Police Interviewers’ Perceptions of Child Credibility in Forensic Investigations

Abstract

During investigative interviews, police practice can influence key aspects of child credibility, namely the accuracy, competency, reliability, and truthfulness of their testimony. To date, police interviewers’ perceptions of how best to assess child credibility at interview, and how practice impacts upon credibility, have been overlooked. We conducted a qualitative study that examined data from focus groups with 16 English police officers who regularly interview children. The focus group transcripts were analysed using thematic analysis, and four main themes were identified – the 4Es: eliciting information, evaluating credibility, empowering the interviewee, and a high-quality end product. Within these themes, police officers acknowledged some responsibility for the perceived credibility of child victims. Poor interviewing practice could decrease the accuracy of the information elicited and cross-examined in court. Registered intermediaries could empower child interviewees and increase their competency. A lack of reliability contributed to evaluating credibility, but this relationship was not straightforward. Finally, obtaining the most truthful account from child victims was not always possible, because there are many barriers to overcome. Our findings suggest the need for a continued focus on interview protocols that facilitate disclosure from child victims and a review of the professional relationship between those who interview children and prosecutors.

Perceived victim credibility can significantly affect the outcome of court cases involving allegations of child maltreatment. Studies have found that when child victims are perceived to be less credible, mock jurors are less likely to find the defendant guilty (e.g. Castelli, Goodman, & Ghetti, 2005; Goodman-Delahunty, Cossins, & O’Brien, 2010), and, in real-life cases, the defendant is more likely to receive a shorter sentence (Lewis, Klettke, & Day, 2014). Voogt, Klettke, and Thomson (2017) developed the first conceptual model of perceived victim credibility (PVC) in child sexual assault cases. A systematic review of previous measures of PVC resulted in 33 items that were then thematically analysed into five sub-constructs: accuracy, competency, reliability, truthfulness and believability. Accuracy refers to children’s ability to provide responses that are both correct and consistent with the events that occurred. Competency refers to children’s ability to remember the event as well as their understanding of the defendant’s behaviour and the law surrounding the legal process. Similar to accuracy, reliability is concerned with consistency. However, it focuses on the internal consistency of children’s testimony and to what extent the court can depend on children’s accounts. Truthfulness refers to how honest children are when giving testimony and the perception that they might be intentionally telling a false story. Finally, believability refers to perceptions of how willing children are to provide a false account, but also the extent to which their stories are believed from a more subjective and emotional standpoint. Previous research has suggested that the first four of these sub-constructs can be influenced by police interviewing practice.

First, the investigative interviewing protocol used and adhered to by a police interviewer can have a significant impact on the accuracy of the information elicited. For example, some European countries use the National Institute for Child Health and Human Development (NICHD) Investigative Interview protocol (Lamb, Orbach, Hershkowitz, Esplin, & Horowitz, 2007). The NICHD protocol does not always elicit longer accounts than a standard interview, but it does result in an increase in appropriate questions, which, in turn, are anticipated to achieve higher accuracy (Brown et al., 2013; Lamb et al., 2009; Sternberg, Lamb, Orbach, Esplin, & Mitchell, 2001). In England and Wales, an official set of guidelines, ‘Achieving Best Evidence’ (ABE, Ministry of Justice, 2011), is used. These include a suggestion to use the Enhanced Cognitive Interview mnemonics (ECI, Fisher & Geiselman, 1992). The limited research on the effectiveness of ECI with child witnesses consistently reports an increase in correct information elicited from child witnesses compared to the standard interviews. Sometimes this also results in an increase in incorrect information, although overall accuracy is not affected (McCauley & Fisher, 1995). At other times this does not increase reporting of erroneous information and accuracy is marginally improved (Akehurst, Milne, & Köhnken, 2003). It is important to investigate whether police interviewers will identify the positive impact of using such interview protocols on the accuracy, and subsequent credibility, of the accounts that they elicit from the child victims.

Second, in England and Wales, police interviewers can refer child interviewees to Registered Intermediaries (RIs), who help to enhance the children’s competency to provide a good-quality account of what they have witnessed (ABE Guidance, MoJ, 2011; Section 29 of Youth Justice and Criminal Evidence Act 1999). Children’s ability to remember and recall past events develops with age (Hershkowitz, Lamb, Orbach, Katz, & Horowitz, 2012), so interviewers should adapt to children’s developing communication skills (Lamb, Malloy, & La Rooy, 2011). Based on a formal assessment of the witness’ communication needs, RIs provide written recommendations on how ‘normal’ interviewing practices should be adapted – sometimes meeting with the interviewer prior to the interview to assist with question formulation, and then being present during the interview to monitor questions asked (Henderson, 2015). Their presence during the interview means that they can intervene and rephrase certain questions to avoid any miscommunication or repeat the child’s answers to avoid any misinterpretation (MoJ, 2011). With regard to cross-examination in courts, RIs, lawyers and judges have all reported the positive impact of RIs on facilitating access to the justice system for vulnerable witnesses (Henderson, 2015). That said, police interviewers’ use of, and perception of the impact of, RIs during investigative interviews have not been previously examined.

Third, police interviewers can, and often do, conduct repeat interviews with child witnesses (La Rooy, Katz, Malloy, & Lamb, 2010). Repeated recall attempts have resulted in some details being lost (oblivescence) and added (reminiscence) over time (Erdelyi, 2010), ultimately affecting the perceived reliability of that information. A repeated interview conducted shortly after an initial interview, soon after disclosure by a child, can result in significant increases in new and accurate information (La Rooy, Pipe, & Murray, 2005). However, there are mixed findings regarding the benefits of repeat interviews after longer delays, with some research showing no increase in recall (La Rooy, Pipe, & Murray, 2007). Following a review of the literature, La Rooy et al. (2010) concluded that, when interviewers adhere to best-practice guidelines by asking more appropriate, open-ended questions rather than inappropriate, closed and leading questions, repeat interviews should be used to obtain more complete accounts. It is unclear whether police interviewers are aware of this recommendation in the literature and how they weigh up the benefits of eliciting a more complete account against the costs of that account becoming more unreliable (in terms of credibility).

