Jail inmates’ perspectives on police interrogation.

Item Type Article

Authors Cleary, Hayley M. D.; Bull, Ray

Citation Cleary, H. M. D. and Bull, R. (2018) 'Jail inmates’ perspectives on police interrogation', Psychology, Crime & Law, DOI: 10.1080/1068316X.2018.1503667

DOI 10.1080/1068316X.2018.1503667

Publisher Taylor and Francis

Journal Psychology, Crime & Law

Rights Archived with thanks to Psychology, Crime & Law

Download date 28/09/2021 17:34:22

Link to Item http://hdl.handle.net/10545/622958 Jail Inmates’ Perspectives on Police Interrogation

By Hayley Cleary (Virginia Commonwealth University) and Ray Bull (Derby University)

(To appear in Psychology, Crime and Law)

Few studies have examined police interrogation strategies from suspects’ perspectives, yet assessing suspects’ views about interviewer approaches could provide important insights regarding confession decision making. The current study is the first American survey to assess a diverse sample of adult jail inmates’ views on police interrogation tactics and approaches. The study explored US jail inmates’ (N =

418) perspectives about how police should conduct interrogations. Potential dimensionality among 26 survey items pertaining to police tactics was examined using exploratory factor analysis. Group differences according to demographic and criminological variables were also explored. Four factors emerged, conceptualized as Dominance/Control, Humanity/Integrity, Sympathy/Perspective-Taking, and

Rapport. Respondents most strongly endorsed Humanity/Integrity and Rapport strategies and were unsupportive of approaches involving Dominance/Control. Gender differences emerged for Dominance/

Control and Humanity/Integrity, and Black respondents were more likely to value strategies related to

Sympathy/Perspective-Taking. Suspects endorsed interrogation strategies characterized by respect, dignity, voice, and a commitment to the truth; they reported aversions to the false evidence ploy and approaches involving aggression. Overall, results from this incarcerated sample suggest that interviewees may be more responsive to rapport-building, non-adversarial strategies.

Keywords: interrogation, interviewing strategies, police interviewing techniques, suspects’ reactions, prison inmates 2

Introduction

While great strides have been made in the identification and classification of police interrogation techniques (Kelly, Miller, Redlich, & Kleinman, 2013), much of the existing work is derived from officers’ self-reported use of techniques (e.g., Cleary & Warner, 2016; Kassin et al., 2007; Wachi et al., 2014) or occurs outside the United States (e.g., Bull & Soukara, 2010;

Kebbell, Alison, Hurren, & Mazerolle, 2010; Wachi, Watanabe, Yokota, Otsuka, & Lamb,

2016a), where legal systems and/or police interrogation practices likely differ. Very few studies have examined suspects’ perspectives on police interrogation (Bull, 2013; Goodman-Delahunty,

Martschuk, & Dhami, 2014), yet obtaining suspects’ views could yield important insights about what the ‘targets’ of interrogation think about the experience and thus provide a critical window into suspects’ interrogation decision making.

Research on police interviewing and interrogation strategies often emerges from police training manuals or methods, such as the Reid Technique (Inbau, Reid, Buckley, & Jayne, 2013), the PEACE model (Milne & Bull, 1999), or the Army Field Manual (U.S. Department of the

Army, 2006). Some studies have focused on broad dichotomies such as humanity versus dominance or information gathering versus accusatory questioning, while others assess usage of specific techniques (e.g., asking questions repeatedly, interrupting the suspect, presenting false evidence to the suspect; Kelly et al., 2013). Cleary and Warner (2016) reported that American police were trained in a wide variety of interrogation techniques; for example, more than half of their sample received training in how to discourage , and over 90% were trained in observing body language and building rapport. Respondents were also trained in what scholars consider more coercive or problematic techniques such as suspect isolation (77%) and presenting false evidence (73%). 3

Surveys with police officials indicate that police report actually using a wide variety of these interviewing strategies, ranging from non-confrontational, information-gathering approaches to more adversarial techniques. The three most frequently used techniques in Cleary and Warner’s (2016) study—building rapport, observing body language, and offering things for comfort—are considered more benign approaches, whereas the top three techniques in Kassin and colleagues’ (2007) survey are more confrontational (isolating the suspect, conducting the interrogation in a small room, identifying contradictions in the suspect’s story). However, police officers in both studies reported using, at least sometimes, a range of both adversarial and nonadversarial approaches.

