Police Interviewing Behaviors and Commercially Sexually Exploited Adolescents’ Reluctance

Abstract

Little is known about the relation between law enforcement interviewing behaviors and commercially sexually exploited children’s (CSEC) reluctance. This study examined the relation between officers’ use of maximization, (references to) expertise, minimization, and support and adolescent CSEC victims’ reluctance in a small sample of police interviews (n = 2,416 question-answer pairs across ten interviews). Twenty-six percent of officers’ utterances contained at least one interviewing tactic. When statements were paired with maximization, they were correlated with more reluctance than when they were not paired with an interviewing tactic. Contrary to predictions, support was also related to greater reluctance. Open-ended (recall) questions and statements were associated with greater reluctance than closed-ended (recognition) questions. The results highlight the importance of understanding the context in which interviewing strategies are employed when assessing the relation between interviewer behavior and interviewee reluctance.

Commercial sexual exploitation of children (CSEC) encompasses a range of crimes involving sexual abuse or exploitation of a minor, including sex trafficking in which minors have been recruited to engage in prostitution or pornography (Lavoie et al., 2019). The true prevalence of CSEC in the United States is difficult to determine due to victims' reluctance to disclose abuse, general underreporting of the crime, and difficulty in identifying and ascertaining victims and perpetrators (Farrell et al., 2019; Lavoie et al., 2019). Thus, when the opportunities arise to interview suspected victims, it is crucial to identify and employ interviewing techniques that minimize their reluctance in order to elicit productive and informative accounts (Lavoie et al., 2019).

Although a great deal has been written about interviewing victims of child sexual abuse (CSA), little attention has been paid to how CSEC victims are interviewed. Research on interviewing CSA victims has focused predominantly on pre-adolescent children (Lamb et al., 2018) whose abuse usually comes to the attention of authorities following the child’s self-disclosure (Lyon et al., 2020). CSEC victims, in contrast, tend to be adolescents (Klimley et al., 2018) whose victimization is usually discovered involuntarily through contact with law enforcement (Farrell & Pfeffer, 2014).

In the following paragraphs, we review why CSEC victims are a uniquely vulnerable victim population, and why they may be particularly reluctant during interviews with law enforcement. Then, we discuss interviewing strategies from both the interrogation literature (i.e., minimization and maximization) as well as the CSA interviewing literature (i.e., support) and how these strategies may be associated with CSEC reluctance.

CSEC Victim Reluctance

Research has noted that CSEC victims’ prior experiences with maltreatment and abuse, the nature of their trafficking experience and relationship with the perpetrator, as well as characteristics of trafficking victims’ encounters with law enforcement all lay a foundation for a high degree of disclosure reluctance (Lavoie et al., 2019). CSEC victims are an especially vulnerable population with complex motivations not to disclose abuse and a heightened distrust of adults and law enforcement (Lacks & Gordon, 2005), thus posing unique challenges for interviewers. Exploited adolescents may be seduced under the guise of a romantic relationship and want to protect the perpetrator (Reid, 2016). In other cases, a trafficker may isolate their victims and coerce them into compliance using fear, violence, and intimidation tactics (Anderson et al., 2014). Victims may become completely reliant on a trafficker for basic needs such as food or shelter (Anderson et al., 2014; Reid, 2016), which entraps the victims, and thus disclosing relevant details may leave them without resources and financial support.

CSEC victims may also be reluctant to cooperate with law enforcement out of fear of being labeled a “snitch,” which potentially poses grave consequences. Even when minors are interviewed as alleged victims receiving assurance of non-prosecution, they are still being asked to reveal information that will aid in the prosecution of the perpetrators (i.e., the “pimps”). Since victims may be arrested and questioned, exploiters might use violence to ensure that they do not snitch, in order to protect the “business” and keep their activities profitable (Williams & Prior, 2009). Research shows that many victims are beaten, experience miscarriages, or are even killed by exploiters because of snitching or suspected snitching (Tidball et al., 2016), further deterring other victims from cooperating with authorities in the future.

We are aware of only two studies exploring CSEC victim reluctance when questioned by law enforcement. Lindholm et al. (2015) analyzed 24 translated interview transcripts with exploited adolescents in Sweden to examine the relation between different question types on the informativeness and evasiveness of CSEC victims’ responses. Evasive responses were quite common (17%). Adolescents’ responses largely consisted of yes/no or confirmatory responses instead of spontaneously produced details and were more evasive in response to open-ended (recall) questions. Henderson and colleagues (2021) examined a sample of police interviews and courtroom testimony of adolescent CSEC victims in Los Angeles, California, and found that 26% of responses in police interviews contained at least one form of reluctance.

CSEC Victims as Suspects

Law enforcement may view any person engaged in prostitution as a perpetrator because CSEC cases necessitate that officers reclassify their perceptions regarding actions previously considered to be criminal behavior (Farrell et al., 2010). CSEC victims fall into a so-called “dual-status” category, in which they are simultaneously the victim and the “offender” because they take an active role (willingly or not) in the illegal activity. In fact, Halter (2010) found that officers viewed 40% of CSEC victims as offenders rather than victims, particularly if the victim was less cooperative during the investigation, failed to identify exploiters or disclose abuse, and had a prior criminal record.

Although minor decriminalization and Safe Harbor Laws have been passed in several American states to protect minors from being prosecuted for exploitation-related activities, including California (CA Penal Code 647, 2020), from which the current sample was drawn, a majority of jurisdictions still treat child victims of sexual exploitation as offenders (Barnert et al., 2016). Even with Safe Harbor Laws, CSEC victims may be arrested for other crimes and placed in juvenile detention facilities, and at any rate, they are likely to be under the jurisdiction of juvenile dependency courts and placed in foster or group homes. Moreover, research suggests there are few institutional resources that offer specialized guidelines for law enforcement officials’ responses to human trafficking (Farrell et al., 2010).

Police Interviewing Techniques: Maximization, Minimization, and Expertise

A large body of literature has explored interrogation tactics used by American law enforcement officers when questioning suspects and attempting to elicit confessions. This literature is potentially relevant for understanding law enforcement questioning of CSEC victims. First, as previously noted, law enforcement may often question suspected CSEC victims as if they are suspects. Second, even if they recognize CSEC victims as such, they will attempt to elicit incriminating information about the victims’ traffickers and are therefore likely to face reluctance, which may lead them to question suspected victims as they would a suspected co-conspirator or accomplice.

