2018 - Social Influence in Investigative Interviews

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 12

Received: 8 April 2018 Revised: 15 October 2018 Accepted: 1 November 2018

DOI: 10.1002/acp.3488

RESEARCH ARTICLE

Social influence in investigative interviews: The effects of


authority on informational elements produced in interviews
and written statements
David Matsumoto1,2 | Hyisung C. Hwang1,2

1
Department of Psychology, San Francisco
State University, San Francisco, California Summary
2
Humintell, El Cerrito, CA 94530 This study examined the effect of a principle of social influence—authority—on the
Correspondence informational elements in an investigative interview. Community participants told
David Matsumoto, Department of Psychology,
San Francisco State University, 1600
the truth or lied about a mock crime in a high or low authority context. Informational
Holloway Avenue, San Francisco CA 94132. elements were coded from their oral responses in the interviews and written responses
Email: dm@sfsu.edu
in a statement prior to the interviews. Rapport was rated by the interviewers and
Funding information
High‐Value Detainee Interrogation Group, participants. The authority condition produced effects on the informational elements
Grant/Award Number: DJF‐15‐1200‐V‐ in both the interviews and written statements but mainly for truth tellers. Rapport also
0009510
had direct effects on the informational elements but authority did not affect rapport
and rapport did not mediate the effect of authority on the informational elements.
These findings had theoretical, empirical, and practical implications.

KEY W ORDS

authority, deception, informational elements, investigative interviews, social influence

1 | I N T RO D U CT I O N of social influence, consciously or unconsciously, through trial, error,


and/or experience.
The positive effects of principles of persuasion and influence (hereaf- Despite the potential import of social influence in investigative
ter referred to as social influence) in facilitating initially uncooperative interviewing, research is still in its infancy in testing and documenting
individuals to become cooperative and compliant are well documented their effects in that context. One study involving interviews of
in the literature (Cialdini & Goldstein, 2004b; Schultz, Nolan, Cialdini, 123 intelligence and investigative interviewers reported that social
Goldstein, & Griskevicius, 2007). The most influential work in the area influence principles of liking and reciprocity were most often used in
is by Cialdini, who described six principles: reciprocity, consistency, establishing rapport with interviewees (Goodman‐Delahunty &
consensus, authority, liking, and scarcity (Cialdini, 2009; Cialdini & Howes, 2016). This study provided indirect evidence for the efficacy
Goldstein, 2004a). Each is the basis of many specific forms of both of social influence principles and tactics.
explicit and implicit pressures that can be applied subtly, or not so Another recent study provided direct, causal evidence for their
subtly, to influence individuals to give information, purchase goods, efficacy by examining how one principle—reciprocity—affected the
or agree with ideas. informational elements produced by interviewees and rapport
Although these principles have received considerable research between the interviewer and interviewee in an investigative interview
attention in applied areas such as business and marketing, they (Matsumoto & Hwang, in press). In that study, a community sample
also have potential import in investigative interviewing. Many inter- received either a bottle of water or not (operationalizing reciprocity)
views and interrogations involve a process of guiding uncooperative prior to an interview in which they either told the truth or lied about
individuals to recall and report information, confess to acts, admit to having committed a mock crime. Three informational elements (rele-
facts, or share previously undisclosed sources; thus, such interactions vant and irrelevant details and plausibility) were coded from their
are ripe with opportunities to leverage principles of social influence. responses and rapport in the interaction was coded from video. Offer-
Many investigators already engage in various strategies and tactics ing water facilitated greater rapport and influenced the informational

516 © 2018 John Wiley & Sons, Ltd. wileyonlinelibrary.com/journal/acp Appl Cognit Psychol. 2019;33:516–526.
MATSUMOTO AND HWANG 517

elements and this effect was moderated by veracity condition: issue commands, use bold language and tone of voice, or even
Offering water produced more relevant details and greater plausibility stand over another while talking. These direct, overt actions may
in the statements for liars but not truth tellers, and this effect was enforce one's authority, status, dominance, power, control, or legit-
mediated by rapport. imacy and can be explicit and conscious (Blass, 1999; Milgram,
These recent studies not only have contributed to notions about 1963; Tyler, 1998).
the potential efficacy of social influence to information elicitation in Another way to conceptualize and operationalize authority
an investigative interview but also have added to a growing literature involves subtle, indirect manipulations of cues in the environment
demonstrating the importance of rapport in those interviews. or context that give hints about authority with no direct actions
Rapport has been a central concept to all kinds of interactions by an individual. Cialdini's (Cialdini, 2009; Cialdini & Goldstein,
pertaining to relationship building and formation, information elicita- 2004b) concept of the “trappings of authority”—involving symbols
tion, and therapeutic alliances for years. In the domain of investigative of authority such as titles, clothes, awards, and images that confer
interviewing, recent studies have documented the effects of rapport associations with authority—is better understood as this type of indi-
on the disclosure of meaningful and complete information earlier in rect, contextual authority manipulation. Unlike direct ways described
an interview (Goodman‐Delahunty, Martschuk, & Dhami, 2014) and above, subtle, indirect methods of expressing authority are more
more accurate information in eyewitness testimony (Kieckhaefer, implicit and unconscious.
Vallano, & Schreiber Compo, 2013; Vallano & Compo, 2011). Reviews Although no study to our knowledge has directly tested the
of the literature (see Meissner, Kelly, & Woestehoff, 2015) have also effects of authority in an investigative interview context, a few studies
reported positive albeit indirect effects of rapport in the interviewing have provided suggestive evidence. For example, two studies have
and interrogation context. Thus, rapport is an important concept in reported that directly authoritative and confrontational interview tac-
investigative interviewing; we contend that it mediates the relation tics negatively affect memory retrievals of accurate, verifiable details
between social influence tactics and interview outcomes. (Meissner, Hartwig, & Russano, 2010; Vrij et al., 2017). And a previous
In this study, we examine another principle of social influence— study demonstrated that subtle, indirect perceptions of authority,
authority (Cialdini, 2009; Cialdini & Goldstein, 2004b). Basic research combined with guided imagery, encouraged the generation of poten-
has long demonstrated the effect of authority on behavior (Blass, tially false details and increased the likelihood that individuals would
1999; Milgram, 1963). Cialdini (2009) argued that authority is effec- believe an event occurred to them even if that event was not true
tive because individuals have a deep‐seated sense of duty or obedi- (Paddock & Terranova, 2001).
ence to authority, are trained from birth to obey others, and learn In this study, we directly tested the causal effects of an author-
that obedience to authority is right and disobedience is wrong. He also ity manipulation on information elicitation in an investigative
argued that individuals are just as susceptible to the symbols of interview context. Specifically, we operationalized authority in a
authority as they are to the content or substance of requests or subtle, indirect way by manipulating the trappings of authority in
commands from authority figures. These symbols are the “trappings” the environment and context of the interviews. These trappings
of authority that include titles, clothes, awards, or images that confer included cues about the credibility, credentials, experience, eligibility,
associations with known authority figures. or qualifications of the interviewers but did not involve any direct,
Relevant to the current study is the fact that authority is a central verbal, or behavioral differences in the manner, style, or words used
concept in many applied settings in which investigative interviewing in the interviews. We opted for a subtle, indirect manipulation of
occurs. Law enforcement officers conduct such interviews and such authority to avoid direct, explicit demands on the interviewees to
contexts abound with symbols and manifestations of authority, from the authority manipulation.
the setting to the understood powers of authority granted to law The use of subtle, indirect manipulations of authority has implica-
enforcement by the state. The same is true for interrogations con- tions for the cognitive mechanisms underlying their potential effects.
ducted within military confines. The concept of authority permeates One cognitive account of high stakes lying (Walczyk, Harris, Duck, &
the private sector as well, as interviews by loss prevention and secu- Mulay, 2014; Walczyk, Roper, Seemann, & Humphrey, 2003) offers a
rity officers in private institutions or by human resource professionals possible framework for understanding such cognitive mechanisms.
(e.g., job interviews) are infused with authority. This framework posits four components (activation, decision, con-
struction, and action), of which the first, activation, is most relevant
to this study. In this component, social context is encoded into work-
1.1 | Possible mechanisms underlying the effects of ing memory. Contextual cues include the setting, the individuals
authority on information elicitation involved, and the nature of the conversation or interaction. These
contextual cues may be primed by explicit or implicit signals. “Implicit
Delineating possible cognitive mechanisms underlying any effects signals include the sight of a police officer approaching the driver side
of authority on information elicitation begins with explicating the window of the car of an intoxicated driver who was pulled over after
concept of authority, which can be conceptualized and operational- leaving a bar. An explicit signal can be a question asked in a serious
ized in different ways. Broadly speaking, one way involves direct, tone during a job interview” (Walczyk et al., 2014, pp. 25–26). The
overt actions by an individual that manifests authority, is explicit, encoding of contextual cues in this first, activation, component allows
and sometimes even confrontational, aggressive, or coercive (or for information about the social environment to frame the nature and
at least perceived that way). For example, an authority figure may quality of the impending interaction.
518 MATSUMOTO AND HWANG

