Rind et al.'s Examination of Consent as a Moderator
One of the major conclusions of the Rind study was that "consent" was an important moderator of adjustment in males. The authors later summarized their findings stating, "We showed that for boys in nonclinical populations, willing relations are generally experienced positively or neutrally and are not associated with maladjustment" (Rind, Bauserman, & Tromovitch, 1999, p. 2185). Our review revealed that this purported finding was based on the mislabeling of a group of subjects as "consenting" when in actuality Rind et al. had no data from male subjects who reported that they had "willingly" participated in sex with an adult.
To understand the seriousness of this problem, consider the following analogy. Imagine that a researcher seeks to test psychological differences between African-American and Caucasians. First they find a group of Caucasians, but they don't have a group of African-Americans handy. They do, however, have some surveys from a primarily White university on which subjects were never asked to indicate their race. They decide that some of the subjects were probably African-American and decide to call this group of subjects the "mixed race" group. Despite the fact that all of the samples took different types of tests, they decide to compare the two groups. They find (or claim to find) that the Caucasian group performed better than the group that they have designated as being "mixed race" (which they also call African-American). Despite the fact that all of the different samples were given different tests and differed in other important ways, they decide to attribute all differences between the two groups to race. They use then use this data to suggest African-American's are less well-adjusted than Caucasians.
It should be obvious to anyone reading this analogy that any conclusions about differences between these two groups can not scientifically be attributed to race. First and foremost, there is no way to know if any African Americans were included in the group designated "mixed race." And even if African Americans were included, there is no way to know whether they comprised enough of the sample to have any effect on group scores. Moreover, because each of the samples took different tests that looked a different outcomes, differences could very well be due to the outcomes being measured. Clearly, it would be both nonscientific and irresponsible to attribute differences between these two groups to race, a variable that was never measured in the second group.
Further, we do not believe that such a study design would make it past peer-review committee; nor do we think that psychologists would attempt to defend researchers that performed such shoddy research. Moreover, if the researchers who conducted this hypothetical study went on to use their "nondata" to recommend sweeping changes in how society treats African-Americans, the public would have every right to be incensed.
Unfortunately, this is pretty much how Rind et al. devised their so-called study of "consent." In their meta-analysis, Rind et al. compared samples who indicated that their sexual abuse was unwanted (the "unwanted" group) with samples on which no data had been collected about whether or not the abuse was wanted (they called this the "all-levels-of-consent" group). In this second group, Rind et al. placed samples who were never asked about whether the abuse was unwanted along with samples in which a portion of the subjects were defined as abused based on other definitional criteria (e.g., forced sex, intrafamilial incest, or large age discrepancies between victim and perpetrator). They then referred to these subjects as "willing participants."
Consider for a moment the types of experiences that were included in the "consent" group. A male who was defined as abused based on his report of molestation by a 38-year-old man when he was 6-years-old would be considered a "willing" participant in his abuse. A woman who is categorized as abused based on her report of rape by her father when she was age 10 is also considered a member of the "consent" group based on Rind et al.'s methodology. (For more information on the definitions of abuse used in the male samples that Rind et al. labled as "all-levels-of-consent," see pdf of table 6).
Despite the fact that each sample was given different tests that looked at different outcomes and criteria for being labeled abused differed in other important ways, Rind et al. attributed all differences between the two groups to "consent" -- the variable that was never measured in the second group. In summary, the fact that Rind et al. never studied "consenting" subjects invalidates any findings that they reported in their paper related to "consent" as a moderator. For more information, read the section of our scientific critique dealing with consent.
The following is an excerpt from Dallam et al.'s paper published in the Psychological Bulletin. These researchers perform the same analysis as Rind et al., using the same exact dataset, and are unable to replicate the findings Rind et al. (1998) reported. They further demonstrate the unsuitablity of the dataset itself to determine whether "willing" participation in sex with adults is as benign as Rind et al. suggest.
[Dallam, S. J., Gleaves, D. H., Cepeda-Benito, A., Silberg, J., Kraemer, H. C., & Spiegel, D. (2001). The Effects of Child Sexual Abuse: Comment on Rind, Tromovitch, and Bauserman (1998). Psychological Bulletin, 127, 715-733.]
Studies which have empirically examined the association between "consent" and psychological adjustment.
Because Rind et al. (1998) had no actual data gathered from subjects who reported having participated willingly in CSA, "consent," simple or otherwise, was never measured. Accordingly, there was no empirical support for Rind et al.'s conclusion that willing participation in sex with adults renders CSA harmless to boys. Recently, Rind et al.'s speculative assertions in this area have been refuted empirically.
Kelly, R. J., Wood, J. J., Gonzalez, L. S., MacDonald, V., & Waterman, J. (2002). Effects of mother-son incest and positive perceptions of sexual abuse experiences on the psychosocial adjustment of clinic-referred men. Child Abuse & Neglect, 26 (4), 425-41.
Kelly et al. (2002) examined the relationship between positive initial perceptions of sexual abuse experiences on adult male psychosocial functioning in victims of CSA. Sixty-seven clinic-referred men with a history of sexual abuse participated. Twenty-seven men recalled positive or mixed initial perceptions of the abuse, including about half of the men who had been abused by their mothers. These men reported more adjustment problems than did men who recalled purely negative initial perceptions. The researchers states:
King, M., Coxell, A., & Mezey, G. (2002). Sexual molestation of males: Associations with psychological disturbance. British Journal of Psychiatry, 181 , 153-157 .
King et al. (2002) found that of nearly 2,500 men attending 18 medical practices, those with a history of sexual abuse (n = 150) were more likely than other men to report mental health, sexual or substance abuse problems. Overall, childhood sex abuse showed the most widespread effect on long-term psychological health, with these men being twice as likely to report disorders like depression and anxiety and nearly four times as likely to have tried to kill or harm themselves. This higher-than-average risk of sexual problems, substance abuse and "self-harm," was also found in the men reported that the experience with the older person had been "consensual." In fact, men who reported consensual sex with an older person before age 16 were more likely to attempt to harm themselves compared with those with no abuse history. They reported self-harm at a rate of that was 70% higher than average.