Research commonly classifies interrogation tactics under a two-pronged maximization and minimization approach and describes them as packages of techniques under which specific tactics may be categorized (Kelly et al., 2013). Maximization is a “hard-sell” technique that uses negative-incentive tactics to overcome resistance through intimidation, where the interviewer expresses a strong belief of guilt in order to pressure the suspect into making a confession and to impress on them the futility of denial (Feld, 2006, 2013; Kassin & McNall, 1991). Because maximization tactics intend to intimidate the suspect into confessing, commonly noted strategies include exaggerating the nature and seriousness of the offense (Feld 2006; 2013; King & Snook, 2009; Leo, 1996; Pearse & Gudjonsson, 1999; Soukara, 2009), identifying inconsistencies in the suspect’s story (Feld 2006; 2013; Leo 1996; Pearse & Gudjonsson, 1999; Soukara 2009), accusing the suspect of lying (Feld 2006; 2013; Pearse & Gudjonsson, 1999; Soukara, 2009), or asking the suspect to tell the truth (Feld 2006; 2013; Pearse & Gudjonsson, 1999). Other strategies include confronting the suspect with real evidence (Feld 2006; 2013; Leo, 1996; Pearse & Gudjonsson, 1999; Soukara et al., 2009) or false evidence of guilt (Feld 2006; Leo 1996; Pearse & Gudjonsson, 1999) in order to impress upon the suspect a perceived futility of denial. We anticipated that maximization would be associated with the highest rate of reluctance from CSEC adolescents, given the relation between “get tough” questioning and reactance (Snook et al., 2014)

Minimization, on the other hand, is a “soft-sell” technique that offers moral justifications or face-saving excuses in order to persuade the suspect to confess by lulling them into a false sense of security (Feld, 2006, 2013; Kassin & McNall, 1991; Pearse & Gudjonsson, 1999). The most frequently identified minimization tactics include officers expressing empathy toward the suspect and appealing to the importance of cooperation (Feld, 2006; 2013; King & Snook, 2009; Leo, 1996; Pearse & Gudjonsson, 1999; Soukara et al., 2009). Minimization tactics often offer moral justifications or psychological excuses (Feld 2013; King & Snook 2009; Leo, 1996; Pearse & Gudjonsson, 1999), or understate the facts or nature of the offense in order to reduce the suspect’s potential shame surrounding the offense (Feld, 2006; 2013, Leo, 1996; Pearse & Gudjonsson, 1999; Soukara et al., 2009). Last, officers also de-emphasize the purpose of the interrogation in order to lull the suspect into a false sense of security (Feld, 2006; King & Snook 2009; Leo, 1996). Because minimization in the context of interviewing CSEC victims entails encouraging the suspected victims to incriminate the perpetrator (or “snitch”), we anticipated that minimization would be associated with increased reluctance, though not to the same extent as maximization.

Previous studies have also noted that officers appeal to their own knowledge, expertise, or authority in order to underscore their position of power within the interrogation setting (Feld 2006; Leo 1996; Pearse & Gudjonsson, 1999). Based on our informal review of the transcripts, expertise could be used at times to minimize the victims’ responsibility or to empathize with the victim, while at other times to emphasize the officers’ own authority or knowledge; thus, it was considered separately from maximization and minimization tactics. Several studies also allude to a potential subset of officers’ expertise, namely officers suggesting a “scenario.” While some studies proposed that the tactic is used to suggest possible scenarios of what the suspect did (Pearse & Gudjonsson, 1999; Soukara et al., 2009), other studies classified it as a minimization tactic used for neutralization and reduction of guilt (Feld 2006; 2013). Because classifications of the scenario tactic were inconsistent across previous measures (Feld 2006; 2013; King & Snook, 2009; Pearse & Gudjonsson, 1999; Soukara et al., 2009), it was considered as a type of officer expertise, separately from minimization and maximization. We anticipated that expertise would also be associated with increased reluctance.

In the literature on confessions, maximization and minimization are designed to have suspects incriminate themselves by both maximizing the futility of maintaining innocence and minimizing the perceived risk and consequences of providing a confession (Hirsch, 2014; Kassin & Gudjonsson, 2004). This is accomplished subtly through pragmatic implication, with maximization tactics serving as a proxy for implied threat and minimization tactics serving as a proxy for implied leniency (Feld, 2006; Kassin & McNall, 1991; Redlich et al., 2019). Maximization and minimization tactics work by changing suspects’ perceptions of the consequences that will ensue depending on the course of the interview, implying positive outcomes following a confession and negative outcomes following continued denial (Leo, 2020).

Maximization and minimization techniques have been widely criticized for significantly increasing the risk of false confessions, especially among youth (Kassin et al., 2010; Kassin & Gudjonsson, 2004; Spierer, 2017). Nevertheless, field research shows that officers widely use the same tactics when interrogating juveniles as they would with adults (Feld 2006; 2013), and self-reports find that officers essentially make no distinction between adolescents and adults when endorsing interrogation techniques (Cleary & Warner, 2016; Meyer & Reppucci, 2007).

Supportive Questioning

In contrast to the limited research on questioning CSEC victims, an enormous amount of research has examined how to effectively interview CSA victims (e.g., Blasbalg et al., 2018; Hershkowitz et al., 2006; 2015; Lewy et al., 2015; Melinder & Gilstrap, 2009). When children disclosing sexual abuse are reluctant, guidelines recommend that interviewers address reluctance through supportive techniques rather than by continuing to ask abuse-related questions (Blasbalg et al., 2018; Hershkowitz et al., 2006; 2017). Unless they are specially trained, forensic interviewers confronted with reluctance tend to decrease their use of supportiveness and appropriate questioning strategies, which is subsequently associated with increased reluctance (Hershkowitz et al., 2006; Lewy et al., 2015; Melinder & Gilstrap, 2009). Conversely, increased support, such as building rapport, acknowledging the child’s feelings, or using positive reinforcements such as thanking the child are associated with decreased reluctance (Hershkowitz et al., 2015). We tentatively anticipated that support would be associated with less reluctance in our sample as well.

