Forensic risk assessment interviews with youth: how do we elicit the most reliable and complete information?

Abstract

When undertaking a forensic risk assessment with a young person, most evaluators complete an assessment interview to elicit autobiographical history as well as perspectives, thoughts and feelings about the individual’s offending behaviour. While forensic risk assessment tools provide some suggestions on interview questions and techniques, there is no empirical research that explores the most effective strategies for eliciting detailed and reliable information in this context. This article reviews existing recommendations from the related fields of investigative interviewing and suicide risk assessment and integrates this with guidance from the forensic risk assessment literature to identify best practice recommendations for evaluators. It is hoped that this review will provide a starting point to explore how research from other fields may be integrated into risk assessment interviews to improve the quality and accuracy of forensic assessment.

Interviewing young people for a forensic risk assessment

In Australia, young people aged 10–17 years account for 11% of the general population, but they make up 13% of the offending population (Australian Bureau of Statistics, 2018). According to the Australian Bureau of Statistics (2018), there has been a 35% increase in sexual assault and related offences over the past decade, and acts intended to cause injury were the second most prevalent youth offence in 2016–2017. These statistics highlight the ever-pressing need for reliable assessment of adolescents at risk of engaging in violent behaviour or sexual offending.

Reliable assessments are informed by reviews of collateral information, as well as assessment interviews with stakeholders and the young person of interest. Given the importance of interviews in a forensic risk assessment, it is imperative that clinicians possess strong skills in eliciting information from young people. It is anticipated that this review will provide a starting point for the development of such skills by integrating literature from related fields to identify key components and strategies required for effective interviewing. To begin, the role of interviewing in a forensic risk assessment is considered, followed by a discussion of the most advanced interviewing literature available, drawn from research on investigative interviews. This is followed by consideration of the emerging literature on clinical interviews for suicide risk assessment. The key points from these bodies of work are then compared with recommendations from the scant literature on forensic risk assessment interviewing. Finally, recommendations for best practice and future research are provided.

The role of interviewing in a youth forensic risk assessment

Currently, most forensic risk assessments of youth are based on the structured professional judgement (SPJ) model. This model supports evaluators to determine future violence risk by utilising assessment tools that explore empirically derived risk and protective factors. Such assessment tools have been developed to evaluate many types of offending risk in youth, including general offending (Hoge & Andrews, 2002; Levene et al., 2007), violent offending (Borum, Bartel, & Forth, 2006) and sexual offending (Prentky & Righthand, 2003; Worling & Curwen, 2001). Further, where a young person may be at risk of engaging in specific targeted violence, emerging research suggests that evaluators should consider eight warning behaviours (Reid Meloy, Hoffmann, Guldimann, & James, 2012; Reid Meloy, Hoffmann, Roshdi, & Guldimann, 2014) and the ACTION framework (Borum & Reddy, 2001). The ACTION framework considers the young person’s attitudes towards violence, capacity for an attack, threshold crossing behaviours, intent, other’s opinions and any non-compliance with risk reduction strategies.

Regardless of the type of risk being assessed, all assessment tools emphasise the importance of gathering information from a wide range of sources in order to corroborate, triangulate and validate facts. An example of this is the Structured Assessment of Violence Risk in Youth (SAVRY; Borum et al., 2006), which includes 24 risk items across historical, social/contextual and individual/clinical factors, and six protective factors. The authors recommended that evaluators obtain information from multiple sources including interviews with the young person, their parents and teachers, and review of reports/records from social workers, police, probation, psychologists/psychiatrists, school and medical services. Similarly, the Estimate of Risk of Adolescent Sexual Offense Recidivism (ERASOR; Worling & Curwen, 2001) highlights that evaluators should collect information from multiple sources through varied methods including ‘clinical interviews, psychological tests, behavioural observation, medical examinations, and reviews of previous case records and reports’ (Boer, Hart, Kropp and Webster, 1997, as cited in Worling & Curwen, 2001, p. 4).

In a forensic risk assessment, an interview with the young person is designed to elicit their autobiographical information as well as their perspective, thoughts and feelings about their offending behaviour. This information is integral to formulating their level of risk, as well as to informing treatment and safety planning targets. The ERASOR manual states that ‘at a minimum, evaluators should collect information directly from the offenders AND official records . . . ’ (Worling & Curwen, 2001, p. 4). Similarly, the Psychopathy Checklist–Youth Version (PCL:YV; Forth, Kosson, & Hare, 2003a) outlines that the purpose of the assessment interview is to obtain information on the young person’s history, current functioning and interpersonal style, as well as to explore possible consistent and inconsistent information with the young person.