Finally, a reluctance of children to talk about alleged events in detail at interview is a recurring issue that police interviewers face (Leander, 2010; Magnusson, Ernberg, & Landström, 2017). For example, eliciting abuse-related information at interview can be very difficult when a child does not wish to talk. Disclosure is important because it increases the likelihood of a case being referred for prosecution (Stroud, Martens, & Barker, 2000) and reduces the likelihood of the case being withdrawn (Christensen, Sharman, & Powell, 2016). A systematic review of the literature on child sexual assault cases (Lemaigre, Taylor, & Gittoes, 2017) identified three barriers to eliciting a disclosure at interview from child victims: perceived negative consequences for self, perceived negative consequences for others, and the emotional response to the abuse (e.g. shame, guilt, self-blame). Rapport-building can facilitate disclosure of stressful or embarrassing events (Larsson & Lamb, 2009). To resolve ongoing issues of non-disclosure, the NICHD interviewing protocol was revised to put a greater emphasis on rapport-building at the start of the interview (Hershkowitz, Lamb, Katz, & Malloy, 2015). For children interviewed using the revised protocol, this resulted in more co-operation and greater willingness to talk about their experience than those interviewed using the original protocol (Ahern, Hershkowitz, Lamb, Blasbalg, & Winstanley, 2014; Hershkowitz, Lamb, & Katz, 2014). By increasing the amount of abuse-related information elicited and making children feel comfortable to share an honest account of what happened, police interviewers can increase the truthfulness of children’s accounts.

The current study was the first to take a qualitative approach to understanding how police interviewers conceptualise child credibility as a whole, with an emphasis on how they assess child credibility at interview, and how they perceive that their practice impacts upon credibility. Previous research studies using qualitative methodologies to investigate police interviewers’ professional experiences (Cherryman & Bull, 2001; Hanway & Akehurst, 2017; Wright & Powell, 2006) have yielded rich data with small samples. It was anticipated that the findings of the current study would highlight whether police interviewers conceptualise child credibility in line with the aforementioned psychological literature and whether they use the evidence-based tools and procedures associated with increasing child victims’ perceived credibility. In England and Wales, child victims’ video-recorded investigative interviews can act as evidence-in-chief in subsequent court proceedings (MoJ, 2011). The impact of police interviewing practice on a child’s perceived victim credibility can, therefore, have serious legal implications.

Method

Participants

Police officers who regularly interview children were recruited from two English police forces: one in the north of England and one in the south of England. Emails about the research were sent to all officers working in the Child Protection Units of the two police forces and stipulated that volunteers should be regularly interviewing children. The emails included contact details for the principal investigator. Three focus groups were organised, each with a maximum of 5 participants. As officers volunteered, they were allocated to a group; when each group was full, recruitment for that group ceased (in the end we over-recruited for one focus group, so 6 officers attended). Therefore, 16 eligible police officers (10 women, 6 men), from two English police forces, who regularly interview child witnesses participated in the study. The age of the interviewers ranged from 31 to 60 years (M = 41.53 years, SD = 8.37). Overall, our interviewers had 3–40 years of experience working for the police (M = 17.09 years, SD = 8.16), and 5 months–22 years of experience in interviewing children as part of their police role (M = 8.56 years, SD = 6.16). The first focus group (FG1) consisted of 6 female interviewers, the second group (FG2) consisted of 5 interviewers (3 men, 2 women), and the third focus group (FG3) consisted of 5 interviewers (3 men, 2 women). There were no differences between focus groups for overall police experience, F(2, 13) = 0.10, p = .90, or for experience interviewing children, F(2, 13) = 0.41, p = .67. All interviewers had undergone a specialist interviewing course. The age of the children whom our interviewers routinely interviewed ranged from 2 to 17 years for the majority of the sample (62.5%). The rest of the sample interviewed adolescents (13- to 17-year-old children) only. When asked the reason for their most recent interviews with children, 70.8% were cases of alleged child sexual abuse. Other recent cases included neglect, physical assault and witness to a murder.

Procedure

Police interviewers, who had received specialist interview training (for interviewing child witnesses) and who regularly interview child witnesses, were contacted via e-mail to invite them to attend a focus group; they were each sent a participant information sheet with details of the study. Participation was on a voluntary basis (i.e. informed consent was obtained from all individual participants) and participants took part in one of three focus group sessions that all followed the same procedure. Upon arrival, police officers were provided with an opportunity to ask any questions before providing informed consent to participate in this study. Following this, the focus group facilitator outlined the structure of the session and the ground rules for the focus group discussions (e.g. to avoid interruptions).

Focus groups

In this study, the focus groups were moderated by the facilitator and conducted in a quiet environment on police premises. Sessions took about 40 min each. The focus groups were semi-structured discussions. Police interviewers were asked: (1) What training do you currently receive to assess child credibility? (2) What techniques do you currently use to assess child credibility? (3) What do you think about techniques informed by psychological research? (4) How does the age of the child affect assessments of their credibility? (5) What has been overlooked by research? All questions were open-ended, with the facilitator asking follow-up questions to clarify participants’ responses if necessary (e.g. ‘explain to me what you mean by that’). The questions were framed to be broad regarding child credibility and its assessment, so that the interviewers’ responses were not influenced by the researchers’ knowledge of the current literature.

