a large database on self-reported reactions and effects, enabling examination of the pervasiveness of negative outcomes.
Review of the college samples revealed that 14% of college men and 27% of college women reported events classifiable as CSA, according to the various definitions used. Results from the college data do not support the commonly assumed view that CSA possesses the four basic properties outlined previously. CSA was associated with poorer psychological adjustment across the college samples, but the magnitude of this association (i.e., its intensity) was small, with CSA explaining less than 1% of the adjustment variance. Further, this small association could not be attributed to CSA for several reasons: (a) family environment was confounded with CSA, (b) family environment predicted adjustment problems better than CSA by a factor of nine, and (c) statistical control tended to eliminate significant relations between CSA and adjustment. Results also revealed that lasting negative effects of CSA were not pervasive among SA students, and that CSA was not an equivalent experience for men and women. These results imply that, in the college population, CSA does not produce pervasive and intensely negative effects regardless of gender. Therefore, the commonly assumed view that CSA possesses basic properties regardless of population of interest is not supported. These findings are consistent with Constantine's (1981 , p. 238) conclusion that CSA has "no inbuilt or inevitable outcome or set of emotional reactions" associated with it. It is important to add that analysis at the population level estimates the typical case and therefore obscures individual cases. That is, the findings of the current review should not be construed to imply that CSA never causes intense harm for men or women-clinical research has well documented that in specific cases it can. What the findings do imply is that the negative potential of CSA for most individuals who have experienced it has been overstated.
The validity of using studies based on the college population to
assess characteristics of CSA in the general population is of particular concern.
Objections to such an approach have included claims that SA college students may be too
young for symptoms to appear, typically experience less severe forms of CSA and
consequently are less harmed, or are better able to cope with their experiences than
persons in the general population (e.g., Briere, 1988 ; Jumper, 1995 ; Pallotta, 1992 ).
Evidence from the current review of similarities in CSA between the college and general
populations, however, contradicts these views. Compared with SA persons in national
samples, SA college students experienced intercourse, close family CSA, and multiple
incidents of CSA just as often, and the overall prevalence of CSA was not lower in the
college samples. The magnitudes of CSA-adjustment relations in the college samples and in
the national samples meta-analyzed by Rind and Tromovitch (1997) were
identical:
[*5] Despite all the empirically based similarities between the college and national populations, it is tempting to speculate that certain differences exist. Persons with extremely harmful CSA episodes may be unable to attend college or remain there once they have begun. In this way, surveys of college students may miss extreme cases of CSA, limiting the generalizability of findings from the college population. Nevertheless, the results of the current review, while not demonstrating equivalence between the two populations, strongly suggest that the gulf between them is narrow, and much narrower than child abuse researchers have generally acknowledged.