In spite of the emphasis on assessment interviews in all risk tools, there is very little guidance on how to conduct the interview. Partial exceptions to this include the Youth Level of Service–Case Management Inventory (YLS–CMI) and the PCL:YV. The YLS–CMI provides an optional-use structured interview guide with 58 questions targeting the risk and protective factors in the tool (Hoge & Andrews, 2002). Use of this guide may reduce clinician errors of omission, but it does not advise users on other key components of an interview such as building rapport, establishing ground rules and the limits of confidentiality, exploring inconsistencies or closure. The PCL:YV goes further and provides both an optional-use structured interview guide and general advice about effective interview techniques (Forth et al., 2003a). Evaluators are encouraged to consider multiple assessment sessions to obtain a representative sample of the young person’s presentation and explore inconsistencies between statements and other collateral information. Additionally, evaluators are encouraged to build rapport by discussing neutral areas and avoiding highly structured interviews. Within the interview guide, evaluators are instructed to ‘avoid yes–no pattern of questions and responses’ (Forth, Kosson, & Hare, 2003b, p. 3) although a majority of the questions in the interview guide are yes–no questions with follow-up probes such as ‘why?’. Suggested questioning techniques are also included in an Appendix in the SAVRY, with an emphasis on using behaviourally specific questions rather than more general language (Borum et al., 2006). For example, rather than ask ‘Have you ever been violent?’ evaluators are advised to use a question such as: ‘Has there ever been a time when you hit, slapped, kicked, pushed, shoved or grabbed someone?’ (Borum et al., 2006, p. 96). While these assessment tools provide guidance on interviews and questioning style, the effectiveness of the interview guides in eliciting valid and reliable information has not been empirically evaluated. Further, some of the suggestions are not compatible with research findings from other fields on how to elicit elaborate and accurate responses (Benia, Hauck-Filho, Dillenburg, & Stein, 2015; La Rooy et al., 2015). Finally, beyond these assessment tools, there is little literature to guide clinicians undertaking forensic risk assessment interviews with either adults or youth (Logan, 2018).

Contribution of research on investigative interviewing to developing best practice protocols

The most advanced research on interviewing comes from the literature pertaining to investigative interviewing, which refers to the process of eliciting accurate and detailed information about events or situations in order to assist decision making. Historically, investigative interviewing literature has focused on police interviews of suspects (Memon, Meissner, & Fraser, 2010; Read, Powell, Kebbell, & Milne, 2009) and vulnerable witnesses (Benia et al., 2015; Goodman-Delahunty, Nolan, & van Gijn-Grosvenor, 2017; Powell, Westera, Goodman-Delahunty, & Pichler, 2016). The findings from both areas of this literature are relevant to youth forensic risk assessment because they are related to interviewing vulnerable young people and people who have caused harm to others.

There is a broad consensus in investigative interviewing literature in regards to what constitutes best practice approaches to eliciting accurate and detailed information from witnesses (La Rooy et al., 2015). The structure of the interview should include a period of rapport building, explanation of the purpose and ground rules of the interview, introducing the topic, eliciting a free narrative account, clarifying critical details and closure (Benia et al., 2015; Fisher & Geiselman, 1992; Read et al., 2009). Further, the specific techniques used by interviewers should include an emphasis on open-ended prompts, which encourage more detailed responses without specifying the type of information expected or required in the response (Powell & Snow, 2007a). These prompts may encourage a free narrative response – such as ‘tell me all about what you did’; prompt depth of detail – such as ‘tell me what happened with Joe’; prompt breadth of narrative – such as ‘what happened after he touched your leg’; or descriptive detail – such as ‘tell me all about the special room’. In contrast, specific questions dictate the type of information sought – such as ‘Who’, ‘What’, ‘When’, ‘Where’ details. Because these latter questions elicit more errors, they should only be asked once narrative detail has been exhausted. There is also an emphasis on avoiding any suggestive questions that may introduce new information into the child’s narrative.

