Perceptions of Testimony Reinforcement and Its' Effects on Juror's Decisions
By: Sheldon Bruno
Introduction
It has always been the goal
of the legal system to make sure that individuals who commit crimes, especially
crimes of a sexually depraved nature and committed against children, be made
to pay for their acts. To do this, the court often calls on the victims of these
crimes to offer their own accounts of the events surrounding their assaults
to persuade juries to find their attacker(s) guilty.
In recent years there has been much controversy with eyewitness testimony, namely
that testimony given from child victims is subject to inconsistencies, either
due to lapses of memory on the child's part or due to information lost during
others' interpretations of the child's testimony (Warren, Nunez, Keeney, Buck
& Smith, 2002). Another issue of concern, identified in the McMartin child
sexual abuse trial, dealt with the pernicious effects that coerced testimonies
have when an overzealous child psychiatrist coaxed and guided the responses
of 20 or more children who were allegedly molested by the owners and proprietors
of the McMartin daycare center. By the end of the trial, it was clear that the
testimonies were merely the words and sentiments of a self-righteous physician
that the children were simply reciting as opposed to sharing their actual experiences.
And due to such instances, research on child testimonies has been performed
in order to determine whether or not some form of coercion was used on a child
in order to elicit a more favorable eyewitness accounts (Garven, Wood &
Malpass, 2000).
One study was conducted to determine which form of testimony juries would find
more believable, either the actual videotaped testimony from a child, a verbatim
transcript of the child's interview (containing both exact questions and answers),
or a summary or recapitulation of the child's interview (Warren, Nunez, Keeney,
Buck & Smith, 2002).
163 female and 71 male college students participated in this experiment. Prior
to the actual experiment, thirty 4 to 6 year old children were interviewed a
week after their classes met with a "stranger" named Sam Stone, who
played 2 games with the children's classes. These interviews involved the children
giving recounts of what games Sam Stone played with them as a basis for the
testimony to be used in the experiment. The testimony provided by a 4-year-old
named Carol was chosen for use in the experiment. Two modalities of testimony
were used; both transcripts (written records of interview) and videotapes (video
records of actual child interview), to examine how each mode of testimony affected
juror interpretation of the interview content. The independent variable of testimony
version was manipulated in three ways. One group was given a videotape of the
child's actual testimony. Another group of participants was given a written,
verbatim version of the child's interview in which both exact questions and
answers were supplied. The final group was given a written, summarized version
of the child's interview that merely gave a shorthand account of the child's
recount. After each group either viewed or read their specific type of testimony,
they were given a 26-item questionnaire that asked whether Sam Stone was "guilty"
or "not guilty" of playing certain games with the child. The questionnaire
also asked the level of confidence the participants had in deciding Sam Stone's
guilt or innocence.
No significant difference was determined for the effects that each type of testimony
had on the level of accuracy and believability. As far as the participants'
judgments of the level of credibility amongst the various types of testimony,
both the verbatim and summary forms of adult witness testimony were regarded
as being more credible than the child's videotaped testimony, but the summary
testimony was rated as being more credible than verbatim testimony. The results
also showed that there was no significant difference in levels of believability,
truthfulness, or accuracy of memory between any of the testimony types. There
was also no significant difference rated by participants on the levels of believability,
truthfulness, or accuracy of memory seen between the two modalities of testimony.
The only dependent variable that the participants scored a significant difference
in was degree of credibility, where the adult summarized testimony was scored
as more credible than either the verbatim or child testimonies, while both the
verbatim and child testimonies were shown to be statistically equivalent.
Although the researchers used well controlled independent variables, similar
to those that would be presented to an actual jury, they failed to take into
consideration the difference in a person's mindset if they were considering
the testimony of an actual sexual abuse trial as opposed to who played what
game. Another issue is the varying physical appearance of children who may have
been subject to some sort of disfiguring sexual assault and how viewing such
a witness could greatly influence testimony perception, whereas just reading
a transcript of the testimony would fail to offer the same exposure. There is
also the issue of the possible difficulties future researchers would face if
trying to replicate this study, due to the unclear manner in which the event
summaries were manufactured.
Whilst this previous study addressed issues concerning jurors' perceptions of
child testimony, there are still many questions left unanswered about the role
that the children in these cases play in giving their accounts. The goal of
this second study reviewed, was to determine how the techniques utilized in
the McMartin child sexual abuse trial can influence the responses children give
to inquiries about their past experiences. Specifically, the purpose of this
study was to determine if positive reinforcement, and/or cowitness information,
had any effects on the accuracy of the recounts given by kindergarten children.
