adevarul si minciuna in camera de ascultare

8/12/2019 Adevarul Si Minciuna in Camera de Ascultare

1/18

Law andHuman Behavior, Vol.23, No. 5,1999

"I'm Innocent ":EffectsofTrainingonJudgmentsofTruthandDeceptionin theInterrogation RoomSaul M. Kassin1,2and Christina T. Fong1

Th e present research examined the extent to which people can distinguish true andfalse denials made in a criminal interrogation,an d testedth ehypothesis that trainingin the use of verbal and nonverbal cues increasest he accuracy of these judgments.In Phase One, 16participants committed one of four mock crimes breaking andentering,vandalism, shoplifting, a computer break-in) or a related but innocent act.Given incentives todeny involvement rather than confess, these suspects were theninterrogated. In Phase Two, 40 observers were either trained in theanalysiso f verbaland nonverbal deception cues or not trainedbefore viewing the videotaped interroga-tionsandmakingtheir judgments. As in past studies conducted in nonforensic settings,observers were generally unableto distinguish betweentruthful and deceptive suspects.In addition, those who underwent training were less accurate than naive controlsthough they were more confident and citedmore reasons for their judgments. Theimplications of these findings are discussed in light of what is known about policeinterrogations,false confessions, and the wrongful conviction of innocent suspects.

When Robert Mooreconfessed to the capital murder of a New York taxi driverin 1996,thecaseappearedto be solved.After beinginterrogated byNassau Countydetectivesfor22hr,Mooresaid thathe and twoacquaintanceshadkilledthe maninarobbery attempt.Moorewaslaterreleased,however, whentherealkillerswhowerearrestedonotherchargesproduced themurder weapon. Mooresaid after-wardthat he hadconfessed because he waslonely, tired,andscared."Iwanted togohome,"he said (Hoffman, 1998).

In recent years, numerous documented cases involving proven or probablefalse confessions havebeen uncoveredbyresearchers,journalists, andlegalscholars(for an excellent review, see Leo &Ofshe, 1998).Nobodyknows the actual rateatwhichthis occursor hasdevisedanadequatemethodofcalculatingitsprevalence.Not surprisingly, then,researchers disagreeintheir estimatesof the magnitudeofthe problem (Cassell,1996,1998;Huff,Rattner, &Sagarin, 1986;Radelet, Bedau, &1Departmento fPsychology, Williams College, Williamstown, Massachusetts.2To whom correspondence should be addressed at the Department of Psychology, Williams College,Williamstown,MA01267 (e-mail: [email protected]).

4990147-7307/99/1000-0499$16.00/l 1999 American Psychology-Law Society/Division41 of theAmerican Psychology Association


2/18

Putnam, 1992; Schulhofer, 1996). Regardless of the precise numbers, which are indispute, analysesof past andcurrent cases have revealed that whereas some falseconfessions are v oluntarily given, others are elicited by the police th rough variousprocesses of social influ ence (Gu djonsso n, 1992; Kassin, 1997; K assin & W righ tsm an,1985; Ofshe & Leo, 1997; Wrightsman&Kassin, 1993).Motivated by the concern that interrogations are "inherently coercive" (Mi-randav. Arizona, 1966), th at certain tactics may increase the risk offalseconfessions(Kassin & Kiechel, 1996; Kassin & McNall, 1991), and that juries are overly influ-enced by confession evidence (Kassin & Neumann, 1997; Kassin & Sukel, 1997),researchers have sought to document what transpires behind the closed doors ofthe interrogation roomthe most private space in an American police station(Kamisar, 1965; Leo, 1996; Simon,1991).In this regard, perhaps the most revealingglimpse isprovided by the manuals tha t are used to train police detectivesin howto extract confessions from crime suspects (Aubry & Caputo, 1980; O'Hara &O'Hara, 1981; Macdonald &Mich aud, 1987; Zulaw ski& Wicklander, 1993). Th emost popular such man ual is Inbau, Reid, and Bu ckley's(1986)C riminal interroga-tion a nd co nfessions. Basedon theapproaches presented inthis book,the Chicago-based firm of John E. Reid and Associates trains law enforcement officials viacourses, seminars, and a videotaped program in the "Reid Technique" of inter-viewing an d interrogation.2Inbau et al. (1986) advise interrogators to seat the suspect in asm all, sound-proof, skeletally furnished roomthus creating a physical environment that isdesigned to promote feeling s of social isolation, sensory dep rivation, and h elp-lessness. They go on to describe a multistep process of social influence in whichth e interrogator confrontsth e suspect withhis or herg uilt, refuses to accept state-mentsofinnocenceanddenial,offerssympathyan dface-saving alternative explana-tions for the crime, and then gets the suspect to recount his or her guilty actionsin a full written confession. According to Jayne (1986), this approach will leadguilty suspects to incrim inate themselves by reducing the perceived negative conse-quences ofconfessing,while increasing the an xiety associated with deception. O n thequestion of wh ether their admittedly powerfulap proach m igh t also elicit confessionsfrom innocent people, Inbau et al. (1986) note that only those suspects who appearguilty in a preinterrogation interview are subjected to these techniques. They goon to advise that by analyzing an individual's verbal and nonve rbal behavior duringthis initial questioning, interrogators can reliably distinguish betw een denials m adeby those w ho are guilty and those w ho are innoc ent and then proceed accordingly.To assist the police in making these judgments, specific training is offered on theanalysisof verbal and nonverbal cues to deception.3Research suggests that this unw avering faith and reliance in the interrogator'sdiagnostic skills is misplaced. Over the years, controlled studies have shown thatpeople are poor intuitive judges of truth and deception (Zu ckerm an, D ePaulo, &Rosenthal, 1981), an d that police investigatorsa nd other so-called experts w ho2According to John E. Reid and Associates, more than 65,000 law enforcement professionals haveattendedtheir 3-dayseminarsover the past25years (see http://www.reid.com).3Therecommendations are summarized in an unpublishedbooklet thataccompaniesthe trainingvideo-tape, which iscommercially available.

