Keywords

In 1983 a mother suffering from serious mental illness alleged that Raymond Buckey, a teacher at the McMartin Preschool in Manhattan Beach, California, had molested her 2½-year-old son (Nathan & Snedeker, 2001). The accusations set off a highly publicized investigation and eventually led to the longest-running criminal trial in United States history. Based on accusations by dozens of children, Buckey, his mother Peggy, and five other preschool workers were charged with multiple counts of sexual abuse . Many children claimed that they had been forced to participate in bizarre events, such as being spirited into tunnels below the school to carry out satanic rituals. Eventually, Raymond and Peggy Buckey were acquitted by a jury and charges against the other defendants were dropped. At the Buckeys’ trial their attorneys presented video-recorded interviews of the alleged victims. Jurors later stated that these recordings, which showed interviewers questioning the children in a highly suggestive manner, heavily influenced their verdict to acquit.

Similar stories played out across America in the 1980s and early 1990s at other daycare centers and schools. In many cases, children made bizarre accusations of satanic abuse against teachers and daycare workers that were eventually discredited, although only after the defendants had spent years in prison. These cases raised important questions: Why would children fabricate such bizarre, detailed stories about child abuse? How could social workers and law enforcement be convinced that the stories were true despite little or no evidence to corroborate them? And perhaps most importantly, what role did suggestive interviewing play in creating these children’s allegations of abuse?

In the following years, social scientists have generated a large body of research that sheds light on the factors that contribute to children’s suggestibility, that is, their acquiescence to suggestions made by interviewers or other individuals. It is now widely recognized that although child witnesses and victims are often capable of giving reliable reports of their experience, suggestive influences and interviewer pressure can alter these reports and even induce children to make false criminal allegations .

This chapter reviews the kinds of missteps that can contribute to children’s suggestibility in cases of suspected sexual abuse . The first part of the chapter explains the dangers of interviewer bias and how it can lead to suggestive questioning. The second part describes the four suggestive questioning techniques that are most commonly used by biased interviewers. The third part discusses additional ways that child abuse investigations can go awry. All three parts of the chapter include recommendations to help interviewers and investigators in sexual abuse cases avoid the pitfalls that we describe. A central task of sexual abuse investigations is to uncover the truth about whether or not abuse has occurred. To achieve this goal, interviewers and investigators must be able to recognize and avoid the mistakes that lead to suggestible responding by children.

Interviewer Bias

According to Maggie Bruck and Stephen Ceci, ground-breaking researchers in the field of suggestibility, interviewer bias constitutes the central and most important characteristic of suggestive child interviews (Bruck & Ceci, 2011). A biased interviewer enters the interviewing room with preconceived ideas—for instance, that the child has been abused. If the child violates these expectations—for instance, by denying abuse —the interviewer may resort to suggestive questioning until the child’s reports become more consistent with what the interviewer expects to hear.

Interviewer bias is an example of what psychologists call “confirmation bias,” the tendency to one-sidedly seek evidence that confirms one’s own preexisting beliefs while ignoring evidence that disconfirms them (Cialdini, 2009; see also Kassin, Dror, & Kukucka, 2013, and following commentaries). Confirmation bias is a common human failing that can be observed every day, for example, when people discuss political topics. However, in a child forensic interview , it can have disastrous effects. A child’s statement that confirms the interviewer’s preconceived ideas may be accepted uncritically even if it is vague or improbable, whereas a statement disconfirming these ideas is likely to be ignored or discounted.

Confirmation bias can easily give rise to many undesirable interviewer behaviors. A study by Bruck, Ceci, Melnyk, and Finkelberg (1999, as cited by Bruck & Ceci (2011)) demonstrated how a biased interviewer can unwittingly encourage children to give false statements. Participants in the study were 120 preschool children, 90 of whom attended a birthday party with a visitor. The remaining 30 children were not at the party but instead spent time coloring with a visitor.

Interviewers in the study were graduate students recruited from social work and counseling programs who knew that the children had participated in an activity with a visitor but not what it was. Each interviewer was asked to individually question four children to discover what they had done with the visitor. Unknown to the interviewer, the first three children he or she questioned had been at the birthday party but the fourth had not.

