So I'm currently reading through Mistakes Were Made (But Not By Me)
, and I've finished the chapter dealing with false child abuse scares.
I think this is pretty interesting but I really like the last paragraph and think it's relevant to the SJL narrative:
For example, in an experiment with preschool children, Sena Garven and her colleagues used interview techniques that were based on the actual transcripts of interrogations of children in the McMartin case. A young man visited children at their preschool, read them a story, and handed out treats. He did nothing aggressive, inappropriate, or surprising. A week later an experimenter questioned the children about the man's visit. She asked one group leading questions, such as "Did he shove the teacher? Did he throw a crayon at a kid who was talking?" She asked a second group the same questions along with influence techniques used by the McMartin interrogators: for example, telling the children what other kids had supposedly said, expressing disappointment if answers were negative, and praising children for making allegations. In the first group, children said "yes, it happened" to about 15 percent of the false allegations about the man's visit; not a high percentage, but not a trivial number, either. In the second group, however, the three-year-olds said "yes, it happened" to over 80 percent of the false allegations suggested to them, and the four- to six-year-olds said yes to about half the allegations. And those results occurred after interviews lasting only five to ten minutes; in actual criminal investigations, interviewers often question children repeatedly over weeks and months. In a similar study, this time with five- to seven-year-olds, the investigators found they could easily influence the children to agree with preposterous questions, such as "Did Paco take you flying in an airplane?" What was more troubling was that within a short time, many of the children's inaccurate statements had crystallized into stable, but false, memories.
Research like this has enabled psychologists to improve their methods of interviewing children, so that they can help children who have been abused disclose what happened to them, but without increasing the suggestibility of children who have not been abused. The scientists have shown that very young children, under age five, often cannot tell the difference between something they were told and something that actually happened to them. If preschoolers overhear adults exchanging rumors about some event, for example, many of the children will later come to believe they actually experienced the event themselves. In all these studies, the most powerful finding is that adults are highly likely to taint an interview when they go into it already convinced that a child has been molested. When that is so, there is only one "truth" they are prepared to accept when they ask the child to tell the truth. Like Susan Kelley, they never accept the child's "no"; "no" means the child is denying or repressing or afraid to tell. The child can do nothing to convince the adult she has not been molested.
We can understand why so many Susan Kelleys, prosecutors, and parents have been quick to assume the worst; no one wants to let a child molester go free. But no one should want to contribute to the conviction of an innocent adult, either. Today, informed by years of experimental research with children, the National Institute of Child Health and Human Development and some individual states, notably Michigan, have drafted new model protocols for social workers, police investigators, and others who conduct child interviews. These protocols emphasize the hazards of the confirmation bias, instructing interviewers to test the hypothesis of possible abuse, and not assume they know what happened. The guidelines recognize that most children will readily disclose actual abuse, and some need prodding; the guidelines also caution against the use of techniques known to produce false reports.
This change, from the uncritical "believe the children" to "understand the children," reflects a recognition that mental-health professionals need to think more like scientists and less like advocates, weighing all the evidence fairly and considering the possibility that their suspicions are unfounded. If they do not, it will not be justice that is served, but self-justification.