Analysis

The focus groups were audio-recorded using a dictaphone and later transcribed verbatim. Interviewers’ responses were anonymised (names were replaced by letters). The focus group transcripts were coded inductively using thematic analysis. Thematic analysis was chosen because it does not prescribe a method of data collection nor a theoretical position (Braun & Clarke, 2006). Its flexibility enabled all the data across all three focus groups to be compiled ready for coding for broader themes. The focus group transcripts were summarised and reduced down to smaller “packets” of information, which were then analysed for themes in the data (Boyatzis, 1998). Emerging themes had to reflect similarities across multiple interviewers’ responses (Braun & Clarke, 2006). Themes were presented by the first researcher to the second researcher in terms of their description and keywords. The second researcher then checked the presence of these themes against the verbatim transcript excerpts to ensure that the findings were representative of the original data. Four main themes were identified – the 4Es of child interviewing – with each structured into two sub-themes: (1) Eliciting information with sub-themes, Techniques and Barriers to disclosure. (2) Evaluating credibility with sub-themes, Indicators of deception and Exceptions to the rule. (3) Empowering the interviewee with sub-themes, Planning and preparation and Flexibility. (4) End product with sub-themes, Preparing the evidence and Cross-examination. (See Supplementary Materials for a table displaying which participants endorsed which sub-themes and the overall proportion of endorsement per sub-theme.)

Results

In this section, we present the interviewers’ thoughts and experiences regarding child credibility with an emphasis on how they assess child credibility at interview, and how they perceive that their practice impacts upon credibility. For the following quotes, to protect privacy, ‘FG’ refers to the focus group number and ‘P’ refers to the participant letter that was assigned to each member of each focus group. To give an overview of the main themes and sub-themes, a map has been provided (see Figure 1).

Figure 1.

Main themes and sub-themes – the 4E’s.

Eliciting information

This theme reflected many interviewers’ views that the interview was first and foremost an information-gathering procedure. Within this theme, police interviewers recognised their own responsibility for eliciting information from child interviewees through the use of specific techniques. They also highlighted potential personal barriers that child interviewees might encounter when asked to disclose abuse.

Techniques

Our interviewers talked about general interview strategies for encouraging child interviewees to talk about their experiences, such as asking them to concentrate and making interviewees feel as comfortable as possible. Interviewers from two focus groups also mentioned a specific technique that they use at the beginning of interviews to elicit longer responses later on:

I’ll give them an example of how much detail I need. I might say to them, ‘so tell me what you see about that cup’ and they’ll say, ‘oh it’s a cup, you know it’s cream in colour’ and I’ll say, ‘but if you look at it now, the detail that I need is that it’s cream in colour, it’s got red writing on it, it’s got ‘wibble wobble’ and a picture of a jelly. Inside it’s dirty’. So, I’d explain to them that that’s the detail that I’d want (FG1, PE).

I often provide an example and I try and tailor it to the individual. So, if it’s a child, often I’ve picked the child up so I will talk about the journey we’ve taken to get there so it’s something they’ll identify with and often they’ll chip in with bits that I’ve missed out but I try to get them to consider all the senses and explain that as well … which gives them an idea of the level of detail I’m looking for (FG3, PD).

Overall, interviewers agreed that anything that increases the chances of a child disclosing more details is good, with fewer questions being particularly beneficial.

The best one is where you don’t have to ask them an awful lot of questions …something that would get people to give more information without asking …too many questions (FG3, PC).

Our interviewers understood that children are able to disclose information without much prompting from the interviewer, but that, sometimes, the level of detailed disclosure required for the investigation needed to be outlined at the beginning of the interview.

Barriers to disclosure

Our interviewers mentioned that while some child interviewees – especially,young children – unintentionally forgot key information, some children also intentionally excluded certain details from their accounts. It was recognised that emotional factors such as shame, self-blame and embarrassment might preclude children from talking about their abuse.

It’s just they’re either ashamed of what’s happened, they realise it was wrong what’s happened to them and they don’t want to talk about it (FG1, PE).

They miss bits out … where they think they may have been part of, you know, partly to blame or something because they put the blame on themselves (FG1, PB).

there’ll just be some words that they can’t say … but we need them to actually say what that body part is and that’s sometimes a real barrier (FG3, PC).

Also, in terms of willingness and readiness to talk, interviewers stated that children who are sometimes not the origin of the disclosure may not wish to talk about what has happened to them.

Often the allegations that come to us haven’t come from the child themselves so it’s not that the child’s been ready to talk about it … so we go out with our tick list of lines of inquiry … but actually for that person, if we go in there too soon … then they’ll just drip feed you a tiny bit of information and they won’t actually be telling you everything that’s happened to them (FG3, PC).

Interviewers mentioned that it can even be difficult in the first instance to get children to assent to an interview;

You’ve got people that crikey it’s enough getting them through the door…they don’t want to be there in the first place (FG3, PA).

A reluctance to be interviewed was considered by our interviewers to be partly attributable to feelings of loyalty towards an alleged offender, who is often a close family member.

I had a teenage girl who was sexually abused by her biological dad … he got convicted but she, all the way through, found it so difficult talking about it ’cause she still loved her dad …very loyal (FG1, PA).

Furthermore, interviewers suggested that children can also understand the consequences for disclosing abuse and may wish to avoid these negative outcomes by keeping quiet.

Younger children, they’ll cover up for their parents …it has happened to them but they don’t want to be removed and they know the consequences if they tell us things so they try and hide as much as possible (FG2, PB).

Interviewers in the current study suggested that children’s loyalty towards their family and peers can also produce a ‘them-and-us’ mentality towards the police, creating negative attitudes about talking to police officers.

They’ll call us pigs and they won’t talk to us and it’s how they have been brought up and the perception they have of police … they won’t tell us what’s happened ‘cause they’ve learnt from a young age, don’t talk to police officers …it’s quite bizarre when we get little ones telling us to f**k off (FG1, PA).

Our interviewers recognised that children might encounter a variety of barriers when it comes to disclosing information about what has happened to them. They highlighted the diversity of motivations to deceive police into believing a false negative (i.e. that nothing has happened when, in truth, it has).

Evaluating credibility

While all interviewers in all focus groups agreed that it was not the role of the police interviewers to judge the veracity of an individual child, there were comments about getting a feeling for whether a child is telling the whole truth or not;

We get a feeling straightaway when we’re interviewing children. I had one the other day and I thought she’s not telling me the truth (FG1, PA).