The study found that, after questioning the first three children who had attended the party, most interviewers wrongly assumed that the fourth child had also been there. The interviewers then engaged in biased questioning to confirm their faulty preconceptions. In response to these suggestive questions, 60 % of children who had not actually attended the party made false claims to have been there, and 85 % of interviewers wrongly concluded that all four of the children they questioned had attended the party. As this study showed, even well-intentioned child interviewers can become biased and then use suggestive techniques to extract false statements.

Another study by White, Leichtman, and Ceci (1997) illustrates the negative effects of bias. Two professionals, a teacher and social worker, were given a list of activities that had supposedly occurred during a play session by a group of preschoolers. Unknown to the professionals, half of the activities were bogus and had not really occurred. The professionals then questioned the children to learn what had happened during the play session. The study found that interviewers repeatedly used suggestive questions to ask the children about the bogus activities. In response , the children falsely agreed that they had engaged in about 30 % of these activities, some of which involved bodily touch. Further, some children who initially denied that the bogus event occurred later changed their accounts and provided false details about it.

Avoiding Interviewer Bias

Avoiding interviewer bias is not a simple matter because, like any kind of confirmation bias, it is typically accompanied by a lack of awareness. Biased interviewers usually do not realize that they have lost their objectivity or engaged in suggestive questioning, and thus a mere admonition to “avoid bias” is unlikely to be effective in altering their behavior. Instead, law enforcement and child protection agencies should establish training requirements and other procedures to prevent bias (Powell, Hughes-Scholes, & Sharman, 2012), as set forth in the following three recommendations:

  1. 1.

    Agencies should ensure that child interviews are conducted only by individuals who have received formal training in the principles and practice of good interviewing . Because child forensic interview ing is a specialized professional skill, it requires more expertise than can be provided by only on-the-job training and brief workshops. Interviewers need to be closely familiar with professional guidelines (Lamb, 1994) and books (Ceci & Bruck, 1995; Lamb, Hershkowitz, Orbach, & Esplin, 2008; Poole & Lamb, 1998) that provide an in-depth understanding of the child interviewing process and the dangers of suggestiveness.

  2. 2.

    Agencies should develop clear and detailed protocols for child interviews that are based on best practices and consistent with the scientific literature. A structured interview developed at the National Institute of Child Health and Development (NICHD; Lamb et al., 2008) is presently the only interview protocol that has received extensive support in the scientific literature. It can be used without charge and agencies should consider adopting it or using it as a model for their own protocol.

  3. 3.

    Agencies should ensure that all child interviews are recorded. Video recording that clearly shows both parties should be the default, although audio recording is acceptable for brief initial interviews conducted in the field. Experts recommend such recording to ensure that all details in a child’s statement are preserved, to reduce the number of times the child is interviewed, and to provide a record which allows assessment of whether or not the interviewer has engaged in biased or suggestive questioning (Lamb, 1994; Poole & Lamb, 1998).

    In some cases investigated by child protection and law enforcement agencies, legitimate doubts can arise regarding the reliability of a particular child’s allegations of sexual abuse , either because the child makes vague or improbable statements or because there is reason to suspect that the child has been coached or unduly influenced to make a false report. A common failing of biased interviewers is to ignore obvious indications that a child’s statements are unreliable. To avoid such bias, we offer the following additional recommendations.

  4. 4.

    When a child makes statements that are vague, confusing, contradictory, or improbable, the interviewer should follow up by using open-ended questions to clarify the child’s statement. Books on interviewing explain how to discuss apparent contradictions with children in a gentle, nonconfrontational way.

  5. 5.

    Sometimes abuse reports are vague and unconvincing because they fail to include a detailed description of any specific incident. For instance, a child may state simply that she was touched by a certain person, but without specifying the surrounding context, the events that led up to the incident, or the words that were spoken. To avoid this problem we recommend that whenever abuse has been alleged, the interviewer should ensure that the child describes at least one specific incident of abuse from beginning to end and with as much detail as possible. For instance, the interviewer might say, “Tell me everything about the last time it happened. Start at the beginning. Tell me where you were and everything that happened.” After the child has given a free narrative response , appropriate follow-up questions should be asked to clarify details of the incident and when it occurred.

  6. 6.