Most research on interrogation techniques has focused on either understanding or classifying the techniques themselves or associating, via field observations or laboratory experiments, the techniques with interrogation outcomes (see, e.g., Meissner et al., 2014 and the studies therein; Walsh & Bull, 2015). Very few studies have assessed suspects’ own perspectives on police interrogation/interviewing strategies (Bull, 2013). Since confession or even cooperation is ultimately the suspect’s decision, it is essential that researchers examine how suspects perceive the interrogation interaction. Such research could potentially challenge long- held assumptions about suspects’ (un)willingness to cooperate during interrogations. For example, Leo (2008) noted the common belief that “confessions, especially to serious crimes, are rarely made spontaneously. Rather they are actively elicited….typically after sustained psychological pressure” (p. 119). However, recent suspect-focused research suggests that not all

‘targets’ enter the interrogation setting with a clear intent to deny the allegations. In Australia,

Kebbell, Hurren, & Mazerolle (2006) found that only half of convicted sex offenders in their sample reported entering the police interview having already decided whether to deny or confess. 4

In fact, less than 20% had planned to deny and about 30% had planned to confess; the other 50% entered the police interview not yet having decided whether to deny or confess. In Japan, almost

30% of subsequently convicted prisoners reported not having decided in advance whether to confess or deny (Wachi, Watanabe, Yokota, Otsuka, & Lamb, 2016b). Not only do these findings cast doubt on the assumption that suspects will unilaterally deny criminal allegations in any interrogation, they also highlight the importance of examining how suspects themselves perceive the interrogation experience. Thus it is critical to expand suspect-focused research by examining their perceptions of police interrogation tactics.

Among the first to explore suspect perspectives of police interrogation techniques was a

Swedish study involving a survey of 83 prisoners convicted of murder or sexual offences.

Holmberg and Christianson (2002) introduced the terms ‘humanity’ and ‘dominance’ based on factor analysis of characteristics of interrogating officers as rated by the prisoners. Offenders interviewed with a humane approach were three times more likely to confess (or provide incriminating information) than offenders who perceived their interrogator as dominant.

Offenders who felt respected were nearly six times more likely to confess than those who felt anxiety during the interview. These early findings highlight two important notions—first, that suspects report a wide range of interrogation experiences, and second, that their perceptions of those experiences are related to their decision making.

Kebbell and colleagues (2008, 2010) conducted a series of studies with 43 convicted sexual offenders in Australia to explore the relationship between interviewing approach and confession decision making. One study employed an experimental design involving vignettes with four different interview styles (dominance, humanity, cognitive distortions [of sexual offenders], and a control condition) and asked offenders to rate the likelihood that the vignette 5 character would confess (Kebbell, Alison, & Hurren, 2008). Similar to Holmberg and

Christianson’s (2002) findings, sex offenders rated vignette suspects who were treated humanely as more likely to confess. A humane interviewing approach was also associated with greater perceptions of fairness and a ‘well conducted’ interview (Kebbell et al., 2008). A second study expanded the interviewing strategies to include strength of evidence, ethical interviewing, minimization, and maximization and asked offenders to rate their own interview as well as an

‘ideal’ police interview—that is, how police should conduct questioning in order to elicit confessions from guilty suspects (Kebbell et al., 2010). Ideal police interviews were associated with presentations of evidence and ethical and humane interview approaches, whereas dominance was the least endorsed approach. Moreover, respondents’ perceptions of their own police interview was associated with confession; confessors were more likely to rate their interviewer as ethical and humane, whereas deniers more frequently perceived their interviewer as dominant (Kebbell et al., 2010).

More recent studies with interviewees have occurred in Japan (Wachi et al., 2016a),

Belgium (Vanderhallen & Vervaeke, 2014), Canada (Deslauriers-Varin, Lussier, & St-Yves,

2011; Snook, Brooks, & Bull, 2015), and a multinational sample (Goodman-Delahunty et al.,

2014). Snook and colleagues’ (2015) survey of 100 incarcerated men was consistent with previous findings (Kebbell et al., 2010) in that evidence-related and humanitarian approaches were most strongly associated with confessions. Humanitarian interviewing was also predictive of suspects’ self-reported cooperation during the interview. By contrast, Goodman-Delahunty et al. (2014) also assessed cooperation (though not confession) and reported that confronting suspects with evidence actually promoted resistance rather than cooperation (but see Bull, 2014, and Leahy-Harland & Bull, 2017, who found gradual disclosure of evidence/information to be 6 associated with keeping suspects of serious crimes ‘on topic’). Goodman-Delahunty et al.

(2014) also reported that rapport-like strategies were not associated with cooperation in their particular sample (see also Beune, Giebels, & Taylor, 2010, for cultural effects).

Wachi et al. (2016a) classified detainee respondents according to the manner in which they characterized their police interviewers. The classifications included ‘relationship-focused’ interviewers, who employed active listening and rapport building strategies, and

‘undifferentiated-high’ interviewers, who employed the same techniques but also confrontation and presentation of evidence. Other classifications included ‘evidence-confrontational’ and

‘undifferentiated-low’ interview styles. Among detainees who had decided prior to the interview to deny allegations or were undecided about confessing or denying, those who perceived their interviewers as relationship-focused or undifferentiated-high were more likely to confess. These findings suggest that a particular subset of interviewing strategies, including but perhaps not limited to friendly rapport-building strategies, may be particularly effectual with suspects who enter the interview undecided about confession.

The present study builds on existing work in several ways. First, the suspect perception studies described above are limited by small sample sizes and/or specialized populations (e.g., sex offenders, terrorism suspects). Interrogators may approach these specialized populations differently or the offenders themselves may differ in their reactions to police interrogations

(Kebbell et al., 2008). Second, almost all studies have occurred outside the United States, where legal procedures, interrogation strategies, and cultural norms differ widely. O’Connor and

Carson (2005) reported on 45 interviews conducted with convicted child molesters in Missouri,

US; respondents discussed their reasons for confessing or not confessing. However, this prison- 7 based study—the only US study to our knowledge—included a small, highly specialized sample and was not peer-reviewed.