5 Kassin andFong


3/18

routinelymake such judgmentsarealsohighlypronetoerror (Ekman&O'Sullivan,1991).Even trainingmay notimprove performanceinthis regard. Although modestgainshave been reported (deTurck &Miller, 1990), many researchers have foundthat judgment accuracyis notappreciably increasedbyrehearsal, training,orperfor-mance feedback (DePaulo, 1994; Zuckerman, Koestner, & Alton, 1984), evenamong experienced police officers (Vrij,1994). Part of theproblem isthat peoplewho stand falsely accused of lying often exhibit patterns ofanxietyan d behaviorthat are indistinguishablefrom those who are really lying (Bond &Fahey, 1987).Thus,asSimon(1991) noted, Nervousness, fear, confusion, hostility,astory thatchanges orcontradicts itselfall ar esigns that the man in an interrogation roomislying,particularly in the eyes ofsomeone asnaturally suspicious as adetective.Unfortunately,thesearealsosignsof ahuman beingin astateofhighstress"(p.219).Inlightof thevital importanceof theinitial, interview-based judgments thata remadebypolice investigators,thepresent studyw asconductedwithtw oobjectivesinmind.The first was toexaminetheextenttowhichpeoplecandistinguish betweentrue an d falsedenials made in thecontext of acriminal interrogation. Thesecondwas to test the hypothesis that people can be trained in the use of verbal andnonverbalcuestoincreasetheaccuracyofthese judgments.Toachieve these goals,we created a set of videotapes that depicted the interrogations of suspects w howereguiltyorinnocentofvarious mock crimesand w howere motivatedtodeny theirinvolvement.Groupsofobservers were then either trainedin theReid techniqueornot trained before viewingth e tapes andmaking judgments.

METHODThis experimentw asconductedin twophases. First,on egroupofparticipantssomew ho hadcommittedamock crime,otherswho hadnotwere apprehendedan dinterrogated insessions that were videotaped. Next,asecond groupo fpartici-pantssome w ho were trained in the Reid Technique of analyzing verbal an dnonverbalbehavior, others w ho were naivewatcheda subset of interrogationtapes an dmade judgmentsofguiltan dinnocence.

Phase One:TheInterrogationsParticipant Suspects

Sixteen male college students,allenrolledinintroductory psychology courses,took part in this phase of the study in exchange fo r extra credit. By randomassignment, halfwere instructed to commit one of four crimes, while th e otherswereinnocent ofthose crimes.44An additional 9 participants were excluded due to problems thatarose during their interrogations.Attestingto theinherent powerof thesituation,5guilty suspects were excludedbecausethey confessedorpartially admittedtoself-incriminating detailsand 2innocent suspects were excluded because theysaid they werein apsychology experimentas a way tojustifytheir actionsto the detective.Twoothersuspects were also excluded,onebecause thesessionwasinterruptedby anoutside partyand theotherbecause thesession was too long.

JudgmentsofTruthand Deception 5 1


4/18

After signing up for a "Communication Skil ls Study"whichw as described,in only general terms, as a test ofpeople's ability to communicate true and falsestatements in high-stakes situationsparticipants were informed that they m ay beasked to commit a mock crime. Before randomly assigning th e participants to aguilty or innocent crime condition, we asked them to read and sign an informed-consent statement.5

Th e Mock CrimesIn order to create a varied set of interrogation tapes, we created, after extensivepretesting, four different mock crimes (each was accompanied by an innocent butrelated act): vandalism, shoplifting, brea king and entering, and a comp uter break-in . In allcases, a female ex perim enter ad ministered the instruc tions for the guiltyor innocent task.Vandalism. Participants assigned to the vandalism condition were given col-