    In many agencies it is a common practice for observers to watch child forensic interview s through closed-circuit television from another room. For instance, a police officer may observe an interview and recommend additional questions that may later help in the investigation or prosecution of the case. We recommend that if a child interview is observed in this way, one of the observers should be formally assigned the task, among their other duties, of identifying any problematic aspects of the child’s statement and recommending appropriate follow-up questions. For example, if the child has made inconsistent statements, or if there is a possibility that the child has been unduly influenced by an adult to make false allegations , the observer should be responsible for identifying these issues and reminding the interviewer to address them during the interview. Similarly, if the interviewer has failed to get the child to describe a specific incident of abuse the observer should remind the interviewer to do so.

  7. 7.

    Before ending an interview in which a child has disclosed abuse , the interviewer should routinely ask if the child has previously talked with anyone else about the abuse and, if so, to whom and under what circumstances. Such questions are included in the NICHD interview mentioned earlier in this section and can help clarify whether the child may have been exposed to suggestive influences before the interview.

  8. 8.

    Child protection and law enforcement agencies should routinely interview the “first confidante,” that is, the first person to whom a child has made a report of abuse (Wood, Nathan, Nezworski, & Uhl, 2009). The first confidante should be asked to describe the circumstances of the disclosure , including his or her recollection of questions posed and what the child said. Recording such interviews is important to preserve details that may later prove essential for evaluating the possibility of undue influence.

Four Suggestive Techniques Commonly Used by Biased Interviewers

Interviewer bias can lead to two characteristic kinds of mistakes: the interviewer may fail to adequately explore evidence that a child ’s statement is unreliable or may engage in suggestive questioning. The first kind of mistake was discussed in the previous section. Suggestive questioning is the topic of the present section.

A study by Nadja Schreiber and her colleagues (2006) provides a framework for our discussion. These researchers examined interview transcripts from the McMartin Preschool case described at the beginning of this chapter and another case and identified four suggestive techniques used by interviewers: Positive and Negative Consequences, Other People, Inviting Speculation, and Introducing Information.

Extensive research has shown that these four suggestive techniques can lead children to make false statements, including false allegations of wrongdoing against adults (e.g., see Finnila, Mahlberg, Santtila, Sandnabba, & Niemi, 2003; Garven, Wood, Malpass, & Shaw, 1998). The following pages describe each technique in detail, with citations to the relevant research and recommendations for avoiding them.

Positive and Negative Consequences

Positive Consequences and Negative Consequences are both suggestive interviewing techniques that involve the use of reinforcement. Positive Consequences involve giving praise or rewards to a child , or indicating that the child can earn such praise or rewards, in return for providing information concerning abuse . Negative Consequences involve expressing disappointment or giving other negative feedback to a child for making statements that the interviewer deems inadequate. Schreiber et al. (2006, p. 27) provide the following examples of Positive Consequences from the McMartin Preschool case:

  • Interviewer: Oh, you’re so smart. I knew you’d remember….

  • Interviewer: So I bet if you guys put on your thinking caps, you can help remember it. Now let’s make a test of your brain and see how good your memories are.

Garven et al. (1998, p. 349) provide the following example of Negative Consequences, also from the McMartin case:

  • Interviewer: “Are you going to be stupid, or are you going to be smart and help us here?”

Negative Consequences can also include expressing doubt about what the child has said (“Are you sure?”) or repeating questions in a way that implies that the child’s prior statements are incorrect.

A study by Garven, Wood, and Malpass (2000) found that reinforcement in the form of Positive and Negative Consequences can have a strong impact on suggestibility. Preschool children were visited at school by a man introduced as Paco Perez and were later questioned about what he had done. Half of the children received reinforcement for making false allegations of wrongdoing against Paco. Within a few minutes reinforced children acquiesced to 35 % of false allegations whereas non-reinforced children acquiesced to only 12 %. A similar study by Uhl, Wood, and Scullin (2014) found that Positive Consequences and Negative Consequences can affect older children as well, with 40 % of reinforced fourth graders acquiescing to false allegations as compared with only 4 % of non-reinforced children.

Avoiding problems with Positive and Negative Consequences. Many researchers agree that Positive Consequences can be helpful at the very beginning of child interviews if they are used to build rapport and encourage the child to talk (Poole & Lamb, 1998; Seidler & Howie, 1999). However, once the topic of abuse is introduced into an interview , it is important to discontinue giving the child any further positive or negative feedback.

