1 Samantha J. Andrews Michael E. LambHow Developmentally Appropriate Are Direct- And Cross-examinations Of Children In Scotland? Samantha J. Andrews Michael E. Lamb Good afternoon, my name is Sam and I wanted to talk a bit about the first of a number of large scale, systematic field studies we’re conducting at the moment looking at the ways in which children are questioned in court in the United Kingdom. Specifically, Scotland. The study that I’m going to talk about today is really the foundational paper to this project, looking at the different types of questions prosecutors and defense lawyers ask, and how children respond. END Click University of Cambridge, UK

2 Context Considerable concern and debateAccused’s right to a fair trial Allow witnesses to give their best evidence Challenge witness credibility Suggestive “He didn’t touch you, did he?” Closed-ended, repetitious Linguistically complex Before I dive into the particulars of this study, I think it’s important to outline the general context of this line of research. There is currently considerable concern and debate in this country, as well as internationally, regarding the ways in which children are questioned in adversarial court systems. The defense’s cross-examination of witnesses is often deemed essential in protecting the accused’s right to a fair trial, and, as well as this, courts have a duty to allow witnesses to give their best evidence. However, in adversarial jurisdictions, such as Scotland, lawyers are motivated to undermine the opponents’ witnesses. In particular, defense cross-examinations aim to challenge witness credibility and persuade children to change details in their accounts. Small-scale and qualitative studies over the years have confirmed that the defense do this by being suggestive (for example, utilizing suggestive tag questions that imply the expected response), and asking closed-ended, repetitious, and linguistically complex questions. Such questioning exploits children’s developmental limitations and has detrimental effects on the responses they provide, be they more inaccurately inconsistent, less productive etc. These techniques consequently violate guidelines, based on, what most of you are contributing to, a now extensive body of experimental and field research, outlining the best ways to elicit truthful testimony. END Click 2

3 Question Types and Children’s Responses in CourtSuggestive (Andrews et al., 2015a; Klemfuss et al., 2014; Zajac et al., 2003) Repetitious (Andrews et al., 2015b) Linguistically complex (Evans et al., 2009; Hanna et al., 2012) Andrews, Lamb, & Lyon (2015) California, N = 120, 6-12 years, Defense – more suggestive (42%); more contradictions Suggestive = more self-contradictions These concerns are further supported by more large-scale quantitative field studies primarily conducted in New Zealand and in California. Particularly relevant to the current research is a study we conducted a couple of years ago looking at the question types asked of children by prosecutors and defense attorneys in California, and how children responded. Overall, attorneys used more closed-ended than open-ended prompts. Defense attorneys used more suggestive prompts and, given that we found suggestive questions elicited more self-contradictions than any other prompt type, the defense elicited more self-contradictions than prosecutors. That said, nearly all prosecutors elicited at least one self- contradiction, and we found no associations throughout with age. Overall, these findings suggest that neither defense lawyers nor prosecutors question children in developmentally appropriate ways. The defense results aren’t a huge surprise, given that the defense are permitted to ask leading questions. But it is clear that the sheer amount of investment we all put into investigating and practicing how to elicit the best testimony from children in the early stages of forensic procedure is largely not being encompassed towards the end of the process when they are questioned in court. END Click

4 Context Cases dealt with justly with current legislation?Training, special measures, Section 28 Use varies widely Effectiveness not evaluated Lord Carloway – “The lack of readily identifiable exemplars elsewhere should not be a reason to draw back from seeking to modernise the justice system. It is the opportunity for Scotland to take a lead...” (Evidence and Procedure Review Report, Section 5.7, March 2015) This raises serious questions about the extent to which cases can be dealt with justly given the current rules and legislation surrounding the questioning of child complainers in court. There has, of course, been recognition of these issues for a while. It is now widely accepted that gathering evidence from young and vulnerable witnesses requires special care. Judges across the UK are required to attend training given by experts, such as Michael Lamb and David La Rooy, on how to question children. In England and Wales, Penny Cooper’s Advocates Gateway provides free access to practical, evidence-based guidance on vulnerable witnesses and defendants. Protective special measures are also available for vulnerable witnesses on application, and England and Wales have recently begun to role out a new procedure essentially using children’s pre-recorded statements in place of testimony in court. However, the extent to which protective measures are used varies considerably within and between jurisdictions, and their effectiveness, as well as the foundation they are meant to assist, has not been systematically evaluated in the UK. The Lord President of the judiciary in Scotland, Lord Carloway, has been making continued calls for the system to better protect vulnerable witnesses against cross-examination that is aggressive, repetitive, and hostile, rather than focused and relevant. These calls have been met with considerable backlash from some advocates concerned that further protections will render complainers in sexual offence cases being given special treatment compared to other complainers, and that the accused's’ right to challenge the evidence put against him will no longer be adequately protected. END Click 4