The present study assessed interrogation perceptions among a large, American sample of jail inmates representing a wide variety of offenses and criminal histories. To our knowledge, it is the first study to assess American jail inmates’ perspectives on police interrogation. The objective of this study was to assess police approaches to interrogation from the perspective of the potential ‘targets’ themselves, given that much of the literature (especially in the US) surveys only police, whose role, experience, and desired outcomes differ from interviewees. Because the decision to confess ultimately rests with the suspect, it is critical to assess his or her perspective on the interrogation process.

Method

Participants

Participants were 444 men and women incarcerated in one of two jail facilities in

Virginia. Sixteen cases were excluded because the questionnaire was left mostly or entirely blank. A further ten cases were excluded because of concerns about data quality. The final sample (N = 418) was predominantly (80.6%) male with a mean age of 35.2 years (SD = 10.8, range = 18-71 years). Individuals self-identified primarily as Black (47.4%) or White (40.0%), and the remainder indicated their race/ethnicity as Hispanic/Latino (2.6%), multiracial (5.7%), or other/missing (4.3%). Current charges included property crimes (19.4%), crimes against persons

(15.1%), weapon or drug offenses (29.9%), probation violations or administrative offenses (e.g., failure to register as sex offender; failure to pay child support; 16.3%), crimes involving fraud or indecency (e.g., distribution of child pornography; 3.3%), and driving or traffic offenses

(including driving under the influence (DUI); 7.2%). Participants reported an average of 8.15 8 lifetime arrests (SD = 6.50; median = 6; range: 1-35 after excluding outliers). Approximately two-thirds of the jail inmates (68.8%) had already been convicted while the remaining one-third

(31.2%) were awaiting trial. Of those convicted who had been sentenced (n = 265), the average sentence term was 4.24 years (SD = 7.86; range: 10 days – 56 years).

Materials and Procedure

A questionnaire, developed for the purposes of this study and informed by the work of

Kebbell et al. (2010) assessed respondents’ opinions about how police should conduct custodial interrogations. It contained 26 statements representing a variety of interrogation techniques and approaches, including maximization and minimization, humanity/dominance, and ethical interviewing. The questionnaire was written at an 8th grade reading level and piloted for clarity; all items appear in their entirety in Table 1. Respondents indicated their level of agreement on a five-point Likert scale (1 = strongly disagree to 5 = strongly agree). Respondents also answered questions pertaining to their offense, arrest, and incarceration histories as well as demographic questions.

Study approval was obtained from the university Institutional Review Board (IRB; the ethics review panel) as well as the jail superintendents. Two jail facilities were visited: a city jail

(2 housing units) and a large regional jail serving several surrounding cities and counties (18 housing units). Incarcerated populations in the United States are quite difficult to access for research purposes, and these facilities were convenience samples, selected due to the superintendents’ willingness to permit the research. Housing unit size ranged from smaller, open-style pods accommodating up to 18 inmates to much larger, traditional cellblock units containing approximately 46 double-bunked cells that accommodated up to 92 inmates. Upon visiting each pod, the first author introduced herself to the group, distributed a study information 9 sheet, and described the research study with an emphasis on voluntariness and confidentiality.

Interested participants were provided a consent form and copy of the questionnaire and encouraged to ask questions. Waiver of documentation of consent had been secured from the

IRB in order to further protect anonymity. Respondents independently completed their surveys seated at tables in the common area or inside their cells. Upon completion, each respondent deposited his or her completed survey in a collection box and was thanked for participating.

Respondents did not receive incentives or compensation for participating. Response rates were computed for each housing unit and ranged from 11% (2/18) to 94% (17/18); the overall response rate was 40.7%.

Results

Mean scores and standard deviations for each item appear in Table 1. Negatively worded items were reversed scored for interpretation purposes, such that higher scores reflect more

“suspect-friendly” practices. Of all 26 items, respondents most strongly endorsed the notion that

‘A police interviewer should give suspects a chance to tell their side of the story’ (M = 4.65, SD

= .80). Seventy-eight percent of respondents strongly agreed with this sentiment, and 91% agreed or strongly agreed. Respondents also strongly agreed that police interviewers should favor fact-finding over obtaining confessions (M = 4.61, SD = .81). The next highest rated items pertained to interviewers refraining from casting insults (M = 4.50, SD = .81) and treating suspects with respect (M = 4.48, SD = .86). The least endorsed items pertained to interviewer aggression and deceit. Of all 26 items, respondents were most opposed to interviewers using the false evidence ploy (M = 1.41, SD = .84); nearly three-quarters of the sample (74%) strongly disagreed that ‘A police interviewer should tell the suspect he has evidence that the suspect is guilty, even if he doesn’t actually have any evidence.’ Respondents also indicated a strong 10 aversion to interviewers yelling at suspects (M = 1.55, SD = .79) and behaving aggressively toward suspects (M = 1.59, SD = .82).