This first component, activation, is most relevant to the authority the previous study demonstrated that social influence effects were
manipulation utilized in this study in that the encoding of contextual moderated by veracity condition, we predicted that veracity condition
cues allows for information about the social environment to aid will moderate the effects of authority as well.
respondents in understanding that accurate information is about to In sum, we tested the following hypotheses:
be sought, facilitating the retrieval and encoding of that information
Hypothesis 1. That higher authority will produce higher
to working memory. That is, subtle, indirect contextual cues to author-
rapport ratings, more relevant details and more plausibil-
ity such as the trappings of authority as operationalized in this study
ity in interviewee responses in both the interviews and
should implicitly prime the concept of authority during activation. That
written statements.
priming should engage an existing neural‐cognitive network related to
authority and associated concepts such as obedience, disclosure, Hypothesis 2. That rapport will have direct effects on
cooperation, or compliance. Coupled with the notion that activation the informational elements and will mediate the relation-
also causes respondents to realize that accurate information is being ship between authority and the informational elements in
sought by others, truthful facts should enter working memory more the interviews and written statements.
easily, facilitating recall and reporting.
Hypothesis 3. That veracity condition will moderate the
relationship between authority and the informational ele-
ments in the interviews and written statements.
1.2 | Overview and hypotheses

The goal of this study was to examine the causal influence of subtle, 2 | METHODS1
indirect, contextual operationalizations of authority on rapport and
informational outcomes in investigative interviews. A community
2.1 | Design
sample participated in a mock crime experiment that was the same
used in the previous study described above that tested the effects The experiment tested a moderated mediational model with authority
of reciprocity (Matsumoto & Hwang, in press), allowing a direct (high vs. low) and veracity (truth vs. lie) conditions as factors, three
comparison of different principles of social influence. Participants informational elements coded from interview transcripts and written
were randomly assigned either to steal a check and lie about it or to statements as dependents, and rapport rated by the interviewers
just look at the check and tell the truth. After committing the mock and interviewees as a mediator. We also included gender as a factor
crime or not, they participated in an investigative interview that was in overall analyses to determine if it moderated any findings although
audio‐video recorded. We coded the same three types of informa- we did not have specific hypotheses about it.
tional elements—relevant and irrelevant details and plausibility—from
the responses to the main question in that interview. We also coded
2.2 | Participants
the same informational elements in a written statement provided
by the interviewee prior to the interviews (this written statement
A community sample responded to online ads and posted flyers
was not coded in Matsumoto & Hwang, in press, as it occurred prior
recruiting participants for a study examining cultural differences in
to the reciprocity manipulation). Rapport was independently rated by
how people feel when going through security interviews. A total
interviewers and interviewees.
N = 120 U.S. born‐and‐raised individuals whose first language was
On the basis of the findings of the previous study described above
English participated (n = 63 females, mean age = 42.47; n = 57 males,
(Matsumoto & Hwang, in press) and on the framework described
mean age = 42.59). All were randomly assigned to the authority and
earlier concerning subtle, indirect manipulations of authority and its
veracity conditions.
possible effects on activation, we predicted that higher authority
would produce more relevant details, less irrelevant details, and more
plausibility in the interviewees' responses. With regard to rapport, 2.3 | Measures
different effects were conceptually possible. On one hand, higher
At the beginning of the experiment, participants completed a series of
authority may result in less rapport between the interactants, as the
questionnaires including a brief demographic questionnaire, the Gen-
authority manipulation may produce perceived constraints on the
eral Ethnicity Questionnaire (GEQ; Tsai, Ying, & Lee, 2000), the Brief
interactions, thereby lowering rapport. On the other hand, higher
NEO‐Five Factor Inventory (Costa & McCrae, 1992), the Social Dom-
authority may produce better rapport if authority is perceived by
inance Orientation Scale (Pratto, Sidanius, Stallworth, & Malle, 1994),
interviewees as signs of qualifications or eligibility and not as
an adapted version of the Schwartz Value Scale (Schwartz, 2006),
constraints. For the purposes of this study and on the basis of the
the Satisfaction with Life Scale (Diener, Emmons, Larsen, & Griffin,
positive findings from the previous study and the conceptual frame-
1985), and the Self‐Monitoring Scale (Snyder, 1974). Participants also
work above, we predicted that higher authority will produce higher
completed an emotion checklist at the beginning and end of the
rapport ratings.
experiment, which included 15 emotion words (guilt, fear, anger,
If higher authority has an effect on informational elements and
rapport, we further predicted that rapport will mediate the relation 1
Portions of the methodology have been previously described in Hwang,
between authority and the informational elements. And given that Matsumoto, and Sandoval (2016) and Matsumoto and Hwang (in press).
MATSUMOTO AND HWANG 519

