Interrogation Of Children

Hollida Wakefield and Ralph Underwager^*

How can children say sexual things happened to them that didn't happen?

This is the fundamental question when a decision must be made as to whether an accusation of child sexual abuse is true or false.  The question is most pointed when there is an uncorroborated statement by a child alleging sexual abuse and the accused perpetrator denies abusing the child.  If the sexual abuse is not real, then how and why is the child telling about sexual behaviors and describing abusive acts?  How can children talk about sexual acts if they didn't experience them?

When there is an accusation of child sexual abuse, children often are the only witnesses offered.  Frequently no corroborating or supporting evidence is found in spite of extensive effort.  Then the accusation rests solely upon a child's alleged words.  The alleged perpetrator is saying, "No!  I did not do that!"  But adults can't believe that a young child would fabricate a story about being sexually abused so when someone says a child is telling about sexual abuse, the story is often unconditionally accepted.  Adults may believe the simplistic maxim, "Children never lie about sexual abuse."  Therefore, whenever a child supposedly makes a disclosure about sexual abuse, those adults immediately believe that the abuse is real.
  

The Wrong Question

But it is mistake to pose the question in the form of whether or not the child has lied.  To lie assumes a conscious, willful, and deliberate purpose and intent to deceive.  Young children are unlikely to have the cognitive capacity or the maliciousness to lie in this way (although some older children and adolescents may).  People, including children, may say things that are not accurate.  Inaccuracy alone does not mean that people are lying.  They may be misinformed, mistaken, or ignorant.

Children do not know what they don't know.  Their experience is limited and their store of available knowledge is small.  When asked questions they don't understand or to which they have no answer they can blithely give an answer without knowing it is mistaken.  Unless there is strong evidence of deliberate, intentional dissembling, it is foolish to spend much time or energy on the question of children lying.  It is almost always the wrong question.
  

The First Answer

When trying to understand a child's behavior, the first place to look is in the environment of the child, and not to inferred internal states or dispositions.  The farther away an interpretation is from observable events, the greater the probability of error.  The more complex the inferred causal chain and the more speculative the inferred internal events, the more likelihood there is of introducing error.

It is surprising that so many mental health professionals apparently think first of dynamic, internal, private events (behaviors said to show shame, fear, anxiety, embarrassment, etc.) when interpreting the behavior of a child rather than considering environmental variables.  We have seen this approach in every almost case where mental health professionals believe the abuse was real.  When behaviors are explained by inferring to unobservable internal events inside a young child, the alternative of causal environmental influences outside the child is seldom considered or even acknowledged.

An example we have seen many times is that a child's silence or lack of response to questions is interpreted by an adult to mean the child is ashamed.  The child must therefore be ashamed of something.  The assumption of the adult is that children are ashamed of sex and of being sexually abused.  Therefore the child's nonresponse is proof that the child was sexually abused.  This interpretation is then offered in testimony as evidence of abuse.

Another variant of interpreting nonresponse is for the adult to assume that the child has a secret, and therefore the child is silent.  Then the adult must, of course, press the child to reveal the secret.  Therefore, questions are repeatedly asked, any denials from the child are ignored, and adult pressure continues until some response is obtained which can be interpreted to show abuse happened.

But it is more parsimonious to ask if the child's nonresponse may be caused by strangeness of the environment, an unfamiliar adult, or other environmental factors.  Or it may simply be that nothing happened and the child can't say anything about a non-event.  It takes much less of a leap of faith, much less credulity, to understand nonresponse to be caused by environmental variables or the absence of any abuse than to infer the convoluted, non-empirical chain of speculative intervening variables required to say nonresponse proves abuse.

Environmental cues and constraints are powerful determinants of behavior for both adults and children and children are more susceptible to environmental influences than are adults.  Therefore, the place to begin understanding a child is in the environment.  This does not exclude consideration of internal, private events but rather is likely to sharpen and focus any discussion of internal events to reduce purely speculative invention of non-empirical intervening variables.

When a mental health professional fails to consider environmental factors when dealing with an accusation of sexual abuse, there are three possible causes: (1) The professional may be ignorant, ill-trained and unaware of the scientific facts about environmental influence; (2) The professional may be committed to a theoretical position (i.e. Freudianism, ego-psychology) that minimizes or denies the influence of the environment and emphasizes internal subjective processes; (3) The professional may unwittingly be projecting onto the child an adult subjective experience of adult internal motivations and intents.  Whatever the cause, failure to consider environmental influences when evaluating an allegation of child sexual abuse based solely upon statements of a young child shows a flawed, mistaken, and biased approach.
  

Adult Social Influence

When examining the environment of a child the first question to ask is what degree, level, and type of adult social influence is exerted upon the child in that specific environment.  The power and centrality of social influence is one of the most solidly established scientific facts in psychology.  To ignore the impact of social influence upon any human interactive behavior reduces the credibility of any explanation offered.  How adults behave toward children during the course of a developing sexual abuse accusation must be considered a possible causal factor in producing statements of a child.  Before assuming that an accusation is true, in the absence of independent, corroborating data, adult social influence as a cause of children's statements must be ruled out.

We have seen again and again, in sexual abuse cases throughout this country and in other countries the massive imposition of adult social influence on suggestible children.  Children are questioned and interrogated repeatedly by adults who believe that the abuse is real.  The adult behaviors toward the children are frequently coercive, leading, suggestive, and punitive.  Under such pressure children may produce statements which adults then believe prove sexual abuse happened.  Through this environment of powerful adult influence the child may be taught an account of sexual abuse that is false.

By the time this process of adult social influence interviews has been repeated, sometimes over a number of months or even years, a non-abused child may come to believe the story.  When a child has been taught a fabricated story, repeated it enough to produce subjective certainty, for that child a non-event has become reality.  This is not a benign or innocuous experience.  It may result in long-term damage to a child.  This also means that the child will appear truthful, credible, and reliable because for the child it is a real story.  Reliance by a fact finder upon the demeanor of a witness to assess credibility is then an error.
  

Victims of Good Intentions

An adult interrogator is seldom deliberately trying to brainwash the child or develop a fabricated account of sexual abuse.  This may happen occasionally in divorce and custody situations, but it probably accounts for a small minority of the cases of false accusations.  Instead, the interviewer is likely to have a preconceived idea about abuse along with an lack of awareness of the suggestibility of children and a lack of understanding of the environmental factors including the stimulus value of adults.

Few persons involved in interviewing children show any awareness of the impact their procedures have on the children being assessed.  This is true of mental health and law enforcement professionals as well as laymen.  Much of the literature on child sexual abuse does not deal forthrightly with this issue but rather proceeds on the assumption that whatever a child says or is alleged to have said must be believed.  This literature does not discuss the impact of the interrogation process upon statements alleged to establish that abuse occurred.  This is surprising and distressing in that any knowledgeable and competent mental health professional knows that in any human interaction involving two or more people factors of social influence are at work.
  