[PLEASE INSERT TABLE 1 ABOUT HERE]

Exploratory Factor Analysis

We next conducted exploratory factor analysis (EFA) to investigate underlying factor structure and compare with prior research. Because the item-level data are ordinal, factor analysis was performed using polychoric correlations instead of the traditional Pearson product- moment correlations. We first screened the data for missing values, outliers, and multicollinearity and examined the factorability of the polychoric correlation matrix. Thirty-nine of the 418 cases were univariate outliers and were excluded from further analyses. Inter-item correlations ranged from r = .00 to r = .37. Several indicators suggested a suitably factorable matrix: numerous pairs of significant inter-item correlations, near-zero values in the anti-image correlation matrix, significant Bartlett’s test for sphericity (χ2 = 3584.77, p < .001), and the

Kaiser-Meyer-Olkin measure of sampling adequacy (KMO = .82; Tabachnick & Fidell, 2013;

Field, 2018).

We conducted EFA on the polychoric correlation matrix of the 26 items using oblique

(oblimin) rotation to account for presumed nonindependence of factors. The scree plot presented two potential points of inflection, suggesting either a four- or five-factor solution. Upon examining eigenvalues in subsequent models we ultimately retained four factors, accounting for

49% of the total variance. Rotated factor loadings are presented in Table 1. Loadings greater than .45 were retained and interpreted (Tabachnick & Fidell, 2013). The first factor, labeled

Dominance/Control, included eight items and accounted for 25.9% of the variance. Items loading onto this factor pertained to interviewing tactics involving aggression, dominance, or 11 manipulation of suspects’ emotions. Example items include ‘A police interviewer should be able to interrupt the suspect if he wants to’ and ‘A police interviewer should tell the suspect that they will feel worse if they don’t confess.’ Means for individual items ranged from 1.41 to 2.45 on a

5-pt Likert scale. A mean composite score computed across the 8 items (M = 1.84, SD = .59) indicated that respondents disagreed/strongly disagreed that police interviewers should use these approaches.

The second factor, labeled Sympathy/Perspective-Taking, included seven items and accounted for 11.3% of the variance. Two items produced cross-loadings less than .45 and were excluded from the composite measure. Items loading onto this factor pertained to interview approaches that show understanding of the suspect’s point of view. Example items include ‘A police interviewer should try to understand how a suspect is feeling’ and ‘A police interviewer should show sympathy toward a suspect.’ Mean scores for individual items ranged from 2.66 to

3.93 (composite score: M = 3.54, SD = .63).

The third factor, labeled Humanity/Integrity, included seven items and accounted for

6.6% of the variance. One item cross-loaded onto multiple factors and was excluded. Items loading onto this factor pertained to interviewing approaches that involve respect, neutrality, or a commitment to the truth. Example items include ‘A police interviewer should not insult the suspect during the interview’ and ‘A police interviewer should be interested in finding out the truth, not just getting a confession from the suspect.’ Mean scores for individual items ranged from 4.17 to 4.65 (composite score: M = 4.44, SD = .50).

The fourth and final factor, labeled Rapport, included four items and accounted for 5.1% of the variance. Items loading onto this factor pertained to strategies involving cooperation, voice, and allowing the suspect to participate in the conversation. Example items include ‘A 12 police interviewer should not bother getting to know the suspect before starting the interview’

[reverse scored] and ‘A police interviewer should rush the interview, and not allow the suspect time for reflection’ [reverse scored]. Mean scores for individual items ranged from 3.25 to 4.35

(composite score: M = 3.98, SD = .69).

Overall, mean composite scores indicate that respondents most strongly endorsed the

Humanity/Integrity component (M = 4.44, SD = .50). Endorsement rates were also high for

Rapport (M = 3.98, SD = .69) and slightly lower, but still positive, for Sympathy/Perspective-

Taking (M = 3.54, SD = .63). Respondents were not supportive of interrogation approaches characterized by Dominance/Control (M = 1.89, SD = .65).

Cronbach’s alpha for the items comprising each component were as follows: Dominance/

Control (.76), Humanity/Integrity (.61), Sympathy/Perspective-Taking (.61), and Rapport (.49).

While lower than conventional guidelines, these values are not unexpected given the relatively small number of items (where α becomes artificially inflated as the number of items increases;

Field, 2018) and that similar findings have been reported in the offender interrogation literature

(Kebbell et al., 2010). Nonetheless, caution in presuming a high degree of dimensionality among the items is warranted.

Group Differences

We next explored group differences according to key demographic and criminological variables using the mean composite scores for the four factors (Table 2). With regard to gender, women (M = 1.64, SD = .54) were significantly less likely to endorse police usage of

Dominance/Control approaches than men (M = 1.88, SD = .59), t(358) = -3.08, p = .002, d = .41.

Additionally, women (M = 4.56, SD = .42) reported greater preference for Humanity/Integrity approaches than men (M = 4.42, SD = .49), t(370) = 2.19, p = .029, d = .29. Women and men 13 did not differ in their perceptions of Sympathy/Perspective-Taking or Rapport in ideal police interrogations (p’s > .05).