ored chalk an d instructed to write aderogatory an d obscene message on thewallof a designated c amp us building, then drop the chalk and w alk away. In con trast,the innocentsuspectsin this condition were instructed to follow a walking paththat le d them to the wallo nwhichthe previous message had been chalked. Theywere told to stop, note the message and color in which it was written, and walkaway. Moments later, both guilty and innocent participants were apprehended bya21-year-old m anposingas anundercover securityofficer. After showingasecuritybadge, th e officer escorted participants to an interrogation room.Shoplifting. In this condition, guilty participants were instructed to enter alocalgift shop an dshoplift eitheran itemo fjewelryo r a stuffed animal. They w erenot told how to commit th e crime, only that they were to remove th e designateditem from th estore without being caught. The cooperating store owner pretendednot tonoticethe theft.Inthiscondition, innocent suspectswereinstructed toenterth estore, browse, and exit. They were told that they would then see a paper bagon th e ground outside and that they should pick it up and inspect itscontents. Atthat point, participants were apprehended for interrogation.Breaking and Entering. In this condition, participants were scheduled duringeveninghours to appear at the siteof acollege bu ilding.Thoseassigned to committh e crime were instructed to break into that buildingan d steal th e answer key toan exam. They w ere told that although the bu ilding wou ld be locked, they couldclimb in throug h an open windo w. The innocent suspects were told simply to enterthrought he frontdoor(whichw asunlocked)t oobtain a questionnaire. When theyarrived, all participants found that all lights were off and that the building wasempty. What they did not realize w as that th e site w as protected by a securitysystem armed with motion detectors. Thus, participants set off an alarmand asynthesized voice wh ich screamed, "Intruder Intrude r Pleaseleaveth epremises "Th e consentform readasfollows:"I havebeen told thatfor the purposesofthisstudy,I may beaskedto commit a mock 'crime'or some related innocentact and then undergoan ' interrogation' aboutt heexperience. I understandthat th e eventsa re simulated,no t real, an d that al lpartiesa re awarethat Iam takingpartin apsychology experim ent.Ialso understand thatthe interrogationwillb evideotapedand that th e tapemay be showna t a later t imeto other research participants... I understandthat Imay withdraw my consentan d discontinue participationin th is project at any time withoutpenalty."

5 2 Kassin andFong


5/18

Thisalarm signaled the security guard, wholocatedthe p articipant (who w as eitherstanding inplace,hiding,or tryingtoescape)an d made th e apprehension.Computer Break-in. In the fourth crime, participants were told to enter ascience compu ter laboratoryand approacha specific terminal.Those assigned to theguilty condition received written instructions on how toaccess the college e-mailsystem and logonto th eaccount of a female student (the account user's namean dpassword we re provided). Once in the account, they were told, they could read thestudent's personal e-mail. By design, the inbox contained three highlyconfidentialmessages. Innocent suspects were told to sit at the designated computer and logonto their ow n accounts. In doing so , however, they discovered that th e otherstudent's inbox was already open and a message was visible on the screen. At thispoint, th e security guard appeared an d took th e participant to the interrogationroom.

The InterrogationsPrior to their commission of the mock crimes or therelated innocent acts,participants w ere instructed that they w ould be arrested and questioned about their

activities. They were told that they w ould be asked to waive their Miranda rights,which they should do, but that they should maintain their innocence and refuse tosigna confession. G uilty suspects were told to come upwithw hatev er alibis, excuses,or stories they needed to fabricate a plausible denial of the charges. Innocentsuspects were instructed to tell the t ruth concerning their w hereabou ts and activi-tiesexcept for the fact that they were taking part in a psychology experiment.All suspects were informed tha t th e interrogator did not knowi fthey were guiltyor innocent, but that he would make a judgment subsequent to the interrogation.In order to increase performance motivation, al l participants were told that,i fjudged deceptive and guilty,they would beescorted across campus to the securityoffice an d detained there for 5min. Participants were told, however, that ifjudgedtruthful and innocent, they would receive a $5 bonus. Although the interrogatorprivatelyrecordedhis actual judgm ents, he publicly judge d all participants innocentand aw arded the m the m oney. Following their apprehension, all participants weretaken b y the securityofficer to an interrogation room and asked to take a seat andwait for the detective in thiscase to arrive. Before leaving, the officer said, "Thinkabout what you've done and how you are going to get out of the situation."As specified by Inbau et al.(1986), the interrogation room w as furnished withtwo straight-back chairs, a plain table, and a one-w ay m irror on the right wall. Inaddition, a videotape camera on a tripod w asstationed behind th e interrogator'schairsothatittaped participantsfrom afrontal view, overthe detective's shoulder.Toenhanceth eaud itory portion of therecordings,aremote microphonewasplacedon the table that separated the interrogator and participant. As is comm on in man yvideotaped interrogations, aclock was mounted on the back wall and wasvisibleon th e tapes.The interrogator in all the sessions was a 48-year-old man dressed in civilianclothing. After the participant had waited for 5 min, the interrogator entered theroom, turned on the videotape camera and microphone, and formally introducedhimself as Detective McCarthy. He carried a clipboard that contained a list of the