Law enforcement officers questioning suspected CSEC victims may attempt to use some supportive techniques, even when they perceive the CSEC victims as akin to suspects. In the United States, the most popular and frequently implemented training model for suspect interviewing and interrogation is the Reid Technique (Meyer & Reppucci, 2007; Gudjonsson & Pearse, 2011), which recommends that officers attempt to build rapport with suspects (Vallano & Compo, 2015) before they engage in maximization and minimization. However, Reid interviewing and interrogation manuals “offer few concrete techniques to aid interviewers when building rapport with cooperative witnesses, and even fewer techniques when building rapport with criminal suspects” (Vallano & Compo, 2015, p. 89). Similarly to the Reid Technique, the PEACE (Preparation and Planning, Engage and Explain, Account, Closure, and Evaluation) interviewing method, which is more commonly used in Europe, also recommends that interrogators attempt to build rapport with suspects; however, unlike the Reid Technique, the PEACE method advocates that interviewers act as neutral fact-finders (Miller et al., 2018; Snook et al., 2014).

Survey evidence has found that police officers recognize the importance of building rapport when questioning crime victims (Compo et al., 2012). However, it is by no means guaranteed that law enforcement will be supportive when questioning CSEC victims. Crime victims often experience negative interactions with police officers (Aviv & Weisburd, 2016), reporting that officers questioned them in an insensitive, mistrustful, or dominant manner (Holmberg, 2004), and rape victims specifically report that law enforcement often questioned them with little displays of empathy (Logan et al., 2005; Maier, 2008; Patterson, 2011; Webster et al., 2020).

The Present Study

This study examined the relation between police interviewing tactics and adolescents’ reluctance in a small sample of commercially sexually exploited adolescents questioned in Los Angeles County, California. A novel measure of police interviewing tactics was developed based on research examining adult and juvenile interrogations (Feld, 2006; 2013; King & Snook, 2009; Leo, 1996; Pearse & Gudjonsson, 1999; Soukara, 2009). The coding scheme includes both interrogation techniques typically used with suspects (e.g., minimization and maximization) and interviewing techniques typically used with victims (e.g., support; Blasbalg et al., 2019, Lamb et al., 2018) to explore how police officers question CSEC victims and how these techniques may influence expressions of reluctance. Reluctance was measured by a coding scheme capturing 16 tactics by which adolescents express reluctance (Henderson et al., 2021).

We first examined the extent to which law enforcement used the different techniques during these interviews, and whether use of the techniques was associated with reluctant responses. We predicted that officers’ use of maximization, expertise, or minimization would be associated with adolescents’ reluctant responses, compared to officers’ utterances containing no tactic. We also predicted that maximization and officer expertise would be more strongly associated with reluctance compared to minimization. Consistent with child interviewing research, we hypothesized that the use of supportive techniques would be associated with a decrease in reluctance. Last, derived from findings in Lindholm et al. (2015), we hypothesized that open-ended questions would be associated with an increase in reluctance compared to closed-ended questions.

Method

Sample

Pursuant to the California Public Records Act (California Government Code 6250, 2010), we obtained trial transcripts of criminal cases involving charges of pimping and pandering of a minor filed under Sections 236 and 266 of the California Penal Code (i.e., human trafficking, pimping/soliciting of a minor). The current sample consists of adolescent CSEC victims who were associated with the same trafficker. The case was tried by a jury in Los Angeles County in 2017, and the defendant was found guilty on 15 counts, including human trafficking of a minor by force or fear (Pen. Code §236.1, subd. (c)(1)); pandering by procuring a minor under the age of 16 to be a prostitute (§266i, subd. (b)(2)); pimping a minor under the age of 16 (§266h, subd. (b)(2)); and assault by means of force likely to produce great bodily injury (§254, subd. (a)(4)). In an appeal in 2018, the convictions were upheld.

We examined police interview transcripts (n = 2,416 question-answer pairs in ten interviews) with eight female CSEC victims who were between the ages of 15 and 17 (M_age = 16.3) at the time of the interview. Prior to coding and analysis, all interview transcripts were anonymized to protect victims’ identities.

Because victims’ individual experiences with both the criminal justice system and the defendant may influence their reluctance, police interview and trial transcripts were examined for information regarding victims’ interactions with both, and particularly for factors that may increase or decrease reluctance to cooperate. In the current sample, one of the ten interviews (Victim E) began with an assurance that the officers were “not here because [they’re] looking at [the victim] for anything criminal.” The other interviews began with brief introductions, primarily consisting of requests for identifying information, followed by direct questioning (i.e., “Let me ask you, do you know a pimp who goes by the name X?”; [Victim A]) or accusations (i.e., “I heard you was with a guy X”; [Victim F]). All of the victims had been previously arrested, either for probation violations, open warrants, and/or solicitation of prostitution (prior to Safe Harbor Laws being enacted), but none of the victims were prosecuted for their actions in the current case. Additionally, all victims had been placed in juvenile halls or group living homes, where recruiting for the defendant often took place. Victims E and H were arrested immediately before their interviews occurred, and Victims H and G were both interviewed within juvenile hall.

Victim F, who exhibited the highest rate of reluctance (51.4%; see Table 2) previously testified in the trial of another trafficker per the request of the interviewing officers and was subsequently called a “snitch.” In her current interview, she expressed her unwillingness to be involved in the trial and her concern over again being labeled a “snitch” in her community again. Victims D, E, and H also expressed concern during the interview about being a “snitch.” Victim E described being raped by a man who claimed to be an undercover cop and said in the interview that “it took [her] a while to even trust police officers again after that.”

Table 2.

Prevalence of Each Tactic in Interview by Interviewer and Victim

Interviewer	Victim	n	Maximization	Expertise	Minimization	Support	Tactic Total	Reluctance
Z		497	6%	9%	5%	4%	21%	21%
	A	96	20%	13%	6%	3%	32%	34%
	B	143	2%	11%	4%	6%	20%	32%
	C	258	4%	7%	5%	2%	17%	10%
Z & Y		905	20%	13%	12%	11%	43%	42%
	D	420	27%	16%	16%	6%	50%	44%
	E	120	3%	16%	12%	16%	36%	16%
	F	208	18%	7%	5%	14%	36%	51%
	G	157	14%	8%	10%	17%	39%	43%
X	H	313	4%	5%	1%	7%	16%	14%
W & V	H	701	10%	8%	7%	8%	14%	13%

Note. Victim C’s interview was conducted over 2 sessions with the same officer. The tactic subcategory percentages do not add up to the tactic total percentage because of overlap among the individual subcategories.

Regarding notable interactions with the defendant, Victim C, who exhibited the lowest rate of reluctance in the sample (10%), worked for the defendant for the longest period of time, but cooperated with law enforcement, disclosing that the defendant beat her while she was pregnant. Victim H unwillingly worked for the defendant for a short amount of time and described being raped by him and exhibited a relatively low rate of reluctance (13% -14% across 2 interviews). Conversely, Victim D claimed that she willingly worked for the defendant and considered them to be a “team,” and thus refused to provide the name of her trafficker and exhibited a high rate of reluctance (44%). These varying circumstances surrounding the victims’ interactions with the legal system and the defendant may partially account for the range of reluctance seen across interviews with different victims.