Belief in the Face of Unpopular Facts

Many professionals are convinced that any alleged disclosures from a child must be unconditionally believed.  Yuille (undated) points out that children's disclosures of abuse are now given so much credibility that it is often sufficient for the child to show sexual symbolism in play for a social agency to remove the child from home.  If an adult interrogator assumes that abuse did, in fact, occur, he will perceive his role as substantiating the abuse so that appropriate action can be taken to protect the child and punish the perpetrator.  Raskin and Yuille (in press) state that "It is common practice for interviewers to assume that the allegations are true and that the purpose of the assessment is to obtain information that can be used to arrive at that conclusion."  An adult who believes the accusation is true will then question the child in a way that shapes, molds, and creates statements about abuse.

Years of established and replicated research in social psychology demonstrate the principles of behavior that are involved in adult social influence on children (see Wakefield & Underwager, 1988, for a summary).  Together with the evidence for suggestibility, these are unpopular facts that raise cautions, disconfirm, and may falsify many allegations of child sexual abuse.  But what happens when there is scientific evidence contrary to strongly held beliefs?

What should happen is described by Skinner (Mahoney, 1976), "Science is a willingness to accept facts even when they are opposed to our wishes."  Hardly anyone would say that we should deliberately ignore facts.  Scientists agree that a good theory is one that can be tested and proven false.  When a theory or concept is falsified, it is reasonable to expect that some attitudes and beliefs will change in accordance with the facts.

What does happen is demonstrated by a large body of research (Lord, Ross, & Lepper, 1979; Ross & Lepper, 1980; Ross, Lepper & Hubbard, 1976; Festinger, Riecken, & Schachter, 1958; Aronson, 1988).  People don't like to see or hear anything that conflicts with deeply held beliefs or wishes.  Some time ago the response to bad news was to kill the messenger — literally.  Today it is more likely that the source of the facts is demeaned and belittled.  When those who ignore the scientific facts that would disconfirm their beliefs are confronted with those facts, the best prediction is that they will demean the source and refuse to change their beliefs or behaviors.  A good example of this process is the continued reality that millions of people smoke in the face of overwhelming evidence that it kills them.

What are these facts from social psychology?  There are several areas of the science of psychology that appear in every introductory psychology textbook.  They are firmly grounded in broad and rich bodies of theory and research literature.  Every psychologist ought to be familiar with them and understand their basic principles.  They cannot be gainsaid or ignored without truncating the body of scientific fact psychology has built.  These are the unpopular facts that must be acknowledged.

The research on the expectancy effect and experimenter bias demonstrates that expectancies about an outcome of an experiment or interaction can influence the outcome itself.  Biased investigators will err in the direction of their expectancies when they summarize, analyze, and interpret their data, and their own attitudes and expectancies will influence the actual behavior of their subjects.

In interviews, the bias of the interviewer can affect both the selection of the information to be recorded and the substance of the information itself.  Subjects respond the way they feel to be most proper in light of the interviewer's expectancies communicated verbally and nonverbally.  Garbarino and Scott (1988) state that in interviews with children, experiences of a professional, unfounded in empirical data, but which the professional strongly believes in, can influence what information is given by the children.  If the professional believes that all or most allegations of abuse are real, he will produce information to validate abuse.

The research on conformity and compliance demonstrates how the desire to get approval from others exerts a powerful influence upon behavior.  In interviews with children there is pressure to conform to the perceived expectations of the adults.  This can only be avoided by a careful effort by interviewers who recognize their own stimulus value and minimize cues on how to respond.

Reinforcement theory describes how behavior is controlled by the consequences that follow the behavior.  Theories on parenting techniques recognize that the best way to change the behavior of a child is to use attention, praise, approval and other social reinforcement.  Children are sensitive to approval from adults and will quickly learn to behave in the way that gets positively reinforced.  If interviewers are not aware of this, they will inadvertently reinforce responses of the child that confirm their prior biases.

In many cases the social reinforcement given to children is obvious. A child may be told that she is brave and that "Mommy will be proud of you for telling the scary secret" or a child is told that he can go for a treat after he tells about the abuse.  But children are also rewarded for making certain statements simply by a smile, a nod, an approving tone of voice, saying "what else," etc.  It is also reinforcing to remove an aversive stimulus.  When an adult stops asking repeated questions when the child produces the desired answer after several repetitions, the cessation of the repeated question reinforces a specific answer and may also reinforce giving an answer approved by the interrogator.
  

The Child is an Object, Not a Person

When an adult relates to a child by ignoring the interactive nature of the relationship, denying any impact of the adult's behavior, intents, and purposes, and refusing to consider environmental variables, that adult has objectified the child.  The child is not treated as a person, is not understood to be a child, but is only an object for the adult's agenda.

This is the same basic problem feminists have objected to in their discussion of male-female relationships.  When a man approaches a woman with nothing in his mind but his own agenda — sexual satisfaction — and shows no awareness of the other person's value, needs, desires, capacities or personhood, the woman is turned into an object.  This demeans, diminishes, and renders powerless the person objectified.  Feminists have shown us that this is a form of oppression and victimization.  In this instance the rhetoric matches the reality.  Male dominance and control of women has generated deep, enduring, and often devastating hurts.  It has produced costs to society, the economy, culture, and the nation states.

If children are to be treated as a persons in their own right, adults must be aware of their own position of power and the ability to dominate and control a child.  Adults must control their own purposes, assumptions, and behavior toward a child and acknowledge the potential for their behavior to influence the child.  If this is not done there is no way the adult can avoid objectifying the child.  It is tragic and odd that an attempt to assure children are treated as persons winds up not treating them as persons.  We can and must do better in discovering ways to relate to children that foster their individuality and their personhood.
  

How Does It Get Started?

The development of the system we now have for responding to allegations of child sexual abuse is a social process.  The main players are medicine, politics, the justice system (including social workers and law enforcement), and psychology.  The relationships among the persons who are in these institutions, guided by shared norms, beliefs, and values, are the social characteristics of the child abuse system.  A model that explains how this started is the "invisible college" concept (Crane, 1972) which shows the powerful informal communication networks linking the justice system, the media, and groups of researchers and political figures.  Studies of bibliographic citations analyzing the communications patterns empirically demonstrate the reality of the "invisible college" (Dunn, 1981; Yokote & Utterbach, 1974; Price & Beaver, 1966; Price, 1965; Lipton & Hershaft, 1985).