5 Present study Child age, lawyer role (prosecution/defence):Question type Frequency of self-contradictions Children’s acquiescence and resistance Defense worse, effects worse, no age effects This brings us to the present project. With the approval and support of the Lord President, we sought to examine systematically, for the first time in the UK, how children are questioned in court. As I said previously, this particular study is foundational in nature, and so takes a relatively similar, basic approach to previous research in beginning to understand some of the key issues. I don’t have time to summarize the whole paper and everything we looked, it’s quite a large study. So specifically for today, child age and lawyer role were examined in relation to the types of questions asked, the frequency of self-contradictions, and children’s acquiescence and resistance to lawyers’ suggestive questions. We made a number of specific predictions, but generally, in light of previous findings, we expected the questioning to be very closed-ended and suggestive from both lawyer roles, but expected that, overall, defense lawyers would perform worse, and this would have the corresponding detrimental effects on children’s responses, such as more contradictions and more acquiescence. We expected there to be no age differences. END Click 5

6 Sample 36 trials, 56 children Scotland, 2009 – 20145 – 17 years, Mage = 13.9 (SD = 2.7) Under 13, 14 – 15, 16 – 17 20% boys Variety of alleged sexual abuse. Special measures Defendants The transcripts comprised of 56 children involved in 36 trials across Scotland between 2009 and 2014, representing the most contemporary sample yet studied. The children were aged 5 to 17 years of age, with a mean age of 14. We categorized them into the 3 age groups you can see here due to the negative skew. The sample comprised of 20% boys, and children testified as victims on a wide range of alleged sexual offences. Some single events, others multiple events, ranging from exposure to penetration, and most of the children received special measures. All defendants were male and almost all of them were known to the children, and most of the trials led to convictions. END Click 6

7 Question Types Substantive only Code Examples 1. Invitation“Earlier you mentioned [person/object/action]. Tell me more about that.” 2. Directive “Who did that to you?” [when ‘that’ was previously mentioned by the child] 3. Option-posing “Did you see his penis?” “Was the light on or off?” 4. Suggestive “Your dad told me that B. touched your private part. Did B. touch your private part?” Child: “He took me into the toilet.” Lawyer: “When did he lock the door?” For the purposes of this presentation, I’m only focusing on the coding for the substantive portions of the direct- and cross-examinations. We coded both lawyers’ questions and children’s responses. First, the question type that elicited the response was categorized. I’m assuming that most of you are familiar with these categorizations. One thing I did want to note, but something I can’t go into, was that, to yield greater insight, this paper is the first to present descriptively the full taxonomy of NICHD utterance types. The full coding scheme consists 15 subtypes, but for inferential analysis we collapsed them down. END Click 7

8 Suggestive Question TypesCode Examples 1. Confrontation “You’re lying, aren’t you?” [when asked a 3rd time] “So did he touch you once or more than once?” [when asked a 3rd time] 2. Supposition Child: “Then I went to meet X.” Lawyer: “You met X. What did she tell you?” (when the child did not mention that X told anything) 3. Introduction Child: “I went to the park…” Lawyer: “You said you went to skate park.” “You saw a gun, didn’t you?” Due to the prevalence of suggestive prompts in criminal court proceedings, we further distinguished among 12 different types of suggestive prompts (that are again presented descriptively in the paper), but then collapsed down into the 3 categories that you can see on the screen for inferential analysis. Vey broadly, confrontation suggestions are instances where doubt is raised and when the veracity of what the child is saying is challenged. Suppositional suggestions are when the lawyers assume information not disclosed by the child, or ignore information provided by the child when asking subsequent questions. Introduction suggestions are mainly when the lawyers introduce information not previously disclosed, or modify, incorrectly summarize, or draw conclusions from the information the child has provided. This coding system is a relatively fresh one, designed by Orbach and colleagues, whose paper is in press. On top of this, we also identified whether or not these suggestions were tagged. END Click Tag questions 8

9 Children’s Responses Self-contradictionsAcquiescent/resistant to suggestion In terms of children’s responses, we identified when children were inconsistent in their responses by contradicting themselves, and coded whether children acquiesced or resisted suggestive questions. Because we can’t assess the accuracy of children’s responses in the transcripts, these last codes should not be viewed as a proxy for accuracy per se, but as an indication of the ‘worst case scenario’ – just how much information is being suggested by lawyers rather than being offered by the children. END Click 9

10 Results – Question Frequency23,325 substantive questions (M = 417) Prosecutors: 14,138 (M = 253) Defence: 9,187 (M = 164) Children under 13  substantive prompts when questioned by prosecutors 16- and 17-year-olds  substantive prompts when questioned by defence We identified just over 23,000 question-response pairs in the transcripts, with around 420 asked of each child. Prosecutors asked about 100 more questions on average than defense lawyers. When looking at interactions in question frequency between lawyer role and age, we found that children under 13, compared to older children, were asked more substantive prompts when questioned by prosecutors, and the oldest children were asked more substantive prompts when questioned by defense lawyers. END Click 10