Other People

The Other People technique involves telling a child what other witnesses have already said concerning the events that the child is being questioned about. Schreiber et al. (2006, p. 28) provide the following example from the McMartin Preschool transcripts:

  • Interviewer: You see all the kids in this picture? Every single kid in this picture has come here and talked to us. Isn’t that amazing? … These kids came to visit us and we found out they know a lot of yucky old secrets from that old school. And they all came and told us the secrets. And they’re helping us figure out this whole puzzle of what used to go on in that place …

The Other People technique can influence a child ’s reports in two ways. First, it can pressure the child to conform with the statements of other witnesses. The child, believing that the other witnesses are correct (Cialdini, 2009) or feeling reluctant to contradict them, may wrongly confirm what they have said. Several studies have shown that both adults and children can be influenced to make false statements if they are exposed to inaccurate information from other witnesses (Carol, 2014; Jones, 2013; Paterson, Kemp, & Ng, 2011; Shaw, Garven, & Wood, 1997).

The Other People technique can also influence a child ’s statements by increasing the plausibility of a false event. A child must consider a false event plausible before he or she can develop a mistaken memory that it occurred (Mazzoni, Loftus, & Kirsch, 2001). Studies have demonstrated that increasing the plausibility of a false event, even one that is highly improbable, can increase the likelihood that children wrongly believe that it occurred (Otgaar, Smeets, & Peters, 2012; Strange, Sutherland, & Garry, 2006). For example, Otgaar, Candel, Merckelbach, and Wade (2009) told children 7–8 years old that they were abducted by a UFO when they were 4 years old and then provided them with false newspaper articles that discussed the prevalence of UFO abductions. Children who read the newspaper article were twice as likely to report having been abducted by a UFO as were children who did not read it.

Similarly, telling a child that his or her peers have reported abuse may increase the child’s belief that he or she could have been abused. As the event becomes more plausible, the child may come to accept suggestions of abuse as actual memories. The studies cited earlier have shown that younger children are more susceptible to this effect than older children.

Avoiding problems with Other People. Interviewers sometimes use the Other People technique because they hope to make a child feel more comfortable about disclosing abusive incidents. However, this technique becomes highly suggestive when child witnesses are told the details of other children’s reports, thus pressuring them to provide information that is consistent with these reports. While it might be acceptable for an interviewer to inform a child witness at the very beginning of an interview that other children have been questioned, no information should be provided about what these children have said.

Inviting Speculation

Inviting Speculation involves asking a child to offer opinions or speculations about what could have happened during an event or to pretend that an event occurred. Schreiber et al. (2006, p. 29) provide the following example from the McMartin Preschool transcripts:

  • Interviewer: Now, I think this is another one of those tricky games. What do you think, Rags?

  • Child: Yep.

  • Interviewer: Yes. Do you think some of that yucky touching happened, Rags, when she was tied up and she couldn’t get away? Do you think some of that touching that—Mr. Ray might have done some of that touching? Do you think that’s possible? Where do you think he would have touched her? Can you use your pointer and show us where he would have touched her? [Emphasis added]

Several studies have examined the effects of Inviting Speculation on children’s memories. For instance, in a study by Ackil and Zaragoza (1998) half of the children interviewed about a video were forced to make up false answers in response to unanswerable questions. For example, when asked “What present did the boy get for his birthday?” these children had to provide a response, even though the video did not depict a boy receiving a present for his birthday. The remaining children in the study were asked the same unanswerable questions but allowed to answer “I don’t know.”

When children were interviewed a second time by a different interviewer, they were informed that the first interviewer had made some mistakes and that the children needed to help the new interviewer find out what really happened in the video. Children who had been forced to generate false answers made twice as many memory errors during this second interview than children who were allowed to say “I don’t know.” Several other studies have found similar negative effects of Inviting Speculation on children’s accuracy, even when they are no longer being encouraged to speculate (Krähenbühl & Blades, 2006; Poole & White, 1991; Schreiber & Parker, 2004; Shapiro & Purdy, 2005; Stolzenberg & Pezdek, 2013). These findings indicate that children will make up responses to questions if asked to do so and may later misinterpret these fabricated details as actual memories.