The "invisible college" connected with sexual abuse is readily evident in the interlocking system of a small number of researchers, prosecutors, mental health professionals, and politicians who revolve through seminars, workshops, lectures, training sessions, media presentations, and literature.  There is a common set of beliefs and values, a common set of attitudes, and a common set of behaviors characterizing this group both in this country and abroad.  The result is the rapid dissemination of views, claimed research findings, putative methodologies through this network and their use in the real world.  Examples are rapid proliferation of the use of dolls, coloring books, prevention programs, unsupported maxims such as children are traumatized by testifying in the presence of the accused.  All of these became standard operating procedure in accusations of child abuse in the absence of any credible, reliable evidence to support their use.

In a typical example of this, a mother brought in her three-year-old daughter for assessment of alleged sexual abuse following the child's return from visitation with the father.  The mother told the pediatrician that she believed the father (with whom she was involved in an acrimonious custody dispute) had sexually abused the child.  The pediatrician saw the child for twenty minutes and conducted a sexual abuse evaluation by visually observing her genitalia.  He used no procedures of measurement or magnification other than his unaided visual observation.  He made a positive diagnosis of sexual abuse although there were no visible signs and the child said nothing to him about abuse in response to his questions.  He stated that he assumed that the history given by the mother was true.

When questioned about the physical examination, he said that the magic number was a hymeneal opening over 4 millimeters.  This was proof of penetration and therefore sexual abuse.  Responding to further questions he said that his base for this unequivocal diagnosis was what he had heard about hymeneal openings from another physician during a grand rounds in his hospital.  He did not know any research or scientific literature about the range of hymeneal openings in normal non-abused children.  He called what he heard in the grand rounds "medical information."  When given research falsifying the assertion that anything over 4 mm means abuse, he refused to accept it saying you can find a study to prove anything you want.

An example of how such a system can function to produce policies and behaviors based upon dubious and mistaken theories and research is medical research on herpes.  Lipton and Hershaft (1985) describe how an article published in a prestigious American Medical Association journal was widely accepted, touted as fact, advanced as a basis for treatment, but is highly vulnerable to scientific criticism and of doubtful value.  The few attempts to analyze systematically the problem of unsound or fake research have raised great anxiety about the quality of scientific research.  Even more fearsome is professional readiness to accept and act upon purported research without any critical acumen or exercise of rational suspension of belief.  Of the few professionals who read research literature, most never read the methodology section but skip to results and discussion.
  

A Common Pattern

We have examined transcripts, audio- and videotapes, charges, psychological, evaluations, and testimony from hundreds of cases (Wakefield & Underwager, 1988) and have found that the way children are interrogated when sexual abuse is suspected shows a common pattern across the nation and in other countries.  The structure of reporting laws, child protection agencies, law enforcement officials, prosecutors, and the laws and regulatory codes governing these agencies shape the common pattern.

An adult usually first suspects possible sexual abuse of a child.  (Although an older child may spontaneously say something to an adult, with younger children, the process most often begins not with the child but with an adult.)  The most frequent trigger for the suspicion is some sort of change in the child's behavior or condition such as bed wetting, nightmares, or a slight redness or soreness in the genital area.

The adult then questions the child and calls the authorities.  If the adult is not a parent, the parents are also informed (although in some instances the first the parent learns of the accusation is when the authorities arrive and begin an investigation).  An initial report is made either to the child protection agency or to the police.  If the first report is made to the police, the child protection group is usually informed.  The first person who has contact with the child or the child's family is usually a social worker.

Sometimes the child is first taken to the family physician or to a hospital emergency room where the adult tells the doctor that abuse is suspected.  The doctor may question the child and then make a report to child protection.  Although there are seldom clear physical signs of sexual abuse, the notes of the physician frequently state "suspected sexual abuse" based on the history given by the adult.

The parent has probably questioned the child before the police or child protection gets involved.  The powerful emotion triggered by a suspicion of sexual abuse may result in intense, repeated, and highly suggestive questioning.  Then when the official talks to the reporting adult, the parent will give his recollection of whatever the child said in response to this questioning along whatever suspicions led to the original report of abuse.  If the investigating official believes that children must always be believed and false allegations are rare, he will accept this account as factual.  The initial official contact with the child will therefore be based upon the assumption that the alleged abuse really happened.  This bias affects the way the official questions the child and the subsequent outcome of the investigation.

What happens in this first official interrogation is important in assessing the credibility of any statements a child makes.  The younger and more suggestible the child is, the greater the significance and effect of this first interrogation.  It will set the direction and the scope for all future contacts with the child.  But it is probably the least documented and most likely distorted of the succession of interrogations.

The first official interrogation of a child may range from a single social worker interviewing the child to several people, including police, social workers and prosecutors, coming unexpectedly to the home and taking the child to the police station (Jordan, Minnesota).  There is often a social worker and a police officer, or two or more officials.  Particularly in divorce and custody cases, the accusing parent may take the child to a child protection worker, physician, or mental health professional for the first official interview and be present for the interview and even participate in it.

The initial interrogation by officials is usually not recorded.  There may or may not be notes or reports but the amount of information available about this first interview is generally minimal.  Most often, the only information is a report summarizing what the child allegedly said in the interview.  However, if the interview is not recorded, there is no way to know what actually went on.  We have found that reports of what supposedly transpired in an interview are often markedly different from what actually took place (which we discover when we are able to view the videotape of the actual interview).

Often interviewers ask a question or make a statement to which the child gives little response.  After the question is repeated several times, the child may finally nod or answer yes.  But in the written report, the child is presented as making the statement rather than only agreeing with the interviewer's statement.  There is seldom mention of any denials which may have preceded the eventual affirmation.  This is probably not a deliberate misrepresentation; instead, the prior beliefs and bias of the interrogator lead to an erroneous recollection of what actually happened.  Herbert, Grams, and Goranson (1987) state that video-or audiotape recordings are essential for accurate knowledge of what went on in an interview.  Without them, the conclusions drawn about the interview by the interviewer are likely to contain significant factual distortions.  They report that without video or audio taping, interviewers reflect their bias by giving inaccurate and mistaken reports about the interview.

Following the first official interrogation, there is a wide variation in what happens next.  Sometimes there is only the initial interview.  There may be an additional interrogation which is recorded in audio- or videotape.  But the child may be questioned repeatedly by social workers, prosecutors, therapists, parents and foster parents, siblings, or others.  Sometimes the child is taken from the parents and placed in foster care where the foster parents ask questions and encourage talk about the alleged abuse.  The child may be placed in sexual abuse therapy where he talks regularly to a therapist about the alleged abuse.  If the issue is brought to adjudication, either in criminal, civil, family, or juvenile court, the child may be questioned frequently by the prosecutor or attorney and brought into the courtroom to "familiarize" him with the environment.

Through this the account of the abuse is further rehearsed.  It is often months, or even years, before the justice system makes a determination about the abuse.