embarrassment, worry, contempt, excitement, disgust, amusement, two posters with the name of the same law enforcement agency.
nervousness, surprise, interest, sadness, pride, and shame) rated on Interviewers also wore a suit and tie with dress shoes and a laminated
9‐point scales (0 = none, 4 = moderate amount, and 8 = extremely photo ID on a lanyard. In the low authority condition, the interview
strong). All measures except the emotion checklist were omitted from room contained the same certificates and photograph but they were
this study. scrambled to be unintelligible (controlling for color and size). The inter-
view room also contained two posters from the neutral images of the
Interactional Affective Picture System (IAPS; Lang, Bradley, &
2.4 | Procedure Cuthbert, 1997); the interviewers wore a plain white shirt with slacks
and walking shoes and no photo ID.
After consenting, participants completed the pre‐session measures The authority condition was operationalized, therefore, using sub-
and then were informed of their assignment to a veracity condition. tle, indirect cues in the context and environment. Importantly, the
The truth condition required participants to look at but not take a interview protocols themselves did not differ according to condition:
check made out to cash and tell the truth about what they did. The The high authority condition did not include more confrontational or
lie condition required them to take the check and lie about it. Partici- aggressive interview questions or tone than the low authority condi-
pants were told that they would be interviewed regarding what they tion. The only differences between the conditions were the subtle
did, would have to persuade the interviewers about their honesty, trappings of authority placed in context. The interviews were checked
and would earn a minimum of $30 for their participation and bonuses twice—in real time and later—by two research assistants blind to the
of anywhere from $0 to $50 depending on their assigned condition hypotheses to confirm that the same questions were administered in
and interviewers' determinations. They were also told that if they all sessions.
were not believed, they would have to stay an extra hour to complete After the interview was completed, participants went back to the
a long questionnaire; if they were believed, they could leave early. initial area to complete the post‐session measures. During this time, a
Actually, no determinations were made, all participants received a randomly selected subsample (n = 40) was asked to rate how much
standard fee of $40, and no one was detained. As a manipulation authority they perceived the interviewer had using a scale labeled 0,
check, participants rated the severity of these expected consequences no authority at all, through 5, moderate authority, to 10, high authority.
if they were judged to be lying using a scale from 0, no consequence, to After completing all post‐session measures, all participants were
10, maximum consequence. The mean was 7.03 (SD = 3.13), suggesting debriefed and paid. Experiments generally lasted between 75 and
that the participants perceived the stakes on at least a moderate level. 90 min.
After participants confirmed their understanding, they partici-
pated in an initial screening interview to ascertain their intent to com- 2.5 | Coding procedures and reliability
mit a crime. They then went alone to a room where the check was
located, stole the check or not, and then exited to wait for the next Participants' responses in the written statement and interview were
step. An interviewer then escorted the participants to an interview transcribed verbatim. We coded their entire response in the written
room and asked them to write a statement about what they did in statement and their response to the following prompt from the inter-
the previous room (participants wrote the statement alone). This view: “I've read your statement but I want to hear in your own words
statement was included because in many real‐life interviews, what you did in the file room. Tell me what you did.” This prompt was
interviewees are often asked to write a statement prior to the the same as was used when participants wrote their written state-
interview and these written statements can serve useful purposes in ment, ensuring that the data were comparable.
the interview. Also, because most previous research focused on the
informational elements generated in an interview, we included 2.5.1 | Informational elements
the written statements as an additional source of information to
examine our hypotheses. After completing the written statement, Participants' responses for relevant details, irrelevant details, and plau-
the interviewer returned and an interview proceeded using a standard sibility were coded using the same procedures as reported in the
protocol (all interviewees were asked the same questions). Three previous study testing reciprocity (Matsumoto & Hwang, in press).
male interviewers, blind to the veracity condition, utilized a standard Relevant details were any reported detail about what the participant
protocol to conduct the interviews (the first author was one of did, that likely occurred during the experiment, and was relevant to
the interviewers). the question being answered or the context (i.e., in relation to the
The authority condition was operationalized using subtle, indirect actual instructions; e.g., “I opened the envelope,” “There was a
trappings of authority (Cialdini, 2009; Cialdini & Goldstein, 2004a) check.”). Irrelevant details were any detail that may have occurred dur-
placed in the context of the interview room. In the high authority con- ing the experiment but was not relevant to the question being asked
dition, a wall in the interview room was adorned with two certificates or the context (e.g., “The carpet was brown”) or was likely not to have
from a federal law enforcement agency in a fictitious name of the occurred during the experiment (e.g., “I met someone”). Plausibility
interviewer, both attesting to the interviewer's competence (one for was defined as the degree to which the events described in the
exceptional service in the public interest, the other for merit). The responses were likely to have occurred in reality for an average person
interview room also included a photoshopped image of the inter- in the situation with the instructions given (i.e., whether the coder
viewer with a well‐known government law enforcement figure and believed the story of the participant). Plausibility was coded on a
520 MATSUMOTO AND HWANG

7‐point scale labeled 0 (not plausible at all), 1 (minimally plausible), 3 interactions were significant, F (14, 1232) = 13.05, p < 0.001,
(moderately plausibly), and 6 (maximally plausible). ηp2 = 0.129 and F (14, 1232) = 4.44, p = 0.001, ηp2 = 0.048, respec-
Two coders blind to the study's hypotheses and participants' con- tively. Neither authority nor gender produced any significant effects.
ditions independently coded all responses. The number of relevant Simple effects analyses indicated that liars increased in guilt, fear,
and irrelevant details in each response for each case was tallied. anger, embarrassment, worry, contempt, disgust, nervousness, sur-
Reliabilities were high and acceptable for all codes: For relevant and prise, sadness, and shame, and the effect sizes were moderate‐high
irrelevant details, ICCs = 0.99 and 0.98 (interview) and 0.99 and 0.98 (ds = −0.84, −0.77, −0.29, −0.56, −0.51, −0.31, −0.44, −0.89, −0.39,
(written statement), respectively; for plausibility, rs = 0.91 and 0.90 −0.27, and −0.65, respectively). Truth tellers also increased in fear,
for interview and written statement, respectively. All codes were anger, embarrassment, contempt, nervousness, and surprise
averaged across coders for analyses. (ds = −0.38, −0.30, −0.35, −0.28, −0.40, and −0.64, respectively) but
decreased in excitement (d = 0.45). Thus, participants were emotion-
ally engaged in the study.
2.5.2 | Rapport