Interviews were conducted by four sets of officers consisting of five interviewers total (see Table 2). Four of the interviews were conducted by Officer Z (n = 497 question-answer pairs), four by Officers Z and Y (n = 905 question-answer pairs), one by Officer X (n = 313 question-answer pairs), and one by Officers W and V (n = 701question-answer pairs). We examined the officers’ testimonies in the subsequent court transcripts for demographic information and determined that Officers Z and Y, who conducted a majority of the interviews, were partners who had been in the police force for 17 years, working as vice detectives, and had extensive experience working in the Human Trafficking Task Force. We were unable to locate descriptive information about the other three officers, but they do not work in the same police department as Officers Z and Y.

Development and Coding of Police Officers’ Utterances

An utterance was defined as each person’s individual speaking turn within a question-answer pair, with an officer’s question constituting one utterance and the victim’s response another. The coding scheme was created by adapting and expanding upon the observations of maximizing and minimizing tactics in previous work (Feld, 2006; 2013; King & Snook, 2009; Leo, 1996; Pearse & Gudjonsson, 1999; Soukara et al., 2009). These coding measures were chosen specifically because the variables were derived from examining actual police interrogations, transcripts, or audiotapes. In addition to “maximization” and “minimization” categories, informed by the previous literature, a novel “expertise” category was created, including the subcategories of “explicit expertise,” “scenario,” and “terms” (see Table 1). Although explicit expertise (e.g., emphasizing an officer's own experience and knowledge) was categorized under maximization by previous work, we considered it separately because it is an authoritative tactic which is not as directly confrontational and instead serves to subtly emphasize the unequal power dynamic in the interview setting. A novel “terms” subcategory was included, referring to when officers demonstrated their expertise by using street slang specific to prostitution (e.g., “the game”) before the adolescent used the term. This is highly specific to commercial sexual exploitation, as the use of street slang allows officers to demonstrate specific knowledge surrounding the trafficking process. Interviewing tactics were not considered mutually exclusive, so an officer’s utterance could contain multiple interviewing tactics (e.g., maximization and expertise).

Table 1.

Police Interviewing Behavior Coding Scheme

Tactic	Definition	Example
Maximization
Truth	Interviewer explicitly asks or requests that the victim tells them the truth	-Tell me the truth. -Will you be honest with me today?
Accuse of Lying	Interviewer accuses the victim of lying to them either explicitly (i.e., using the words lying/not being honest/truth), OR by directly referencing the victim’s statement, OR contradicting the victim's answer.	-Stop lying, you’re not telling me the truth. -But you want to sit here and tell us you don’t know what he looks like.
Deny/Reject	Interviewer denies or rejects assertions made by the victim	-A: I don’t know him. Q: You know exactly who I’m talking about.
Maximization	Interviewer exaggerates or emphasizes the nature and seriousness of the offense or the interview (including blaming language—agent of the action is the victim; action committed is negative)	-Prostitution is a felony crime. You can go to jail for prostitution. -You chose to give him your money.
Confrontation	Interviewer confronts the victim with an accusation of guilt either by introducing direct evidence, presenting hypothetical evidence, or making an accusation without presenting evidence	-Those girls are your hoe partners. That’s what they said. -If one walked up to you on the street, you’d know who that pimp is right away, right? -So you did pay his pockets.
Inconsistencies	Interviewer explicitly identifies inconsistencies or contradictions in the victim’s story (you said x, now you’re saying y) OR points out an inconsistency directly following the victim’s statement	-You said 50s or 60s. Now you're in the 40s. -You said now I kick it with him. You said one time, but you kick it with him all the time.
Expertise
Terms	Interviewer uses street terms/slang that the victim has not previously introduced (not echoing the child, or any subsequent use) *Excluding vernacular and non-prostitution related street slang	E.g., Automatic, Back Page, Choose Up, Division, Folks, the Game, Gorilla/Romeo Pimp, Hoe Partner, John, the Lifestyle (“the Life”), Out of Pocket, Paying Pockets, Pimp Partners, Renegade, the Street, Trap, Wifey
Scenario	Interviewer proposes a possible scenario or theme based on available information, summarizing information and/or building a storyline. A scenario can be case-specific or built off an officer's general knowledge *Must include at least a) 1 assertion followed by a question, OR b) 2+ assertions that may or may not be followed by a question)	-You guys are working together. You post a back page together. And you said you paid your folks right… Okay now so if you paid your folks and she’s paying her folks, that means your folks have to be pimp partners, right?
Explicit Expertise	Interviewer draws attention to their personal or general expertise (excluding knowing names of places), emphasizing their authority and knowledge	-I’m a square but I understand this game. -I got ears. -He’s one of the more well-known pimps.
Minimization
Empathy	Interviewer expresses empathy towards the victim; aimed at expressing “I know how you feel” to the victim	-I hear you. -I understand you think he treated you well.
Cooperation	Interviewer requests cooperation OR suggests that cooperation is in the victim’s self-interest	-Talk to me, let’s have this conversation and be done. -Don’t you think that this person that did this to you, you should be able to show him that what he did to you was wrong?
Justifications
External	Interviewer offers statements that minimize the victim’s responsibility by targeting external forces (i.e., third parties: pimp, circumstances) *The actor of the verb is not the victim	-He treated you the way he needed to treat you so that you would stay. -He’s exploited you for what you are.
Internal	Interviewer offers justifications or psychological excuses aimed at reducing guilt or feelings of shame regarding the trafficking OR protecting the perpetrator *The subject is the actor of the verb	-You had feelings for him. -So that’s why you want to protect him, because in your mind he wasn’t bad for you.
Minimizing	Interviewer minimizes the seriousness/nature of either the incident, the interview or both	-Did you pay his pockets? If you did you did, it’s not that big of a deal. -We come, we talk to you, you tell your story to us. -You’re not going to get him in trouble.
Support Variables
Establishing Rapport	Including welcoming the victim or apologizing to them, small gestures of goodwill	-How are you today? -[Victim], do we go way back? -Sorry about that.
Emotion	Interviewer acknowledges or asks about the victim’s emotions or attitude	-We asked you some tough questions today, right? Probably got you a little frustrated. -I know this is boring to you, and it’s not something you want to talk about.
Thanks/Appreciation	Interviewer reinforces victim’s positive behavior during the interview by thanking them	-Thank you very much for giving us your time. -I appreciate you talking to me.
Motivation	Interviewer explains the motivation behind the interview or the interviewer’s actions	-We truly want to listen to what you have to say. -We just want to know whether you know somebody in this photograph. -We have an issue with you being exploited as a minor.
Victim as Expert	Interviewer emphasizes their ignorance, and that the victim is the source of the knowledge/information	-School me up on the game. -Correct me if I’m wrong about this game. -What does that mean?
Victim’s Discretion	Interviewer emphasizes that the victim has the ability to choose whether to cooperate and/or what to disclose	-So, if it’s okay with you, we’d like to know your story. -You’re not obligated to pick anyone out. -Okay, can I tell you what I heard?
Reassurance	Interviewer reassures the victim, discussing the anticipated consequences of the interview or expressed fears	-We’re not looking at you for anything criminal, so you can rest assured on that. -You ain’t no snitch. -I won’t lie to you.

In addition to interrogation-based tactics, the coding scheme identified techniques which are considered to be “supportive,” based on child interviewing research (Blasbalg et al., 2019; Lamb et al., 2018). It is important to note that because law enforcement officers often fail to adequately build and maintain rapport (Abbe & Brandon, 2014), and because some “support” strategies may overlap with police interviewing tactics (e.g., “I’m just trying to understand this. [support: explaining motivation] You guys are working together. You post [an advertisement] together. [maximization: confront with evidence] And you said you paid your folks, right?”), the supportive strategies appearing here are somewhat different from the supportive strategies recommended in child interviewing literature. We provide a comparison of the supportive categories used here with a common measure of support in the child interviewing literature (Lamb et al., 2018) in Table A1 of the Supplemental Material.

Officers’ utterances were first coded according to the subtypes of the four tactic categories (e.g., “see, you told me you would tell me the truth”; [maximization: accuse of lying]), which were not mutually exclusive, and then dichotomized for descriptive analysis (e.g., maximization: presence/absence). In order to make tactics mutually exclusive for statistical models so that they could be included as one fixed effect, for instances in which tactics co-occurred, they were categorized from most to least likely to elicit reluctance (i.e., maximization, expertise, minimization, support). This means that some maximization tactics may have included other interviewing tactics, but support strategies did not contain any other interviewing tactics. Finally, officers’ utterances were also coded for utterance type: open-ended question (wh-questions and invitations), closed-ended question (yes/no, forced-choice, tag, statement, and negative-term questions), or statement (no question was asked) (Ahern et al., 2018).

Reluctance in victims’ utterances was coded according to Henderson and colleagues (2021), which identified 16 ways in which CSEC adolescents can express reluctance, including refusals to answer, denying assertions, or challenging the question. In order to capture the more sophisticated ways in which adolescent victims convey resistance, both subtle and overt expressions of reluctance were noted. Reluctance and cooperation were not considered mutually exclusive, highlighting the fact that adolescents can express reluctance while also being informative (e.g., “Yes, he was there. Why do you keep asking me about this?” [reluctance: complaining about repetitive questioning]). Reluctance was coded by utterance, and the 16 categories were then collapsed into a binary variable (present/absent) during analyses (see Table A2 of the Supplemental Material).

Inter-rater reliability for utterance type was high (K = 0.97). Three independent raters coded 100% of the sample according to the reluctance and interviewing tactics coding schemes. Reliability was calculated among the three coders throughout development of the coding scheme and during final coding of transcripts. Cohen’s Kappa may indicate poor reliability in studies with a low prevalence of individual codes, reflecting the nature of the sample rather than poor inter-rater agreement. Thus, researchers recommend that kappa interpretation should be qualified by prevalence bias (Byrt et al., 1993). Prevalence-adjusted bias-adjusted kappa (PABAK) exceeded 0.90 for all police interviewing variables and percent agreement (i.e., the absolute difference between the agreements on the positive classification and the agreements on the negative classification, divided by the sum of agreements and disagreements) exceeded 95% for all variables. For adolescent reluctance, PABAK exceeded 0.90 for all variables and percent agreement exceeded 94% for all variables. All discrepancies were discussed and resolved with 100% mutual agreement.

Analysis Plan

Descriptive analyses examined the prevalence of interviewing tactics, utterance type, and CSEC victims’ reluctance in the police interviews. Next, analyses examined the relationship between interviewer tactics, utterance type, and reluctance at the utterance level within each question-answer pair. One Generalized Linear Mixed Model (GLMM) explored whether interviewers' use of tactics (five levels: maximization, expertise, minimization, and support, no tactic [baseline]; comparing each tactic type against the baseline), utterance type (three levels: open-ended question, closed-ended question [baseline], statement); comparing each utterance type against the baseline), and an interaction between tactic and utterance type were associated with CSEC victims’ subsequent reluctance (two levels: present/absent). In all models, the victims were included as a random effect in order to control for the repeated nature of the questioning and specific individual response proclivities. For example:

Reluctance ~ Tactic + Utterance Type + Tactic x Utterance Type + (1 ∣ Victim)

Then, four individual GLMMs (one for maximization, minimization, expertise, and support) were conducted to explore whether interviewers’ subsequent use of specific tactics was associated with adolescents’ reluctance. For example, a model examined whether officers were more likely to respond with maximization after either a reluctant or non-reluctant response:

Maximization ~ Reluctance + (1 ∣ Victim)

Lastly, a GLMM was conducted that was identical to the first model but also included a variable for reluctance in the adolescents’ previous utterance (present/absent) in order to control for previous reluctance, as well as an interaction between utterance type and previous reluctance. For example:

Reluctance ~ Tactic + Utterance Type + Tactic x Utterance Type + Previous reluctance + Previous reluctance x Utterance Type + (1 ∣ Victim)

Models were cross-validated regarding all fixed, random, and interaction effects in order to identify the best fit model. When a random effect for the interviewing police officer was included in the model, this did not improve model fit. The best-fit models are described above, and the results are reported below. All statistical output for models can be found in tables A3-A5 of the Supplemental Material. Model-fitting was computed using the anova function in the R stats package (R Core Team, 2013). Analyses were performed using the glmer function in the R package lme4 with the bobyqa optimizer (Bates et al., 2015). Pairwise comparisons with Tukey HSD and adjusted means were computed using the emmeans function in the R package emmeans (Lenth et al., 2020). GLMMs combine the properties of linear mixed models (which incorporate random effects) and generalized linear models, and are preferable to traditional analysis of variance (ANOVA) models because they have fewer assumptions, handle response variables from different distributions (e.g., binary), and maximize power while simultaneously estimating between-subject variance (Bates et al., 2015). The results from the best-fit models are reported below, accompanied by the unstandardized fixed effect estimates (β), standard errors of the estimates (SE), and estimates of significance (Z and p values).