Avoiding problems with Inviting Speculation. Garven et al. (1998) note that Inviting Speculation was typically used by the interviewers in the McMartin Preschool case after other techniques had failed to elicit accusations from children. However, rather than yielding accurate information, this technique yields fabricated information that children may later come to believe is true. Therefore, interviewers should refrain entirely from asking children to speculate or pretend. If a child witness has repeatedly denied the occurrence of abuse , interviewers should normally accept the denials at face value rather than resorting to a technique that is likely to produce false memories.

Introducing Information

The technique of Introducing Information is similar to what lawyers call “leading questions.” It involves the introduction into an interview of accurate or inaccurate information that has not been previously mentioned by the child . Schreiber et al. (2006, p. 29) provide the following example from the McMartin Preschool transcripts:

  • Interviewer: How about Naked Movie Star? You guys remember that game?

  • Child: No.

  • Interviewer: Everybody remembered that game. Let’s see if we can figure it out.

As noted by Schreiber et al. (2006), Introducing Information overlaps with some of the other suggestive techniques that have already been discussed here. Referred to by some researchers as suggestive questioning or postevent misinformation, it has been studied extensively with both adults (see Garry & Loftus, 1994 and Ayers & Reder, 1998 for reviews) and children (Otgaar, Candel, Smeets, & Merckelbach, 2010; Sutherland & Hayne, 2001; see Bruck & Ceci, 1999 for a review). For example, in a study by Leichtman and Ceci (1995) a fictional character named Sam Stone visited nursery schools for a brief staged event. Children were interviewed about true and false aspects of the visit using one of four types of interviews : (a) the control interview , which used nonsuggestive questions; (b) the stereotype interview, which repeatedly introduced children to a stereotype about Sam as clumsy and used nonsuggestive questions; (c) the highly suggestive interview, which used suggestive questions (i.e., “When Sam Stone got that bear dirty did he do it on purpose or was it an accident?”); or (d) the stereotype plus highly suggestive interview, in which children were given a stereotype about Sam and interviewed using suggestive questions. Children given the stereotype plus highly suggestive interview were most likely to agree with inaccurate details about Sam’s visit. Children given the suggestive interview were more likely than children given the control interview to agree with these inaccurate details. Other studies have generally produced similar results, showing that Introducing Information reduces the accuracy of children’s reports and that younger children are especially vulnerable to its negative effects.

Avoiding problems with Introducing Information. Interviewers rarely know “ground truth ” when questioning a child about abuse . Thus, when they introduce information during an interview , they cannot be sure whether or not the information is accurate. Inaccurate information can become part of the child’s report because some children, particularly preschoolers, misinterpret it as real memories (see Introducing Imagery for a description of reality monitoring). To avoid creating inaccurate memories, interviewers should refrain from using this technique.

Speaking more generally, the same agency procedures recommended earlier in this chapter to prevent interviewer bias can also help prevent the use of Introducing Information and other suggestive techniques: (1) rely on well-trained interviewers who are familiar with professional guidelines and standard textbooks on child interviewing , (2) formally adopt a scientifically tested interview protocol, and (3) make video or audio recordings of all interviews .

Other Problems in Child Abuse Interviews and Investigations

Interviewer bias and suggestive questioning, the topics covered in the preceding sections, account for a large proportion of the problems that occur in child abuse investigations. However, other less common mistakes sometimes lead child abuse investigations astray, as discussed in this section.

Failure to Consider Disconfirming Evidence

Some child abuse cases go awry because investigators have uncritically accepted children’s statements as true without considering other physical, medical, and documentary evidence . For example, in the McMartin Preschool case some children claimed to have been taken into tunnels underneath their school and made to participate in satanic rituals. Extensive searches by law enforcement officials, including digs conducted by teams of archeologists, failed to uncover any tunnels. Despite this disconfirming evidence, officials and many members of the community steadfastly maintained their belief in the children’s reports (Nathan & Snedeker, 2001), an example of confirmation bias as discussed earlier in this chapter.

In more than half of child sexual abuse cases, the child’s statement is the only evidence of abuse . However, in a substantial minority of cases there is additional evidence to confirm the child’s statement. In a review of 894 cases of alleged child abuse, Herman (2010) found that when a child made an allegation of abuse there was a 40 % probability that it was corroborated by external evidence such as a confession by the perpetrator , statements from another eyewitness , or medical evidence.