Different methods of assessing rapport in an interaction exist


(Bernieri, 1988; Bernieri, Davis, Rosenthal, & Knee, 1994; Bernieri &
3.1.2 | Relevant details, irrelevant details, and
plausibility
Gillis, 1995; Bernieri, Reznick, & Rosenthal, 1988; Grahe & Bernieri,
1999; Tickle‐Degnen & Rosenthal, 1987, 1990; Trout & Rosenfeld,
We computed univariate, three‐way, between‐subject ANOVAs using
1980), each with advantages and disadvantages. In this study, we were
authority, gender, and veracity conditions as factors on the coded
interested in perceptions of rapport from the interactants (as opposed
informational elements and rapport, separately for the interview and
to third‐party observer ratings) because interactants' perceptions may
written statement. The veracity condition main effects served as a
have been differentially related to interview outcomes. To facilitate
manipulation check on the coding procedures for the dependents
the ease by which the interviewers and interviewees rated rapport,
and were significant for relevant details and plausibility for both the
they independently rated rapport in the interactions they had with
interview, F (1, 108) = 76.42, p < 0.001, ηp2 = 0.414 and F (1,
each other using a 10‐point scale labeled 0, none at all, 5, a moderate
108) = 94.2, p < 0.001, ηp2 = 0.466, and written statement, F (1,
amount, and 10, maximum amount. We used a single, overall rating
108) = 86.10, p < 0.001, ηp2 = 0.441 and F (1, 108) = 142.41,
of rapport because a factor analysis of multiple dimensions of rapport
p < 0.001, ηp2 = 0.566. Truth tellers' responses had more relevant
coded in our previous work in investigative interviews (Matsumoto &
details and were more plausible, which were to be expected given
Hwang, in press) produced a single factor solution.
the coding rules (see Table 1 for descriptives by authority condition).
Interviewers rated rapport twice, immediately after the initial
This main effect was not significant for irrelevant details for either
screening interview (not part of this study) and after the second,
the interview or written statement, F (1, 108) = 0.75, p = 0.389,
investigative interview; participants rated rapport once, immediately
ηp2 = 0.007 and F (1, 108) = 0.77, p = 0.383, ηp2 = 0.007, respectively.
after the investigative interview and before debriefing. These ratings
Identifying veracity by authority condition interactions allowed
occurred after the interviews, rendering interpretations concerning
us to select the appropriate moderated mediational model to test
when and how rapport was built impossible, which is a limitation. Mul-
tiple ratings of rapport during the interviews, however, would have
TABLE 1 Means and standard errors for the dependent variables
been logistically difficult, cumbersome, and artificial. Because there
(coded informational elements) and mediator (rapport) by authority
were significant mean differences among the interviewers' rapport rat-
condition
ings, F (2, 115) = 46.74, p < 0.001, ηp2 = 0.448 and F (2, 114) = 12.72,
High authority Low authority
p < 0.001, ηp2 = 0.182, for the first and second ratings, respectively, Truth tellers Liars Truth tellers Liars
we standardized the ratings within each interviewer for use below.
Interview Relevant 16.45 2.99 10.01 3.62
details (1.15) (1.12) (1.14) (1.13)
Irrelevant 19.07 16.50 8.66 15.46
details (2.47) (2.41) (2.46) (2.44)
3 | RESULTS Plausibility 4.25 0.98 3.90 1.38
(0.30) (0.29) (0.30) (0.30)
Written Relevant 11.32 1.65 8.28 2.61
3.1 | Manipulation checks and preliminary analyses
statement details (0.85) (0.82) (0.83) (0.81)
Irrelevant 7.48 6.20 5.24 8.61
3.1.1 | Emotion ratings details (1.23) (1.18) (1.19) (1.17)
Plausibility 4.57 0.82 4.08 1.49
(0.27) (0.26) (0.27) (0.26)
We examined whether participants were emotionally engaged by
Rapport
computing a five‐way, mixed ANOVA on their emotion ratings using
Interviewer first 0.14 −0.42 −0.21 −0.42
gender (2), veracity condition (2), and authority condition (2) as rapport rating (0.20) (.20) (.20) (.20)
between‐subject variables and pre‐post (2) and emotion (15) as Interviewer second 0.64 −0.18 0.32 0.12
repeated measures. (As mentioned above, gender was included in rapport rating (0.23) (.22) (.22) (.22)
the overall analyses to determine if it moderated any results.) The Participants' 6.04 6.32 6.14 5.71
rapport rating (0.50) (0.49) (0.49) (0.49)
pre‐post by emotion and pre‐post by emotion by veracity condition
MATSUMOTO AND HWANG 521

the hypotheses. This interaction was significant or marginally signifi- 3.1.5 | Intercorrelations
cant on relevant and irrelevant details from the interview and all
three informational elements from the written statements, F (1, We computed intercorrelations among the informational elements and
108) = 9.68, p = 0.002, ηp2 = 0.082; F (1, 108) = 3.68, p = 0.058, rapport ratings (Table 2). These analyses were informative because
ηp2 = 0.033; F (1, 108) = 5.87, p = 0.017, ηp 2
= 0.051; F (1, interpretation of any findings from univariate analyses on the informa-
108) = 3.83, p = 0.053, ηp2 = 0.034; F (1, 108) = 4.80, p = 0.031, tional elements would be informed by intercorrelations among those
ηp2 = 0.042, respectively. That interaction was not significant on plau- three variables. Also, because we included multiple ratings of rapport,
sibility in the interviews, F (1, 108) = 1.55, p = 0.216, ηp2 = 0.014. elucidating their relations would be helpful for understanding the main
Thus, we incorporated veracity as a moderator in five of the main anal- analyses below. Identifying which rapport ratings were associated with
yses below. Gender did not interact with authority and was thus the informational elements also guided decisions concerning which to
dropped from further analyses. use in the main analyses. The lack of associations between any rapport
rating and the informational elements would also be informative.
3.1.3 | Perceived authority Unsurprisingly, relevant details were correlated with plausibility in
both the interview and written statements. Relevant details in the
We computed a two‐way ANOVA on the perceived authority ratings interviews were also correlated with irrelevant details. With one
using veracity and authority conditions as independent variables. No exception, relevant details and plausibility from the interviews and
effect involving authority was significant, F (1, 36) = 0.01, p = 0.934, written statements were correlated with the interviewers' first
ηp2 = 0.000 and F (1, 36) = 0.42, p = 0.521, ηp2 = 0.012, for the main and second rapport ratings. Interestingly, participants' rapport ratings
effect and interaction, respectively, indicating that participants did not were not correlated with any of the informational elements or the
consciously perceive a difference between the authority conditions. interviewers' rapport ratings.
Inspection of the cell means of the veracity by authority interaction,
however, suggested differential trends for effects of authority by 3.2 | Moderated mediational analyses testing
veracity condition (albeit nonsignificant). Truth tellers tended to rate Hypotheses 1–3
authority higher in the high authority condition (Mhigh = 8.50,
SD = 2.59; Mlow = 7.78, SD = 3.11), whereas liars tended to rate We tested hypotheses using SPSS PROCESS with 1,000 bootstraps
authority higher in the low authority condition (Mhigh = 7.73, SD = 3.55; (Hayes, 2013). For relevant and irrelevant details from the interview
Mlow = 8.29, SD = 2.14). These trending differential effects for truth and relevant and irrelevant details and plausibility ratings from the
tellers and liars may inform interpretations about veracity condition written statements, we used Model 5 (veracity moderating the relation
differences below. between the independent variable and the dependent), based on the
preliminary analyses above. For plausibility in the interviews, we used
3.1.4 | Rapport Model 4. In all analyses, we included the interviewers' second rapport
rating as the mediator as this was the only rapport rating correlated
Veracity by authority ANOVAs on the rapport ratings indicated that with the dependents.
the veracity main effect was marginally significant and significant
for the interviewers' first and second ratings, respectively, F (1, 3.2.1 | Interview
110) = 3.77, p = 0.055, ηp2 = 0.033; F (1, 109) = 5.27, p = 0.024,
ηp2 = 0.046; both were higher in the truth condition than the lie The overall models were significant for relevant details, irrelevant
condition. No other effects were significant for the interviewers' details, and plausibility, R(115) = 0.69, F (4, 110) = 25.46,
ratings. Interestingly, no effects were significant for participants' p < 0.001; R(115) = 0.42, F (4, 110) = 5.79, p = 0.001; and
rapport ratings (Table 1). R(115) = 0.30, F (2, 112) = 5.46, p = 0.005, respectively. Authority