Results

Officers’ Use of Interviewing Tactics

Interview length varied, ranging from 96 to 705 question-answer pairs (M = 268 question answer pairs). Of the police officers’ utterances, 73% were questions (48% closed-ended (recognition) and 25% open-ended (recall)), and 27% were statements. The proportion of interrogation tactics used in each interview varied considerably, ranging from officers using tactics in 14% of their total utterances to 50% of their total utterances (M = 26%, 63 utterances per interview; see Table 2). There was also considerable variation in the amount of reluctance that victims exhibited, ranging from 10% to 50% of their responses (M = 26%, 62 utterances per interview).

Maximization was employed most frequently, occurring in 10% of all officers’ utterances (M = 23 utterances per interview), with officers most often confronting the victim with evidence (42% of total maximization, M = 9.9 utterances per interview) or accusing the victim of lying (31% of total maximization, M = 7.3 utterances per interview). Officers alluded to their expertise in 8% of their utterances (M = 18.8 utterances per interview). Minimization tactics were present in 7% of officer’s utterances (M = 17.5 utterances per interview), most often through offering justifications (56% of total minimization, M = 9.8 utterances per interview). Support was present in 8% of the officers' utterances (M = 19.4 utterances per interview) (see Table 3).

Table 3.

Prevalence of Police Interviewing Tactics

Tactic	n	%
Maximization	235	10%
Confrontation	99	42%
Accuse of Lying	73	31%
Deny	39	17%
Maximize	36	15%
Truth	19	8%
Inconsistent	2	1%
Expertise	188	8%
Officer Expertise	83	44%
Terms	66	35%
Scenario	56	30%
Minimization	175	7%
Justifications	98	56%
Minimize	42	24%
Cooperation	36	21%
Empathy	28	16%
Support	194	8%
Motivation	65	34%
Reassurance	50	26%
Rapport	38	20%
Victim’s Discretion	29	15%
Victim as Expert	21	11%
Appreciation	19	10%
Emotion	15	8%
TOTAL	633	26%

Note. The tactic total is less than the sum of n’s because 19.5% of police interviewing tactics overlapped, containing multiple tactics within the same utterance.

Relation between Tactic and Subsequent Reluctance

Because adolescents’ interruptions of officers’ utterances could prevent officers from using a tactic and/or asking a question, we excluded interruptions to officers’ utterances that did not request information or contain a tactic (n = 67). A GLMM was conducted to examine whether police tactic (five levels; maximization, expertise, minimization, support, no tactic [baseline]), utterance type (three levels: open-ended question, closed-ended question [baseline], statement), and an interaction between police tactic and utterance type were associated with adolescents’ subsequent reluctance (reluctant/non-reluctant).

Controlling for the random effect of victim (Variance = 0.87, SD = 0.93), the model found a main effect of support (B = 1.01, SE = 0.43, Z = 2.32, p = .02), such that utterances with support (27%, SE = 7.7) were associated with significantly more reluctance than utterances with no tactic (16%, SE = 4.2). A main effect of utterance type (open-ended question vs. closed-ended question: B = 0.66, SE = 0.16, Z = 4.02, p < .001; statement vs. closed-ended question: B = 0.47, SE = 0.19, Z = 2.45, p = .01), found that open-ended questions (27%, SE = 7.3) and statements (32%, SE = 7.3) were linked to more reluctance than closed-ended questions (15%, SE = 4.5); but there was no significant difference between open-ended questions and statements. The model also found an interaction between maximization and utterance type (B = 1.35, SE = 0.46, Z = 2.95, p = .003), which showed that when statements were paired with maximization, they were significantly related to more reluctance (48%, SE = 8.9) than when they were not paired with an interviewing tactic (17%, SE = 4.9).

Relation between Reluctance and Subsequent Tactic

Next, we explored how officers responded to the CSEC victims’ reluctance. A series of GLMMs examined whether reluctance was associated with officers’ subsequent use of tactics. Victims’ reluctant responses were positively associated with officers’ subsequent use of support (11%, SE = 2.3, vs. 7%, SE = 1.3; B = 0.61, SE = 0.17, Z = 3.60, p < .001; Random effect: Variance = 0.28, SD = 0.53), maximization (10%, SE = 3.2, vs. 6%, SE = 2.0; B = 0.52, SE = 0.15, Z = 3.35, p < .001; Random effect: Variance = 0.98, SD = 0.99), and expertise (11%, SE = 2.5, vs. 7%, SE = 1.4; B = 0.56, SE = 0.17, Z = 3.32, p < .001; Random effect: Variance = 0.37, SD = 0.61) but not with officers’ subsequent use of minimization (7%, SE = 2.0, vs. 5%, SE = 1.4).

Relation between Tactic and Subsequent Reluctance, Controlling for Prior Reluctance

Because support and maximization were both associated with reluctance and with officers’ responses to reluctance, we further examined the relation between tactics and reluctance, controlling for prior reluctance. Figure 1 shows the prevalence of reluctance associated with each tactic, controlling for prior reluctance. The model examined whether reluctance in the prior response (reluctant/non-reluctant), police tactic (five levels: maximization, expertise, minimization, support, no tactic [baseline]), utterance type (three levels: open-ended question, closed-ended question [baseline], statement), and interactions between previous reluctance and utterance type as well as police tactic and utterance type were associated with subsequent reluctance (reluctant/non-reluctant). The interaction between police tactic and prior reluctance did not significantly improve the fit of the model, so it was not included.

Figure 1. Reluctance Elicited by Tactic and Utterance Type.

Note. Error bars correspond to +/− 1 standard errors.