To examine potential differences according to race/ethnicity, this variable was collapsed into the categories Black (n = 174; 46% of sample), White (n = 157; 41%) and other/multiple races/ethnicities (n = 48; 13%). Across the four factors, significant differences by race emerged only for Sympathy/Perspective-Taking, F(2, 363) = 7.24, p = .001. Specifically, Games-Howell post hoc tests, selected to account for unequal sample sizes, indicated that Black respondents (M

= 3.67, SD = .61) were more likely to endorse this approach than White respondents (M = 3.41,

SD = .60), t(361) = 4.00, p < .001, d = .43.

When examining criminological variables, group differences emerged only with respect to conviction status. Individuals who had already been convicted and were currently serving their sentence (M = 1.88, SD = .59) reported higher mean ratings for Dominance/Control than individuals who were awaiting trial (M = 1.74, SD = .58), t(356) = -2.12, p = .034, d = .25.

Respondents’ ratings on the four factors were consistent across offense types and interrogation status (i.e., whether respondent was interrogated in reference to current charge), and neither respondent age nor arrest history was correlated with mean factor scores.

[PLEASE INSERT TABLE 2 ABOUT HERE]

Discussion

This study is the first to our knowledge to assess a large sample of American jail inmates’

(and by extension criminal suspects’) perspectives on police interrogation practices. Overall, findings indicated that suspects in the sample strongly endorse rapport-building strategies and techniques that promote respect, humanity, and neutrality. Ratings were largely consistent 14 across groups, suggesting that these preferences for ‘ideal’ police interrogations are stable across a range of offender characteristics and experiences.

In particular, the most highly rated of all 26 items pertained to interrogators giving suspects an opportunity to explain their perspectives. That respondents ranked this item above all others speaks to the importance of procedural justice in police interrogations. While procedural justice as a psycholegal construct has received tremendous empirical attention, almost no research has focused on procedural justice in a police interrogation context.

Sivasubramaniam and Heuer (2012) suggested that procedural justice evaluations in police interrogations are moderated by the individual’s role; specifically, they argued that the - presumptive nature of police interrogations means that interrogators’ perceptions of justice are moderated by their focus on the interrogation outcome (as opposed to the process) and that their confidence in the suspects’ guilt means they perceive suspects as less deserving of respectful treatment. Our data corroborate Sivasubramaniam and Heuer’s (2012) argument that the subordinates (i.e., suspects) value fairness in the interrogation process itself. A more definitive

—but difficult to execute—future study to explore procedural justice differences according to interrogation role might involve interviewing or surveying both interrogators and suspects about their procedural justice judgments immediately following a custodial interrogation. Goodman-

Delahunty et al. (2014) did interview both law enforcement officials and suspected terrorist detainees, but the officers in the sample did not interview the detainees in the sample and procedural justice was not assessed.

From a practice perspective, encouraging interviewees to provide a narrative account of the event is considered an investigative interviewing best practice (Powell, Fisher, & Wright,

2005). The ‘A’ in the PEACE model, England and Wales’s model for police interviewing, 15 stands for the ‘Account’ phase in which interviewers used open-ended questioning strategies that encourage suspects to provide uninterrupted accounts in their own words (Milne & Bull, 1999).

Numerous studies confirm that this approach is associated with more accurate and detailed statements (see, generally, Bull, 2014). Eliciting narrative accounts via open-ended questions offers a host of advantages, including obtaining more numerous and accurate details and reducing the potential for contamination (Read & Powell, 2011). A survey of UK investigators revealed that obtaining suspects’ uninterrupted accounts was perceived as both frequently used and highly effective (Dando, Wilcock, & Milne, 2008). Thus, adopting a narrative framework may be mutually preferable for both investigators and suspects.

It is noteworthy that several of the items that loaded onto the Sympathy/Perspective-

Taking component were initially included in the questionnaire as examples of minimization tactics. Kassin et al. (2010) described minimization tactics as “moral justification and face- saving excuses….[the interrogator] normalizes and minimizes the crime, often suggesting that he or she would have behaved similarly; and offers the suspect a choice of alternative explanations”

(p. 12). The loading of minimization-related items in our questionnaire (e.g., , suggesting that other people have done worse) along with seemingly less nefarious strategies

(e.g., showing sympathy toward the suspect, trying to understand how he or she is feeling) raises the question of whether suspects are particularly vulnerable to minimization tactics—that is, they may misinterpret deliberate minimization as genuine feelings of sympathy from an interrogator.

