Mistakes Were Made - Carol Tavris [56]
Therapists who have not been trained to think scientifically will probably not wonder about the invisible cases—the children they don’t see as clients. They probably will not think to ask how common the symptoms of bedwetting, sex play, and fearfulness are in the general population of children. When researchers did ask, they found that children who have not been molested are also likely to masturbate and be sexually curious; temperamentally fearful children are also likely to wet the bed and be scared of the dark. 20 Even children who have been molested show no predictable set of symptoms, something scientists learned only by observing children’s reactions over time instead of by assessing them once or twice in a clinical interview. A review of forty-five studies that followed sexually abused children for up to eighteen months found that although these children at first had more symptoms of fearfulness and sexual acting-out than nonabused children, “no one symptom characterized a majority of sexually abused children [and] approximately one third of victims had no symptoms…. The findings suggest the absence of any specific syndrome in sexually abused children.”21
Moreover, children who have not been abused do not appreciably differ from abused children in how they play with anatomically detailed dolls; those prominent genitals are pretty interesting. Some children do bizarre things and it doesn’t mean anything at all, except that the dolls are unreliable as diagnostic tests.22 In one study headed by two eminent developmental psychologists, Maggie Bruck and Stephen Ceci, a child pounded a stick into the doll’s vagina to show her parents what supposedly had happened to her during a doctor’s exam that day.23 The (videotaped) doctor had done no such thing, but you can imagine how you would feel if you watched your daughter playing so violently with the doll, and a psychiatrist told you solemnly it meant she had been molested. You would want that doctor’s hide.
Many therapists who began to specialize in child abuse in the 1980s often feel extremely confident of their ability to determine whether a child has been molested; after all, they say, they have years of clinical experience to back up their judgments. Yet study after study shows that their confidence is mistaken. For example, clinical psychologist Thomas Horner and his colleagues examined the evaluations provided by a team of expert clinicians in a case in which a father was accused of molesting his three-year-old daughter. The experts reviewed transcripts, watched interviews of the child and videotapes of parent-child exchanges, and reviewed clinical findings. They had identical information, but some were convinced the abuse had occurred while others were just as convinced it had never happened. The researchers then recruited 129 other mental-health specialists and asked them to assess the evidence in this case, estimate the likelihood that the little girl had been molested by her father, and make a recommendation regarding custody. Again, the results ranged from certainty that the child had been molested to certainty that she had not. Some wanted to forbid the father to see his daughter ever again; others wanted to give him full custody. Those experts who were prone to believe that sexual abuse is rampant in families were quick to interpret ambiguous evidence in ways that supported that belief; those who were skeptical did not. For the unskeptical experts, the researchers said, “believing is seeing.” 24
Other studies of the unreliability of clinical predictions, and there are hundreds of them, are dissonance-creating news to the mental-health professionals whose self-confidence rests on the belief that their expert assessments are extremely accurate.25 When we said that science is a form of arrogance control, that’s what