TABLE 2 Intercorrelations among the dependent variables and rapport ratings


Interview Rapport ratings
Relevant details Irrelevant details Plausibility Participants' ratings Interviewers' first Interviewers' second
(A) (B) (C) (D) ratings (E) ratings (F)

A 0.22* 0.82** −0.08 0.20* 0.39**


B −0.13 −0.02 0.11 0.29**
C −0.09 0.16 0.30**
D 0.04 0.00
E 0.51**
Written statement
A 0.06 0.87** −0.06 0.22* 0.37**
B −0.14 −0.10 −0.04 0.05
C −0.05 0.21* 0.37**

*p < 0.05. **p < 0.01.


522 MATSUMOTO AND HWANG

had significant, direct effects on relevant and irrelevant details; statements. But a marginally significant veracity by authority interac-
higher authority produced more of both types of details (see tion again qualified this interpretation and conditional direct effects
Table 3). Significant and marginally significant veracity by authority indicated that authority had significant effects on relevant details for
interactions, however, qualified these interpretations and conditional truth tellers but not liars.
direct effects indicated that authority had significant effects on rele- Rapport again had significant, direct effects on relevant details
vant and irrelevant details for truth tellers but not liars. Authority did and plausibility (see Table 4), indicating that interviewers' perceptions
not affect plausibility ratings. of greater rapport were associated with more relevant details and
Rapport had significant, direct effects on all three informational higher plausibility in the written statements. As reported above,
elements; interviewers' perceptions of greater rapport were associ- authority did not affect interviewer rapport ratings and rapport did
ated with more details and higher plausibility in those details. But not mediate the effects of authority on the informational elements
authority did not affect the interviewers' rapport ratings in the written statements.
R(110) = 0.01, F (1, 113) = 0.004, p = 0.951, and rapport did not medi-
ate the effect of authority on the informational elements.
3.3 | Post hoc analyses
3.2.2 | Written statement
We examined whether the different interviewers affected the infor-
The overall models were significant for relevant details and plausibility, mational elements in the interviews or written statements by comput-
R(115) = 0.70, F (4, 110) = 26.05, p < 0.001 and R(115) = 0.77, F (4, ing three‐way ANOVAs involving interviewer, veracity, and authority
110) = 41.12, p < 0.001, respectively, but not for irrelevant details, conditions as between‐subject factors. No effect involving interviewer
R(110) = 0.20, F (4, 110) = 1.18, p = 0.322. Authority had significant, was significant in any analyses.
direct effects on relevant details, indicating that perceptions of In order to examine possible differences between the interviews
authority produced more relevant details even in the written and written statements in the amount of details produced, we

TABLE 3 Moderated mediational analyses on the informational elements coded from the interviews
Coefficient SE t p LLCI ULCI
Relevant details
Direct effects
Interviewer rapport rating 1.71 0.49 3.46 0.001 0.73 2.68
Authority 11.71 3.65 3.21 0.002 4.49 18.94
Veracity −0.05 3.62 −0.01 0.990 −7.23 7.14
Veracity by authority −5.96 2.30 −2.59 0.011 −10.51 −1.40
Conditional direct effects
Truth tellers 5.76 1.63 3.53 0.001 2.53 8.98
Liars −0.20 1.61 −0.12 0.902 −3.40 3.00
Indirect Effects
Interviewer rapport rating 0.02 0.39 −0.86 0.73
Irrelevant details
Direct effects
Interviewer rapport rating 3.47 1.00 3.48 0.001 1.49 5.44
Authority 17.08 7.37 2.32 0.022 2.48 31.67
Veracity 15.99 7.32 2.18 0.031 1.48 30.50
Veracity by authority −7.82 4.65 −1.68 0.096 −17.03 1.40
Conditional direct effects
Truth tellers 9.26 3.29 2.81 0.006 2.74 15.78
Liars 1.44 3.26 0.44 0.659 −5.02 7.90
Indirect effects
Interviewer rapport rating 0.05 0.77 −1.64 1.48
Plausibility
Direct effects
Interviewer rapport rating 0.53 0.16 3.29 0.001 0.21 0.85
Authority −0.13 0.39 −0.34 0.735 −0.90 0.64
Total and indirect effects
Total effects of authority −0.12 0.40 −0.31 0.759 −0.92 0.68
Interviewer rapport rating 0.01 0.13 −0.24 0.29
MATSUMOTO AND HWANG 523

TABLE 4 Moderated mediational analyses on the informational elements coded from the written statements

