Handbook of Psychology of Investigative Interviewing: Current Developments and Future Directions
Page 32
during cross - examination with that used during direct examination (Goodman
et al. , 1992 ; Davies & Seymour, 1998 ; Zajac et al. , 2003 ). Problematic gram-
matical constructions that are common to cross - examination include multifac-
eted questions (e.g., ‘ So he picked you up and then the two of you went to
the movies and then he dropped you at the bus stop – is that correct? ’ ), ques-
tions phrased in the negative (e.g., ‘ Did you not leave the house at 3 pm? ’ ),
and questions that are inappropriately grammatically or semantically linked
(e.g., ‘ But you realize, I suggest, that he was there, didn ’ t you? ’ Davies &
Seymour, 1998 ). These types of cross - examination questions are diffi cult for
children to repeat verbatim, let alone fully comprehend (Brennan & Brennan,
1988 ). Second, many cross - examination questions require children to under-
stand and employ complex relational concepts (e.g., height, weight, age, time,
date and distance; Brennan
& Brennan,
1988 ; Walker,
1993 ), or recount
highly peripheral aspects of the alleged event (Glaser
& Spencer,
1990 ).
Finally, during cross
- examination, the topic of questioning often changes
abruptly. This is likely to confuse children (Cashmore, 1991 ; Saywitz, 1995 ),
who expect conversations to fl ow in a logical sequence (Grice, 1975 ).
Empirical research indicates that the mismatch between cross - examination
questions and children ’ s abilities may prevent children from responding to
questions in a truthful and meaningful way. For example, children often believe
that they have understood a question when in fact they haven ’ t (Markman,
1977; 1979 ; Singer & Flavell, 1981 ; Carter et al. , 1996 ). Furthermore, even
when children are aware that they have not understood a question, they rarely
ask interviewers for clarifi cation (Markman, 1977; 1979 ; Saywitz & Snyder,
1993 ; Kebbell & Johnson, 2000 ). Children will even attempt to answer ques-
tions that do not make sense (e.g., ‘ Where do circles live? ’ Hughes & Grieve,
Investigative Interviewing in the Courtroom
165
1980 ; Pratt, 1990 ; Waterman et al. , 2000 ), especially when questions require
only a yes or no answer (e.g., ‘ Is a stone slower than an ear? ’ Waterman, Blades
& Spencer, 2001 ). As such, children ’ s willingness to provide an answer should
not be taken as evidence that they have understood and correctly interpreted
the question (Walker, 1993 ). Not surprisingly, children ’ s accuracy is signifi -
cantly compromised when responding to complex questions (Carter et al. ,
1996 ).
Why do cross - examining lawyers ask children questions that exceed their
linguistic and cognitive capabilities? There are two possible explanations, and
they are not mutually exclusive. The fi rst is that lawyers use complex language
unwittingly (Henderson, 2002 ). It is possible, for example, that some poorly
constructed questions or sudden changes in topic merely refl ect the relatively
spontaneous nature of cross - examination. Complex questions could also be
accounted for by limited knowledge of what constitutes a developmentally
appropriate question. This point pertains especially to the criminal defence
lawyer who, in the context of representing adult defendants, may only occa-
sionally be required to deal with a child witness.
On the other hand, it is possible that the use of complex questions may, in
some situations, be intentional (Glaser & Spencer, 1990 ; Davies et al. , 1997 ).
Davies & Seymour, for example, raise the possibility that ‘ some lawyers are
using diffi cult questions in a deliberate attempt to confuse young complain-
ants ’ ( 1998 : 7). Support for this argument comes from the observation that
legal textbooks encourage some practices (e.g., abrupt changes in the topic of
questioning during cross - examination; Stone, 1988 ; Eichelbaum, 1989 ), but
discourage them when instructing lawyers on how to question their own
witnesses.
Problem 3: Challenges to c redibility
Casting doubt on the witness ’ s testimony is one of the defi ning characteristics
of cross
- examination. In doing so, cross
- examining lawyers frequently put
forward an alternative version of events, the plausibility of which the witness
is pressured to acknowledge (Davies et al. , 1997 ). Unfortunately, all of what
we know about interviewing children advises that interviewers utilize a neutral
manner so as to avoid contaminating children ’ s reports with their own beliefs
or knowledge. Biased interviewing has been shown to exert a considerable
negative infl uence on children
’ s reports (Pettit
, Fegan
& Howie,
1990 ;
Thompson, Clarke - Stewart & Lepore, 1997 ).