Judgments ofTruthandDeception 5 3


6/18

Miranda rights, a Miranda waiver form to be signed by the participant, a set ofinterrogation questions designed to fit each crime, and a han dw ritten confessionstatement. Prior to each session, he was informed of the crime that w as beinginvestigated but wasblindas tow h e th er th ep art icipant w asguilty or innocent.The interrogation consisted of a brief five-part sequence of events. First, theinterrogator read the Miranda rights and had the participant sign a waiver form.Second, he recapped th e crime and situation leading to the arrest an d asked th eparticipant, in an open-end ed m anne r, to account for his wh ereabo uts and beha viorduringthat period oftime. Third, th e interrogator refused toaccept th e statementsofde nial contained w ithin the participa nt's story. At this point, he p ounde d his fiston the table an d said, "Dammit, youwere caught red-handed, sostop lying to me.I want to know exactly, minute by minute, where you were and what you weredoing before you were picked up." This statement prompted th e participants toreaffirm their innocence and restate their alibis,often in greater detail. Fourth, theinterrogator tried to break down each participant's story by repeating it , notingomissions an d inconsistencies, characterizing th e story as implausible, and de-manding further explanation. Fifth, the interrogator reacted to this final account

by presenting a handwritten confession to the crime, reading it aloud, and tellingthe participant to sign i t ("I admit that I . I am now unde r my own free willand I am sorry fo r wh a tIdid").As instructed, th e participant refused to sign th estatement, at which point the interrogator tried again.After a second refusal, theinterrogator exited, leaving the participant to await his jud gm ent. Overall , thesessions ranged from 3.5 to 6 min in duration, with a mean of 4 min and 35 sec.Postinterrogation Questionnaire

After each interrogation, the experimenter administered a brief questionnairedesigned to assess participant reactions to the experiment. First, they rated thestressfulnesso f theprearrest experience (the crime situation)an d postarrest experi-ence (the interrogation) bothon a10-point scale, where1 = not at allstressfuland10 = very stressful. Also on 10-point scales, the participants rated the amountofpressure theyfeltto confess, how intimida ting the interrogator w as, and how nervoustheywere dur ingth equestioning. Participants were then asked to predict whetherthe interrogator would judge them guiltyorinnocent, rate their confidence inthatprediction, an d then estimate th e percentage ofobservers w howould judge themto be guiltyo r innocent. Finally, participants estimated both th e amoun t of t imethey spent waiting for the interrogator to arrive and the amount of time they spentbeing questioned. Afterward, all participants were told that th e interrogator hadjudged them to be innocent , whereupon they were given the $5 reward, fullydebriefed, and thanked.

Phase Two: Observations and JudgmentsParticipant Observers

Forty introductory psychology students (11 male, 29 female) from a localuniversity took part in this second phase of the study as part of their course

5 4 KassinandFong


7/18

requirem ent. To ensure th at they w ould not recognize the suspects, this group wasrecruited from a neighboring institution.To further ensure that observers did notself-select into th e naive an d training conditions, they al l signed up for a 2-hrexperim ent. For those rand om ly assigned to the naiv e (no-training) condition,however, the session lasted only 1 hr. All sessions were conducted in 3- to 6-person groups.

Training Manipulation an d AssessmentParticipants in the training condition were shown tw o 15-min videotape seg-ments from The Reid Techniq ue: Interviewing an d interrogation produced byJohn E. Reid andAssociates(1991).The entire seminar consists offour videotapedunits. The first unitpresents the psychology of interviewing, the fourth presentsth e nine steps of interrogation,and the second an d thirdunitswhichwere usedin this studypurportto enhance one's abilityto distinguish between truth an ddeception throughan analysisof verbalan d n onve rbal behavior. Everyone inthiscondition w atched these two tapes, took a 10-min break, and were the n given 10m into studyawritten summ aryof keypoints thatform partof the tra inin g packet.The firsttraining tape trains policein theanalysisofverb al behavior.Accordingto the manual,truthful suspects are said to be direct, spontaneous, helpful, an dconcerned about the charges. Their denials arebroad,sweeping, and unequivocal.They tend to use first-person pronouns and descriptive verbs ("I did not steal th emoney") and unqualified lang uage ("Absolutely not I would never do such athing "). By contrast, deceptive suspects are said to be guarded, unhelpful, an dunconcerned about th e charges. They often hesitate before stating their denials,shake the ir headsormu mb le insteadofvocalize clearly, respond tochargesinwaysthat ar e overly general or evasive, omit details, and use phrases that are weak,narrowly defined, or qualified ("As far as I recall ...").The second tape focused on nonverbal cues that betraytruth and deceptioncues pertaining to asuspect'sposture, protective gestures, grooming gestures, barri-ers, illustrators, and eye contact. According to the training manual, for example,truthful suspects te nd to sit comfortably upright,facethe interrogator, lean forwardas a sign of interest, use their han ds and arms to em phasize a point, and m aintainappropriate eye contact when answering key questions. By contrast, deceptivesuspects are said to exhibit a rigid body posture, slouch backward rather thansi tforward, align themselves nonfrontally with th e interrogator, cross their arms orlegs, exhibit various grooming gestures to relieve anx iety , cover the ir eyes andm outh, and eithe r stare or avoid eye contact w hen answering key questions. Impor-tantly, the manual cautions that no single behavior can be used to indicate truthordeception,whichis why it isnecessarytoconsider "behavioralclusters."Themanual further states that the analysis of behavior can produce an accuracy rateof80-85%,tha t most errors involve a judgm ent thatsomeoneinnocent islying,andthatall evaluations of truth and deception should take oth er evidence into account.