Controlling for the random effects of victim (Variance = 0.72, SD = 0.85), the model found a main effect of prior reluctance (B = 0.86, SE = 0.21, Z = 3.98, p < .001), such that when the previous utterance contained reluctance, the subsequent response was significantly more likely to contain reluctance (32%, SE = 6.8) compared to when the previous utterance did not contain reluctance (21%, SE = 5.0). A main effect for support (B = 0.91, SE = 0.45, Z = 2.01, p = .045), showed that support (29%, SE = 7.6) was significantly associated with more reluctance compared to utterances with no tactic (17%, SE = 4.2). A main effect for utterance type (open-ended vs. closed-ended: B = 0.79, SE = 0.19, Z= 4.23, p < .001; statement vs. closed-ended: B = 0.50, SE = 0.22, Z = 2.29, p = .02) found that open-ended questions (29%, SE = 7.3) and statements (34%, SE = 6.9) were linked to more reluctance than closed-ended questions (17%, SE = 4.7), but there was no significant difference between open-ended questions and statements. Finally, an interaction between maximization and utterance type (B = 1.43, SE = 0.47, Z = 3.07, p = .002), showed that when statements were paired with maximization, they were significantly correlated with more reluctance (50%, SE = 8.4) than when they were not paired with an interviewing tactic (18%, SE = 4.8)

Exploring the Positive Relation between Support and Reluctance

Due to the unanticipated finding that support was significantly associated with reluctance, a post-hoc exploratory analyses was conducted to examine whether the support finding was a product of changing base rates of reluctance and supportiveness across the interviews, such that interviewers typically started out being supportive and the adolescent victims typically started out reluctant. However, we did not find any indication that this was true. Reluctance exhibited little variance, such that 21% of the responses were reluctant in the first half of the interviews and 25% in the second half. Support increased slightly from the first half (4%) to the second half (6%).

Discussion

This study examined police officers’ interviewing behaviors and their relation to adolescent CSEC victims’ subsequent reluctance. We coded for the presence or absence of 16 types of reluctance (Henderson et al., 2021), and based on our review of police questioning behavior with suspects and victims, identified four interviewing techniques: maximization, (reference to) expertise, minimization, and support. We found that about a fourth of officers’ utterances contained one or more interviewing techniques. About half of officers’ utterances were closed-ended (recognition) questions, a fourth were open-ended (recall) questions, and about a fourth were statements. We also found that prior reluctance was correlated with subsequent reluctance and was related to increased use of some interviewing techniques, so prior reluctance was included in a subsequent model.

When combined with statements, maximization was associated with higher rates of reluctance, and regardless of utterance type, support was associated with greater reluctance. Although we predicted the positive relation between maximization and reluctance, the positive relation between support and reluctance was contrary to our predictions. With respect to utterance type we found that statements and open-ended questions were associated with greater reluctance than closed-ended questions. In the following paragraphs, we discuss possible reasons for the findings regarding maximization, support, and utterance type as well as implications of the results for further research and policy regarding interviewing CSEC adolescents.

Maximization

The positive relation between maximization and reluctance reflected the fact that when officers made statements, reluctance followed maximization 50% of the time compared to 17% of the time when no tactic was used. Given that CSEC victims are a population known to be distrustful of law enforcement and adults in general (Lacks & Gordon, 2005) and are highly reluctant to disclose details of the abuse (Lavoie et al., 2019), it is unsurprising that the accusatory style of maximization was associated with greater reluctance. Rather than being impartial, maximization is often controlling and dogmatic, and uses imperatives (e.g., “Just be honest with us,”), allegations (e.g., “So you did work for him,”), derision towards other perspectives (e.g., challenging the interviewee’s response by presenting others’ statements as evidence), or threatening statements with warnings (e.g., “You better not be lying to us”), which may be perceived as combative, thus provoking reactance or anger (Moss, 2016).

Support

Unlike findings in the child interviewing literature, which tend to find that supportive techniques are related to decreased reluctance in sexually abused children (Hershkowitz et. al, 2015), the officers’ use of tactics considered to be supportive in the current study were associated with an increase in CSEC victims’ reluctant responses. Overall, the use of support was followed by reluctance 29% of the time compared to 17% when no tactic was used.

There are a number of possible explanations for this finding. First, given their age and attitudes toward their questioners, CSEC adolescents may react differently to support than younger victims of child sexual abuse. Second, support in the CSEC interviews studied here might be qualitatively different from support in interviews with sexually abused children. For example, one form of support is emphasizing that the victim is the source of the information. Forensic interviewers may convey this message by expressing, “I am asking you these questions because I was not there and do not know what has happened to you,” the officers conveyed that same notion by asking the victims to “school them up on how the game works.” Another example of this kind of support is giving the interviewee a choice to cooperate. In a forensic interview this could be expressed as “It's your choice whether to tell me or not, and it is my job to let you choose, (Lamb et al., 2018)” whereas in the police interviews, officers sometimes asked the victims “if it’s okay with you, we’d like to know your story” or “do you want to see the pictures." However, an important distinction is that officers rarely took “no” for an answer and continued to press the victim to choose, thus invalidating the notion that the victim genuinely had any discretion to cooperate (e.g., Officer: “Okay. Can I put [the initials down for you];” Victim: “No, you better not;” Officer: “I’ll put [initials] and I’ll just say I put [initials] for you”). Thus, supportive tactics may have been perceived as disingenuous by the victims.

Third, combining support with maximization and other techniques frequently used with suspects may reverse support’s positive effect. If officers are using a high baseline of maximization tactics such as accusing the victim of lying or rejecting claims made by the victim, victims may view support as a manipulative ploy designed to placate the victim and elicit information. For example, reluctance was common in response to officers’ offering reassurance (e.g., Officer: “I’m not going to have you do anything;” Victim: “You always say that" [reluctance: questioning interviewer’s motivation]).

Statements and Open-ended Questions

Statements and open-ended (recall) questions were significantly associated with increased reluctance compared to closed-ended (recognition) questions. While closed-ended questions were associated with reluctance in 17% of the interactions, open-ended questions and statements were associated with reluctance 29-34% of the time. This is similar to Lindholm and colleague’s (2015) finding that open-ended questions were linked to more evasive responses (17%) than closed-ended questions (7%). This may be due to the fact that closed-ended questions, which typically call for a yes or a no, can easily be answered without elaboration. The ease with which interviewees can provide reticent responses may reduce their tendency to explicitly signal their reluctance.

Limitations

As discussed above with respect to the surprising findings regarding support, a type of interviewing strategy might look quite different in various interviewing contexts. A related problem is that what might be characterized as manipulative behavior in one context may appear supportive in another context. For example, in the interrogation literature expressing empathy towards a suspect is considered a form of minimization, used to lull the suspect into a false sense of security (Feld, 2006; 2013; King & Snook, 2009; Leo, 1996; Pearse & Gudjonsson, 1999; Soukara et al., 2009). Alternatively, in victim interviews it could be considered a supportive technique. There are likely to be subtle differences in the two types of interviews; here, the officer who empathized with the victim by stating “I understand you think he treated you well” was suggesting information rather than reflecting something the victim had said. Regardless of how empathy is categorized, that because officers expressed empathy in only 28 utterances, this is unlikely to have affected the results.