The literature on what constitutes good interviewing is so advanced, that research can now focus on how granular differences in question wording or approach may impact responses. For example, studies have explored the ideal length of time for building rapport with vulnerable witnesses, so interviewers may effectively reduce an interviewee’s anxiety without unnecessarily taxing their capacity for sustained attention (Teoh & Lamb, 2010; Whiting & Price, 2017). Similarly, a recent study demonstrated that changing one word in an introductory prompt (from ‘why’ to ‘what’) leads to a more informative response by children (Earhart, Danby, Brubacher, Powell, & Sharman, 2018).

Empirical studies on investigative interviewing have also focused on evaluating the effectiveness of interview training for police and other professionals (Cederborg, Alm, Lima da Silva Nises, & Lamb, 2013; Powell, 2008). Due to inadequate training, it appears that most professionals do not adhere to best practice interviewing and instead over-rely on specific questions to elicit narratives (Powell, 2005). This has led to a focus on enhancing the efficacy of training programmes (Powell & Snow, 2007a) with empirically supported programmes emerging (Cederborg et al., 2013).

Overall, the principles of investigative interviewing and effective training are well established, yet these techniques have not been well integrated into different areas of clinical practice, such as forensic risk assessment. Forensic risk assessments tools are designed to achieve the similar goals to investigative interviews (Forth et al., 2003b); the problem seems to lie in the guidance or education to support clinicians. Given that research shows that most professionals without adequate training over-rely on a more directive approach, it can be hypothesised that there may be a similar skills deficit for clinicians who interview young people for forensic risk assessment. This highlights the need to begin integrating the findings from investigative interview literature into forensic risk assessment practice.

Contribution of suicide risk assessment literature to developing best practice protocols

There is emerging interest in what constitutes effective interview techniques to support suicide risk assessment (Shea, 2017a). Suicide risk assessment and forensic risk assessment are a similar process in many respects. Both assessments involve identification of risk and protective factors, as well as exploring current ideation, planning, behaviours and/or intent to undertake the action of concern. The information is then combined to make a clinical formulation of risk (Shea, 2009). The context of the assessment interviews are also similar. Bantjes and van Ommen (2008) highlighted that when assessing suicide risk, clients often present when they are already highly distressed, so a clinician may need to assess risk at their first meeting when they have not yet established therapeutic rapport. This is a similar context for forensic risk assessment, which is usually conducted by a non-treating clinician. The need to breach confidentiality is more likely in both suicide and forensic risk assessment, where confidentiality will be breached if the client is identified to be at risk of harming themselves or others. Given these many parallels, developments in suicide risk assessment may also inform violence and threat assessment.

While there are a range of tools designed to guide clinicians in determining the severity of suicide risk, accurate clinical judgement requires reliable information. One method to support this is through the use of an interview schedule that may guide clinicians to explore relevant information and avoid errors of omission (Bantjes & van Ommen, 2008). However, even with an interview schedule, clinicians may miss vital information if they do not know how to ask the question. An analysis of interviews between 77 patients and psychiatrists demonstrated that (for enquiries about suicidal thoughts) 75% of questions were leading closed questions with an assumption of the patient having no suicidal thoughts, such as ‘You don’t have thoughts of harming yourself?’ When the question was phrased like this, patients were significantly more likely to deny suicidal thoughts. Conversely, when the question was positively phrased, such as ‘Do you have thoughts of suicide?’, there was no significant increase in patients reporting suicidal thoughts. These results suggest that careful consideration needs to be given to the structure of questions as this aspect can markedly influence outcomes. Further, the results highlight the importance of training clinicians in how to ask questions, as well as what information to explore when evaluating risk.

The Chronological Assessment of Suicide Events (CASE) approach is an interview strategy designed to elicit a patient’s suicidal ideation, planning, behaviour, desire and intent (Shea, 2009) and has been implemented across a range of organisations (Shea, 2017a). Shea (2017b) also adapted this approach for eliciting violent ideation and intent, known as the Chronological Assessment of Dangerous Events (CADE). Although the CADE approach is less well known, it follows the same principles as the CASE approach. The overall principles of the approach are described below with examples from suicide risk assessment (examples from violence risk assessment are discussed in the next section on forensic risk interviews).