Stimulus TapesPhase O ne of our study yielded 16 videotaped interrogations: 8 true denialsand 8 false denials (2 of each per m ock crime). Thus, the re w ere 2 guilty suspects

Judgments of TruthandDeception 5 5


8/18

and 2 innocents in the vandalism , shoplifting, breaking and entering, and com puterbreak-in crime conditions.The 16 segments were then randomly divided intotwoeight-segment stimulus tapes (tapes 1 and 2), each containingone guilty suspectan done innocent suspect per crime.Within the trainingand naive conditions, groups were randomly assigned toobserve stimulus tape1 or 2.Both tapesopenedwithawrit tenset oftask instruction ssimilarto those that were given orally. Each interrogationwas thenprecededbya general, one-paragraph description of the crime and the circumstances underwhich the suspect had been arrested (there were a total of four descriptions, oneper crime; these were identical for guiltyand innocen t suspects). Each interrogationwas followed by a blank screen that directed participants to respond by giving ajudgment , a confidence rating, and a listof reasons.

ProcedureScheduled in small groups, participants were told that th e experiment w asdesigned totest their abilitytodetermine wh en others, particularly crime suspects,

are tellingthe truthor lying.Those assignedto the training condition underwentthe procedures previously described; the others did not. Prior to being tested,participant observers completed a brief questionnaire in which they rated (1) howconfident they were in their abilityto succeed at the task, and (2) how easy theythought it was for people ingeneral to tell when a stranger is tellingthe truth orlying. Both ratings were madeon ascale from 1 to 10.In both the train ing and naive conditions, participants were th en ad ministereda 20-item o bjective test designedtoassesstheir knowledgeof theReid recomm enda-tions concerning th e verbal and nonverbal cues that signal truth an d deception(e.g., "Wh atis anillustrator?"; "Wh atare qualifyingphrases,and whousesthem?";"Are truthful suspectsor deceptive suspects more likelyto make sweeping deni-als?"; "Are truthful suspects more likely to use first-person pronouns, second-person pronouns,orthird-person pronouns whe n answeringquestions?").This testwas designed with tw o purposes in mind: to check on the effectiveness of thetraining manipulation and to examine the correlation between knowledge of thematerialandperformance. Participants were credited onepointper correct answer(there were three 2-part questions for which participants could receive partialcredit), so scores could range from 0 to 20.Participants were instructed that they would see a videotape of eight malesuspects being questioned by a detective about their possible involvement in variouscrimes.As in past studies (Frank&Ekman, 1997),wesoughtto deter participantsfrom assuming that al l suspects were either guilty or innocentbyindicatingwithinthe instruction that between one fourth and three fourths are lyingto protectthemselves. Participants were then told tha taftereach segment th ey would be askedto m ake a judgm ent by circling either "tru thfu l" or "lying," rate their confidence inthat judgment on a scale from 1 to 10, and w rite in their own words the bases forthat judgment. After responses were obtained for alleight segments, participantscompleted a posttask questionnaire that w asidentical (except for the use of pasttense)to thepretask questionnaire. Thus, they rated th eir confidenceintheirownperformance and the generalease or difficulty of the task.

5 6 Kassin an d Fong


9/18

RESULTSSuspectSelf Reports

Suspects were administeredaquestionnaire inorder to assess thesubjectiveimpact oftheir experiencesand toexamine whether those experiences weredifferentas a function of thecrimefo r which th e suspects were arrested an d their guilt orinnocence. On 10-point scales, suspects rated their prearrest (crime) and post-arrest (interrogation) experiences asmoderately stressful (overallM s = 4.56 an d5.69, respectively). As revealed via 2(guilt status) X 4 (crime type) analyseso fvariance on these measures, there were no significantdifferences between guiltyand innocent suspectsoramong those accusedof the differentcrimes (allFs< 1).Although self-reported stress w ashighest amongguiltyshoplifters,guiltyvandals,an dinnocent e-mail intruders,th esuspect guilt xcrime interactionwas not signifi-cant,F(3,8) = 3.44,p


10/18

5 8 Kassin and FongTable1.Summa r yof Results

Main measuresTraining test scoresJudgm ent accuracyConfidence ratingsNumber of reasons cited

Trained13.283.65

6.553.96

Naive5.83***4.40*5.91*2.98***

Note:*p


11/18

Judgments of TruthandDeception 5 9

training materials used in this study. To examine the effects of training on thetendency to make judgmen t sof truth versus deception, w e conducted a 2 x 2ANOVA on the number of"lying" judgments out of eight that were made. Thisanalysis indicated that w hile there was a nonsign ificant tenden cy for participantsto make judgmen t sof deception rather than truth (55.37% vs.44.63%), trainin ghad no effect on the number of such judgments that were made (M s = 4.35 and4.50 among th e trained an d naive participants, respectively), F (1 , 36) < 1.Therew as also no maineffect for tape and no interaction.

Confidence RatingsAfter each interrogation, participants rated their confidence in the judgmentthey madeon ascalefrom 1 to 1 0. Atwo -way ANOV A on thetotalo fthese ratingsrevealed a significant main effect of training, F (1 , 36) = 5.57, p < .05, with th etrained observers reporting more confidence than naive observers (Ms = 6.55 and5.91, respectively; R 2 = .134). There was no significant main effect fors t imulustape, F (1 ,36) = 2.50,p


12/18

responsesb ycountingth etotal numberofdiscrete verbalandnonverbal cues thatwere cited, regardless ofwhether they argued for or against th e judgment given(e.g., "He kept changinghisstory,""He isvery sureofhimself, "Doesn'tknowtimes," Wouldn't lookat thequestioner," "Very nervous").Overall, participantscited a mean of 3.47 reasons per judgment. Interestingly,a two-way ANOVArevealed a highly significant main effect for training,F(1, 36) = 15.14,p


13/18

ratings and the n u m b e r of reasons cited, r(40) = .42, p < .05. Turning to thepretask and posttask confidence ratings, we found that these m easures w ere highlyintercorrelated, as participants who w ere m ore con fident in their abilities beforethe task also tended to be more confident after the task, r(40) = .62,p


14/18

theydid not commit.Basedon areviewofover3 0studies, Zuckerm an et al.(1981)concluded that there is a mismatch between th e behaviors that actually signaldeception an d those thatare used by layperceivers. They noted, fo rexample, thatwhile certain vocal cues (e.g., rises in pitch, hesitations) are often revealing ofdeception, people tend to focus instead onchannelsofcom m unication associatedwith th e face and the bod y. Sim ilarly, it is clear tha t the trai ning materia ls used inthis study explicitly direct observers to usecues no t reliably linked to deception.The forensic importance of our results is bolstered by two unique aspectsofth epresent study. First, observers m ade the ir judg m entson thebasisof videotapedinterrogations of moderately anxious suspects who were guilty or innocent of oneof fourmock crimes, and all of whom were motivated to deny involvement and bejudged innocent. Second, train ing participants were instructed via videotaped andwritten materials published byJohnE . Reid an d Associates, an approach that hasbeen used over th e years in the trainingof lawenforcement officials.Our third m ajor finding pertains to the m etacognitive aspects of performan ce.Recent studies have shown thatpeople in general cannot accurately assess theirow n lie-detection skills,but rather exhibit high levels ofco nfidence regardless ofwhether their judgments ar e correct or incorrect (DePaulo, Charlton, Cooper,Lindsay, & Mu hlenb ruck, 1997). Sim ilar results have emerged in the litera ture oneyewitness testimony, where the correlation between identification accuracy andconfidence is often not reliable (Wells & Murray, 1984; Penrod & Cutler, 1995;Sporer, Penrod, Read, & Cutler, 1995). Consistent with these observations, w efound among both trained and naive participants tha t judg m ent accuracy and confi-dence were no t significantly correlated, regardless ofwhether the measure ofconfi-dence w as taken before, after, or during the task. Further demonstrating the meta-cognitive problems in this domain is that confidence ratings were positivelycorrelated withthe number of reasons (includ ing Reid-b ased reasons) articulatedas a basis for judgments, another dependent measure not predictive of accuracy.Training had a particularly adverse effect in this regard. Specifically, those whowere trained compared to those in the naive condition were less accurate in theirjudgments of truth and deception. Yet they were more self-confident and morearticulate about the reasons for their often erroneous judgments.In light of the practical implications, it is important to consider the possiblelimitations an dboundary conditions of ourresults. With regard to the findingthattrainingin the Reid techniquedid not increase accuracy,the resultsareunambigu-ous. Trained participantsscored higher than controls whentested fo r their knowl-edge of the training material, and they were more likely to use that material intheir reasoning, yet their judgments were less accurate. In addition, test scoresacrossconditions w ere not significantly correlated with accuracyrates.Th epresent study suggests thatthe Reid techniquem ay not beeffectivea nd,indeed, may be counterproductiveas a method of distinguishingtruth anddecep-tion. At this point, however,thegeneralityof our conclusion shouldbe consideredtentative. To begin with, all of our participant suspects were male, a feature thatbest models a normative interrogation. It could be argued that th e exclusive useof male suspects might have impacted upon the difficulty of the judgment task.However, research suggests that while men and women betray their lie telling in

512 Kassinand Fong


15/18

different ways, there is little or no difference in the accuracy with which peopledetect lies in male versus female senders (B. M. DePaulo, Personal comm unication,April16,1999).There are three possible bases for the argument that our experiment did notoptimally model th e interview an d interrogation process. First, on e could arguethat training would benefit police investigators m ore th an naive college students,perhaps because training requires past experience or follow-up practice to be mosteffective. To address this point, future research should consider such variationsinthe subject population, particularly in light of observed individ ual differencesinth e ability to judge truth an d deception (Frank & Ekman, 1997). In this vein,it isimportant to keep in mind that police and other experienced la w enforcementprofessionals have not outperformed novices in related past studies (DePaulo &Pfeifer,1986; Ekman & O'Sullivan, 1991; Kraut & Poe, 1980), in part because theyapproach the task in asuspicious frameofm ind, biased tow ard perceptions ofguiltan d deception (Burgoon, Buller, Ebesu, & Rockwell, 1994).9A second possible argumentisthat interrogators are better equipped tojudgetruthan d deception than observers are byvirtueoftheirfirsthand and more exten-sive interactionwith the targets ofcriminal investigations.It isclear, for example,that police detectives, unlike the observers in our study, are in a position to controlthe flow ofinformation throughthequestions they choose to ask and thewaystheyreact to the suspec t's responses. It is not clear, how eve r, tha t this more interactiv erole helps or hinders their ability to interpret accurately a suspect's behavior.Experiments in nonforensic settings have shown that preinteraction expectationsinfluence the kindsofquestions that interviewersas k (Snyder&S w ann , 1978), an dthat this confirmatory hypothesis-testing bias ca nproduce false support fo r theseinitial beliefs (Zuckerman, Knee, Hodgins, & Miyake, 1995; for reviews, see Hil-ton & Darley, 1991; S nyde r, 1984). Mo reover, rese arch suggests th at judg m en ts oftruth and deception are m ore accurate wh en made by observers than by conversa-tional interactants. Buller, Strzyzewski, and Hunsaker (1991) had observers watchvideotaped conversations between tw o other participants, one of whom was in-structed to lie or tell the tru th, and found that the observers w ere more acc urate intheir judgmentsof thetarget tha n were thosew howere engaged in theconversation.Third, it isimportant to keep inm ind that although trainingin the influentialReid technique did not improve judgments of interrogative truth and deception,other methods of credibility assessment m ay u ltimate ly prove effective. Focusedspecifically on content analysesofverb al cues, several inv estigators havefound thatit is possible to make such judgments at better-than-chance levels of accuracy(Landry & Brigham, 1992; Sporer, 1997). Porter and Yuille (1996), for example,noted that people giving truthful accounts of a personal experience can be signifi-cantly distinguishedfrom deceiversby the amountofdetail they report, th e logicalcoherence oftheir statements, an d their willingnessto admita lackofmemoryfor9Horvath, Jayne, andBuckley(1994)reported tha t lawenforcement professionals exhibited high levelsof accuracy in making judgments of interrogative t ru th anddeception,but they had hand-picked fortesting only 4subjects, with unspecified affiliations, all of whom were "trained an d experienced" intheReid technique.

Judgmentsof Truthand Deception 513


16/18

certain aspects of the event. Others have similarly argued that it ispossible toseparatetrue and false confessions through linguistic analyses (Shuy,1998).Notwithstandingthepossible limitationsof thepresentstudyand the need foradditionalresearch,the practical implicationsof ourresultsareclear.The investiga-tor who makes an incorrect judgment of guilt and deception, and who does sowith confidence and a well-articulated rationale, is likely to employ th e powerfultechniques of interrogation on innocent suspects (Inbau et al., 1986), techniquesthatmayincreasethe risk ofelicitingacoerced false confession (Gu djons son, 1992;Kassin, 1997).10 In light of the numerous documented instances in which police-induced false confessions have led to the wrongfu larrest, prosecution, conviction,and incarceration of innocent people (Leo & Ofshe, 1998), further research isneeded to examine th e ways inwhich la wenforcement officials make theseinitial,potentially self-perpetuating, judgments.

REFERENCESAkeh urst, L., Koh nken G., Vrij, A., & B ull, R. (1996). Laypersons' and police officers' beliefs regardingdeceptive behavior. Applied Cognitive Psychology, 10,461-471.Aubry , A. S., & C aputo, R. R. (1980). Criminal interrogation(3rd ed.).Springfield,IL: Charles C. Thomas.Bedau, H., & Radelet, M. (1987). Miscarriages of justice in potentially capital cases. Stanford La wReview,40,21-179.Bond, C. F., & Fahey, W. E. (1987). False suspicion and the misperception of deceit. British Journalof Social Psychology, 26 ,41-46.Borchard, E. M. (1932). Convicting the innocent: Errors of criminal justice. New Haven, CT: YaleUniversity Press.Buller,D. B.,Strzyzewski,K. D ., &H unsaker,F. G.(1991). In terperson al deception:II. The inferiorityofconversational particip ants as deception detectors. Communication Monographs, 58,25-40.Burgoon,J. K.,Buller,D. B.,Ebesu,A . S., &Rockwell ,P.(1994). Interpersona l deception:V .Accuracyin deception detection. Communication Monographs, 61,303-325.Cassell, P. G.(1996).M iranda s social costs: An empirical reassessment. Northwestern University La wReview, 90, 387-499.Cassell, P. G.(1998). Protecting the innocent from false confessions and lost confessionsand fromMiranda. Journal of Criminal Law and Criminology,8 8,497-556.Cohen, J.(1988). Statistical power analysis for the behavioral sciences.Hillsdale, NJ: Erlbaum.DePaulo, B. M. (1994). Spotting lies: Can hum ans learn to do better? Current Directions in PsychologicalScience, 3, 83-86.DePaulo, B. M.,Ch arlton,K.,Cooper,H., Lindsay,J. J., & Muhlenbruck,L. (1997). The accuracy-confidence correlation in the detection of deception. Personality an d Social Psychology Review,1,346-357.DePaulo, B. M., & Friedman, H. (1998). Nonverbal communication. In D. Gilbert, S. Fiske, & G.Lindzey (Eds.), The handbook of social psychology (4th ed.,p p.3-40). Boston: McGraw-Hill.DePaulo, B. M.,Lanier, K., &Davis,T .(1983). Detecting the deceit of the motivated liar. Journal ofPersonality and Social Psychology, 45,1096-1103.DePaulo, B. M., & Pfeifer, R. L. (1986). On-the-job experience at detecting deception. Journal ofApplied Social Psychology, 16,252-267.deTurck,M. A., &Miller,G. R.(1990). Training ob serverstodetect deception: Effectsofself-m onitoringand rehearsal. Human Comm unication Research, 16,603-620.Ek m a n ,P.(1985).Telling lies: Clues to deceitin the marketplace, politics, and marriage.Ne wYork: Norton.Ekm an, P., &Friesen, W . V. (1974). Detecting deception from the bodyo r face. Journal of Personalityan d Social Psychology, 29,288-298.10Although it is considered a less costly, more tolerable error, the investigator who makes an initialincorrect judgment of truth and innocence is also at riskto abandon the interrogation and lose thepossibility of obtaininga true confession (Cassell,1998).

514 Kassin and Fong


17/18

E kma n ,P., &O 'Sull ivan,M.(1991).Who can catcha liar? American Psychologist,46,913-920.Frank,M. G., & E kma n ,P. (1997). The abilityto detect deceit generalizes across different typesofhigh-stake lies. Journal of Personality and Social Psychology, 72,1429-1439.Oeller, W. A.(1993).Videotap ing interrogations and confessions. National Institute o f Justice: Researchin brief. Washington, DC: U.S. Department ofJustice.Gilovich,T., Savitsky,K., &Medvec ,V. H. (1998).The illusionof transparency: Biased assessmentsof others' ability to read one's emotional states. Journal of Personality and Social Psychology,75,332-346.Gudjonsson, G. H. (1992). The psychology of interrogations, confessions, and testimony.L ondon: Wiley.Hilton,J. L., &Darley,J. M.(1991).The effectsof interaction goalson person perception. Advancesin Experimental Social Psychology, 24,235-267.Hoffman,J.(1998).Police refinem ethodssopotent , eventheinnocent hav e confessed.New York Times,March30, 1998,pp. Al, B4.Horvath, F., Jayne, B., & Buckley, J. (1994). Differentiation of tru thful and deceptive crim inal suspectsin behavior analysis interviews.Journal of Forensic Sciences, 39,793-807.Huff, C. R., Rattner, A., & Sagarin, E.(1986). Guil tyuntil proven innocent: Wrongful conviction an dpublic policy. Crimeand Delinquency,32,518-544.Inbau, F. E., Reid, J. E., & Buckley,J. P. (1986). Criminal interrogation and confessions (3rd ed.).Balt imore, MD : W ill iams & Wilkins.Jayne, B. C. (1986). The psychological principles of criminal interr oga tion. App endix . In F. Inba u, J.Reid, & J . Buckley, Criminal interrogation and confessions (3rd ed.). Balt imore, MD: Wil-liams& Wilkins.John E. Reidand Associates (1991). The Reid Technique: Interviewingand interrogation [Videotape].Kamisar, Y.(1965). Equal justice in the gatehouses and mansions of American criminal procedure. InA. Howard (Ed.) , Criminal Justicein our Time (pp.11-38). Charlotesville, VA. UniversityPressof Virginia.Kassin,S. M. (1997). The psychology of confession evidence. American Psychologist,52,221-233.Kassin,S. M., &Kiechel ,K. L.(1996).Thesocial psychologyoffalseconfessions: Com pliance, internaliza-tion, and confabulation. Psychological Science, 7,125-128.Kassin, S. M., & McNall ,K.(1991). Police interrogations an d confessions: Communicating promisesan d threats bypragm atic implicat ion. Law and Human Behavior, 15,233-251.Kassin, S. M., & Neumann, K. (1997). On the power of confession evidence: An experimental test ofthe "fundamental difference" hypothesis .Law and Human Behavior,21,469-484.Kassin, S. M., & Sukel, H. (1997). Coerced confessions and the ju ry: An e xpe rime ntal test of the"harmlesserror" rule.Law and Human Behavior,21,27-46.Kassin, S. M., & Wrightsman, L. S.(1985). Confession evidence. In S. Kassin & L. Wrightsman(Eds.),Th e psychology of evidence and trial procedure (pp.67-94). Beverly Hills, CA: Sage.Kraut ,R. E., &Poe,D.(1980).On theline:Thedeception judgmentsofcustom s inspectorsandl aymen .Journal of Personality and Social Psychology, 39,784-798.Landry, K. L., & Brigham, J . C.(1992).T he effects of trai ning in criteria-based conten t analysis on theability of detecting deception in adults. