[Page 749 continued]

Dallam et al. (2001) questioned our qualitative analyses.

Similarly, Ondersma et al. (2001) challenged self-reports and argued that short-term harm is well documented. We now rebut these claims.

Temporary Versus Lasting Harm

Dallam et al. (2001) claimed that our conclusion that negative effects are often temporary is severely limited because only three studies examined this issue (Landis, 1956; Nash & West. 1985; West & Woodhouse, 1993), including very little in the way of serious abuse.

It is true that only these three studies directly asked how long any effects might have lasted. Nevertheless, all 11 samples (from eight studies) with data on self-reported effects (see Rind et al., 1998, Table 8) are relevant to lasting versus temporary harm. If a certain percentage in any study reported no negative effects, then by definition negative effects were not lasting for those cases. Taken together, self-perceived lasting harm was rare in these samples.

Additionally, Dallam et al.'s (2001) claim that these samples included little in the way of serious (i.e., oral sex and intercourse) CSA is incorrect. For example, as previously discussed, Condy et al. (1987) reported that 68% of the male students had intercourse CSA, yet only 16% of the entire sample reported negative effects. West and Woodhouse (1993) had a notable proportion of trivial homosexual advances, as Dallam et al. (2001) noted, but also a fair amount of serious CSA, as we discussed previously, with virtually no reports of any lasting negative effects. A college study we failed to include in the table was that of Fromuth and Burkhart (1987); we relied on their 1989 publication for our meta-analysis. In their sample, 30% of males reported oral sex and another 25% reported intercourse, yet only 15% reported negative effects (Fromuth & Burkhart, 1987).

In a previous rebuttal (Rind, Tromovitch, & Bauserman, 2000a) to Dallam et al.'s (1999) argument that Landis (1956) included too

[Page 750]

many noncontact cases, we adjusted his results on the basis of the assumption that all negative effects came from contact sex. We found that lasting self-reported negative sexual effects at a maximum would have occurred in 4% of male and 6% of female contact cases, and lasting general negative effects would have obtained in 0% of male and 7% of female contact cases. These results suggest that the Landis study is indeed informative about lasting effects of CSA.

Finally, Dallam et al. (2001) endorsed Nash and West (1985) as the "most relevant of the three studies" (p. 727; the other two were Landis, 1956, and West & Woodhouse, 1993). They noted that Nash and West used abroad definition of CSA and found that 22% of their female participants reported they were still being adversely affected.

It is unclear why the Nash and West study is more relevant than the West and Woodhouse study. Similar definitions of CSA were used in both, English college students formed the samples in each, and West oversaw both studies. The only important difference was that the first study involved women, 22% of whom felt lasting negative effects, whereas the second involved men, with only 2% feeling lasting negative effects. In comparing the female and male reactions in these two samples, West and Woodhouse noted the "particularly striking contrast" (p. 122) between them. Both female and male narratives are relevant, and it is unclear why one source should be taken as more relevant than the other, unless relevance is confused with the degree to which a study indicates harm.

Effects on Current Life

Dallam et al. (2001) claimed that we misreported Fishman's (1991) results on negative general and sexual effects. They noted that we reported these as 27% and 13%, respectively, but claimed, citing Fishman (1991, p. 162), that the correct figures are 47% and 23%, respectively. In fact, the latter figures refer to homosexua1 encounters only. The corresponding figures for heterosexual encounters were both 0%. When both homosexual and heterosexual encounters are combined, these percentages turn out exactly as we reported them.

Initial Emotional Reactions

In Dallam et al.'s (1999) initial critique, they stated that we were "ill responsible [sic)" with seeming "intent on misleading the reader" (per personal communication from R. Fowler, June 4,  1999) in how we coded the reaction data. In rebuttal, after arguing that our coding was valid, we showed that removing the disputed studies had virtually zero effect on the percentages of positive, neutral, and negative reactions (Rind, Tromovitch, & Bauserman, 2000a).

In their most recent critique, Dallam et al. (2001) renewed this claim by citing three different examples (Brubaker, 1991, 1994; Fishman, 1991). They claimed that Fishman reported 53% negative reactions at the time, not the 30% value we reported.