The CASE approach outlines both a recommended interview structure and strategies to elicit reliable and detailed information. Shea (2017a) recommended that interviews are structured to explore recent, past and then current suicidal incidents. Specifically, the interviewer starts by exploring incidents that have occurred in the past 48 hours and then incidents over the past two months. The interviewer then elicits a general history and finally asks about thoughts, feelings or intent that are present at the time of the interview. This temporal sequence is recommended in order to support interviewers in building rapport and trust with the individual before asking about the most sensitive risk relevant information.

In addition to outlining a recommended structure, the CASE approach also guides interviewers on specific techniques to facilitate open disclosure by individuals at risk of suicide. First, when raising the topic of concern, Shea (2009) recommends prefacing the topic with a statement that normalises suicidal or violent thoughts or attenuates the potential shame associated with those thoughts. An example of a normalising statement may be: ‘Sometimes when people are being bullied a lot at school, they think about killing themselves. Have you thought about killing yourself?’ In contrast, a shame-attenuating statement reflects on the person’s pain as a gateway into the topic, such as ‘With all the stress you are experiencing at home and school, have you had thoughts of killing yourself?’. Shea suggests that these strategies indicate to the patient that it is safe to discuss the topic, so they may be more likely to disclose suicide-related thoughts or events. The clinician can then explore their risk using the chronology of events outlined above.

The CASE approach highlights a further four questioning techniques to support the exploration of suicidal incidents. The first technique is to focus the patient on behavioural descriptions of events and facts, rather than ask for opinions. For example, ‘When you say that you ‘taught your son a lesson’ what did you actually do?’ The second technique is to use gentle assumptions and word questions in a way that assumes that suicidal or violent thoughts and preparatory acts have occurred, such as ‘what other ways have you thought of killing yourself?’ As this question is leading, Shea cautions its use with patients that may be likely to acquiesce easily. The third technique is to ask specific questions after a patient has denied a generic question. For example, after a client has denied thinking of any other way to kill themselves or hurt others, an interviewer may ask: ‘Have you thought of overdosing?’ Shea argues that these questions are useful as either they may trigger a memory in a patient or their non-verbal response (such as a pause or twitch) may provide insight into false denials. The final strategy is symptom amplification, where an interviewer will provide extreme estimates of symptom frequency or severity if they are concerned a patient may minimise in their response. For example, ‘How often do you think about suicide when you are awake, would you say 70%, 90% . . . of the time?’ Again, Shea cautioned against the use of this technique with anyone who may be motivated to ‘maximise’ their response, such as an adolescent boasting about their quantity of alcohol intake. Despite the increasing use of the CASE approach in practice, there is no empirical research exploring its effectiveness for eliciting reliable and valid information.

Current recommendations from literature on interviewing youth during a forensic risk assessment

Empirical research exploring interviewing in the context of youth forensic risk assessment is scant; however, there is some literature summarising recommendations from experts in the field. Recently, Davies (2019) proposed a model to support the development of evidence-based clinical forensic interviews. The model emphasised four key overlapping domains that must be considered when reviewing interview practice. These domains include the interview process, the content of information collected, the interpersonal skills of the interviewer and the personal style of the interviewer. Each of these domains is influenced by factors including the purpose and context of the interview, training and experience of the interviewer and interviewee factors. Finally, there are four overarching aspects of the interview including ethics, consent and confidentiality, oversight of interview practice and other interested parties. It is beyond the scope of this article to provide a critique of the model; however, the four core domains of process, content, skills and style are used as a helpful way to organise recommendations in the current literature. This is outlined below.

Interview process

Davies (2019) defined the interview process as ‘the manner in which the interview is conducted’ (p. 6), which includes the interview structure, role of the interviewer and communication. Given the overlap between effective communication and the domains of interpersonal skills and personal style, this section focuses on reviewing the literature related to interview structure and interviewer role. Effective communication is considered in the section exploring interpersonal skills.

Of the limited literature available from the forensic field, there is general consensus on the structure of a risk assessment interview. Broadly, the phases include preparation, introductions and greetings, eliciting information and closure (Medoff & Kinscherff, 2006; Van Der Meer & Diekhuis, 2014), which broadly aligns with the investigative interviewing process outlined above. During the preparation phase, interviewers are advised to review all collateral material to identify information that may be missing, inconsistent or require corroboration and to prepare the challenge to any mistruths, minimisation or denials (Medoff & Kinscherff, 2006; Van Der Meer & Diekhuis, 2014). This preparation will also potentially help the interviewer with rapport building by identifying sensitive topic areas that may need careful exploration and empathy (McCuish, Hanniball, & Corrado, 2019).