This pattern is not limited to the United States.  We are familiar with similar techniques and similar problems from cases in Canada, New Zealand, Australia, The United Kingdom, and the Netherlands.  Hayes (1987) describes interviews in England in which great pressure is put on the child to disclose the alleged abuse, including the use of directed play with dolls and highly leading and coercive questions.  Interviews with leading questions and directed interactions with the dolls were used in Cleveland, England with the result that two hundred children were falsely identified as having been sexually abused by their parents.  As in the United States, the problem was with interviewers "who commence an interview with a preconception that abuse has taken place" (Enright, 1987, p.672).
  

Interrogation as a Learning Experience

Children may be interviewed dozens of times before a legal determination is made about abuse.  In every interview the child learns more about what the interrogator expects and learns what to say or do that will get a positive response from the interrogator.  The child learns the language of the sexual abuse literature, such as the distinction between "good touch" and "bad touch."  The child learns about sexual behavior, including deviant behavior, and learns to equate sexual touch with touch that hurts.  The child learns the victim role and learns to express anger towards the alleged abuser.  The child learns the story and may come to believe it happened, even if the allegations are false.

The interviewer must know something about the allegations in order to ask any questions.  He bases his questions on his own assumptions about what has happened and the direction of the interview is determined by these questions.  This results in a bias in the interview procedures of even the most skillful investigators.  If the interrogator is unaware of this and has strong and certain beliefs, this bias will be very powerful.  It frequently results in an interview directed not towards determining the truth but towards substantiating abuse.

The bias results in the interviewer attending to information that supports his beliefs and ignoring details which don't support his assumptions or which suggest a different direction.  Statements of the child that do not fit into the interviewer's beliefs are seen as evasions or confusions.  When a child says that nothing happened the interrogator keeps repeating the question and asking other questions until he finally gets the child to affirm the abuse.  If the interrogator is is persuaded that he is right, he is very likely to falsely confirm his theory through this process.

In a study about this process of interrogation of children, it was found that the belief of the interrogator about the truth of the allegations was predictive of the outcomes of the interrogation (Dent, 1982).  If the initial first interrogation involves an adult who has the belief that abuse occurred and who the abuser is, that prior belief will affect the outcomes of the interrogation.

The child tries to figure out and produce what he believes the adult wants to hear.  The desired responses are cued to the child by tone of voice, inflections, small body movements and postures as well as by suggestive and repeated questions.

This is particularly true when the child has an inadequate memory of what is being asked.  Children are apt to add material when they do not remember and the practice of asking children "what else" is likely to increase the number of errors of adding extraneous and contradictory information (Saywitz, 1987).

The variables of power, authority, status, and credibility of the adult interrogator interact with the limited capacity and competencies of the child to produce a powerful confounding of this interrogation process.  This entire process contaminates, confuses, and lowers the reliability of statements made by children.

There is no evidence that this is a reliable process for assessing possible sexual abuse in children.  The reality that is completely overlooked is that each of these experiences of interrogation is a learning experience for the child.  The younger the child the more powerful the teaching and learning experience.  But the persons interrogating children seldom show any awareness of their own stimulus value or of the impact of their procedures as a learning experience upon the children and the reliability of statements made by them.
  

Types of Questions

Different kinds of questions will elicit different responses.  An open-ended question calls for spontaneous, free recall.  For example, a parent might ask a weeping child "What happened?"  If open-ended questioning does not produce sufficient information, the interviewer may turn to more specific questions, such as "Did he hit you?"  At this point the questioner has taken a more active role and the witness a more passive one.  Research has shown that while specific questions result in an over-all increase in the number of statements a witness makes in comparison to free recall, the increase is due to a rise in both accurate and inaccurate statements (Dent & Stephenson, 1979; Lipton, 1977).  Thus the memory for an event can be made more elaborate, but the greater detail will include more false memories as well as more truth.  Child witnesses are more subject to this error than adults because they give fewer answers in free recall (Kobasigawa, 1974; Mandler & Johnson, 1977; Perlmutter & Ricks, 1979) and therefore may cause interviewers to turn sooner to specific, closed questions and to use proportionately more of them.  The predominant method of obtaining information from children is to use leading and suggestive questions (Thomas, 1956).

Turtle and Wells (1987), commenting on the recent research on children as witnesses, observe :hat the paucity of children's recall:

...can lead to an inordinate amount of subsequent questioning from various agents throughout the legal proceeding and hence to a greater exposure to possible misleading information.  Unfortunately for the system ... children suffer from greater susceptibility to having their testimony distorted by such misleading information (p. 240).

Adults are more suggestible when an authoritative rather than a nonauthoritative person asks leading questions (Eagly, 1983, Loftus, 1979). Ceci, Ross and Toglia (1987) state that the young children's suggestibility could be partially accounted for by the fact that they are likely to conform to what they believe to be the expectations of the adult.  It may well be that young children are especially affected by suggestion and leading questions simply because so many people are generally authoritative in relation to them.  This would be particularly pronounced if the child is being interrogated by someone identified as a doctor, a therapist, or a police officer.  Parents are also authority figures to their children.

In a more active line of questioning, the interrogator is supplying information to the witness.  "Did Allen hit you on the arm?" and similar questions can give shape and content to the recall of a memory that is, in fact, vague.  There have been many studies that demonstrate how the memories of both children and adults can be distorted by the introduction of false information into questions (see Loftus & Davies, 1984, for a review).  When an unsure or reluctant witness causes the questioner to guess at what might have occurred and thereby provide information, perhaps true and perhaps false, for the witness to affirm or deny, the resultant testimony may be the truth or it may be a fabrication that is mutually agreed upon and believed to be true by both parties.
  

Research on Memory and Suggestibility in Children

Children are the main witnesses in sexual abuse cases.  Their suggestibility and susceptibility to influence as well as the accuracy of their memories must be considered in assessing their testimony.

After the turn of the century there were many studies on children's memory and suggestibility.  Many professionals concluded from this research that "children are the most dangerous of all witnesses" and demanded that children's testimony be excluded from the court record (Goodman, 1984).  At the same time, the spontaneous account of an event by children was thought to be reliable.  The overall picture from the early studies is of a potentially accurate witness, who can recount events and answer non-leading questions reasonably correctly, but whose report can easily be contaminated by suggestion.

Later studies compared the memory and suggestibility of children to that of adults.  Much research demonstrates that adults' memories are influenced by suggestion.  Loftus (1979) and Loftus and Davies (1984) report the results of studies in which subjects are presented with a film of a complex event, and afterwards are asked a series of questions.  Some of the questions are designed to present misleading information.  The subjects presented with the misleading question are afterwards more likely to "recall" having seen something that was not present in the film.  This false information will be integrated into the memory.  Once the alteration occurs, it becomes entrenched and it is difficult to induce a witness to retrieve the original memory.  The question that has been investigated in studies of children's memory and suggestibility is how suggestible they are in comparison to adults.