In cases of alleged child abuse , it is important to consider evidence that tends to either confirm or disconfirm the child’s statement. For example, in a high-profile case in Bakersfield, California, two parents were accused of bizarre abuse (Nathan & Snedeker, 2001). At trial their children described being hung from hooks impaled in their backs. If these allegations were true there should have been scarring on the children’s backs, but no such evidence was presented by prosecutors. This lack of corroborating physical evidence should have raised grave questions concerning the reliability of the children’s testimony . Especially in cases that involve bizarre or improbable accusations, investigators should actively seek and evaluate corroborative and noncorroborative evidence before evaluating the allegations.

Multiple Interviews

There is no doubt that multiple interviews with a biased interviewer can lead to inaccurate statements and create false memories (see Ceci, Huffman, Smith, & Loftus, 1994, and Ceci, Loftus, Leichtman, & Bruck, 1994 for early research on this issue). In fact, research has demonstrated that just one interview with a biased interviewer can lead to false accusations, especially when some of the other techniques identified in this chapter, such as reinforcement, have been used (Garven et al., 2000).

However, some research suggests that multiple interviews may have neutral or even positive effects on memory if they are conducted without any bias or suggestive techniques. For example, La Rooy, Pipe, and Murray (2005) found that repeated interviews of 5- and 6-year-olds with open-ended prompts produced a reminiscence effect, that is, new information was obtained in subsequent interviews. More research is necessary to clarify the impact of repeated interviews. Goodman and Quas (2008) suggest that merely examining the impact of repeated interviews may not be enough, and that other factors, such as bias, social pressure, and delays must also be considered in order to understand the impact of repeated interviews on children’s accuracy.

Suggestive Therapy

In a small proportion of cases seen by protective agencies, children deny victimization to interviewers even though there is strong suspicion that abuse has occurred (e.g., the child may have privately made allegations to a friend). Occasionally, the children in such cases are assumed to be “in denial” and referred for therapy to help them “disclose.”

There are several reasons why it is inadvisable to put “strongly suspected” children in therapy for the purpose of encouraging disclosure . First, many of these children may not have been abused. A strong suspicion is not the same as certainty, and thus in some unknown proportion of suspected cases, even the “strongly suspected” ones, the suspicion is wrong.

Second, therapy to encourage disclosure is questionable from an ethical point of view. Such therapy typically has the covert goal of producing evidence that can be used in legal proceedings. However, this forensic goal can interfere with therapeutic goals, such as making the child felt understood and providing emotional support in difficult circumstances.

Third, a therapist who sets out to elicit a disclosure is by definition engaging in a form of confirmation bias and thus is likely to engage in suggestive techniques. Suggestive questioning by a therapist is particularly problematic from a legal perspective because therapy sessions, unlike child forensic interview s , are not usually recorded.

For these reasons, we strongly recommend against referring children to therapy in order to encourage disclosure . Instead, we recommend an alternative procedure adopted by many agencies: if a child is strongly suspected of having been abused but has not made an allegation to investigators, a referral should be made to a well-trained therapist for supportive counseling. The goal of therapy should be to provide a secure relationship in which the child can discuss important problems and worries. The therapist should not attempt to elicit a disclosure from the child, initiate discussions of abuse , or imply that the child may have been victimized. Instead, by focusing on the child’s current concerns, a skilled therapist can provide practical guidance and emotional support, thus creating a safe space in which the child may eventually feel comfortable enough to disclose abuse, if abuse has in fact occurred.

If a child has not made allegations , it is also important to protect her or him from other suggestive or biased influences. For example, parents and other caregivers should be advised to provide emotional support to the child but without initiating discussions of abuse . If the child initiates a discussion of abuse, parents should arrange for the child to discuss the topic with the therapist.

Encouraging Children to Create Imagery

Child interviewers and child therapists sometimes ask children to imagine, draw, or spend time thinking about an event in order to help them remember its details. All these activities create mental imagery (Ceci, Huffman et al., 1994; Ceci, Loftus et al., 1994; Quas, Schaaf, Alexander, & Goodman, 2000).

Reality monitoring refers to a person’s ability to distinguish between events they have actually experienced versus events they have only imagined. Research has demonstrated that the reality monitoring skills of young children are much poorer than those of older children and adults (Markham, Howie, & Hlavacek, 1999; Welch-Ross, 1995). For instance, a study by Foley and Johnson (1985) had children imagine, pretend, or “think real hard” about performing some actions and actually perform other actions. For example, children imagined making a sad face but actually did a jumping jack. The children also imagined and watched other people performing certain actions. For example, they imagined a person touching his or her own nose but actually witnessed a person running in place.