The question is, therefore, what protocol do interviewers follow when they get this ‘gut feeling’? Although interviewers reported that they did not receive any official training on assessing children’s credibility, they did talk about general cues that might indicate a deceptive account. Our interviewers also believed that there were ‘exceptions to the rule’: factors regarding a child witness or an incident that would explain the presence or absence of certain perceived indicators of deceit.

Indicators of deception

Interviewers suggested that a reliable indicator of deceit was a lack of detail in children’s responses. This could be a less detailed statement overall or few details regarding a specific aspect of the incident:

It is hard to get detail from someone that may be lying and they do sort of tie themselves up in knots and you will come out with not a lot of information (FG1, PB).

They can tell you details about everything around it, but actually when you come to the act, that’s where you can often detect the lie because the detail within that is usually small (FG3, PD).

Interviewers suggested that the use of probing questions following the free recall phase of an interview could be particularly revealing.

People who prepare a story that they’re prepared to give you do not anticipate us going into the depth we do and the finer grain detail and you sometimes see elements of vagueness arising …and it does then start to initiate your thought process …. I find some of the techniques about detail are where I get indicators, where I start to become uncomfortable about what somebody’s saying (FG2, PA).

Another possible indicator of deception mentioned by our interviewers was the age-appropriateness of the language used by children.

They might use a word that you think they wouldn’t normally use at their age, you know, that’s come from the mum or the dad or whoever …it’s normally quite obvious (FG1, PB).

Children might also provide inconsistent information across the course of an interview. Interviewers reported that they were taught to take brief notes during interviews of inconsistencies that they should clarify, not challenge, at the end of the interview.

If there’s an issue it would be …‘I don’t quite understand, I’m a little bit confused about what you said earlier on so can we just go through that again’ and try and clarify it, never challenge (FG1, PB).

Our interviewers pointed out that they cannot directly accuse a child of lying and cannot be biased at interview. Interviewers have to strike a delicate balance between sufficiently clarifying the information provided sufficiently to satisfy the defence team but not so much that the prosecution can accuse them of treating the interviewee as if s/he was deceitful.

We’re trying to test [their] account at the very end without saying, ‘you haven’t told me the truth or you’ve told me different things’. We’ve got to be very careful (FG1, PA).

Another difficulty highlighted was when children’s statements appeared to contain indicators of both truths and lies. Our interviewers mentioned that teenage girls sometimes report that an ex-boyfriend, who they have recently split up with, has raped them.

They’ll give you information about a time when they had sex, which you know is all truthful, and it’s the ‘how it was forced or ‘how they didn’t consent” … which is the difficult bit (FG3, PC).

One interviewer recalled a case of a girl who reported a catalogue of sexual offences against her, some of which were confirmed and some of which were not confirmed. However, the addition of false information discredited her entire account.

She used elements of truthfulness and confabulated lies around that … and what exposed her was when she was telling us about events we were then able to check those events and recover CCTV footage. When we viewed the CCTV what she was saying happened, which came across very plausible and understandable, was actually something completely different from the CCTV … the frustrating thing is we believed that some of it was truthful but she did so much damage to the investigation by telling us stuff we were able to show didn’t happen … (FG2, PA).

Interviewers recognised that inconsistencies could also arise after interviews have finished. Interviewers were inclined, when making credibility assessments, to take a step back and look at all the information elicited during interviews with multiple people. They reported that this allows them to see where corroboration occurred.

That’s about getting accounts from different people. From the suspect, from the witness, stuff like that and then it builds a bigger picture rather than trying to work out if that individual’s being truthful or not (FG1, PD).

In summary, our interviewers identified multiple potential indicators of deception as well as ways in which they might actively test these indicators – for example, asking more specific questions and comparing the information provided by a child with other case information.

Exceptions to the rule

Our interviewers put forward alternative explanations for brief statements that lack detail (that might originally be construed as fabrications). If children were young, they were expected to give shorter statements and would need prompting more often than older children. In other words, it was suggested that it might take more questioning to elicit the same amount of detail from very young child truth-tellers compared to when older child truth-tellers were being interviewed. Interviewers also considered it unlikely that very young children would lie in the first place, because they have yet to develop an understanding of the concept of lying. It was felt that, as children become more capable of lying as they get older, they would be more likely to lie.

They [young children] are often brutally honest at that age. It’s not until they are sort of four/five that their minds, you know, allow them to [lie] and I think a lot of that comes from maybe being at school with older children that are able to lie about things (FG1, PF).

On their own, very young children were considered to be incapable lie-tellers who required more mature models through whom they could learn vicariously the ability to lie. This idea of innocence for younger children was also related to them reporting topics that were age-inappropriate and this being an indicator of truthfulness.

You know, if they’re describing seeing … their dad’s penis or something and they’re saying, ‘well it was pointing up and doing this’, the only way they would know that information is if they’d seen it (FG3, PC).

Another reason for a lack of information was suggested to be the effect of the trauma of the incident on the ability of the children to recall everything that happened:

Given that we’re often dealing with serious sexual offences that [saying they can’t remember part of the incident] might be as a result of trauma. So actually, they can’t piece together things in a structured, chronological order, that it twists things in their mind or they can’t talk about it because it’s purely too traumatic (FG3, PD).

Finally, interviewers also spoke about how lies could contain lots of details. For example, children coming from negative home environments may embellish or exaggerate their stories to escape their current living situation.

Every time she’s been interviewed she’s added more and added more … she’s developed this thing where six different men have apparently abused her to make sure she never goes back in and, I’m not saying it’s a definite lie, but it’s not really that plausible and she’s done it because, I suppose, in her life that’s her way to get out (FG2, PB).

Here, interviewers demonstrated a sensitivity towards contextual factors that might explain a lack of detail and were also able to highlight examples of where a lot of detail might not necessarily reflect a truthful account. These ‘exceptions’ combined with the ‘barriers’ listed under the previous theme highlight the sheer complexity of assessing credibility in child maltreatment cases.

