SHOULD CHILDREN BE ALLOWED TO TESTIFY IN COURT?
Over the past ten years, more research has been done involving children's testimony
than that of all the prior decades combined. Ceci & Bruck (93) have cited four reasons for this :
• The opinion of psychology experts is increasingly being accepted by courts as testimony,
• Social research is more commonly being applied to the issues of children's rights,
• More research into adult suggestibility in accordance with reason naturally leads to more
research into child suggestibility,
• Children are more commonly being used as witnesses in cases where they are directly involved
(i.e. sexual abuses cases), requiring the development of better ways for dealing with them as
Some psychologists deem children to be "Highly resistant to suggestion, as unlikely to
lie, and as reliable as adult witnesses about acts perpetrated on their bodies" (Ceci & Bruck
1993). However, children are also described as "Having difficulty distinguishing reality from
fantasy, as being susceptible to coaching by powerful authority figures, and therefore as being
potentially less reliable than adults" (Ceci & Bruck 1993). The suggestibility of child witnesses,
the effects of participation on children's reports, and the effects of postevent information on a
prior memory representation must be taken into account when it comes to seeking answers to
the reliability of their testimony, especially because sexual abuse and sexual assault cases are a
big part of children's testimony and they are often the only witness.
Those psychologists who feel that children can be rated as "Highly resistant to
suggestion...." etc. seem to have a good argument, whereas those who take the opposite view
also seem to have just as valid an argument. Which psychologists are right? Maybe both. It
seems that without outside influences, social encounters, or other interference's, children's
testimony has the potential to be quite valid. This is under ideal situations, however, which
unfortunately rarely occur.
One of the major problems when assessing the validity of child witnesses is the
suggestibility of the child. Ceci & Bruck (1993) define suggestibility as "The degree to which
children's encoding, storage, retrieval, and reporting of events can be influenced by a range of
social and psychological factors." A child's perception of events may be manipulated by many
factors with misleading questions being the most common way to assess a subjects suggestibility
(Smith & Ellsworth, 87). A misleading question according to Smith et al, is one that "provides
information that is inconsistent with the event witnessed, suggesting, for example, the existence
of an object that was in fact not present." After being asked leading questions, a subject is much
more likely than a person not asked leading questions to report the presented false information
This statement was validated by Kaufman and Richter's 1990 study. In this study a
number of young children (4 - 7 year olds) saw a short film featuring a circus performance. A
few weeks after watching the film, the children were split into two groups. They were then asked
(individually) a number of questions relating to the film. The first group were asked leading
questions i.e. "What colour was the clowns hat" (where in fact the clown had not been wearing a
hat), while the other group was simply asked "Was the clown wearing a hat". Kaufman and
Richter found that regardless of age, children often answered the leading questions and
accepted the fabricated information as being the truth.
This study clearly shows that children can be manipulated by clever questioning about a
witnessed event. However, this study did not involve the child interacting with the event i.e. the
child did nit participate on any emotional level by simply watching a video.
Rudy & Goodman (1991) conducted an experiment involving the effects of participation
on children's testimony. The main purpose of Rudy & Goodman's work being to see whether the
factors of age or participation influenced the recall of a child. Their experiment involved thirty-six
children (eighteen 7-year olds and eighteen 4-year olds) going in pairs into a parked trailer with a
male stranger (a confederate). One child played games with the confederate while the other child
watched closely. Positive verbal and physical interaction took place between the confederate and
the participant, and positive verbal interaction took place between the confederate and the
bystander. The events were videotaped and lasted about ten minutes. The children then returned
individually ten to twelve days later for a memory test and were asked to recall everything he or
she could remember about the day in the trailer. Various questions were asked, which included
specific and misleading questions. Rudy and Goodman concluded that although the children's
participation level in an event did not have a pervasive effect on their memory, it did serve to
increase the child's resistance to suggestion. Thus, children are more likely to resist suggestion if
they are somehow involved with the event.
Robins et al (92) criticised this investigation on a number of levels and concluded that it
could not be applied to support the use of child witnesses in a courtroom situation with any
validity. Firstly, Robins noted that the majority of cases in which children were used as witnesses
(i.e. sexual abuse and sexual assault cases), the events that the witnesses are asked to recall
are often far from positive and indeed, where a child has been allegedly assaulted/abused, the
witness may experience a number of emotional factors that alter the perception of the event i.e.
guilt, fear, hatred, confusion etc, none of which were present in the Rudy and Goodman study.
Although, of course, the introduction of these factors would have been a serious breach of ethics.
Another criticism is that the time frame involved may not have been sufficiently long enough to
emulate the period from event to testimony in a child witness case. Finally, the situation under
which they were asked to recall the event was not nearly the same as being asked to recall
details in a courtroom (Vickers & Fuller, 92).
There are many other types of suggestibility that can affect the reliability of a child's
recall of events.
Experiments that involve the effects of postevent information on a prior memory
representation were performed by Rovee-Collier et al (1993) involving three-month-old babies.
Rovee-Collier et al stated that ".....in eyewitness testimony research, postevent information
impairs retention of the original event and increases the probability that interpolated [new]
information will be identified as part of the original event." The infants used in the experiments
were taught to kick to cause a crib mobile to move. They were then exposed to information on a
novel mobile for a short amount of time. The information received by the babies after the novel
event impaired their recognition of the original mobile when it immediately followed their training.