Causality

Due to the observational nature of the data, we cannot say that the relation between officer’s behavior and CSEC victims’ responses is causal. Of course, this is true of any observational study examining the relation between interviewer utterances and victim responses, since the questions are not systematically varied. Questions may influence responses, but responses may influence questions. Furthermore, other aspects of interviewer utterances are likely to influence victim reluctance, such as the topic of the utterance.

In this study, we examined the relation between interviewee response and subsequent interviewer utterance, and indeed found that greater reluctance was associated with greater subsequent use of some techniques. We took a step toward addressing this problem by including prior reluctance in our analysis of the relation between interviewer technique and subsequent reluctance. The key findings remained unchanged, which increases confidence in these results. Nevertheless, it is easy to imagine how interviewer utterances and interviewee reluctance may influence each other over longer (and difficult to identify) intervals, making any causal claims tentative.

With respect to question topic, we observed that the interviews were quite pointed and focused on eliciting information about the victims’ interactions with the defendant, making it obvious to the victims that the officers were single-mindedly seeking incriminating information. It thus seems unlikely that topic would make a difference in this sample. We suspect that topic will matter if interviewers attempt to elicit different types of information, particularly when some of it is relatively innocuous or at least less obviously incriminating. Indeed, Lindholm and colleagues (2015) observed this with respect to their sample of commercially sexually exploited youth, who were most evasive to disclose crime-specific details (e.g., names of people and places). This deserves further exploration in samples of interviews that range more widely in topic, particularly because an effective strategy for eliciting useful information from CSEC victims may be to seek information about their perceived “positive” aspects of the relationship with their exploiter.

Sample Size and Generalizability

Another limitation concerns the limited number of interviews and the fact that the interviews came from a single case, with a small group of victims and an even smaller group of interviewers. Certainly, one cannot use these findings to make generalizable statements about how all law enforcement officers question CSEC victims or how all CSEC victims respond. Nevertheless, the results shed some light on how different dynamics within interviews are related to reluctance. Even within this small sample, the frequency of interviewer techniques and reluctance varied widely. Moreover, it is remarkable that we identified frequent use of interviewing techniques identified in the literature on police interrogations, supporting future examination of the use of approaches associated with suspect interrogation in police interviews with crime victims.

Given the small size of our sample, other factors that may have influenced both officers’ use of techniques and victims’ rates of reluctance could not be reliably tested. Our anecdotal exploration of what we knew about the individual victims highlighted one potentially important fact: the victims’ personal experiences with the perpetrator likely affected their willingness to cooperate with law enforcement, understanding that their responses would affect the success of prosecution as well as the risks of retaliation by the perpetrator. Similarly, Lindholm and colleagues’ (2015) qualitative assessment of trafficked adolescent interviews found that the extent and recency of violence appeared to affect the rates of evasiveness.

Future replication work with larger samples can increase confidence in the relation between interviewing techniques and reluctance. Furthermore, that work can explore how interviewing techniques may interact with case characteristics in predicting reluctance. For example, minimization may reduce reluctance in cases in which the victims no longer side with the perpetrator, and thus view their culpability as a separate concern, or in cases in which they believe they can be protected. Support may reduce reluctance when investigators are attempting to obtain background information that is less obviously incriminating, especially if it is unaccompanied by other interviewing tactics.

Implications for Practice

We view this study as a first step to understanding the unusual dynamics of interviews with adolescent CSEC victims, rather than as a road map for interviewing. Clearly, one would not want to conclude from these findings that interviewers should ask fewer open-ended questions and use less support as a means of reducing reluctance. As we discussed, the findings may be better understood as revealing the inefficacy of best practice interviewing methods when embedded in a coercive context. Furthermore, the results suggest caution in assuming that adolescent CSEC victims will respond to questions and interviewing techniques in the same way that younger victims of child sexual abuse do.

The fact that maximization is associated with greater reluctance does not mean that all types of maximization are contraindicated. Presentation of evidence is considered a form of maximization but is widely accepted as a legitimate method for questioning suspects (Oleszkiewicz & Watson, in press), is used in questioning child sexual abuse victims (Ahern et al., 2019; Lytle et al., 2019), and has been found in experimental work with children to increase true disclosures of transgressions without increasing false allegations (Evans & Lyon, 2019). There are obvious distinctions that should be drawn between presenting true evidence and presenting false evidence, and between maximization that is overtly coercive and matter-of-fact references to other evidence in the case.

Practitioners might dismiss reluctance on the grounds that what matters most is eliciting incriminating information, and that reluctance is not incompatible with informativeness (Blasbalg et al., 2019). This highlights the need to consider the goals of interviews beyond information-seeking. Research has consistently shown that victims of crime harbor significantly more negative attitudes towards police than individuals who have not previously been victimized (Aviv & Weisburd, 2016). Additionally, negative interactions with police officers often aggravate the initial trauma of victimization or are considered a traumatic experience on their own, dubbed as a “second injury” following abuse (Aviv & Weisburd, 2016; Parsons & Bergin, 2010). These experiences with police officers influence perceived legitimacy of the police, which in turn play an important role in promoting compliance with the law and willingness to cooperate with police, as well as help-seeking and crime-reporting behaviors (Murphy et al., 2008; Tyler & Fagan, 2008). For example, in the sample studied here, greater reluctance predicted a refusal to appear as a witness at the defendant’s trial (Henderson et al., 2021).

Conclusion

Commercial sexual exploitation of children and adolescents is a phenomenon we unfortunately know too little about. Victims are often the key to understanding not only why and how it happens, but also in identifying traffickers and other victims, so that at-risk youth can be protected. Because law enforcement officers are usually the first to interview victims, and the content of these interviews are likely to be used in court, it is imperative that officers use effective and appropriate interviewing approaches not only to elicit productive, accurate, and informative accounts, but also to avoid re-traumatizing victims.

This study examined the relation between interview techniques and reluctance in a small sample of adolescent CSEC victims questioned by law enforcement and found that victims exhibited higher rates of reluctance when officers used maximization with statements, used support, and asked open-ended questions. The results highlight the difficulty of questioning adolescent CSEC victims and the need to develop strategies for questioning that recognize their special vulnerabilities.