Children typically recall less than do adults (Johnson & Foley, 1984).  But this free recall is generally accurate.  However, since the typical interview with children contains a large proportion of leading questions, in evaluating their ability to serve as witnesses, memory is only one consideration.  What is their suggestibility compared to adults?  It is well established that adults are influenced by leading questions.  Are children influenced in the same way?

There has been a debate concerning whether children are more susceptible than adults to distortions of memory caused by leading questions.  The older studies concluded that children were more suggestible.  Newer studies found that both adults and children are influenced by leading questions but were inconsistent as to whether children were more suggestible.  However, recent studies have found young children to be more suggestible than adults and younger children to be more suggestible than older children (Goodman & Reed, 1986; Goodman, Hepps & Reed, 1986; King & Yuille, 1987; Ceci, Ross, & Toglia, 1987; Goodman, Aman & Hirschman, 1987).  Young children are particularly bad at making eyewitness identifications, especially when the target individual is not present in the lineup.  In such cases, the child makes a very large number of false identifications (Peters, 1987, 1988).

Part of what happens with younger children is that the less a child remembers the more he can be misled and the younger a child is, the less he will remember.  The less a child reports in free recall, the sooner the interviewer will turn to using leading questions.  Also, children may have a different perception of tasks than do the adults.  Children are likely to draw upon all available information in the interview situation to provide the interviewer what they believe the interviewer wishes to hear.  Cole and Loftus (1987) state that " ... the demand characteristics of being given certain information by an adult, and even of being questioned by an adult are powerful components of suggestibility in young children." (p.199) and Ceci et al. (1987) indicate that the young children's suggestibility could be partially accounted for by the fact that they are especially likely to conform to what they believe to be the expectations of the adult.  To avoid this, King and Yuille (1987) stress that the child be told that the interviewer is only interested in what the child remembers and that admissions of memory failure and memory gaps are expected.

A problem with all of this research in suggestibility is its ecological validity.  Actual situations faced by the child witness cannot be reproduced in a laboratory study.  For example, in a typical study, the children are presented the misleading information once and may be given two or three leading questions and/or misleading information.  Our research suggests that leading questions and other types of error-inducing questioning occurs from half to four-fifths of the time in the typical interview of the child witness.

In the research studies, the children are tested immediately or after several days.  But in the sex abuse cases in which children are required to testify as witnesses, the children are often interviewed many times by a variety of people over a long period of time.  Ornstein and Gordon (1988) stress that we have no information in the literature on the ability of children to remember salient events over months or sometimes years.  In situations where a child will eventually testify, the memory will consist of a combination of recall and reconstruction influenced by all of the interrogations, conversations, and sexual abuse therapy that have occurred during the delay.  The longer the delay, the greater the possibility of social influence and the more the memory may consist of reconstruction rather than recall.

Most of the experiments are on children's recollections of events that they have observed.  But when the child is a witness in a sex abuse case, he is alleged to have been involved in a traumatic event.  Although two recent studies have used actual events — a visit to the dentist (Peters, 1987) and a shot at the doctor's office (Goodman et al., 1986) — these events are likely to be less stress-producing than sexual abuse.  In addition, as Raskin and Yuille (in press) point out, providing testimony in a sexual abuse case may have profound effects on the child's life.  A disclosure of sexual abuse may lead to the break-up of the home or placement in a foster home.  In research studies, the children are unaffected by the events they are reporting and their testimony has no consequences.  Raskin and Yuille state that "these differences ... render tenuous any conclusions that may be drawn from the published literature."

In summary, children's memories in a free recall situation may be quite accurate.  If the child is interviewed carefully and leading questions and suggestion are avoided, a child witness can be competent to provide testimony.  However, the suggestive and coercive nature of the interrogations commonly imposed upon children means that the credibility of what they say must be carefully assessed.
  

Frequently Used Interrogation Techniques

There are several techniques that are commonly used in the interrogations.  These include the use of "anatomically correct" dolls, books such as "Red Flag Green Flag People," puppets, drawings, role play, establishing rapport with the child, establishing the credibility of the interrogator, and simplistic attempts to determine the competency of the child.  But there is no evidence establishing that these procedures are reliable or valid techniques in assessing possible sexual abuse in children.  Their use is likely to contaminate and influence the statements children may make and therefore raise serious questions about the reliability of the statements.
  

Drawings

Children's drawings are often used in assessing possible sexual abuse.  The rationale for this is that the drawings of children with emotional problems are believed to differ from the drawings of normal children.  The assumption is that qualitative features of the drawings may be used as "signs" that indicate the child has been sexually abused.

However, the research does not support this assumption.  In surveys of the DAP (Draw A Person), Harris (1963) and Roback (1968) both conclude that there is very little evidence to support the use of "signs" as valid indicators of personality characteristics.  There is no research supporting the claim that qualitative "signs" (elongated or squat figures; smoke or no smoke from chimneys; absence or presence of windows; hands in front of the genital area, and so on) have any relationship to sexual abuse.

There is so much variability in children's drawings that nothing reliable can be said about the particular features of any one drawing.  There are no reliable or valid criteria for using the drawings to assess possible sexual abuse.  There are major shortcomings in the few research studies which claim that the drawings of abused and non-abused children differ (Wakefield & Underwager, 1988).  We know of no good research establishing that the drawings can be used diagnostically, especially to "substantiate" sexual abuse.

The only valid use of drawings is in opening up a line of communication between the evaluator and the child.  But if they are used this way, the interviewer must remain objective and impartial and avoid cues and selective reinforcement of responses.
  

Books\

There are a number of books that are used in assessing sexual abuse.  A typical book is the coloring book "Red Flag Green Flag People" (Rape and Crises Abuse Center, 1985).  In this book, the child is led through a series of pages that present good touch and bad touch.  After several pages a child is asked to color portions of a figure that may have been touched.  When the child colors a genital area this is regarded as evidence that the child has been abused.

When used in this way, a book becomes a programmed learning text.  The progression of stimuli are arranged in the fashion of programmed texts used to teach students about biology, math or geography.  Children's responses in this situation do not represent a true account but rather the effectiveness of the book as a programmed text.

None of these books have been validated for diagnosing child sexual abuse.  They should not be used as diagnostic devised to conclude that a suspected case of sexual abuse is real.
  

Play Therapy

Although there is no evidence that play therapy has any efficacy or utility as a therapeutic procedure for sexual abuse (Wakefield & Underwager, 1988), children are frequently given therapy before there has been any determination by the justice system that sexual abuse has, in fact, occurred.

The play therapy sessions focus on reenactments and discussions of the alleged abuse and can serve to model and encourage statements about abuse.  Young children may be given dozens of sessions of play therapy over many months and may begin believing that they have been sexually abused even if there has been no abuse.  The play therapy is often combined with interviews and questioning by the therapist about the alleged abuse.  The behavior of the child in the play therapy sessions is used to form conclusions about abuse and material from the play therapy sessions is often reported to the prosecution, child protection workers and the courts as support for the accusation.  This material is often given without any empirical basis.