11 Results – Question TypesWhen looking at question types, we found that, overall, prosecutors and defense lawyers prompted children less with invitations than directive, option-posing, and suggestive prompts, less with directives than option-posing and suggestive prompts, and less with suggestive prompts than option-posing prompts. Prosecutors prompted children with more invitations, directive, and option-posing prompts, whereas defense lawyers prompted children with more suggestive prompts. There were a few effects with age. Children under 13 were asked fewer option-posing prompts than the older children when questioned by prosecutors. When questioned by defense lawyers, under 13’s were asked more directives, and option-posing prompts. 14 and 15 year olds were asked more suggestive prompts than younger and older children when questioned by defense lawyers. END Click Under 13 – prosecutor, fewer OP Under 13 – defence, more directives and OP 14 and 15 – defence, more suggestive 11

12 Results - Suggestive Question TypesTurning to the suggestive question subtypes, suggestive introduction questions were asked more than suggestive confrontation and suggestive supposition questions. Defense lawyers asked more confrontation questions than prosecutors, whereas prosecutors made more suggestive suppositions than defense lawyers. Children under 13 were asked significantly more suggestive confrontation and supposition questions, and significantly fewer suggestive introduction questions, than older children. 14 and 15 year olds were asked proportionally fewer suggestive confrontation questions than 16- and 17-year-olds. Defense lawyers asked proportionally more tag questions than prosecutors. Importantly, there was no significant interaction between children’s age and the proportion of suggestive tag questions asked by prosecutors and defense lawyers. END Click Under 13 – more confrontation and supposition, fewer introduction 14 and 15 – fewer confrontation than oldest 12

13 Results - Self-contradictionsn = 973, 4.4% of all responses Defence more – 6.5% vs 2.7%  invitations than OP  suggestive than all other prompt types Prosecutors  from oldest than 14 and 15 Defence  from youngest  supposition In total, 973 contradictions were identified, constituting four and a half percent of all children’s responses. All children contradicted themselves at least once, and the defense elicited more self-contradictions than prosecutors. Invitations elicited fewer self-contradictions than option-posing questions, but, of course, suggestive questions elicited more self-contradictions than any other question types. Prosecutors elicited significantly more self-contradictions from 16- and 17-year-olds than 14- and 15-year-olds. Defense lawyers were more likely to elicit self-contradictions from children under 13 than older children. Suggestive suppositions elicited fewer contradictions than suggestive confrontation and introduction questions, and there was no difference in the proportion of contradictions elicited by tagged and untagged questions. END Click 13

14 Results – Acquiescence/ResistanceAcquiesced 68%, resisted 29%, unclear 2% Defence   confrontations  suppositions than introduction  resistant to tagged No age effects Of the six and a half thousand suggestive question-response pairs, children acquiesced to the suggestion 68% of the time of the time, and resisted the suggestion 29% of the time. Children resisted defense lawyers’ suggestive questions more than prosecutors’. Children resisted suggestive confrontation questions more than suggestive supposition and introduction questions, and children resisted suggestive supposition questions more than suggestive introduction questions. Children were less resistant to tagged questions than untagged questions. We found no age differences in children’s propensity to acquiesce or resist suggestion. END Click 14

15 Conclusions Suggestive questions adversely affect responsesMore serious during cross-examination Lawyers were not sensitive to developmental capabilities and limitations Implications Further research The present study provides further evidence that suggestive questions adversely affect the consistency of children’s responses and that these problems were more serious when defense lawyers questioned children. Both prosecutors and defense lawyers asked predominantly closed-ended questions, including suggestive questions, and, although unlike previous research, we did find some age effects, perhaps due to our older sample, the results seem to suggest that neither prosecutors nor defense lawyers were sensitive to children’s developmental capabilities and limitations. Interestingly, controlling for child age, the proportion of question types used by prosecutors in the present study was extremely similar to the pattern found in our previous Californian studies. The proportion of question types used by defense lawyers was also similar between the studies, although the defense in the present study actually used around 10% more suggestive questions. This is incredibly worrying considering that the Californian trials were conducted between 1997 and 2001, whereas the trials in this study were conducted much more recently – 2009 to 2014, and it has been in these recent years that there has been a real push to improve questioning practices in court. Why in terms of question types then, are Scottish defense lawyers now worse than American defense lawyers almost 20 years ago? For real improvements to be made, there needs to be a strong evidence-based foundation from which to work from and continual timely assessments and readjustments of training and procedures. Also, in terms of research, it’s not just question types that impact opportunity to give best evidence. There are many aspects that have not received as much attention, even in the interviewing literature. So, on top of this study, we are also looking into the impact of question repetition, linguistic complexity, and question content, as well as taking a more qualitative approach in examining procedural issues and questioning techniques. So, I think it’s safe to say we have a long way to go yet. END Click 15

16 Thank you! Samantha J. Andrews [email protected] 2nd year PhD CandidateDepartment of Psychology University of Cambridge, UK Thank you! Andrews, S. J. & Lamb, M. E. (under review). How developmentally appropriate are direct- and cross-examinations of children in Scotland?