Empowering the interviewee

Many of our interviewers commented on how the usefulness of interview techniques for eliciting information depended on each child interviewee. In other words, one size does not fit all. Interviewers stressed that it was very important to tailor their interviews to each child interviewee to get as much information as possible:

Ultimately, we need to safeguard the children that we’re interviewing, and that’s the most important thing. So, if they want to tell [us] something, it’s got to be the way they want to tell us (FG1, PB).

Interviewee empowerment referred to tailoring their techniques to avoid false denials or the omission of key abuse-related information for a prosecution. Interviewers stated that enabling interviewees required much preparation prior to interviews to ascertain the best interview strategy, but flexibility was also required during interviews to adapt to any unforeseen individual differences that may arise.

Planning and preparation

Interviewers mentioned the importance of supporting child interviewees by preparing prior to their interviews so that they could tailor their interview techniques to get as full and as accurate an account as possible from interviewees. They considered that the best way to conduct pre-interview preparation was by facilitating intermediary assessments. There was overall consensus that intermediary assessments are a good source of information for assessing each child’s suitability for interview and how best to phrase questions.

When the Intermediaries do the assessment … we’re often present … so you get a feeling as to how they respond to certain types of questions and, you know, the number of sort of key words you put in a question (FG3, PC).

Our interviewers felt that an intermediary assessment was also beneficial for highlighting a child’s suggestibility and for recommending how best to interview a child with learning disabilities.

I had one the other day and there was a picture of summer and she [the Intermediary] kept saying to the child, ‘in the winter so this is winter’ and the child was just agreeing (FG1, PA).

You [have] got to assess the individual … and depending upon the learning disability, and perhaps in consultation with an Intermediary as well, you might only do five minutes ‘cause their attention span is so short (FG3, PD).

External input from a registered intermediary helps to shape their questions to be more specific to individuals. Avoiding suggestive questioning was considered key for preventing the elicitation of false information (i.e. false allegations) from child interviewees who may simply acquiesce to what an interviewer says.

It was also mentioned by our interviewers, however, that a delay while waiting for the right registered intermediary (i.e. one who is specialised in assessing a particular child interviewee) can have a negative effect on a young child’s recall ability. One interviewer discussed this issue in relation to interviewing very young children.

For little children, they [Intermediaries] can’t help them remember. So, we will still have to interview them a lot later down the process and if they’re struggling to remember something then the Intermediary, although they can give us how we should be forming our questions, it can’t help them remember (FG1, PB).

This relates back to the aforementioned unintentional withholding of truthful information (i.e. false denials) due to young children’s forgetfulness. If interviewers considered this particular population of children to already be at risk of forgetting, then interviewers suggested that they might choose to forego an intermediary assessment.

Flexibility

Despite preparing for the interview, our interviewers discussed the need to be flexible and to be able to adapt to any individual differences that are not covered during the intermediary assessment. For example, a child’s background may make it difficult to predict how s/he will behave during an interview.

The kids we’re going to get, whether they’ve been abused or not, whether they’re truth- or lie-tellers, are going to be chaotic people who are in deprived backgrounds, don’t know where their next meal’s coming from, no discipline (FG2, PC).

It can be predicted that a generally chaotic child might behave unpredictably at interview, so it is up to interviewers to use their expertise and experience to react accordingly. The best advice mentioned by our interviewers was just to ‘roll with it’, demonstrating that, to some extent, the interviewer should just go with how the interviewee wants to present themselves and their story. Interviewers also mentioned that a child’s personal disposition may also mean adapting the flow of the interview to that child.

We might be going in and going out of the subject, you know, if you’re seeing that you’re upsetting them (FG3, PA).

Interviewers recognised that the sensitive subjects of many interviews with child witnesses may have an effect on their willingness to recall a target incident. However, as much as interviewers reported trying to construct a safe and encouraging environment to increase the likelihood of disclosure, sometimes this is not always possible. One child interviewer stated,

I had one years ago and I did three attempts ’cause they were just struggling to talk about it and they wouldn’t talk about the offence. Like you say they would talk about everything around it but it was just too horrific for them to talk about. So, three times we tried and, in the end, we never got it (FG1, PF).

The interviewer must, therefore, accept that being flexible for an interviewee also includes terminating an interview before sufficient evidence has been gathered, which may mean that a case cannot be progressed. Interviewers must sometimes accept that, despite their best efforts, a child will continue to falsely deny that any abuse has occurred.

End product

The final theme was the importance of ensuring a good ‘end product’ (i.e. a high-quality video-recorded interview) that can be played in court to ensure that children can be safeguarded and potentially achieve justice.

The whole idea is we’re trying to help that child get this to the courtroom and get that offender convicted. If we do something wrong … they’ll throw the interview out and the case is lost and we haven’t protected that child (FG1, PB).

In other words, interviewing child interviewees is a ‘high stakes’ business and interviewers in our study recognised that if they put a foot wrong, then there can be serious legal and safeguarding implications. Interviewers must carefully prepare the evidence for the Crown Prosecution Service (CPS) so that it is admissible in court and will stand up to cross-examination from the defence team.

Preparing the evidence

Across all focus groups, comments regarding evidence were mainly about the pitfalls of interviewing children that may render evidence inadmissible in court. Our interviewers noted that police interviewers must do what is right for the children at the same time as doing what is right for the court. Interviews, therefore, need to be in sync, and not at odds, with all of the components of the judicial system. Interviews need to be conducted according to national guidelines:

We have to show that we did everything to the book” (FG1, PA)

But, taken to the extreme, this resulted in some interviewers showing resentment towards the system;

We avoid transporting children to interviews and having conversations with them in the car because … you have to write down the whole conversation in the white book so that they know. Obviously, we’re police officers, we can’t be trusted. We have to write down to show what topics we’ve spoken about so that we’ve not been coaching them in the car (FG3, PC).