A major difficulty with the use of play therapy as an investigative tool is the unexamined assumption that play reflects reality.  For example, in one case, a child who probed toy animals with a tinkertoy was said to be "reenacting" the alleged insertion of objects into her vagina while in her day care center.  Children have been said to be showing signs of abuse when they play with certain figures at a sand table, when a boy playing with male dolls has them wrestle, when a girl pretends she is putting diapers on a doll, when a doll the adult labels at "Daddy" is placed in a dollhouse the adult labels as "jail."  But there is no evidence that behaviors occurring in play therapy can be used as signs to establish the truth of events that are believed to have happened in the past.  Nevertheless, the behaviors of the child in play therapy together with the statements the child is alleged to have said are given often used as evidence that the child has been abused by the person accused.
  

Dolls

The "anatomically correct" dolls are widely used in the assessment of alleged child sexual abuse.  They have been criticized and recently there have been several studies concerning their use.  Some of the studies report differences between abused and non-abused children; others report no differences.  However, the studies claiming to show differences between the responses of sexually abused and non-abused children have major methodological shortcomings which limit any conclusions that can be drawn from them.  Taken as a whole, there is no evidence from the research on dolls supporting their use as valid methods for diagnosing sexual abuse.

A major difficulty in the unquestioned acceptance and use of these dolls is that until recently there has been little information about how normal children respond to the dolls.  Baseline information about the antecedent probability of a child's response to the dolls is necessary in order to interpret what responses to such dolls mean.  But several recent studies indicate that normal, non-abused children may respond to the dolls in a way that is likely to trigger suspicions of abuse in the investigator (Gabriel, 1985; Mclver, Wakefield & Underwager, 1989; Boat & Everson, undated; Herbert, Grams, & Goranson (1987).  King and Yuille (1987) point out that "... the dolls serve the function of a suggestive question with young children.  The genitals and orifices of the dolls suggest a play pattern to children, and that play may be misinterpreted as evidence for abuse" (p.31).

Herbert, et al. (1987) and Mclver, et al. (1989) also found that the children were very suggestible in response to leading questions with the dolls and could easily be led to produce statements and behaviors which could be interpreted as indicating sexual abuse.  Although Aman and Goodman (1987) reported that the dolls in their study did not lead to false reports of molestation, the three year olds in the doll condition were significantly less accurate in answering the objective questions than were the three year olds in the no doll condition.  Also, false alarm errors, that is, errors that could lead to false reports of abuse, were significantly greater for three year olds than five year olds and one in five three-year-old children made such errors.

White, Strom, and Santilli (1985) and White, Strom, Santilli, and Halpin (1986) developed a protocol for interviewing preschoolers with the "anatomically correct" dolls and reported that with their protocol, non-sexually abused children interacted differently with the dolls than did the abused sample.  Jampole and Weber (1987) and August and Forman (1986) also reported differences between abused and non-abused children.  However, in all three of these studies there is no information concerning the content of any interviews about sexual abuse, any therapy given to the sexually abused children, or what discussions were held with these children about sexual abuse by foster parents or social workers.  Without this information, no conclusions can be drawn from their results.  Thus these studies do not really compare abused and non-abused children but rather children who have been interrogated by the system and children not interrogated.  Also, Jensen, Realmuto, and Wescoe (1986), using White's protocol, found no differences between abused and non-abused children.

Interviewers using the dolls assume that the children will identify the dolls as male or female and then use the dolls to symbolize actual people to demonstrate what has happened to them.  However, these assumptions are not supported.  Children do not identify gender on the basis of the symbolic genitalia, rarely use the dolls to symbolize persons in their lives, and seldom spontaneously choose to play with them (Mclver, et al., 1989; Herbert, et al., 1987; Sivan, Schor, Koeppl, & Nobel 1988).

The use of the dolls can provide modeling and teaching to the child.  Interviewers model handling the dolls, suggest that they be undressed, and label them for the child.  They may say "Let's pretend that this is you and this is Daddy," which invites the child to confuse fact and fantasy.  They ask the child to show what Daddy did with the dolls and they may place the dolls in the sexually explicit positions for the child.  This is a teaching experience for the child.

When an adult gives support and structure to a child in using dolls to tell a story the child produces a story much different and at a higher developmental level than when left to produce a spontaneous story with little adult support.  When actions are modeled by an adult using dolls the effect is to facilitate pretending and fantasy behavior (Watson & Fischer, 1977).  This falsifies any claim that the dolls can be used with young children because then they can show something real with dolls that they can't talk about.

There has been no research to establish standardized procedures for using the dolls in an investigation.  Although White, et al. (1986) describe a protocol, there are no normative data to standardize it.  Standardized procedures that can be repeated by others are an absolute requirement before anything sensible can be said or any conclusions drawn.  A California Appeals Court ruled in 1987 that the use of the dolls was not supported by the scientific evidence and their use did not meet the Frye test for admissibility.  Testimony based upon the use of the dolls was therefore ruled inadmissible (Law Week, 1987).

The problem is exacerbated by the fact that persons actually using the dolls are untrained, unsophisticated, ignorant of child development, and widely variable in their interpretation of behavior with the dolls (Boat & Everson, undated).  The American Psychological Association's Committee on Psychological Testing and Assessment determined in its March, 1988, meeting that the dolls "are considered to be a psychological test and are subject to the standards when used to assess individuals and make inferences about their behavior" (Landers, 1988).  This means that a psychologist who uses the dolls and reports conclusions based on their use without including appropriate cautions about their reliability and validity is behaving unethically.

A basic issue is the use of the dolls to elicit behaviors from young children that can be said to show abuse.  Children are in a process of developing.  They show various levels of capacity and ability throughout that process.  This puts limits on what children are able to do at any given level of development.  Adults want to break through developmental limits in order to attain their objectives.  This means that whatever techniques adults use to overcome the limits of a child's developmental ability are high risk for producing false and mistaken information reflecting the adult agenda rather than the truth.  If such techniques are to be used, they must be shown to avoid this risk.  If they do not demonstrate validity and reliability, they must not be used.

The use of the dolls in interviews must not be viewed as a pursuit of truth but rather as a learning experience.  To date there are no data that support a differential behavior of abused and non-abused children when the dolls are used to assess sexual abuse.  The data that is available suggests that they cannot be used to distinguish abused from non-abused children.
  

Video- and Audiotape Analysis Research Project

We are engaged in an ongoing research project of analyzing audio- and videotaped interviews from actual cases of alleged sexual abuse (Wakefield & Underwager, 1988).  To date, we have analyzed 35 cases.  We have reviewed additional videotapes in many other cases; the cases where we performed the analysis are typical of the ones we have seen.

