Handbook of Psychology of Investigative Interviewing: Current Developments and Future Directions

by Ray Bull, Tim Valentine, Dr Tom Williamson

  conducted by Zajac, Jury & O ’ Neill (in press) , who examined the role of

  several psychosocial variables on fi ve - and six - year - old children ’ s performance

  under cross - examination. Although cross - examination compromised accuracy

  in the vast majority of participants, children with low self - esteem, self - confi -

  dence and assertiveness performed particularly poorly. Given that child abuse

  has been associated with low scores on these variables (Martin & Beezley,

  1977 ; Oates, Forrest & Peacock, 1985 ; Kaufman & Cicchetti, 1989 ; Howing,

  Wodarski, Kurtz & Gaudin, 1990 ), our fi ndings raise the concerning possibil-

  ity that the same factors that may make children targets for abuse, or may be

  the consequences of abuse, could also make them particularly susceptible to

  the cross - examination process. Similar research examining the role of cognitive

  factors (e.g., IQ, memory and language ability) on children ’ s responses to

  cross - examination is underway, but preliminary fi ndings indicate that, within

  the normal range of functioning, these variables contribute little to cross

  -

  examination performance (O ’ Neill, Jury & Zajac, 2005 ).

  The i mpact of d elay

  Even in countries making concerted efforts to expedite trials involving child

  complainants, cross - examination typically occurs long after an allegation has

  been made. In New Zealand, for example, child witnesses can expect to wait

  an average of eight months between making an allegation and appearing in

  court (Lash, 1995 ). Child witnesses testifying in other countries also encounter

  170

  Handbook of Psychology of Investigative Interviewing

  lengthy delays (Goodman et al. , 1992 ; Plotnikoff & Woolfson, 1995 ; Eastwood

  & Patton, 2002 ), with some as long as three years (Eastwood & Patton, 2002 ).

  In an attempt to model the conditions in actual forensic settings, the cross -

  examination interview in our original analogue studies (Zajac & Hayne, 2003;

  2006 ) occurred eight months after direct evidence was pre - recorded. In light

  of research showing that suggestibility increases with delay (e.g., Zaragoza &

  Lane, 1994 ), we hypothesized that eliminating the delay between direct exami-

  nation and cross - examination might facilitate fi ve - and six - year - old children ’ s

  accuracy during cross - examination. To test this hypothesis, we employed the

  same basic paradigm, but cross - examined children either 1 – 3 days or eight

  months after their direct - examination interview. Despite highly accurate initial

  reports, children ’ s performance during cross - examination was very poor, even

  when they were cross - examined shortly after the target event. In fact, chil-

  dren ’ s cross - examination accuracy scores did not differ as a function of delay,

  suggesting that reducing the delay between the allegation and the trial or

  conducting pre

  - trial cross

  - examination may do little to facilitate children

  ’ s

  performance. Furthermore, when the direct examination interview questions

  were repeated one week following cross

  - examination, children

  ’ s accuracy

  returned to pre - cross - examination levels. These data suggest that children ’ s

  poor cross - examination performance cannot be solely attributed to memory

  impairment, and that the changes that children make during cross - examination

  do not necessarily result in memory impairment (Righarts, Zajac & Hayne,

  2009 ).

  Of course, delays during criminal investigations are not restricted to those

  occurring between allegation and trial. Child victims of sexual abuse, for

  example, may not disclose until many months or even years after the abuse

  has occurred. It is possible that, by conducting our direct examination inter-

  views very soon after the target event, we have effectively inoculated children

  against the impact of delay. Consequently, our fi ndings using short delays

  between each phase of the experimental paradigm might well be considered a

  best - case scenario of children ’ s cross - examination performance.

  Preparing c hildren for c ross - e xamination

  The fi nding that children ’ s responses to cross - examination cannot be attrib-

  uted to memory impairment suggests that pre - trial interventions for children

  may hold promise, to an extent that they would not if memory was the primary

  issue. In light of this possibility, our most recent research has been exploring

  potential interventions to facilitate children ’ s accuracy during cross - examina-

  tion questioning.

  Many jurisdictions have implemented formal preparation programmes for

  children who are required to testify in court. These programmes generally

  involve familiarizing children with their role as a witness and with courtroom

  personnel and procedures. Although these programmes are not specifi cally

  Investigative Interviewing in the Courtroom

  171

  designed to facilitate accuracy, it is anticipated that reducing distress and con-

  fusion associated with testifying will indirectly help children to provide

  complete and accurate testimony. While systematic research evaluating these

  programmes is scarce, existing evaluations suggest that the preparation sessions

  increase children

  ’ s understanding of the trial process, reduce their anxiety

  about giving evidence and help them to testify with more confi dence (Dezwirek -

  Sas, 1992 ; Davies et al. , 2004 ).