Interviewers reported that they regulated their behaviour to try to avoid the mistrust of court officials at trial and to ensure that the children’s evidence was not considered tainted. However, interviewers did express how unnatural this regulated behaviour felt when confronted with children in obvious distress.

They’ll [the defence lawyers] say things like … ‘it’s not a fair trial’, you’ve got to be so careful … and it’s awful when you’ve got a five-year-old or six-year-old that’s sitting there crying, especially being a female as well, you want to comfort that child (FG1, PA).

Indeed, this lack of comforting also extended to avoiding verbal encouragements and neutralising all aspects of their behaviour;

We’re not even allowed to say, ‘well done, thank you’ nothing. They sit in that interview and we are like a robot (FG1, PD).

These comments reflect the constraints that interviewers perceived the court system to impose upon their practice. There was also a sense of ‘walking on eggshells’, especially when interviewers did not agree with changes in procedural guidelines. One child interviewer reflected on the contrast between previous and current practice in terms of empathic responsivity.

Before if a victim used to cry, we’d be able to put our arms around the victim and we’d given them a bit of comfort. Now, it’s just about getting them a box of tissues, ‘are you okay to continue? Would you like a break? Would you like to go and see whoever is downstairs?’ (FG 1, PA).

Interviewer behaviour aside, difficulties also arose from the politics of getting evidence to court. One child interviewer described a case where a boy told a story that was only half supported by eyewitness evidence; other parts of the child’s account were not supported by other evidence. They described that, in the end, the courts only looked to prosecute the mother for what was supported, ignoring all the other detail given by the child, about other crimes that were not supported. They could not understand what decision rules the CPS had used to make their judgements:

CPS … what are their qualifications to make these decisions? ’Cause that’s what they do, day in day out, is make decisions about whether a kid is telling the truth based on pretty much nothing I can establish (FG2, PC).

To summarise, the interviewers acted proactively when preparing evidence for court; they attempted to pre-empt the criticism that the CPS would make and tried to resolve these issues earlier on in the investigative process rather than reacting to CPS feedback when the evidence had already been submitted. Their responses suggested a discordance between how police interviewers evaluate credibility and how the CPS might evaluate credibility.

Cross-examination

Our interviewers spoke about their experiences in court when they are cross-examined about how they have elicited evidence from their child witnesses. This experience was resoundingly negative across all focus groups. In particular, they expressed a great deal of negativity regarding defence barristers.

You’ve got your horrible defence barristers who will just throw anything. They haven’t got to prove anything. They’ve got that luxury where they can just stand there and make a suggestion (FG3, PD).

The interviewers explained how the defence team are more likely to question their interview techniques rather than the actual content of the interviews themselves.

They look at procedure, not necessarily the content and product of the interview itself because invariably there is good material in that … the last thing the defence team want to do is touch the evidence because it’s usually pretty compelling (FG2, PA).

Interviewers’ understanding of the court dynamics explained their caution when preparing their evidence for trial. If they are the most likely to come under fire when the court is examining child witness’ testimonies, they need to be able to demonstrate that they made defensible decisions throughout the investigative process. Interviewers talked about trying to avoid criticisms, which would later occur in court, at the time of interviewing children. Anything that could be seen to be leading child interviewees was avoided because it might taint the information elicited from the children:

We have to be totally transparent and impartial and we could get criticised that we’re goading or that we’re influencing or that we’re preparing them for the interview (FG1, PA).

In addition, our interviewers explained how often nowadays they cut the length of the recorded interview material down before it is sent to the CPS, as ‘lengthy interviews’ were perceived to be another constant criticism from the courts;

We get criticised enough for how long our interviews are (FG3, PC).

However, shorter interviews are not always possible due to the nature of the incidents that children are talking about:

We always argue you know if somebody has come to talk to you about a sequence of events that have been happening to them over a period of years, you can’t get that person to give you all that information in a half an hour interview, ‘cause somebody that’s talking about one event would come and take twenty, twenty-five minutes to tell you that information (FG3, PC).

In other words, it was perceived by our interviewers that the courts were not sympathetic to the diversity of child witnesses; that they didn’t understand that one size (or duration, in this case) did not fit all. Overall, it was very clear that the cross-examination process in courts means that police interviewers have to conform and adjust to the standards of the courts and not the other way around. With regard to children’s credibility, it was suggested that this judgment was more likely to be based on whether the interviewer had adhered to best practice guidelines rather than on the cross-examination of the child witnesses themselves.

Discussion

The aim of the current study was to explore police officers’ perceptions of child credibility, with an emphasis on how they assess child credibility at interview and how they perceive that their practice impacts upon credibility. In particular, our interviewers highlighted the importance of eliciting information, evaluating its credibility, empowering the interviewee and having a high-quality end product.

Eliciting information

Our interviewers agreed with researchers (e.g. Milne & Bull, 2006; Vrij, Hope, & Fisher, 2014) that investigative interviews are primarily an information-gathering procedure. They recognised that it is possible to facilitate detailed disclosure through the use of specific interview techniques and that, sometimes, the level of detail disclosed by a child interviewee was a function of the expectations that the interviewer did or did not outline. Although interviewers reported being able to elicit more detail from child interviewees using a certain technique, giving an example of a detailed statement, they did not comment on whether this increase in detail impacted upon the accuracy of the additional information elicited. When gleaning information from interviewees, it is important to ensure that there is not a trade-off between quantity of detail and accuracy (Koriat, Goldsmith, Schneider, & Nakash-Dura, 2001; Koriat & Goldsmith, 1996). The specific technique mentioned has yet to be researched with child interviewees, so its impact on increasing correct and incorrect information is unclear.