Coefficient SE t p LLCI ULCI


Relevant details
Direct effects
Interviewer rapport rating 1.14 0.37 3.10 0.002 0.41 1.86
Authority 5.24 2.63 1.99 0.049 0.02 10.46
Veracity −2.80 2.58 −1.08 0.281 −7.92 2.32
Veracity by authority −2.92 1.66 −1.76 0.081 −6.21 0.37
Conditional direct effects
Truth tellers 2.32 1.18 1.97 0.051 −0.02 4.65
Liars −0.61 1.16 −0.52 0.603 −2.90 1.70
Indirect effects
Interviewer rapport rating 0.07 0.26 −0.44 0.58
Irrelevant details
Direct effects
Interviewer rapport rating 0.26 0.53 0.48 0.631 −0.80 1.31
Authority 6.27 3.82 1.64 0.104 −1.31 13.84
Veracity 7.73 3.75 2.06 0.041 0.31 15.16
Veracity by authority −4.35 2.41 −1.80 0.074 −9.12 0.43
Conditional direct effects
Truth tellers 1.92 1.71 1.12 0.263 −1.47 5.31
Liars −2.43 1.68 −1.44 0.150 −5.76 0.91
Indirect effects
Interviewer rapport rating 0.02 0.13 −0.15 0.43
Plausibility
Direct effects
Interviewer rapport rating 0.36 0.12 3.12 0.002 0.13 0.59
Authority 1.20 0.83 1.44 0.151 −0.45 2.84
Veracity −1.67 0.81 −2.05 0.042 −3.28 −0.06
Veracity by authority −0.88 0.52 −1.68 0.096 −1.91 0.16
Conditional direct effects
Truth tellers 0.32 0.37 0.87 0.389 −0.41 1.06
Liars −0.56 0.37 −1.52 0.131 −1.28 0.17
Indirect effects
Interviewer rapport rating 0.02 0.079 −0.14 0.18

computed four‐way, mixed ANOVAs using authority, gender, and the interviews. Interviewers' second rapport ratings were positively
veracity conditions as between‐subject factors and source (interviews correlated with increases in both relevant and irrelevant details,
vs. written statements) as a repeated measure, separately on the infor- r(101) = 0.28, p = 0.005 and r(101) = 0.32, p = 0.001, respectively.
mational elements. For these analyses, we were only interested in Correlations using participants' rapport ratings were not significant.
effects involving source. The source main effect was significant for Thus, better rapport as perceived by the interviewers was associ-
both relevant and irrelevant details, F (1, 106) = 61.61, p < 0.001, ated with more information elicited in the interviews than in the
ηp2 = 0.368 and F (1, 106) = 61.09, p < 0.001, ηp2 = 0.366, respec- written statements.
tively; more of both were produced in the interviews than written We then examined whether authority produced increases in details
statements. This interpretation was qualified by a significant source by computing two‐way authority by veracity condition ANOVAs on the
by veracity by authority interaction on relevant details and a signifi- change in relevant and irrelevant details. Authority condition main
cant source by authority interaction on irrelevant details, F (1, effects were significant for both, F (1, 98) = 8.29, p = 0.005, ηp2 = 0.078
106) = 7.20, p = 0.008, ηp2 = 0.064 and F (1, 106) = 7.68, p = 0.007, and F (1, 98) = 9.41, p = 0.003, ηp2 = 0.088; the authority condition pro-
ηp2 = 0.068, respectively. Cell means (Table 1), however, indicated dif- duced increases in both types of details. The interpretation for relevant
ferences in the degree of difference. There were no significant effects details was qualified by a significant authority by veracity condition
involving source on plausibility ratings. interaction, F (1, 98) = 6.80, p = 0.011, ηp2 = 0.065; simple effects
Because the interviews produced more details, we examined indicated that the authority condition produced increases in relevant
whether rapport was associated with increases in details by correlat- details for truth tellers but not liars, F (1, 46) = 9.68, p = 0.003,
ing rapport and the change in details from the written statement to ηp2 = 0.174 and F (1, 52) = 0.07, p = 0.796, ηp2 = 0.001, respectively.
524 MATSUMOTO AND HWANG

4 | DISCUSSION case. Liars can be cooperative (and compliant) but not entirely forth-
right or credible in the contents provided. The above argument
Hypotheses 1 and 3 were partially supported: The authority condition depends on whether the concept of cooperation requires truthfulness
produced more relevant and irrelevant details in the interview and more as a component. Conceptualizing cooperation as a “working alliance”
relevant details in the written statements but just for truth tellers and or “operational accord,” for example, does not necessarily require
not liars, suggesting that the influence of authority was different for truthful responses (although such alliances make truthful responses
them. Authority did not affect plausibility. Hypothesis 2 was also more likely). The nondifferences in participants' rapport ratings further
partially supported: Rapport had direct effects on all informational raise questions about whether cooperation may have been the reason
elements from the interview and on relevant details and plausibility in why authority did not affect liars.
the written statements. But authority did not affect rapport and rapport Authority did not affect plausibility but these nonfindings likely
did not mediate the effect of authority on the informational elements. occurred because of the increased amount of irrelevant details
These results added to a growing literature demonstrating the produced by the authority operationalization, reducing the overall plau-
effects of social influence in investigative interviews. For truth tellers, sibility of the responses. In fact, ratios of irrelevant to relevant details
the effects were fairly strong and robust; essentially, the trappings of were correlated with plausibility ratings, r(97) = −0.60, p < 0.001 and
authority facilitated them to talk more. The same was true for relevant r(93) = −0.55, p < 0.001, for interviews and written statements, respec-
details in the written statements, which was interesting because these tively, and these correlations were likely underestimates because they
were produced without the interviewer present (although the effects did not include data with zero values for either type of detail.
were marginally significant and the CIs overlapped with zero). The The effects of interviewer‐rated rapport were fairly robust on all
trappings of authority also facilitated truth tellers to provide even informational elements from the interviews and on relevant details
more details in the interviews compared with their written statements, and plausibility from the written statements (all CIs did not overlap
and all effects occurred across different interviewers and interviewee with zero). That rapport had direct effects on the interviews was not
genders. These findings were somewhat remarkable given that the surprising, given that the informational elements produced by the
authority operationalization was subtle and implicit and that the inter- interviewees and the interviewers' rapport ratings were based on their
viewees themselves did not report a difference in perceptions of interaction. That interviewers' rapport ratings also had direct effects
authority between the conditions. on the written statements was somewhat surprising given that the
Theoretically, the trappings may have primed the concept of statements were produced at the beginning of the interviews when
authority during the activation phase of a four‐component framework interviewees were alone whereas rapport ratings were obtained after
of the cognitive mechanisms associated with deceptive contexts the interviews were completed. There are at least two interpretations
(Walczyk et al., 2003; Walczyk et al., 2014). Such priming can infuse for this finding: The developing rapport between the interactants prior
judgment and decision‐making processes with concepts related to to the request to write the statements may have affected production
authority such as obedience, disclosure, cooperation, or compliance. of relevant details in the statements written thereafter. The positive
Given that activation facilitates respondents in realizing that truthful correlations between the interviewers' first rapport ratings and the
information is being sought, truthful facts should have been more relevant details and plausibility in the written statements produced
easily recalled and reported. subsequently are suggestive of this possibility. Or the greater produc-
But the findings suggested that different components of the con- tion of relevant details in the statements influenced better rapport
cept of authority may have been primed for truth tellers and liars. For outcomes in the subsequent interviews. Unfortunately, the way the
truth tellers, the authority priming may have activated the cooperation rapport ratings were obtained did not allow for any conclusive inter-
component of authority. If truth tellers are akin to cooperative eyewit- pretations of when and how rapport was built. Regardless, these
nesses, the findings for truth tellers make sense and were likely not at findings contributed to a growing literature on the importance of
odds with their preexisting framework of cooperation, facilitating rapport in investigative interviews (cited earlier).
information retrieval and reporting. Liars, however, may have been Participant rapport ratings were not related to any of the informa-
uncooperative to begin with in order to conceal their wrongdoing. tional elements or interviewer rapport ratings, raising questions about
As such priming the cooperation component of authority may have differential effects of different sources of rapport in investigative inter-
been at odds with their preexisting, noncooperative cognitive frame- views. These nonfindings suggested that the effects of authority on the
work and the authority manipulation may have primed a component informational elements occurred regardless of interviewees' percep-
other than cooperation, such as compliance; thus, priming authority tions of the quality of the interactions. For them, the interview context
appeared to have different effects on liars. This speculation is sup- may have been sufficiently charged so that the press of the context
ported by the differential albeit nonsignificant trends in authority rat- diminished effects of their perceptions of rapport. Interviewers had a
ings for truth tellers and liars reported in the manipulation check. Also, different perspective and may have been attuned to greater operational
if cooperation is considered a component of rapport, the differences in accord in the interaction, which may account for the positive effects
interviewers' ratings of rapport between veracity conditions lend involving their ratings. Alternative interpretations may also exist.
some credence to this interpretation because their rapport ratings The current findings urge practitioners to consider subtle manipu-
were higher in the truth condition. lations in their operational contexts that may facilitate information
But characterizing truth tellers as cooperative and liars as not elicitation. Anecdotally, most investigators want their interviewees to
assumes that liars are not cooperative, which is not necessarily the talk more so that claims can be verified and ground truth more easily
MATSUMOTO AND HWANG 525