Later the children were asked to identify which actions actually happened and which were imagined. Compared with adults, 6- and 9-year-old children were equally poor in discriminating between performed and imagined actions. In other words, the children confused real actions with actions they only imagined. This and other studies on reality monitoring indicate that asking a child to imagine or pretend that a false event occurred can cause the child to believe that the event actually occurred. Other techniques involving the creation of imagery, such as drawing pictures or providing children with photographs of an event, can also cause children to believe that imagined events really occurred (Bruck, Melnyk, & Ceci, 2000; Strange, Garry, & Sutherland, 2003; Strange, Hayne, & Garry, 2008). For this reason, it is inadvisable to gather forensically relevant information from children by encouraging creation of imagery.

Anatomical Dolls

The use of anatomical dolls in forensic interview s with children has generated considerable controversy. On the one hand, experts advocating the use of anatomical dolls have pointed to their use in aiding memory retrieval and making children feel comfortable disclosing abuse (Everson & Boat, 1997). On the other hand, researchers opposed to the use of anatomical dolls express doubts whether the dolls are useful for generating accurate information and argue that the dolls may increase children’s suggestibility, particularly if combined with suggestive questioning methods (Ceci & Bruck, 1994; Everson & Boat, 1997).

Research on the usefulness of anatomical dolls has yielded mixed results (see Aldridge, 1998; Everson & Boat, 1997; Salmon, 2001 for reviews). Most studies have found that nonabused children rarely use the dolls for explicit sex play, although they may explore the dolls in a manner that could be misconstrued as sexual , for example, by exploring the doll’s anus. Some studies have found that children interviewed using anatomical dolls reveal more information about events than children interviewed without the dolls (i.e., Goodman, Quas, Batterman-Faunce, Riddlesberger, & Kuhn, 1997), but other studies have not found such an effect (i.e., Lamb, Hershkowitz, Sternberg, Boat, & Everson, 1996).

Part of the disagreement regarding the use of anatomical dolls stems from the disparate ways in which they are used. In a study examining actual forensic interview s with anatomical dolls, Thierry, Lamb, Orbach, and Pipe (2005; see also Santila, Korkman, & Sandnabba, 2004) found that different interviewers introduced the dolls at different points in the interview and used different approaches when questioning children about them. These variations in use create difficulty in assessing the reliability and validity of anatomical dolls as a tool for identifying abuse .

Despite the mixed findings, research on anatomical dolls tends to converge on one important point: the dolls should not be used with children ages 3½ years or younger (Ceci & Bruck, 1994; Everson & Boat, 1997). These young children are especially likely to produce inaccurate reports when anatomical dolls are used (Thierry, Lamb, Orbach, & Pipe, 2005).

There is no conclusive evidence that anatomical dolls elicit more information, or more accurate information, than can be gathered during a well-conducted child interview , nor is there convincing evidence that abuse can be diagnosed by watching children play with the dolls. On the other hand, very little research shows that the dolls can cause a child to report inaccurate sexual details, unless the child is questioned suggestively or is 3½ years old or younger. In addition, the available research indicates that the dolls may be helpful for very limited purposes, such as a memory stimulus for older children (Everson & Boat, 1997). Given the limited usefulness of anatomical dolls and the controversy surrounding them, we recommend against their routine use in child interviews , especially with younger children. If the dolls are used, interviewers should be careful not to engage in suggestive questioning and should not attempt to diagnose abuse based on the child’s doll play.

Social Contagion , Community Panic, and “Reporter Zero”

Social contagion and community panic occur when highly disturbing and “contagious” false ideas are spread among members of a community or other social group (Harrigan, Achananuparp, & Lim, 2012). Such contagion often occurs in high-profile sexual abuse cases that involve schools, day care centers, multiple alleged victims, or multiple alleged perpetrators. Although such cases occur only infrequently, they typically have devastating consequences for everyone concerned, including children, parents, and the accused.