Another focus of this experiment was to determine if either of these two techniques
or possibly both had any prolonged effects on the veracity or lack thereof of
the recounts made by the children at later interviews.
Thirty-nine 5-year olds, sixty-seven 6-year olds, and fourteen 7-year olds participated
in this study. There were fifty-one girls and sixty-nine boys in the study.
The children from five schools attended a story-time session at their own school
where they were introduced to the "stranger" Paco Perez, whose name
was repeated several times by the children's teachers. Paco proceeded to introduce
himself using a planned dialogue that described some of the odd clothing he
was wearing and the story he was planning on telling the children. He continued
by telling the children the story of the Hunchback of Notre Dame in an animated
fashion, and after completing the story, which took approximately twenty minutes,
Paco placed stickers on the hands of each of the children and provided them
with treats, and then left. The children were individually interviewed first
one week after their meeting with Paco Perez, and then once again anywhere from
fourteen to twenty-one days after the first interview. The independent variable
for this experiment was the type of interview the child received, either using
reinforcement (some verbal or non-verbal reward for giving the desired response),
using cowitness information (claims that another source offered corroborating
information, either true or false, prior to the actual interview), both of these
techniques, or a control method, in which neither of the aforementioned techniques
were employed. The dependent variable for the study was defined as the number
of "yes" responses a child gave to any of the eight misleading-mundane
(false, real-life situations), four leading (questions with obvious correct
answers), and four misleading-fantastic (questions involving extravagant falsehoods)
questions asked to them in the interview. The second set of interviews was performed
using the same type of interview questions.
For the misleading-mundane questions, the results showed that children answered
positively more often when reinforcement was used than when no reinforcement
was used. Children were also more likely to answer positively to misleading
questions when cowitness information was provided, rather than when no cowitness
information was used. For the misleading-fantastic questions asked to the children,
the results showed that the children were far more likely to answer "yes"
when reinforcement was present, and scarcely answered "yes" when no
reinforcement was used. Finally, the results for the leading-correct questions
showed that slightly more children answered, "yes" when reinforcement
was given than those who answered "yes" when no reinforcement was
given. While on the other hand, slightly fewer children answered "yes"
when cowitness information was used, as opposed to the number of children that
answered "yes" when no cowitness information was offered. Results
from the second interview showed that children who had received reinforcement
during the first interview were far more likely to respond "yes" regardless
of the use of reinforcement in the second interview.
Possible areas for improvement in this experiment are to consider the fact that
only kindergarten aged children were used in this experiment so there is still
much uncertainty as to whether or not children of other ages or even adults
for that matter would fall into similar trends. Also there is the issue of no
actual sexual abuse being involved in the study, which hinders the overall utility
of the findings, since other forms of mental duress may be involved with the
recount capability of a child who has been afflicted by a sexual assault.
When comparing the two aforementioned studies there are several differences
between the experiments in both the issues being manipulated and measured. The
first study deals with juror responses to various types of testimony, while
the second study measures how the veracity of a child's story can be altered
by the use of various forms of enticement and coercion. But what the two studies
do have in common, is that they both used kindergarten age children and determined
some of the possible ramifications that their responses to staged events could
have on the outcome of an actual trial, especially a sexual abuse trial, which
was the initial motivation for both lines of research. Although both experiments
performed a useful task by shedding some light onto the topics they set out
to research, there is still much lacking from these experimental designs. The
major quality missing from these two previous experiments, is the realism of
participants actually being tested on an actual sexual assault where the physiological
and psychological effects of witnessing or experiencing such a depraved and
mentally-jarring event could play significant roles on the level of testimony
believability, or the degree of accuracy an individual's memory would dedicate
to such an event. Furthermore, there is the question of the effects there would
be on a jury being presented with testimony from an abused child, whom they
believed to have been coached by either legal professional or psychologist,
and how this information could alter the level of veracity they attribute to
that child's testimony.
The present research proposal combines aspects from both of the previously mentioned
studies in that it will investigate how potential jurors weigh the believability
and credibility of three versions of a child's testimony (Warren et al., 2002)
based on information the jurors receive prior to judging the testimonies on
whether or not the child witness' statements were attained through reinforcement
techniques or not (Garven et al., 2000).