obtained. Even subtle, implicit authority operationalizations may aid in RE FE RE NC ES


gathering information from truth tellers, which may render clearer Bernieri, F. J. (1988). Coordinated movement and rapport in teacher–
indicators of who should be ruled out in certain investigative contexts, student interactions. Journal of Nonverbal Behavior, 12(2), 120–138.
https://doi.org/10.1007/BF00986930
especially those dealing with multiple persons of interest. At the same
Bernieri, F. J., Davis, J. M., Rosenthal, R., & Knee, C. R. (1994). Interactional
time, our findings suggest that practitioners be wary of the possible
synchrony and rapport: Measuring synchrony in displays devoid of
effects of similar authority manipulations because more irrelevant sound and facial affect. Personality and Social Psychology Bulletin, 20,
information can also be produced. To be sure, irrelevant information 303–311. https://doi.org/10.1177/0146167294203008
may not necessarily be false; the information may actually have Bernieri, F. J., & Gillis, J. S. (1995). The judgment of rapport: A cross‐
cultural comparison between Americans and Greeks. Journal of Nonver-
occurred but was just not related to the question or context. The fact
bal Behavior, 19(2), 115–130. https://doi.org/10.1007/BF02173170
that irrelevant information increased from the written statement to
Bernieri, F. J., Reznick, S. J., & Rosenthal, R. (1988). Synchrony,
the interview for liars suggested that they were more compliant but pseudosunchrony, and dissynchrony: Measuring the entrainment
not necessarily more cooperative (as suggested above). process in mother‐infant interactions. Journal of Personality and
These findings were not produced without limitations, the first Social Psychology, 54(2), 243–253. https://doi.org/10.1037/0022‐
3514.54.2.243
concerning the operationalization of authority using the trappings of
Blass, T. (1999). The Milgram Paradigm after 35 years: Some things we
authority. This operationalization did not differentiate between status
now know about obedience to authority. Journal of Applied Social Psy-
and power nor did it require any direct intervention by the inter- chology, 29(5), 955–978. https://doi.org/10.1111/j.1559‐1816.1999.
viewers. Also, authority did not affect rapport and rapport did not tb00134.x
have a mediational effect. Future studies should examine other direct Cialdini, R. B. (2009). Influence: Science and practice (5th ed.). Harlow;
London: Pearson Education.
and indirect operationalizations of authority and their effects on
Cialdini, R. B., & Goldstein, N. J. (2004a). Social influence: Compliance and
both information produced and rapport. Different operationalizations
conformity. Annual Review of Psychology, 55, 591–621. https://doi.org/
including changes in questions and interviewer demeanor that 10.1146/annurev.psych.55.090902.142015
directly demonstrate authority, status, or power may influence infor- Cialdini, R. B., & Goldstein, N. J. (2004b). Social Influence: Compliance and
mation and rapport in either positive or negative ways and these conformity. Annual Review of Psychology, 55(1), 591–621. https://doi.
potential influences deserve research attention. Future studies will org/10.1146/annurev.psych.55.090902.142015