Once again, the figure they cited refers to homosexual encounters only, not all encounters that the men had; Fishman reported 0% negative reactions at the time to heterosexual encounters. The figures we reported were based on all experiences.

For the Brubaker samples, they cited her findings from a uni-polar scale of degree of upset feelings. This is not the measure of reactions relevant to our analysis because it did not classify reactions into positive, neutral, and negative categories, as we stipulated a measure must do for inclusion.

We used Brubaker's (1991, 1994) scale of pleasure, interest, surprise, shock, and fear, where the first two are labeled as positive reactions, the third is neutral, and the last two are negative. This classification began with Landis (1956) and was subsequently used by other researchers with college samples (e.g., Finkelhor, 1979a; Goldman & Goldman, 1988). In short, our coding was neither erroneous nor misleading, but was consistent with practices followed by other researchers.

Ondersma et al. 's (2001) Critique of Self-Reports

Related to the issue of harm, Ondersma et al. (2001) argued that harm does not require that the victim perceive the experience negatively. For example, the possibility that a child might learn from an abuser that such experiences are normal and positive is one of the most concerning possible outcomes of CSA. (p. 709)

We believe this assertion is unscientific because it sets up the problem of un-falsifiability (Popper, 1961). All outcomes become evidence for harm, negative and positive alike. By the same logic, organizations that regard homosexuality as pathological (e.g., National Association for the Research and Therapy of Homosexuality) can define it as such even, and perhaps especially, if the gay person has a positive self-identity. To repeat from our original article, the wrongfulness and harmfulness of a behavior are separate issues - and the kind of harm that mental health professionals should be concerned with is mental health (e.g., depression), not attitudinal deviance.

Ondersma et al. (2001) went on to cite Kendall- Tackett et al. (1993) for well-documented evidence of short-term harm and equivalence of outcomes for boys and girls. But even Kendall-Tackett et al. did not maintain that their finding of equivalence should be seen as robust, given the "bias in identification of male victims, [where] only the most symptomatic boys end up in clinical samples" (p. 170). In addition, as we discussed previously, the Kendall-Tackett et al. review has internal validity problems concerning dubious data, measures of harm, and definitions of harm, all of which weaken the authority of this review as a basis for claiming that short-term harm has been well documented.

Ondersma et al. (2001) also criticized our treatment of boys, reports of positive experiences by stating that we failed to consider alternative explanations for such reports, such as the possibility that some men and boys may refuse to recognize themselves as victimized because of male socialization or successful indoctrination by the "abuser."

This criticism represents a curious double standard. Our critics, and many CSA researchers, have had no difficulty accepting reports of negative experiences at face value, but have selectively denied any validity to reports of positive experiences and seek to discount them as the result of processes such as denial (Rind, Bauserman, & Tromovitch, 200la). A more objective stance would hold that neither type of experience can be selectively dismissed or explained away. In our review, we simply summarized and reported how both men and women said they perceived their experiences, without assuming that positive or negative perceptions were inherently suspect.

To our knowledge, no one has systematically studied the sorts of pressures that might exist on individuals to redefine experiences initially seen as positive, as well as those initially seen as negative.

[Page 751]

Although it may be true that denial plays a role in reporting some experiences as positive, it is at least equally plausible to suggest that those who initially perceive their experiences as positive are subjected to massive social pressures to redefine them as negative. Certainly, everyday messages in society about CSA experiences are uniformly negative, and anyone who perceives his or her own experience otherwise is exposed to constant messages about how damaging and traumatizing such experiences are.

In the absence of well-conducted research on the types and relative strength of pressures to redefine contacts experienced as positive versus negative, there is no empirical justification for singling out self-reported positive experiences as uniquely unreliable.

It is clear that boys were significantly more likely than girls to report their experiences as positive, but it is quite possible that this simply reflects the well-documented finding that males are primed for casual sex far more than females, a difference present from adolescence on. A vast and growing body of research documents such differences in male and female sexual attitudes and behaviors (Sommers, 2001).