On what grounds do lawyers attempt to discredit children
’ s testimony
during cross - examination? Davies et al. (1997) interviewed 14 New Zealand
defence lawyers and studied court transcripts of child cross
- examinations,
uncovering many common techniques used by lawyers to discredit children ’ s
evidence in alleged cases of sexual abuse. In 73% of the transcripts studied,
children were accused, either directly or indirectly, of lying about the alleged
abuse. Lawyers ’ justifi cations for these accusations included inconsistencies or
166
Handbook of Psychology of Investigative Interviewing
delays in abuse disclosure, the child ’ s previously happy relationship with the
accused, ulterior motives, an inability to distinguish between real and imagined
actions, and pressure from an adult to fabricate an allegation. Westcott & Page
(2002) note that lawyers who cross
- examine child complainants of sexual
abuse often rely on commonly held, but empirically unsupported, notions
about sexual assault (e.g., that genuine victims of abuse will make a disclosure
immediately).
Not surprisingly, children who have testifi ed in adversarial trials consistently
report fi nding the confrontational nature of cross - examination highly distress-
ing (Prior, Glaser & Lynch, 1997 ; Eastwood & Patton, 2002 ; Wade, 2002 ).
Cross - examination and, in particular, the attitude and behaviour of the defence
lawyer were the ‘ overwhelming area of concern ’ for child complainants in
Eastwood & Patton ’ s Australian study ( 2002 : 59). Flin and colleagues (1992)
noted many anecdotal reports of children being so distressed during cross -
examination that they ‘ clam up ’ or refuse to speak. A child witness from Wade ’ s
(2002) study reported:
They were pushing me to say the opposite thing. … And they shouldn ’ t really
do that. They should, like, ask you to say what happened and then ask you ques-
tions. But they didn ’ t. They kept … pushing me to say summat else. And as
soon as I got out I started crying ’ cos they pushed me and they scared me.
(Wade, 2002 , p. 225)
It is important to note, however, that not all child cross - examinations are
conducted in an aggressiv
e manner. Ten of the 14 New Zealand defence
lawyers surveyed by Davies et al. (1997) considered that the best manner in
which to cross - examine a child complainant was to be non - threatening and
gentle, at least initially. These authors suggest that a friendly, charming ques-
tioning manner can render a child more trusting, while preventing the lawyer
from alienating the jury. Some have even proposed that children ’ s evidence
can be discredited more readily by interviewing in a supportive manner than
it can be using aggression (Flin, 1993 ).
How do t hese p roblems i mpact on c hildren ’ s t estimony?
Although many laboratory studies have considered the three problems out-
lined above in isolation, research evaluating their combined effect on children ’ s
courtroom evidence is limited. As the only empirical study that has examined
how child witnesses respond to the questions that they are posed during cross -
examination, our court transcript analysis suggests that children do not cope
well with this questioning style. Specifi cally, despite the fact that defence
lawyers asked a high proportion of complex and grammatically unsound ques-
tions, child complainants of sexual abuse rarely requested clarifi cation. Children
in this study were also highly likely to comply with leading questions. Most
notably, under cross - examination, over 75% of children changed at least one
Investigative Interviewing in the Courtroom
167
aspect of their earlier testimony. On average, children made three of these
changes, with one child making 16 changes during a relatively short question-
ing period. The changes that children made ranged from peripheral details of
the alleged event(s) to entire allegation retractions (Zajac et al. , 2003 ).
Although these fi ndings give cause for concern, using court transcript data
only allowed us to evaluate the effect of cross
- examination on children
’ s
testimony, as opposed to children ’ s accuracy. Given that cross - examination
is often promoted as a truth - fi nding mechanism, the issue of accuracy is a
critical one.
Cross - e xamination and c hildren ’ s a ccuracy
One widespread assumption within the legal profession is that cross - examina-
tion will not pose any problem for a witness who is telling the truth. If this
were the case, then the only children to change their stories during cross -
examination would be children whose primary evidence contained errors. In
this scenario, the changes that the children in our transcript study made during
cross - examination would have increased their overall accuracy levels. However,
given that cross - examination language exhibits several characteristics likely to
dramatically reduce resistance to suggestion, there is another possibility: the
changes that children make during cross - examination may have little to do
with accuracy, and may have even rendered children ’ s testimony less accurate.
Laboratory research is the only avenue by which to comprehensively evaluate
these possibilities.
The fi rst laboratory - based study of cross
- examination was conducted in
1988 by Turtle & Wells. In this study, eight - and 12 - year - old children and
adults watched a fi lm clip of a simulated child abduction. The next day, they
were asked 10 direct examination and 10 cross - examination questions about
the fi lm. Although participants in all age groups were less accurate in response
to the cross - examination questions than the direct examination style questions,
eight - year - old children ’ s overall accuracy rate (41%) was signifi cantly lower
than that of the adults and the 12 - year - olds (61% and 65%, respectively).