Group comparisons indicated that preferences for humane and rapport-building interrogation strategies, as well as an aversion to dominance or control-oriented strategies, were fairly universal. Women did report a greater sensitivity to Dominance/Control and a slight preference for Humanity/Integrity approaches compared to men. Although we do not know 16 whether the women in this sample were interrogated by male officers, it is statistically likely, as only 10% of supervisory officers in local police departments (Reaves, 2015) and 12% in local sheriff’s offices (Burch, 2016) are female. Because these findings speak to the manner in which interrogating officers relate to suspects within a specific social interaction, it is perhaps useful to consider them in the context of the extremely high rates of prior victimization reported by incarcerated American females. For example, in one study of women in jail, 60% reported experiencing abuse as a child, 77% reported experiencing intimate partner violence (e.g., physical abuse or partner rape), and 86% reported experiencing some kind of sexual violence in their lifetimes (Lynch, DeHart, Belknap, & Green, 2012). While a detailed discussion of

American social and cultural gender roles is beyond the scope of this article, it is possible that female interviewees perceive the power differential between interrogator and suspect more acutely than male interviewees. Feminist scholars have argued that harsh, rigid criminal justice policies and practices in some ways actually exacerbate women’s victimization experiences by virtue of removing women’s autonomy in the decision making process (e.g., mandatory arrest in domestic violence calls; Cantrell, 2013). At the very least, the gender differences reported here suggest there is value in further exploring gender roles in suspects’ perceptions of the interrogation process and context.

A significant difference also emerged based on respondents’ race/ethnicity; Black respondents were more likely to endorse Sympathy/Perspective-Taking as a desirable interview strategy than White respondents. While we do not know interviewer race/ethnicity for the respondents in our sample, Bureau of Justice Statistics data show that approximately three- quarters of local law enforcement officers are White (Reaves, 2015). Psychological and criminological research has consistently demonstrated that Black Americans are less trusting of 17 police and legal institutions than White Americans (Tyler, 2005), and recent work has confirmed the role of race-based stereotype threat in citizen-police encounters (Najdowski, Bottoms, &

Goff, 2015). Najdowski (2011) argued that Black individuals are sensitive of the American cultural stereotype of Black criminality, and this stereotype threat might even cause them to exhibit nonverbal behaviors associated with , leading interviewers to further assume their guilt during an interrogation. Kahn and colleagues (2017) reported that citizens who felt they physically resembled a ‘typical’ member of their racial/ethnic group were especially concerned about negative treatment from police. In our study, a heightened preference for interviewer sympathy or perspective taking might indicate that Black suspects feel alienated or differently situated and want interviewers to attempt to understand their situation.

Finally, perspectives on ‘ideal’ interrogation strategies differed according to conviction status in a surprising way. Respondents who had already been convicted showed less aversion to dominance/control strategies than pretrial detainees. One interpretation is that suspects whose police and court experiences concluded with a jail sentence were ‘disillusioned’ by the experience, such that the data are tapping into a hopelessness of sorts among convicted inmates.

However, it is important to note that Dominance/Control ratings were low among all respondents, regardless of conviction status; additionally, a modest effect size suggests caution in interpretation. Overall, these findings suggest that specifically measuring procedural justice variables among criminal suspects and police interviewees is important to understanding their perspectives on the process.

Several limitations of this study should be noted. First, the present analysis does not directly connect suspect perspectives to their actual confession decision making. Suspect perspectives on police interrogation are both practically important and empirically underexplored 18

(Bull, 2014), and the present study is only one of only a small number to investigate offender preferences for police treatment during interrogation. An important next step would be to explore whether those preferences are directly related to suspects’ behavior or decision making during police questioning. As discussed above, perceptions of procedural justice may be a key mediator in that relationship and should be measured explicitly. Second, questions of clarity in labeling emerge when conceptualizing, operationalizing, measuring, and reporting on police interviewing strategies. This study adopted component labels that we felt best captured the essence of the strategies or behaviors described in each component. However, our terminology, like the terminology adopted in prior work (e.g., ‘humanity/dominance’; Holmberg &

Christianson, 2002; Kebbell et al., 2008; ‘ethical interviewing’; Kebbell et al., 2010), is linguistically imperfect. Third, it is important to note that all of the negatively worded items loaded onto one factor. Several interpretations are possible. One is that the resulting factor is a

‘method factor,’ meaning that the items load together because they share a differing methodological feature. This occasional outcome of factor analyses has been attributed to respondents’ confirmation bias, acquiescence, or careless responding (Weijters, Baumgartner, &

Schillewaert, 2013). Another is that the outcome is merely a coincidence. Reverse-wording items does not in itself guarantee the presence of a method factor, and the psychometrics literature is divided on the issue of reverse-wording items (Weijters et al., 2013). We labeled this factor Rapport because the items are consistent with conceptualizations of interviewer rapport-building elsewhere in the literature, including cooperativeness, a reasonable interview pace, and general ‘harmoniousness’ (see, generally, Vanderhallen & Vervaeke, 2014). However, we agree that “the construct suffers from a conceptual weakness, which generates methodological restrictions that hamper research in the field” (Vanderhallen & Vervaeke, 2014, 19 p. 77). Finally, caution is warranted when extending these findings to the general population of

U.S. jail inmates, given that jail environments can differ substantially and the study did not include a comparison group.

In sum, our findings indicate that suspects may be sensitive to a variety of interviewing strategies within the context of police interrogation. They show a preference for strategies predicated on respect and cooperation and an aversion to aggressive and manipulative approaches. When considered within the broader literature on the relationship between suspect attitudes and decision making, these findings suggest police may have success across a wide variety of interrogation contexts and suspects if they adopt a noncoercive interview approach.