Infants treated postevent data as part of the original training procedure. However, the postevent
information did not impair their recognition of the event if it was delayed by one day. Rovee-
Collier et al (93) proposed that "......postevent information displays conflicting information
coactive with it in primary memory and creates a new, updated memory token of the event"
This seems to suggest that children are more susceptible to alter their perceptions of an
event only if postevent information is supplied directly after the event. However, Summ &
Girston (94) suggested that "........with the brains inability to perceive every single piece of
information of an event, new information will be immediately supplied through the act of
'assumption'" Summ & Girston defined 'assumption' as being the process by which information
which has not been directly perceived (i.e. attenuated information (Triesmann, 64)), is
reconstructed using prior knowledge. This theory, therefore, leads us to believe that there is a
source of unconscious postevent information after every even, even though assumption may
only cause a very small amount of new information.
The fact that Rovee-Collier et al received no indication of postevent information clouding
the recall of the infants they used (if the information was presented a little time after the event)
can be explained by their own justification of using infants in the first place. They suggested that
"..... infants are choice subjects for investigation and study on memory involvement because the
babies are not exposed to problems linked with social or task demands and other circumstances.
It has often been observed that these problems interrupt customary research on eyewitness
testimony. Also, an infants' memory is not crowded with numerous other prior associations". This
infers that the infants could not unconsciously supply their own postevent information as they do
not have enough knowledge of their situation to make assumptions about new event.
Furthermore, children who give testimony in court are most likely to be older that the
ones that Rovee-Collier et al used in their study, and also it is likely that they would have to
recall far more detailed information than the colour or arrangement of a mobile.
Supporters of using children to give testimony in court have claimed that the use of
anatomically correct dolls when interviewing children involved in sexual abuse cases allow the
child to express themselves easier for a number of reasons. These are:
• The dolls simulate a critical event which could spark recall and help to overcome the language
• They help to overcome any shyness or embarrassment that the child might feel,
• They can be used in a projective manner, where a preoccupation with the genitalia may
indicate sexual abuse.
Ceci & Bruck (93) refute these arguments with the following reasons: Firstly, the dolls
themselves are suggestive. They might actually encourage a child to play in a manner appearing
to indicate abuse because they provide a freedom from inhibition that might not be normally
present or could indicate an exposure to sexuality, rather than actual abuse. Second, there is
almost no normative data to support the use of dolls. Until fairly recently, no significant research
had been done which compared patterns of play with the dolls between abused and non-abused
A further issue when deciding whether children should be allowed to testify in a
courtroom is that of factors not directly related to the child. For example how do jurors react
when a child is asked to testify. i.e. Do jurors believe children? This question has been
investigated by researchers in complex detail. Luus, Wells & Turtle (1995) conducted studies on
jurors who were misinformed of the interviewee's age, then asked to judge the information
garnered under cross examination. Surprisingly, it was discovered that transcripts of interviews
with 8-year olds were judged to be as credible as those of adults. Similarly, when the juror was
told the incorrect age, it did not affect the judgment. The jurors assessed the credibility of the
testimony based on its content and not on the age of the interviewee. Interestingly, Luus, Wells
and Turtle point out that this is actually a problem for jurors. Although they judged the children to
be more accurate, they, in fact, were not. yet, the researchers are quick to point out that the
research conditions and a real courtroom setting are very different and this should be accounted
for. Overall, Luus, Wells and Turtle found that jurors may enter the courtroom with a negative
bias against the child witness, but the actual testimony is judged more on the quality than on the
Some psychologist's, however, have refused to believe that children can give an
accurate testimony. By understanding what cognitive issues can affect memory, researchers
have been able to develop an interviewing technique which specifically seeks to maximise the
completeness and accuracy of information reported by and interviewee. This type of interview is
known as a 'cognitive interview' (Saywitz, Geiselman & Bornstein, 1992) showed that cognitive
interviewing with children significantly increased the number if correct facts recalled.
Additionally, they discovered that practice cognitive interviews using an improvised event could
improve recall. These findings have recently been supported by McCauly & Fisher (1995). In
their research, the cognitive interview increased the amount of information accurately recalled by
children. They also confirmed that new, accurate information was recalled ion subsequent
interviews. Uniquely, they found that using the cognitive interviewing technique as a follow-up to
a standard interview increased the amount of accurate information gained over using the
standard interview a second time. This is especially applicable to field work where the first
interview is often conducted hurriedly, without using cognitive procedures (McCauly & Fisher,
In conclusion I have found that the underlying aspect of child testimony is suggestibility.
Researchers today are concentrating most of their efforts in this area. It is easy to suggest that
no children should be allowed to testify on account of the malleability of their recollection.
However, children can play a vital role in the legal system, and indeed there are many cases in
which a child is the only witness to a crime, but until the time that sufficient research has been
done to achieve a system of questioning that will eliminate the suggestibility and social aspects
of a child's testimony, all such testimonies should be treated with caution.
Bernstein, D. A., Roy, E. J., Srull, T. K., Wickens, C. D. (1994) Psychology, 3rd edition.
Houghton Mifflin Company, MA.
Ceci, S & Bruck, M. (1993). Suggestibility of the Child Witness: A Historical Review and
Synthesis, Psychological Bulletin. 113, 403 - 439
Lefrancois, G. R. (1992). Psychology, 2nd edition. Wadsworth Publishing Company.
Luus, C. A. E., Wells, G. L., & Turtle, J. W. (1995). Child eyewitnesses: Seeing is
believing. Journal of Applied Psychology, 80, 317 - 326
Rovee-Collier, C. et al. (1993). Infants Eyewitness Testimony: Effects of Postevent
Information on a Prior Memory Representaion, Memory and Cognition, 21, 267 - 279