The role of the interviewer is established at the beginning of the interview. Medoff and Kinscherff (2006) recommended that evaluators start with explaining the purpose of the interview and the limitations of confidentiality, which may include an explanation of what will occur if the young person discloses additional information that could lead to further charges. Van der Meer and Diekhuis (2014) also recommended stating ‘all questions asked during the interview are relevant for the assessment and that some topics will be personal and frustrating or difficult to speak about’ (p. 63). At this time, interviewers have an opportunity to establish rapport by carefully explaining these details to demonstrate that they are professional and trustworthy (Van Der Meer & Diekhuis, 2014).

Following a few minutes of building rapport, interviewers move into eliciting information relevant to the assessment. Before introducing the topic of concern, Medoff and Kinscherff (2006) suggested that interviewers begin with broad, non-leading, open-ended questions about general functioning to establish a common ground and observe cognitive capacity, language and interpersonal style. When all topics have been explored, the interviewer moves to closure (Van Der Meer & Diekhuis, 2014). Van der Meer and Diekhuis (2014) highlighted that an interview presents an opportunity for the interviewee to understand how they can avoid getting into more trouble with the police and how to keep themselves and others safe, so it may be appropriate to explore this more explicitly during the closure phase. Finally, McCuish et al. (2019) suggested asking the young person if there are any important topics that were not addressed and then ending the interview with brief informal questions to transition the conversation from sensitive to neutral topics, thereby assisting them with the move back into their daily routine.

Interview content

The content of the interview relates to the type of information collected and is linked to the purpose of the interview (Davies, 2019). In a forensic risk assessment, this is primarily driven by the information required for the risk assessment tool. With regards to exploring the index offence, Medoff and Kinscherff (2006) suggested that evaluators ask the young person to describe the events in the lead-up to the offence, then moving through the details of the offence and post-offence conduct. When information is omitted by the young person, the interviewer can introduce these details in order to observe the young person’s response and capacity to integrate the information into their narrative. Similarly, if the young person does not include the perspective of the victim in their narrative, the interviewer can ask them to offer an account of the events from the victim’s perspective.

Interpersonal skills

Interpersonal skills encompass the methods by which interviewers build trust, develop rapport and establish a working alliance with the interviewee in order to elicit accurate and detailed information (Davies, 2019). This topic is addressed in the existing literature with recommendations that interviewers adopt an ‘objective and neutral, but nevertheless friendly, understanding, and a non-judgemental and subtly supportive approach’ (Van Der Meer & Diekhuis, 2014, p. 61). While Medoff and Kinscherff (2006) encouraged interviewers to be ready to challenge inconsistencies, they warned against adopting an interrogative style that may trigger shame and limit the willingness of the young person to engage.

Strategies for building rapport with incarcerated young people are considered in depth by McCuish et al. (2019) who outlined insights gained from interviewing nearly 1,000 young people as part of the Incarcerated Serious and Violent Young Offender Study (ISVYOS). These authors suggested that interviewers should begin by explaining that they value hearing the perspective and opinion of the young person, and they hoped the young person would benefit from the experience. In the context of forensic risk assessment, interviewers may be able to emphasise their role in helping the young person to identify situations that may increase their risk of further trouble and that the interview can assist to develop a plan to reduce police contact and meet their personal goals. McCuish and colleagues then discussed how an emphasis on confidentiality can encourage trust and openness, particularly if interviewers emphasise the legal consequences for breaching confidentiality. While confidentiality is limited in a forensic risk assessment, it may be helpful to emphasise how some confidentiality is maintained – for example, the interviewer will not discuss the case outside their professional role.

While most interview protocols begin with a rapport building phase (for example, see Benia et al., 2015), McCuish et al. (2019) highlighted that it should be considered an ongoing process throughout the interview and that rapport building efforts should increase alongside the intrusiveness of questions. Strategies for achieving this may include empathising when the young person is frustrated with questions and providing opportunities for young people to be a valued source of information, such as providing explanations for jargon. Rapport-building is also key to overcoming reactance, which occurs when a young person may feel that their personal freedom or agency has been reduced or eliminated (Place & Meloy, 2018). In their research on reactance, Place and Meloy (2018) recommended that interviewers be mindful of language, avoid excessive eye contact, acknowledge feelings of resistance, provide choices where possible and seek gradual disclosure of information.