  Further to familiarization programmes, some researchers have explored

  interventions aimed specifi cally at facilitating children

  ’ s accuracy. Many of

  these interventions have been remarkably simple, such as telling children that

  interview questions might be diffi cult, or that saying ‘ I don ’ t know ’ is prefer-

  able to guessing an answer. Although these interventions have met with at

  least some success (e.g., Warren, Hulse - Trotter & Tubbs, 1991 ), their effi cacy

  does not necessarily generalize to cross - examination, during which leading and

  complex questions are often delivered in a highly persuasive manner.

  What does a warning about cross

  - examination need to encompass? As

  described earlier, several aspects of the language used during cross - examination

  are likely to promote compliance. First, reliance on leading questions (e.g.,

  ‘ Your mother was there at the time, wasn ’ t she? ’ ) can suggest to children that

  the interviewer has fi rsthand knowledge of the event, a problem likely to be

  exacerbated by the fact that adults, especially parents and teachers, often know

  the answers to the questions that they ask children. Anecdotally, when we ask

  our child participants leading cross - examination questions, they frequently ask

  us whether we accompanied them on the memory event. Second, because

  children tend to assume that adults are genuine conversational partners (Grice,

  1975 ), they are unlikely to anticipate that the questioning style used during

  cross - examination might be unsupportive. Specifi cally, children are unlikely to

  expect attacks on their credibility or understand that complex language struc-

  tures
can be used as a means to confuse them. Finally, children are often

  reluctant to disagree with adults, particularly adults in positions of authority.

  Children ’ s willingness to correct adults ’ incorrect statements is likely to decrease

  even further when interviewers adopt a confrontational approach.

  Could a brief verbal warning targeting these three areas increase children ’ s

  resistance to misleading cross - examination questions? If so, does it matter who

  delivers the warning? These questions formed the basis for a recent study that

  we conducted with fi ve - and six - year - old and nine - and ten - year - old children

  (Righarts & Zajac, 2009 ). The study followed the same general paradigm used

  in our past research, but some children were given a brief verbal warning

  immediately prior to their cross - examination interview. This warning informed

  children that they would be asked some more questions about the police

  station, that the interviewer did not visit the police station and therefore did

  not know what happened there, that the questions might be tricky and that

  it was OK to disagree with the interviewer if she got things ‘ muddled ’ . The

  warning was either delivered by an unfamiliar experimenter or by the experi-

  menter conducting the cross - examination interview. Children in the control

  172

  Handbook of Psychology of Investigative Interviewing

  group were not given a warning. As in previous studies, children were highly

  accurate during the direct examination interview, but accuracy scores during

  cross - examination were signifi cantly compromised. The warning did not facili-

  tate children ’ s cross - examination performance, regardless of who delivered it.

  It therefore appears that the brief interventions that have proved successful in

  child suggestibility research may be insuffi cient to buffer children from the

  negative effects of cross - examination (Righarts & Zajac, 2009 ).

  Could a more comprehensive intervention be successful? Following the

  unsuccessful warning intervention, we turned our attention to developing a

  means of preparing children for cross - examination questioning. One or two

  days prior to the cross - examination interview, half of the fi ve - to six - year - old

  and nine - to ten - year - old children in the sample were given practice at answer-

  ing cross

  - examination - style questions about a short fi lm (unrelated to the

  memory event), with feedback on their responses. The duration of the prepara-

  tion session was approximately 20 minutes. Control children watched the fi lm,

  but were not given the practice and feedback. During the subsequent cross -

  examination interview, the children who received preparation made fewer

  changes to their earlier responses and changed a smaller proportion of their

  correct responses relative to the control children. Furthermore, overall accu-

  racy levels during the cross - examination interview were signifi cantly higher in

  the preparation group than in the control group. In short, the intervention

  was successful (Righarts & Zajac, 2009 ).

  While this research is in its very early stages, our preparation intervention

  appears to have several advantages. First, during the preparation session,

  children were asked questions that were entirely unrelated to their

  ‘ testi-

  mony ’ , making allegations of coaching less feasible. Furthermore, the inter-

  vention was effective despite being delivered by an unfamiliar interviewer, as

  would be the case in real - life situations where court preparation would be

  conducted by an independent third party. Finally, for the nine - to ten - year - old

  children, the success of the intervention was unrelated to their performance

  during the preparation session. That is, for these older children, mere partici-

  pation in the preparation session was suffi cient to increase accuracy during

  cross - examination.