The project does not attempt to determine the truthfulness or untruthfulness of the statements of the children.  It examines the behaviors, statements, and questions of the participants in actual interviews.  We therefore obtain information on the interviewing process and the responses of children in a real world situation.  The video- and audiotapes are interrogations in actual cases.

Six categories of interviewer behaviors were defined as error-inducing: closed questions, modeling, pressure, rewards, aids and paraphrase.  Closed questions and modeling can give information to the child on how to respond.  Along with pressure and rewards, paraphrasing can reinforce the child's response.  Aids such as the anatomically-correct dolls, which are used in most of the interviews, can provide a modeling effect to the child and can potentially generate false information.

The proportion of adult behaviors that are error-inducing are fairly similar in the cases we have analyzed.  The behavior of the adults that potentially convey information to the child on how to respond range from half to four-fifths of the interviewer behavior (the average is two-thirds).  The adults also do most of the talking in the interviews.  In most cases, the adults are from one and a half to three times more active than the child.  The picture that emerges is one in which the adult is active and task oriented and the child plays a more passive role.

The behaviors of the adults seldom are geared to encourage free recall and allow the children to tell their own accounts free from pressure and suggestion.  The categories of closed questions, pressure, rewards, use of aids, and modeling are adult behaviors that teach a child what is expected, what story to tell, and what pleases the adult.  The adult interrogators often appear trying to extract testimony that will substantiate abuse that they have already concluded is real.

This picture is more disturbing in that the children often had been interrogated several times before the interview was recorded.  It is likely that knowing the interview was being taped would result in the interrogator trying to avoid obvious questionable behavior.  The undocumented and unknown interrogations that precede the documented one are likely to be more coercive and suggestive than the ones we analyzed.

Other professionals are recognizing the problem of suggestive and rehearsed interviews and the importance of a videotaped documentation.  DeLipsey and James (1988) report on the Texas experience with videotaping.  Although the purpose of the videotaping procedure was to protect child witnesses by providing an alternative to the child testifying in court, these authors state that they soon came to realize that the videotaped procedure protected the rights of the accused as well.  They report that they have observed numerous examples of bribery and coercion during their review of videotaped interviews.  Raskin and Yuille (in press) state that "a videotape is the only means whereby the procedures and data obtained during the interview can be fully documented."

Studies on eyewitness testimony and memory and suggestibility of children have a much smaller proportion of leading questions than we found in the tapes we analyzed.  In the real world children being interviewed are given much more error-inducing information than in laboratory research.  There is a much higher level of coercion, pressure and reward for responding in the desired manner.  In some tapes children were told that they couldn't play with toys (or go to the bathroom, or go home or get a treat) until they gave the desired response.  Following a statement about abuse, children were frequently told that they were brave, courageous and that their parents would be proud of them.  In one tape, the interviewer asked the child the same question eighteen times.  In a day care case, the interviewer told several children that another child had already told him about being abused.  But when tapes of the earlier sessions with the child named as telling about abuse were examined, such statements were not there.  Children who do not cooperate in affirming abuse have been called "fraidy cat" or told "you must be stupid."

Our analysis of actual interrogations in sexual abuse cases shows that the real world is much tougher on children than is the research laboratory.  Even the research studies that attempt to duplicate real life situations fail to come close to the amount of pressure and coercion put on children in the tapes that we analyzed.  Research manipulations that used the interview procedures of many of the tapes we have studied would be unethical.  Therefore, the applicability and generalizability of the research studies is limited.  (The problem of ecological validity is acknowledged by several researchers in Ceci, Toglia & Ross, 1987.)  The findings of the research studies most likely greatly underestimate the effects of adult behaviors in actual interviews of children.

While these findings do not invalidate the children's statements, they raise serious questions about the possible role of adult influences upon children's behavior.  The interrogation process cannot be accepted as neutral, objective, or unbiased.  In each case, what has actually been done with a child by all of the people involved in talking to the child, including other children, must be carefully scrutinized as a possible source of error.  Our conclusion from our analyses is that it is not possible to interrogate children to get at the truth unless every effort is made to control contaminating influence.

Conclusions

The way children are interrogated when sexual abuse is suspected shows a common pattern across the nation.  There are common techniques used to investigate the abuse and interview the children.  But the way children are interviewed has a high potential for introducing error and reducing the reliability of statements that children make.  Children, interrogated with suggestive approaches by adults who have no conception of their influence, are being taught.  This raises serious questions about the possible role of adult social influence upon children's behavior.  When there is no corroborating data or no admission from the alleged perpetrator, children's statements standing alone must be viewed with great caution.
  

REFERENCES

Aman, C. & Goodman, G. S. (1987, September). Children's use of anatomically detailed dolls: An experimental study. Symposium on Interviewing Children, Washington, D.C.

Aronson, E. (1988). The Social Animal (). New York: Freeman and Company.

August, R. & Forman, B. (1986, May 16). Differences between sexually and non-sexually abused children in their behavioral responses to anatomically correct dolls. Paper presented at the Fourth National Conference on the Sexual Victimization of Children, New Orleans.

Boat, B. W. & Everson, M.D. (undated) The Anatomical Doll Project: An Overview. Department of Psychiatry, University of North Carolina, Chapel Hill, NC.

Buros, O.K. (1972). The Seventh Mental Measurements Yearbook (Vol. 1) (). Highland Park, NJ: Gryphon Press.

Ceci, S. J., Toglia. M. P. & Ross, D. F. (1987) (Eds.) Children's Eyewitness Memory (). New York: Springer-Verlag.

Ceci, S. J., Ross, D. F., & Toglia. M. P. (1987). Age differences in suggestibility: Narrowing the uncertainties. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp. 36-52). New York: Springer-Verlag.

Cole, C. B. & Loftus, E. F. (1987). The memory of children. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp.178-208). New York: Springer-Verlag.

Dent, H. R., & Stephenson, G. M. (1979). An experimental study of the effectiveness of different techniques of questioning child witnesses. British Journal of Social and Clinical Psychology, 18, 41-51.

Dent, H. R. (1982). The effects of interviewing strategies on the results of interviews with child witnesses. In A. T. Trankell (Ed.) (pp.279-298). Reconstructing the Past (). Deventer, The Netherlands: Kluwer.

Dunn, D. R. F. (1981) Dissemination of the published results of an important clinical trial: An analysis of the citing literature. Bulletin of the Medical Library Association, 69, 301-306

Eagly, A. H. (1983). Gender and social influence: A social and psychological analysis. American Psychologist, 38, 971-981.

Enright, S. (1987). New Law Journal, 137, 633-634.

Festinger, L. Riecken, H. W. & Schachter, S. (1958) When prophecy fails. In Maccoby, Newcomb & Hartley (Eds.) Readings in Social Psychology. New York: Holt, Rinehart & Winston.