Strategies that display respect, dignity, and neutrality and that allow and encourage suspects to speak freely in their own words may be more successful than strategies that reproach, manipulate, or insult interviewees. Recent data on interrogation outcomes has consistently shown that rapport-building, narrative-oriented interviewing improves detail and accuracy in reporting (Bull, 2013) and it is such information that can be skillfully compared with what the interrogator/interviewer knows but has not yet disclosed (Bull, 2014). The present study further shows, from interviewees’ point of view, that such interviewing styles are valued and preferred.

Thus an open, respectful, nonadversarial approach is a ‘win-win’, beneficial for both investigators seeking accurate and reliable information and interviewees seeking the opportunity to be heard and respected. 20

References

Beune, K., Giebels, E., & Taylor, P. J. (2010). Patterns of interaction in police interviews: The

role of cultural dependency. Criminal Justice and Behavior, 37, 904-925. doi:

10.1177/0093854810369623

Bull, R. (2013). What is ‘believed’ or actually ‘known’ about characteristics that may contribute

to being a good/effective interviewer? Investigative Interviewing: Research and Practice,

5, 128-143.

Bull, R. (2014). When in interviews to disclose information to suspects and to challenge them?

In R. Bull (Ed.), Investigative interviewing. New York: Springer.

Bull, R., & Soukara, S. (2010). Four studies of what really happens in police interviews. In G. D.

Lassiter & C. A. Meissner (Eds.), Police interrogations and false confessions: Current

research, practice, and policy recommendations (pp. 81-95). Washington, DC: American

Psychological Association.

Burch, A. M. (2016, June). Sheriffs’ Office Personnel, 1993–2013. U.S. Department of Justice

Office of Justice Programs Bureau of Justice Statistics. Retrieved from:

https://www.bjs.gov/content/pub/pdf/sop9313.pdf

Cantrell, D. (2013). Re-problematizing in domestic violence advocacy. Journal of

Gender, Social Policy, & the Law, 21, 837-863.

Cleary, H. M. D., & Warner, T. C. (2016). Police training in interviewing and interrogation

methods: A comparison of techniques used with adult and juvenile suspects. Law and

Human Behavior, 40, 270-284. doi:10.1037/lhb0000175 21

Dando, C., Wilcock, R., & Milne, R. (2008). The cognitive interview: Inexperienced police

officers' perceptions of their witness/victim interviewing practices. Legal and

Criminological Psychology, 13, 59-70. doi:10.1348/135532506x162498

Deslauriers-Varin, N., Lussier, P., & St-Yves, M. (2011). Confessing their crime: Factors

influencing the offender's decision to confess to the police. Justice Quarterly, 28, 113-

145. doi:10.1080/07418820903218966

Field, A. (2018). Discovering statistics using IBM SPSS statistics. Los Angeles: Sage.

Goodman-Delahunty, J., Martschuk, N., & Dhami, M. K. (2014). Interviewing high value

detainees: Securing cooperation and disclosures. Applied Cognitive Psychology, 28, 883-

897. doi:10.1002/acp.3087

Holmberg, U., & Christianson, S. A. (2002). Murderers' and sexual offenders' experiences of

police interviews and their inclination to admit or deny crimes. Behavioral Sciences &

the Law, 20, 31-45. doi:10.1002/bsl.470

Inbau, F. E., Reid, J. E., Buckley, J. P., & Jayne, B. C. (2013). Criminal interrogation and

confessions (5th ed.). Burlington, MA: Jones & Bartlett Learning.

Kahn, K. B., Lee, J. K., Renauer, B., Henning, K. R., & Stewart, G. (2017). The effects of

perceived phenotypic racial stereotypicality and social identity threat on racial minorities'

attitudes about police. Journal of Social Psychology, 157, 416-428.

doi:10.1080/00224545.2016.1215967

Kassin, S. M., Drizin, S. A., Grisso, T., Gudjonsson, G. H., Leo, R. A., & Redlich, A. D. (2010).

Police-induced confessions: Risk factors and recommendations. Law and Human

Behavior, 34, 3-38. doi:10.1007/s10979-009-9188-6 22

Kassin, S. M., Leo, R. A., Meissner, C. A., Richman, K. D., Colwell, L. H., Leach, A., & La

Fon, D. (2007). Police interviewing and interrogation: A self-report survey of police

practices and beliefs. Law and Human Behavior, 31, 381-400. doi:10.1007/s10979-006-

9073-5

Kebbell, M., Alison, L., & Hurren, E. (2008). Sex offenders' perceptions of the effectiveness and

fairness of humanity, dominance, and displaying an understanding of cognitive

distortions in police interviews: A vignette study. Psychology, Crime, and Law, 14, 435-

449. doi: 10.1080/10683160801950523

Kebbell, M., Alison, L., Hurren, E., & Mazerolle, P. (2010). How do sex offenders think the

police should interview to elicit confessions from sex offenders? Psychology, Crime, and

Law, 16, 567-584. doi:10.1080/10683160902971055

Kebbell, M., Hurren, E., & Mazerolle, P. (2006). An investigation into the effective and ethical

interviewing of suspected sex offenders. Final report to the Australian Criminology

Research Council.