The McMartin case described at the beginning of this chapter illustrates the way that social contagion can occur (Wood et al., 2009). After receiving an initial allegation against Raymond Buckey, police mailed letters to more than 200 parents urging them to question their children about possible molestation at the preschool. The letters set off a panic within the community. Many distraught parents engaged in suggestive questioning of their children that generated numerous false accusations of abuse against Buckey and the other McMartin teachers. These accusations were shared among parents, who then engaged in even more suggestive questioning and generated even more false accusations. Thus, through a process of social contagion, many parents and law enforcement officials became convinced that the children were victims of sinister organized abuse. The fact that the children’s reports were often bizarre, inconsistent, and lacked corroborating evidence did little to reduce the effects of the contagion. The panic was further inflamed by a flurry of sensationalist media reports about the prevalence of ritual satanic abuse (Nathan & Snedeker, 2001; see also Beckett, 1996).

Epidemics involving cholera or other infectious diseases can often be traced back to a single person, referred to by epidemiologists as “patient zero,” who is the initial source of infection. Similarly, community panics involving sexual abuse accusations can usually be traced back to a single adult of questionable credibility who reported the first false allegations or induced a child to make them. We will refer to this person as “reporter zero.” For instance, “reporter zero” in the McMartin case was a parent suffering from schizophrenia. In other high-profile cases, the role of “reporter zero” has been played by a parent involved in a bitter custody dispute, a woman who held a long-standing grudge against the accused perpetrator , or a “serial accuser” who previously had made ill-founded allegations of abuse against other individuals.

Panics involving sexual abuse , like deadly epidemics, are relatively rare. However, because of their devastating consequences they call for a thoughtful and effective response . We offer six recommendations to child protection and law enforcement agencies to help prevent panics and minimize the effects of social contagion.

  1. 1.

    Before broadening an investigation of sexual abuse accusations involving multiple victims or multiple perpetrators, agency officials should carefully evaluate the credibility of the initial allegations . Extreme caution is appropriate if (a) the initial victim has supposedly reported abuse outside the interviewing room but failed to repeat the allegations when formally interviewed; (b) the alleged victim has sometimes denied abuse or made seriously inconsistent statements; (c) the accusations are vague, bizarre, or improbable; (d) there is reason to suspect that the accusations are the result of coaching, suggestive questioning, or other undue influence by an adult (“reporter zero”); or (e) the accusations lack corroborating evidence or are inconsistent with known facts.

  2. 2.

    We have already recommended that the first confidante—that is, the first person to whom a child discloses abuse —should always be interviewed about how the disclosure was made and the exact wording of what was said, including the questions of the first confidante and the statements of the child. The same recommendation applies to “reporter zero” in cases that have a potential for creating community panic. Investigators should consider the psychosocial history and credibility of “reporter zero” when evaluating evidence in the case.

  3. 3.

    If investigators deem it necessary to broaden the investigation and interview additional children regarding the accusations, these new interviews should be carried out individually by well-trained interviewers and within a very short period of time. Established agency procedures should be carefully followed and interviews should be recorded. Parents should be notified only after interviews are complete .

  4. 4.

    In such interviews , interviewers should systematically inquire not only about the alleged abuse but also about possible social contagion, for instance, by asking “Who else have you talked with about [the alleged perpetrator ]? Have you heard about [the alleged perpetrator] doing bad things to other people? How did you hear? Tell me everything you heard.”

  5. 5.

    Information released to parents and the media should be free from specific details about the allegations , including any inflammatory information that could create a panic. Additionally, when releasing information, investigators should request that parents not interview their children about the abuse .

  6. 6.

    Any additional accusations that are made after the release of information should be evaluated with special care because they are likely to have been influenced by social contagion.

Conclusions

In this chapter, we have tried to inform interviewers and investigators in child abuse cases about factors that can create serious errors in sexual abuse investigations. Interviewer bias was identified as the central characteristic of suggestive interviewing , and four suggestive techniques used by biased interviewers were discussed. Several other problematic procedures were described that can lead to erroneous decisions in child sexual abuse investigations.

This chapter presented examples from high-profile abuse investigations in which a convergence of poor investigation methodology, biased interviewers, suggestive interviewing , and social contagion led to false accusations of abuse against multiple perpetrators. Such cases are fairly rare. However, investigators and interviewers can learn from the mistakes made during these investigations and become better equipped to protect children from those who would do them harm.