Interviewers identified other barriers to disclosure that might prevent child victims from being completely truthful and disclosing everything that happened, such as negative emotions (e.g. shame, self-blame), concerns around negative consequences for others, and non-disclosure before interview. These perceived barriers were consistent with the findings of previous research (e.g. Collin-Vézina, De La Sablonnière-Griffin, Palmer, & Milne, 2015; Lemaigre et al., 2017; Magnusson et al., 2017) and continue to demonstrate an awareness that children can consciously make false denials about serious crimes, such as child sexual abuse (Leander, 2010). Interviewers reported that they tried to make interviewees feel as comfortable as possible, which could refer to rapport-building. The sample in the current study were all from English police forces where ABE Guidance (MoJ, 2011) and the ECI (Fisher & Geiselman, 1992) form the basis of conducting interviews with children. Although both of these protocols include the use of rapport-building, it could be that further revisions are required to better emphasise the importance of rapport-building when breaking down barriers to disclosure.

Evaluating credibility

Even though our interviewers reported that they receive no official training regarding the detection of children’s deception, they held certain beliefs about what cues could indicate deceit but were mindful that these beliefs should not bias their practice. Their belief that a low level of detail indicates deceit is consistent with research that shows that truth-tellers do tend to include more detail in their accounts than lie-tellers (DePaulo et al., 2003). The use of age-inappropriate language by child witnesses was also perceived to be a likely indicator that a child had been coached by an adult or an older child. To our knowledge, there are no studies that have looked directly at the effects of coaching by adults on the vocabulary used by child witnesses of different ages. Our interviewers also suggested that children describing events that they do not understand (e.g. describing an erect penis without understanding why it is erect) would only occur if the children had actually witnessed the event they are describing (because such an event could not be fabricated by a child who has no knowledge that penises can become erect). This finding relates to the Criteria-Based Content Analysis (CBCA) criterion, ‘Accurately reported details misunderstood’ (Steller & Köhnken, 1989). This suggests that child’s lack of competency to understand the defendant’s behaviour indicates their truthfulness.

Interviewers also referred to low consistency of child victim testimony as a sign of potential deceit. This is consistent with previous research (DePaulo et al., 2003; Global Deception Research Team, 2006). A lack of internal consistency within an interview or across repeated interviews (i.e. poor reliability) and/or a lack of external consistency between the testimony and other evidence (i.e. low accuracy) could indicate deceit. In these instances, it was important to clarify, rather than challenge, any inconsistencies. This investigative, rather than interrogative, approach is important because memory is subject to change over time (Erdelyi, 2010). Indeed, our interviewers understood that the relationship between cues to deceit and credibility was not straightforward and that there could be exceptions to the assumption that a lack of consistency equates to a lack of credibility.

Our interviewers also took into consideration some of the barriers to disclosure identified in the first theme when evaluating whether a child is purposefully withholding information, or whether they simply cannot remember any more information about the target event. Young children were considered to provide shorter statements compared to older children, which is in line with previous findings (Myklebust & Bjørklund, 2010). Interviewers were also sensitive to the negative effects of the trauma of events on the level of detail in children’s statements. Indeed, forgetfulness may not be the result of a memory error but, rather, deliberately triggered to avoid negative emotions related to the incidents in question (Leander, Christianson, & Granhag, 2007; Leander, Granhag, & Christianson, 2005). Furthermore, maltreatment can result in poor encoding of the traumatic event (Gordon, Baker-Ward, & Ornstein, 2001) resulting in children not being able to recall many details during a free recall (Eisen, Goodman, Qin, Davis, & Crayton, 2007). This demonstrates that our interviewers identified factors that might impact on children’s competency to recall what happened.

Empowering the interviewee

Our interviewers’ experiences with registered intermediaries (RIs) mirrored the benefits outlined in the ABE Guidance that RIs help interviewers to understand children’s use of language and assist them to design appropriate questions to improve the quality of the children’s evidence (MoJ, 2011). In particular, it was noted that RIs were able to highlight a child’s suggestibility and provide guidance on how to word questions so that the interviewer did not elicit any incorrect information, which would reduce the accuracy of their testimony. This contributes to the previous research on RIs by demonstrating that they are perceived to be beneficial during the investigative process as well as during cross-examinations in court (Henderson, 2015). It was also reported that an intermediary is not always readily available, and as a result a time delay may occur. Delays between a target incident and an investigative interview can result in less accurate and less complete accounts of what happened (Akehurst et al., 2003; Larsson, Granhag, & Spjut, 2003). It can also decrease the effectiveness of interview protocols (Lamb, Sternberg, & Esplin, 2000; Memon, Meissner, & Fraser, 2010). It is reassuring that interviewers did identify the potential negative effects of a delay on children’s ability to recall (i.e. their competency) and sometimes, for this reason, this resulted in them foregoing an intermediary assessment.

General flexibility to account for individual differences not assessed by RIs, such as a chaotic home life, emotional predisposition and non-responsiveness, was also highlighted. Interviewers described how they change their own behaviour dependent on the characteristics of the witness (e.g. adjusting to chaotic behaviour, going in and out of the topic if it is upsetting) to allow children to tell their stories in their own way. Facilitating disclosure is important because it can reduce the likelihood of cases being withdrawn from the criminal justice system (Christensen et al. 2016; Stroud et al., 2000) and, even in the absence of any supporting evidence, can result in defendants being convicted (Walsh, Jones, Cross, & Lippert, 2010). Nevertheless, it was highlighted that, despite all attempts to tailor an interview to empower an interviewee and repeated opportunities to disclose, some children will simply deny, sometimes falsely (i.e. affecting truthfulness), that abuse has occurred.

End product

Our interviewers focussed on avoiding any negative effects that their interviewing practice could have on the admissibility of children’s testimony. Despite their awareness of good practice guidelines, interviewers still found it difficult to understand exactly what the CPS considered a credible end product. Interviewers were very aware of the need to be transparent about their interviewing practice to explicitly demonstrate that they had remained impartial and had not asked inappropriate questions that might lead to inaccurate accounts. However, our interviewers also highlighted the lack of constructive criticism received from the CPS. The Association of Chief Police Officers (ACPO, 2013) have published Advice on the Structure of Visually Recorded Witness Interviews that aims to strike a balance between the investigative needs of the police officers and the judicial needs of prosecution services. However, it would seem that more exploration is required to understand how the investigators and the CPS work together to increase the likelihood that a case will see its day in court. Our interviewers wished to know more about how the CPS decides whether or not a case is prosecuted.