Personal style – questioning style

Personal style is the final area described by Davies (2019); this includes wording of questions, tone of voice, use of humour and pace and flow of the interview. Again, there is somewhat of an overlap with interpersonal skills, and the term ‘personal style’ may be misleading when an interviewer is following a protocol that stipulates how questions should be worded. For the purpose of this paper, the focus of this section explores guidance from the literature in relation to questioning style.

After forensic interviewers have gathered general background information, they need to raise the topic of concern with the young person, which may be the young person’s index offence or any threats they have made. Here, Shea’s (2017b) suggested approach from the CADE model included prefacing the questions with a normalising statement, such as ‘Sometimes when people are being bullied a lot at school, they think about hurting the bullies. Have you thought about hurting the bullies?’ or a shame-attenuating statement, such as ‘With all the stress you are experiencing at school, have you thought about getting back at them?’. Such an approach is consistent with strategies recommended in the investigative interview literature, whereby interviewers may introduce the topic of concern by stating known information and asking the interviewee to confirm whether this information is true (Powell & Snow, 2007b). In this case, the known information is that the individual has been bullied, and sometimes bullied people think about hurting the bullies. The interviewer then invites the interviewee to confirm whether this is true for them as well.

Once the topic has been raised, authors regularly recommend that interviewers use open-ended questions to elicit free-recall responses and explore the interviewees’ perspective (Medoff & Kinscherff, 2006; Van Der Meer & Diekhuis, 2014). When the interviewer has elicited as much information as possible with open-ended questions, they may need to ask specific questions to clarify details. Shea (2017a, 2017b) suggested that specific questions should include behaviour descriptors, for example, ‘when you kicked Joe, did it leave a mark or bruise?’ rather than ‘when you kicked Joe, was he hurt?’. Shea also recommended using gentle assumptions such as ‘what other ways have you thought of killing the bullies?’. This advice contrasts with the investigative interviewing literature as it is an example of a suggestive question. Finally, Shea recommended using symptom amplification when asking about frequency or severity of sensitive issues – for example, ‘how often do you fantasise about shooting kids at school −80%, 90% of the day?’ – although care should be taken with this strategy as it is another example of a suggestive question.

At times, an interviewer may need to question a young person about inconsistent information or challenge an answer they have provided. In addition to recommendations about adopting a non-interrogative, non-judgemental style (Medoff & Kinscherff, 2006; Van Der Meer & Diekhuis, 2014), McCuish et al. (2019) encouraged interviewers to consider whether it is necessary to challenge the young person in the first place. If a challenge is warranted, they discuss the difference between a low- and high- intensity challenge. A low-intensity challenge avoids suggesting that the young person was intentionally misleading the interviewer and uses the ‘Columbo’ curious/confused approach – for example, ‘I am a bit confused, you have said you only had one drink that night but in the police report a witness said you were slurring your words. Can you help me understand that a little more?’. A higher intensity challenge may be required when information is particularly sensitive, the honesty of the young person is relevant to the question or a low-intensity challenge has been unsuccessful. In this instance, McCuish et al. recommended that interviewers wait until the end of the interview when rapport may be strongest and any detrimental impact of damaging rapport is reduced.

Questioning young people involved in the youth justice system can be particularly challenging due to higher levels of speech and language issues (Snow & Powell, 2012). For this reason, McCuish et al. (2019) recommended, avoiding ‘why’ questions, which may be difficult for young people who lack insight into their internal emotional world or the vocabulary to express complex internal drivers. Where a young person answers ‘I don’t know’, McCuish and colleagues suggested that interviewers consider whether the young person understood the initial question and then try reframing the question with reference to its context. They provided the following example ‘We were talking about how you see yourself, and youth can sometimes talk about themselves as quiet or loud, happy or sad, and shy or outgoing. Do any of these words describe yourself?’ (McCuish et al., 2019, p. 41).