While Turtle
& Wells
’ fi ndings confi rmed many researchers
’ suspicions
about the effect of cross - examination on children ’ s accuracy, two main prob-
lems restrict their forensic applicability. First, participants in this study were
cross - examined just 24 hours after viewing the fi lm; most child witnesses in
real trials are cross - examined long after their allegation is made (Goodman et
al. , 1992 ; Lash, 1995 ; Plotnikoff & Woolfson, 1995 ; Eastwood & Patton,
2002 ). Second, because little detail was provided regarding the types of ques-
tions that were asked, it is impossible to assess whether the cross - examination
questions were an accurate refl ection of courtroom questioning.
In order to address these issues, we used court transcripts to develop a
standardized laboratory analogue of cross
- examination, which we used to
interview fi ve - and six - year - olds about a staged event (Zajac & Hayne, 2003 ).
168
Handbook of Psychology of Investigative Interviewing
Children visited the local police station, where they participated in four unique
activities (e.g., getting their fi ngerprints taken). Six weeks later, they were
interviewed about the visit. This interview mimicked the process adopted in
several countries (e.g., New Zealand, Australia and England) which allows
direct evidence to be pre - recorded for presentation in court. The interview
consisted of general, open - ended prompts (e.g., ‘ Tell me everything that you
can remember ’ ), as well as specifi c questions about two true activities (e.g.,
‘ Did you see the police car? ’ ) and two false activities (e.g., ‘ Did you get to try
on handcuffs?
’ ). After a forensically relevant eight
- month delay, children
viewed their direct examination videotape and were then interviewed using the
cross - examination analogue. The aim of this cross - examination interview was
to talk children out of their responses to the specifi c questions posed during
direct examination, regardless of response accuracy. The questions used to
achieve this were modelled after the kinds of questions asked by defence
lawyers in Zajac et al. (2003) .
Several important fi ndings emerged. As in the courtroom, a large proportion
of children (85%) changed at least one of their direct examination responses
under cross - examination. In fact, one third of children changed all four of
their previous responses. Crucially, the changes that children made during
cross - examination were by no means limited to correcting earlier mistakes. In
fact, children in this study were just as likely to change a correct response as
they were to change an incorrect one. Overall, cross - examination style ques-
tioning signifi cantly decreased the accuracy of children ’ s reports, to a point
where accuracy scores during the cross - examination interview did not differ
signifi cantly from chance (50%). This fi nding held even when only considering
the substantial number of children whose direct examination reports were
100% accurate.
When the study was repeated with a sample of nine - and ten - year - olds, some
age differences in cross
- examination performance were evident. The older
children made fewer changes during cross - examination than the younger chil-
dren, and were more likely to change an incorrect response than a correct one.<
br />
Nonetheless, older children still changed 43% of their correct responses,
leading to a considerable decrease in accuracy (Zajac & Hayne, 2006 ).
The fi ndings from these three laboratory studies raise serious concerns about
the suitability of the cross - examination process for children. Specifi cally, they
suggest that not only might cross - examination be an ineffective method for
ascertaining the truth, but in some cases it might even create the types of errors
that it aims to uncover. The potential for trial outcome to be impaired in these
cases is immense. Distressed and confused child victims of abuse who are
subjected to rigorous cross - examination could end up retracting their allega-
tions altogether. Alternatively, these children could be coerced into admitting
that they have mistaken the identity of their abuser. Child witnesses for the
defence could end up (falsely) implicating an accused person. Even seemingly
inconsequential inconsistencies in a child ’ s testimony may be seen by jury
members as a sign of unreliability.
Investigative Interviewing in the Courtroom
169
Our attention has now turned to examining the factors that infl uence chil-
dren ’ s responses to cross - examination questioning. Some of these are factors
over which the justice system has little control, but which may help to identify
children who may be particularly vulnerable to cross - examination; others are
procedural factors that may help to shed some light on a way forward.
Individual d ifferences in c ross - e xamination p erformance
As mentioned at the beginning of this chapter, it is now widely acknowledged
that children ’ s eyewitness accuracy hinges primarily on the way in which they
are questioned. However, even when external factors are held constant, chil-
dren do not respond uniformly to forensic questioning (Bruck, Ceci & Melnyk,
1997 ). In light of this fi nding, a recent body of research has focused on factors
within the individual that may infl uence the reliability and accuracy of chil-
dren ’ s reports.
Despite an increasing number of studies examining the individual differ-
ences that infl uence children ’ s primary testimony (see Bruck & Melnyk, 2004 ,
for a review and synthesis), little is known about how individual factors might
affect children ’ s responses to cross - examination. Preliminary research has been