Gabriel, R. M. (1985). Anatomically correct dolls in the diagnosis of sexual abuse of children. The Journal of the Melanie Klein Society, 3(2), 41-51.

Garbarino, J. & Scott, F. (1988, August 15). Adults as recipients of information from children. Paper presented at the 96th Annual Convention of the American Psychological Association in Atlanta, Georgia.

Goodman, G. S (1984). Children's testimony in historical perspective. Journal of Social Issues, 40(2), 9-31.

Goodman, G. S., Aman, C., & Hirschman, J. (1987). Child sexual and physical abuse: Children's testimony. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp. 1-23). New York: Springer-Verlag.

Goodman, G. S., Hepps, D., & Reed, R. S. (1986). The child victim's testimony. In A. Haralambie (Ed.). New Issues for Child Advocates (pp.167-177). Phoenix, AZ: Arizona Association of Council for Children.

Goodman, G. S., & Reed, R. S. (1986). Age differences in eyewitness testimony. Law and Human Behavior, 10, 317-332.

Harris, D. B. (1963). Children' s Drawings as Measures of Intellectual Maturity: A Revision and Extension of the Goodenough Draw a Man Test. New York: Harcourt, Brace and World.

Hayes, M. (1987). Child sexual abuse and the civil courts. Civil Justice Quarterly, 9-16.

Herbert, C. P., Grams, G. D., & Goranson, S. E. (1987). The Use of Anatomically Detailed Dolls in an Investigative Interview: A Preliminary Study of "Non-Abused" Children. Department of Family Practice, University of British Columbia, Vancouver, B.C.

Jampole, L., & Weber, M. K. (1987). An assessment of the behavior of sexually abused and non-sexually abused children with anatomically correct dolls. Child Abuse & Neglect, 11, 187-194.

Johnson, M. K., & Foley, M. A. (1984). Differentiating fact from fantasy: The reliability of children's memory. Journal of Social Issues, 40(2), 33-50.

King, MA., & Yuille, J. C. (1987). Suggestibility and the child witness. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp. 24-35). New York: Springer-Verlag.

Kobasigawa, A. (1974). Utilization of retrieval cues by children in recall. Child Development, 45, 127-134.

Landers, S. (1988) Use of "detailed dolls" questioned. APA Monitor, 19, 24-25.

Lindsay, D. S., & Johnson, M. K. (1987). Reality monitoring and suggestibility: Children's ability to discriminate among memories from different sources. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp.92-121). New York: Springer-Verlag.

Lipton, J. P. (1977). On the psychology of eyewitness testimony. Journal of Applied Psychology, 62, 90-95.

Lipton, J. P. & Hershaft, A. M. (1985) On the widespread acceptance of dubious medical findings. Journal of Health and Social Behavior, 26, 336-351

Loftus, E. F. (1979). Eyewitness Testimony (). Cambridge: Harvard University.

Loftus, E. F. & Davies, G. M. (1984). Distortions in the memory of children. Journal of Social Issues, 40, 51-67.

Loftus, E. F., & Loftus, G. R. (1980). On the permanence of stored information in the human brain. American Psychologist, 35, 409-420.

Lord, C. Ross, L., & Lepper, M. (1979) Biased assimilation and attitude polarization: The effect of prior theory on subsequently considered evidence. Journal of Personality and Social Psychology, 37, 2098-2109

Mahoney, M. (1976) Scientist as Subject (). Cambridge MA: Ballinger.

Mandler, J. M., & Johnson, N. S. (1977). Remembrance of things parsed: Story structure and recall. Cognitive Psychology, 9, 111-151.

Mclver, W., Wakefield, H., & Underwager, R. (1989). Behavior of Abused and Non-Abused Children in Interviews with Anatomically-Correct Dolls. Issues in Child Abuse Accusations, 1(1), 39-48.

Ornstein, P. & Gordon, B. (1988, August 15). Children's memory for visits to the doctor. Paper presented at the 96th Annual Convention of the American Psychological Association in Atlanta, Georgia.

Perlmutter, M., & Ricks, M. (1979). Recall in preschool children. Journal of Experimental Child Psychology, 27, 423-436.

Peters, D. P. (1988, August 15). Arousal on eyewitness evidence in children. Paper presented at the 96th Annual Convention of the American Psychological Association in Atlanta, Georgia.

Peters, D. P. (1987). The impact of naturally occurring stress on children's memory. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp. 122-141). New York: Springer-Verlag.

Price, D. J. (1965). Networks of scientific papers. Science, 149, 510-515.

Price, D. J. & Beaver, D. (1966) Collaboration in an invisible college. American Psychologist, 21, 1011-1018.

Raskin, D. C. & Yuille, J. C. (in press). Problems in evaluating interviews of children in sexual abuse cases. In S J. Ceci, D. F. Ross & M. P. Toglia (Eds.), New Perceptions on Child Witness Research. New York: Springer-Verlag.

Roback, H. B. (1968). Human figure drawings: Their utility in the clinical psychologists' armamentarium for personality assessment. Psychological Bulletin, 70, 1-19.

Ross, L. & Lopper, M. (1980) The perseverance of beliefs: Empirical and normative considerations. In R. Schweder, (Ed), Fallible Judgment in Behavioral Research; New Directions for Methodology of Social and Behavioral Science (). San Francisco; Jossey-Bass.

Ross, L., Lopper, M. & Hubbard, M. (1976). Perseverance in self-perception and social perception; Biased attributional processes in the debriefing paradigm. Journal of Personality and Social Psychology, 32, 880-892.

Saywitz, K. J. (1987). Children's testimony: Age-related patterns of memory errors. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp. 36-52). New York: Springer-Verlag.

Sivan, A. B., Schor, D. P., Koeppl, G. L., & Noble, L. D. (1988). Interaction of normal children with anatomical dolls. Child Abuse & Neglect, 12, 295-304.

Turtle, J. W. & Wells, G. L. (1987). Setting the stage for psychological research on the child eyewitness. In S. J. Ceci, M. P. Toglia, & D. F. Ross (Eds.) Children's Eyewitness Memory (), (pp. 230-248). New York: Springer-Verlag.

Wakefield, H. & Underwager, R. (1988). Accusations of Child Sexual Abuse ()(). Springfield, IL: C. C. Thomas.

Watson, M. W., & Fischer, K. W. (1980). Development of social roles in elicited and spontaneous behavior during the preschool years. Developmental Psychology, 16, 483-494.

White, S., Strom, G. S., Santilli, G., & Halpin, B. M. (1986). Interviewing young sexual abuse victims with anatomically correct dolls. Child Abuse & Neglect, 10, 519-529.

Yokote, G. & Utterbach, R. A. (1974). Time lapses in information dissemination: Research laboratory or physicians office. Bulletin of the Medical Library Association, 62, 251-257.

Yuille, J. C. (undated). The Systematic Assessment of Children's Testimony. Unpublished paper.