also need to examine operationalizations that produce positive Costa, P. T., & McCrae, R. R. (1992). Revised Neo‐Personality Inventory
(NEO‐PI‐R) and Neo Five Factor Inventory (NEO‐FFI). Odessa, FL:
effects in both truth tellers and liars as well as different levels of
Psychological Assessment Resources.
intensity of authority.
Diener, E., Emmons, R. A., Larsen, R. J., & Griffin, S. (1985). The Satisfaction
The different effects of interviewer and interviewee rapport we with Life Scale. Journal of Personality Assessment, 49, 71–75. https://
obtained suggested that researchers incorporate multiple measures doi.org/10.1207/s15327752jpa4901_13
of rapport so that the effects of rapport can be differentially examined Goodman‐Delahunty, J., & Howes, L. M. (2016). Social persuasion to
develop rapport in high‐stakes interviews: Qualitative analyses of
using different assessments of rapport. Future research should
Asian‐Pacific practices. Policing and Society, 26(3), 270–290. https://
examine the effects of different sources of rapport on the elements, doi.org/10.1080/10439463.2014.942848
effectiveness, and efficiency of interviews. Goodman‐Delahunty, J., Martschuk, N., & Dhami, M. K. (2014).
Finally, the findings were limited to one type of crime (mock crime Interviewing high value detainees: Securing cooperation and disclo-
of theft) in a laboratory context; the effects of social influence need to sures. Applied Cognitive Psychology, 28(6), 863–897. https://doi.org/
10.1002/acp.3087.
be tested in other contexts as well. We also tested participants who
Grahe, J. E., & Bernieri, F. J. (1999). The importance of nonverbal cues in
probably had no previous experience with actual investigative situa-
judging rapport. Journal of Nonverbal Behavior, 23(4), 253–269.
tions; thus, the findings were limited to those who were relatively https://doi.org/10.1023/A:1021698725361
naïve about such contexts (although people may have ideas about Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional
such experiences through media). Future studies should examine the process analysis: A regression based approach. New York: Guilford Press.
possibility that the findings might differ for people with experience Hwang, H. C., Matsumoto, D., & Sandoval, V. A. (2016). Linguistic cues of
deception across multiple language groups in a mock crime context.
in actual investigative contexts.
Journal of Investigative Psychology and Offender Profiling, 13(1), 56–69.
https://doi.org/10.1002/jip.1442
ACKNOWLEDGEMEN TS Kieckhaefer, J. M., Vallano, J. P., & Schreiber Compo, N. (2013). Examining
the positive effects of rapport building: When and why does rapport
The authors thank Vincent Sandoval for his suggestions about the building benefit adult eyewitnesses. Memory, 22(8), 1010–1023.
potential effect of principles of persuasion and influence in investiga- https://doi.org/10.1080/09658211.2013.864313
tive interviews. Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (1997). International
This work was funded by the High‐Value Detainee Interrogation Affective Picture System (IAPS): Technical manual and affective ratings.
Washington, DC: NIMH Center for the study of emotion and attention.
Group contract DJF‐15‐1200‐V‐0009510. Statements of fact, opinion,
Matsumoto, D., & Hwang, H. C. (2018). Social influence in investigative
and analysis in the paper are those of the authors and do not reflect the
interviews: The effects of reciprocity. Applied Cognitive Psychology,
official policy or position of the FBI or the U.S. Government. 32, 163–170. https://doi.org/10.1002/acp.3390
Meissner, C. A., Hartwig, M., & Russano, M. B. (2010). The need for a pos-
itive psychological approach and collaborative effort for improving
ORCID
practice in the interrogation room. Law and Human Behavior, 34(1),
David Matsumoto https://orcid.org/0000-0002-3186-1503 43–45. https://doi.org/10.1007/s10979‐009‐9205‐9
526 MATSUMOTO AND HWANG

Meissner, C. A., Kelly, C. E., & Woestehoff, S. A. (2015). Improving the effective- Tsai, J. L., Ying, Y.‐W., & Lee, P. A. (2000). The meaning of “being Chinese”
ness of suspect interrogations. Annual Review of Law and Social Science, 11, and “being American”: Variation among Chinese‐American young
211–233. https://doi.org/10.1146/annurev‐lawsocsci‐120814‐121657 adults. Journal of Cross‐Cultural Psychology, 31(3), 302–332. https://
Milgram, S. (1963). Behavioral study of obedience. Journal of Abnormal and doi.org/10.1177/0022022100031003002
Social Psychology, 67(4), 371–378. https://doi.org/10.1037/h0040525 Tyler, T. R. (1998). The psychology of authority relations: A relational per-
Paddock, J. R., & Terranova, S. (2001). Guided visualization and suggestibil- spective on influence and power in groups. In R. M. Kramer, & M. A.
ity: Effect of perceived authority on recall of autobiographical Neale (Eds.), Power and influence in organizations (pp. 251–260).
memories. The Journal of Genetic Psychology, 162(3), 347–356. Thousand Oaks, CA: Sage Publications. https://doi.org/10.4135/
https://doi.org/10.1080/00221320109597488 9781483345291.n11

Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social dom- Vallano, J. P., & Compo, N. S. (2011). A comfortable witness is a good wit-
inance orientation: A personality variable predicting social and political ness: Rapport‐building and susceptibility to misinformation in an
attitudes. Journal of Personality and Social Psychology, 67(4), 741–763. investigative mock‐crime interview. Applied Cognitive Psychology,
https://doi.org/10.1037/0022‐3514.67.4.741 25(6), 960–970. https://doi.org/10.1002/acp.1789

Schultz, P. W., Nolan, J. M., Cialdini, R. B., Goldstein, N. J., & Griskevicius, Vrij, A., Meissner, C. A., Fisher, R. P., Kassin, S. M., Morgan, C. A., &
V. (2007). The constructive, destructive, and reconstructive power of Kleinman, S. M. (2017). Psychological perspectives on interrogation.
social norms. Psychological Science, 18(5), 429–434. https://doi.org/ Perspectives on Psychological Science, 12(6), 927–955. https://doi.org/
10.1111/j.1467‐9280.2007.01917.x 10.1177/1745691617706515

Schwartz, S. H. (2006). Value orientations: Measurement, antecedents, and Walczyk, J. J., Harris, L. L., Duck, T. K., & Mulay, D. (2014). A social‐
consequences across nations. In R. Jowell, C. Roberts, R. Fitzgerald, & cognitive framework for understanding serious lies: Activation‐
G. Eva (Eds.), Measuring attitudes cross‐nationally—Lessons from the decision‐construction‐action theory. New Ideas in Psychology, 34,
European social survey (pp. 161–193). Long: Sage. 22–36. https://doi.org/10.1016/j.newideapsych.2014.03.001

Snyder, M. (1974). Self‐monitoring of expressive behavior. Journal of Per- Walczyk, J. J., Roper, K. S., Seemann, E., & Humphrey, A. M. (2003). Cogni-
sonality and Social Psychology, 30(4), 526–537. https://doi.org/ tive mechanisms underlying lying to questions: Response time as a cue
10.1037/h0037039 to deception. Applied Cognitive Psychology, 17(7), 755–774. https://doi.
org/10.1002/acp.914
Tickle‐Degnen, L., & Rosenthal, R. (1987). Group rapport and nonverbal
behavior. In C. Hendrick (Ed.), Review of personality and social psychol-
ogy, Vol. 9. Group processes and intergroup relations, (pp. 113‐136).
Thousand Oaks, CA: Sage Publications, Inc. How to cite this article: Matsumoto D, Hwang HC. Social
Tickle‐Degnen, L., & Rosenthal, R. (1990). The nature of rapport and its influence in investigative interviews: The effects of authority
nonverbal correlates. Psychological Inquiry, 1(4), 285–293. https://doi.
on informational elements produced in interviews and written
org/10.1207/s15327965pli0104_1
statements. Appl Cognit Psychol. 2019;33:516–526. https://
Trout, D. L., & Rosenfeld, H. M. (1980). The effect of postural lean and body
congruence on the judgment of psychotherapeutic rapport. Journal of Non- doi.org/10.1002/acp.3488
verbal Behavior, 4(3), 176–190. https://doi.org/10.1007/BF00986818
Copyright of Applied Cognitive Psychology is the property of John Wiley & Sons, Inc. and
its content may not be copied or emailed to multiple sites or posted to a listserv without the
copyright holder's express written permission. However, users may print, download, or email
articles for individual use.

You might also like