  Naturally, there are aspects of the intervention that require more compre-

  hensive investigation. For example, although the intervention did not reduce

  the number of prior mistakes that children corrected during cross - examination,

  the absolute number of errors that children made during direct examination

  was very small, making the statistical power to observe a signifi cant difference

  relatively low. It is also important to bear in mind that the preparation session

  in this study was conducted just one or two days prior to the cross - examination

  interview. Because such short timeframes may not be practical in actual cases,

  further research is necessary to ascertain the role that the timing of the inter-

  vention plays in its effectiveness.

  Most important to note is that while the intervention facilitated children ’ s

  cross - examination performance, it did not eliminate the negative effects of this

  questioning style in either age group. That is, even children in the preparation

  Investigative Interviewing in the Courtroom

  173

  condition made changes to their earlier testimony that decreased their overall

  accuracy levels. This fi nding is further testament to the robust nature of the

  cross - examination effect.

  Postscript: Adults u nder c ross - e xamination

  While the consistent fi nding that cross - examination questioning decreases the

  accuracy of children ’ s reports is concerning on its own, it also raises the ques-

  tion of whether cross - examination could affect adults ’ testimony in a similar

  way. Zajac & Cannan ( 2009 ) note several reasons to suspect that this may be

  the case. Anecdotally, for example, we know that children are not the only

  witnesses who fi nd cross - examination confusing and stressful. Many adults,

  including police witnesses report the same types of concerns (Flin, 1993 ), and

  much has been written to assist adult witnesses to cope with challenging cross -

  examination questions (e.g., Brodsky,

  2004 ). Second, as described below,

  laboratory research has demonstrated that adults ’ eyewitness reports are sus-

  ceptible to the same kinds of contaminating infl uences as those of children.

  Like children, adults are vulnerable to the way in which a question is

  worded. As questions move from open to closed to leading, adults ’ reports

  become less accurate (Poole & White, 1991 ), and when faced with a question

  requiring only a yes or no answer, adults too are more likely to answer ‘ yes ’

  than ‘ no ’ (Kebbell et al. , 2001 ). Adults will also attempt to answer nonsensical

  questions (Pratt, 1990 ), especially when they require only a yes or no answer

  (Waterman et al. , 2001 ). Even subtle changes in the wording of a question

  (e.g., replacing

  a with

  the ) can impair adults

  ’ accuracy (Loftus

  & Zanni,

  1975 ). Like children, adults also fi nd it more diffi cult to accurately answer

  linguistically complex questions relative to questions phrased in simple terms

  (Perry et al. , 1995 ; Kebbell & Johnson, 2000 ), and adults ’ susceptibility to

  misinformation increases with increased question complexity (Loftus

  &

  Greene, 1980 ).

  Furthermore, adults are far from immune to soci
al pressure during the

  course of an interview. Social psychologists have reported for decades that

  adults are vulnerable to persuasion (see, for example, studies on conformity,

  Asch, 1956 ; Cialdini, 1988 ; and on obedience, Milgram, 1963; 1974 ). Adults ’

  suggestibility increases markedly when they are under pressure or in an intimi-

  dating environment (Hyman, Husband & Billings, 1995 ; Kassin & Kiechel,

  1996 ; Loftus, 1997 ; Horselenberg, Merchelbach & Josephs, 2003 ; Redlich &

  Goodman, 2003 ). Like children, adults are more susceptible to leading ques-

  tions when the interviewer is of higher status (Roper & Shewan, 2002 ). In

  fact, adults are more likely to comply with another adult who is dressed in

  uniform as opposed a shabby outfi t, suggesting that the mere emblems of

  authority can elicit compliance (Bushman, 1984; 1988 ).

  In addition to the empirical fi ndings described above, we also know that

  many adult victims of crime exhibit specifi c vulnerability factors that could

  174

  Handbook of Psychology of Investigative Interviewing

  render them particularly susceptible to cross - examination questioning. These

  factors may include testifying about traumatic or highly sensitive events, older

  age, or learning or communication problems. Adults who have been sexually

  victimized are likely to be a particularly vulnerable group of witnesses (Frazier

  & Haney, 1996 ; Baker, 1999 ; Edward & MacLeod, 1999 ; Lees, 2002 ), as

  sexual assault has been linked with low self - esteem and low self - confi dence,

  both of which are associated with vulnerability to suggestion (Gudjonsson &

  Singh, 1984 ; Singh & Gudjonnsson, 1984 ).

  Not only are adult witnesses likely to exhibit factors that may increase their

  vulnerability to cross - examination questioning, the questioning itself is likely

  to be qualitatively and quantitatively different from that used to cross - examine

  children. For example, we know that, compared to children, adult witnesses

  are asked more cross - examination questions, a higher proportion of which are

  complex and credibility challenging (Zajac & Cannan, 2009 ). Furthermore,

  when cross - examining adults, lawyers are likely to use different reasons for

  challenging a witness ’ s story. In cases of a sexual nature, for example, adult