Overall, the current interviewer guidance from the forensic risk assessment literature appears sound on face value, yet there is little empirical research that explores the effectiveness of these recommendations. Further, the extent to which clinicians adhere to these principles when undertaking risk assessment interviews is unclear. If there is variability in the application of these principles, it may be owed to either gaps in interview skills or inconsistent perceptions regarding what constitutes best practice. Further investigation and discussion in this are warranted.

Summary of research

The research on investigative interviewing is exceptionally advanced when compared with other areas of forensic clinical practice, and the findings may inform interview protocols for forensic risk assessments. Research on suicide risk interviews is less well advanced, but techniques such as the CASE/CADE approach highlight that more consideration should be given to strategies for eliciting information in a risk assessment context. Currently, there are contradictions between the investigative interviewing literature and the CASE/CADE approach. In particular, the use of ‘gentle assumptions’ and ‘symptom amplification’ in the CASE/CADE approach would be considered inappropriate in the investigative interviewing literature due to the suggestive nature of these question types. Therefore, careful consideration must be given to how the research and literature from these fields may best inform forensic risk assessment interviews. Notwithstanding this, there are some clear principles that could be considered best practice in risk assessment interviews, which are outlined below.

Principles for best practice risk assessment interviews

Based on the literature outlined above, there are some key principles that can be derived to inform clinicians undertaking forensic risk assessment interviews with young people:

1. Preparation. Preparation is key to an effective interview. This should begin with considering which risk assessment tool or framework may be applied, followed by a thorough review of all available collateral information with consideration of what information is missing or needs further clarification/corroboration. Interviewers should also consider developing a list of key questions or referring to an interview schedule during the meeting to reduce errors of omission.

2. Introductions. Interviewers should ensure that they begin the interview by outlining the purpose of the interview and how the information will be used, including the limits of confidentiality. Interviewees should also be assured that only relevant information will be sought.

3. Rapport. Interviewers should avoid a confrontational or interrogative approach and instead aim to build trust with a non-judgemental, curious approach throughout the interview.

4. Questioning techniques. Interviewers should utilise non-leading, open-ended questions where possible to elicit and clarify information. Where possible, specific questions should be posed in behavioural terms such as ‘Did you punch, kick, hit or slap the victim?’ or ‘Did you have a weapon with you?’ rather than questions that may be interpreted subjectively such as ‘Did you hurt the victim?’ or ‘Did you threaten them?’.

These principles are consistent with the limited literature on interviewing for the purpose of a risk assessment (McCuish et al., 2019; Medoff & Kinscherff, 2006; Place & Meloy, 2018; Shea, 2017b; Van Der Meer & Diekhuis, 2014), as well as the literature in both the investigative interview and suicide assessment fields. While clinicians can confidently rely on these principles, there are some areas that require further consultation and research. These are outlined below:

1. Raising the topic of concern. Investigative interviewing literature discourages assumptions when raising the topic of concern, yet the CASE/CADE approach recommends using gentle assumptions for more sensitive topics. It may be possible to integrate the two recommendations by clearly stating the assumption and then asking the interviewee whether it is true; however, clinician’s need clear guidance on whether this is the best way to raise the topic of concern in the context of a forensic risk assessment.

2. Closure. There is minimal guidance in the literature regarding closure of interviews. Given that forensic risk assessment with youth is frequently focused on managing risk, closure may be an opportunity to discuss risk management strategies with a young person to keep them safe while awaiting the next step in the assessment. The appropriateness and viability of this needs to be considered in research.

3. Neurodevelopmental issues. There is a disproportionate rate of neurodevelopment issues in the forensic youth population, such as attention-deficit hyperactivity disorder (ADHD) and autism spectrum disorder (ASD). Further research is needed to understand the impact of interview techniques with these groups, particularly those prone to confabulation.

Conclusion

Assessment interviews with young people are a key component to a sound forensic risk assessment, yet there is little research exploring the most effective strategies and techniques to elicit accurate and detailed information. The field of investigative interviewing and suicide risk assessment may provide some guidance for forensic evaluators; however, there are some inconsistencies between these areas that must be resolved. Further research is needed to support the development of risk assessment interview strategies and protocols for this context. Once a developmentally appropriate protocol is developed, consideration should be given to adaptions for specific cohorts such as young people with ADHD or ASD.

Ethical standards

Declaration of conflicts of interest

Chelsea L. Leach has declared no conflicts of interest

Martine B. Powell has declared no conflicts of interest

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.