Kelly, C. E., Miller, J. C., Redlich, A. D., & Kleinman, S. M. (2013). A taxonomy of

interrogation methods. Psychology, Public Policy, and Law, 19, 165-178.

doi:10.1037/a0030310

Leahy-Harland, S., & Bull, R. (2017). Police strategies and suspect responses in real-life serious

crime interviews. Journal of Police and Criminal Psychology, 32, 138-151.

doi:10.1007/s11896-016-9207-8

Leo, R. A. (2008). Police interrogation and American justice. Boston: Harvard University Press.

Lynch, S. M., DeHart, D. D., Belknap, J., & Green, B. L. (2012, September). Women’s

pathways to jail: The roles and intersections of serious mental illness and trauma. 23

Washington, DC: U.S. Department of Justice, Office of Justice Programs, Bureau of

Justice Assistance.

Meissner, C. A., Redlich, A. D., Michael, S. W., Evans, J. R., Camilletti, C. R., Bhatt, S., &

Brandon, S. (2014). Accusatorial and information-gathering interrogation methods and

their effects on true and false confessions: A meta-analytic review. Journal of

Experimental Criminology, 10, 459-486. doi:10.1007/s11292-014-9207-6

Milne, R., & Bull, R. (1999). Investigative interviewing: psychology and practice. West Sussex,

UK: John Wiley & Sons.

Najdowski, C. J. (2011). Stereotype threat in criminal interrogations: Why innocent black

suspects are at risk for confessing falsely. Psychology, Public Policy, and Law, 17, 562-

591. doi:10.1037/a0023741

Najdowski, C. J., Bottoms, B. L., & Goff, P. A. (2015). Stereotype threat and racial differences

in citizens' experiences of police encounters. Law and Human Behavior, 39, 463-477.

doi:10.1037/lhb0000140

O’Connor, T., & Carson, W. (2005). Understanding the psychology of child molesters: A key to

getting confessions. The Police Chief, 72, 1-7.

Powell, M. B., Fisher, R. P., & Wright, R. (2005). Investigative interviewing. In N. Brewer & K.

D. Williams (Eds.), Psychology and law: An empirical perspective (pp. 11-42). New

York: Guilford Press.

Read, J. M., & Powell, M. B. (2011). Investigative interviewing of child sex offender suspects:

Strategies to assist the application of a narrative framework. Journal of Investigative

Psychology and Offender Profiling, 8, 163-177. doi:10.1002/jip.135

Reaves, B. A. (2015, May). Local Police Departments, 2013: Personnel, Policies, and 24

Practices. U.S. Department of Justice Office of Justice Programs Bureau of Justice

Statistics. Retrieved from: https://www.bjs.gov/content/pub/pdf/lpd13ppp.pdf

Sivasubramaniam, D., & Heuer, L. (2012). Procedural justice evaluations in interrogations. In B.

Cutler (Ed.), Conviction of the innocent: Lessons from psychological research (pp. 79-

102). Washington, DC: American Psychological Association.

Snook, B., Brooks, D., & Bull, R. (2015). A lesson on interrogations from detainees: Predicting

self-reported confessions and cooperation. Criminal Justice and Behavior, 42, 1243-

1260. doi:10.1177/0093854815604179

Tabachnick, B. G., & Fidell, L. S. (2013). Using multivariate statistics (6th ed.). Boston, MA:

Pearson.

Tyler, T. R. (2005). Policing in black and white: Ethnic group differences in trust and confidence

in the police. Police Quarterly, 8, 322–342. doi:10.1177/1098611104271105

United States Department of the Army. (2006, September). Human intelligence collector

operations. (Field Manual 2-22.3; FM 34–52.) Washington, DC: Author.

Vanderhallen, M., & Vervaeke, G. (2014). Between investigator and suspect: The role of the

working alliance in investigative interviewing. In R. Bull (Ed.), Investigative

interviewing (pp. 63-90). New York: Springer.

Wachi, T., Watanabe, K., Yokota, K., Otsuka, Y., Kuraishi, H., & Lamb, M. (2014). Police

interviewing styles and confessions in Japan. Psychology Crime & Law, 20, 673-694.

doi:10.1080/1068316x.2013.854791

Wachi, T., Watanabe, K., Yokota, K., Otsuka, Y., & Lamb, M. E. (2016a). Japanese

interrogation techniques from prisoners' perspectives. Criminal Justice and Behavior, 43,

617-634. doi:10.1177/0093854815608667 25

Wachi, T., Watanabe, K., Yokota, K., Otsuka, Y., & Lamb, M. E. (2016b). Japanese suspect

interviews, confessions, and related factors. Journal of Police and Criminal Psychology,

31, 217-227. doi:10.1007/s11896-015-9179-0

Walsh, D., & Bull, R. (2015). The association between interview skills, questioning

and evidence disclosure strategies, and interview outcomes. Psychology, Crime and

Law,

21, 661-680. doi: 10.1080/1068316X.2015.1028544

Weijters, B., Baumgartner, H., & Schillewaert, N. (2013). Reversed item bias: An integrative

model. Psychological Methods, 18, 320-334. doi: 10.1037/a0032121.