When the CPS did put cases forward to prosecution, interviewers were already prepared for the cross-examination that they would receive in court. Our interviewers noted that defence lawyers were more likely to try to discredit a child’s testimony by demonstrating poor interviewing practice (which would decrease the accuracy of the testimony) rather than by directly accusing the child of lying (i.e. targeting the reliability and truthfulness of the testimony). To avoid criticism from the defence team, interviewers tried to ensure that their interview style was impartial. They were aware of interviewer bias and that knowledge of the wider investigation could influence their questioning style (Brown & Lamb, 2015). Overall, they felt responsible for the investigation, as was the case for the interviewers in Hanway and Akehurst’s (2017) investigation, and the need to protect children from potential revictimisation. The emphasis in court on interviewers’ questioning suggests that the perceived credibility of child witnesses might rest primarily on the shoulders of the officers.

Limitations

Our sample may not be generalisable to both national and international populations of investigative interviewers. The use of three focus groups from a sample of police officers based in two English police forces does not necessarily represent the thoughts and experiences of all English police officers who interview children. Furthermore, the themes highlighted in this study may not all be relevant for other countries, where different interview protocols, such as the NICHD protocol, are used to interview children. As noted in the Introduction, the NICHD protocol has already been revised to enhance child cooperation and increase disclosure (Hershkowitz et al., 2015); however, ABE Guidance and the ECI have not. That said, in Sweden, where the NICHD protocol is used, similar barriers to disclosure are still being found for cases of child sexual abuse (Magnusson et al., 2017), which suggests that the revised NICHD protocol has yet to eliminate all shortfalls. Furthermore, the use of intermediary assessments is a UK-specific initiative, so other countries might cite other resources that empower the interviewee and facilitate communication for vulnerable witnesses.

Variability within the data (e.g. years of experience interviewing children was wide-ranging across the participants) may have had an impact on findings. For ethical reasons, to protect anonymity, the demographic data of the participants were not linked to their transcribed data. It was, therefore, not possible to explore how certain demographic characteristics, such as years of experience interviewing children or whether they interviewed children of all ages or just adolescents, impacted the study’s findings. The sub-themes with higher levels of endorsement across participants and focus groups – Indicators of Deception, 75%; Barriers to Disclosure, 63% (see Supplementary Materials for all sub-themes) are likely to be less affected by the heterogeneity of the sample.

Future research and practical implications

The findings of the current study present many opportunities for future research. It seems that there may be some discrepancies in how police interviewers and the CPS evaluate credibility. Future studies should explore their differing perceptions of perceived victim credibility and what makes a credible high-quality end product for presentation in court. A recent review of the CPS Rape and Serious Sexual Offences (RASSO) units by Her Majesty’s Crown Prosecution Service Inspectorate (2016) found that the poor quality of files put forward by police officers is the biggest contributory factor to the reworking of cases of serious sexual offences. This could have negative implications for having to re-interview witnesses who find it difficult to talk about what has happened to them. The review reported that a trial process of embedding police officers in RASSO teams or lawyers in police investigative units was underway. It would be of interest for future research to report on how police interviewers and CPS lawyers working together might resolve some of the issues mentioned in this study. The comments of our interviewers suggest that it would be helpful for the police to have further guidance on the decision rules that the CPS use when deciding whether to prosecute a cased of alleged child maltreatment. Ultimately, both investigators and prosecutors are aiming for the same goal: an opportunity to achieve justice for child victims. A more specific and constructive feedback loop between investigators and prosecutors might help to make this legal process more effective.

Conclusion

Despite our interviewers acknowledging that assessing credibility is not the primary focus of their investigative role, they highlighted many ways that their practice could impact on the perceived credibility of the child victim. Accuracy could be negatively affected by the techniques and types of questions that the interviewers posed. Participants suggested that it was important to avoid poor interviewing practice, particularly because this would be targeted during cross-examination and could potentially have an impact on the court’s perceptions of the child victim’s testimony. Competency could be facilitated by the use of RIs, although our interviewers highlighted that there are some factors, such as trauma and time delay, which negatively impact on children’s ability to recall what happened, that cannot be overcome through intermediary assessment. Reliability, in terms of the internal consistency of a child’s testimony, was identified as a sign of potential deceit. That said, our interviewers understood that they should clarify these inconsistencies, rather than jumping to the conclusion that a child was not telling the truth. Finally, truthfulness (i.e. eliciting an honest account) was not always possible, despite our interviewers’ best efforts. They suggested that when child interviewees did disclose information, it was important to reflect on whether this information was elicited according to best practice guidelines and whether there were any potential indicators of false information. If abuse-related information could be elicited, our child interviewers reported taking steps to ensure that the evidence would be both admissible in court and unlikely to come under fire from defence lawyers who might suggest that the children’s evidence had been falsely elicited through poor interviewing practice. Overall, our interviewers’ responses demonstrate that they do acknowledge some responsibility for the perceived credibility of child victims.

Disclosure statement

The authors report no conflict of interest.

Ethical standards

Declaration of conflicts of interest

Author A (Hannah Cassidy) has declared no conflicts of interest.

Author B (Lucy Akehurst) has declared no conflicts of interest.

Author C (Julie Cherryman) has declared no conflicts of interest.

Ethical approval

All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendment or comparable ethical standards.

Informed consent

Informed consent was obtained from all individual participants included in the study.

Supplementary material

Supplementary content and Table are available via the “Supplementary” tab on the article’s online page (https://dx.doi.org/10.1080/13218719.2019.1687044).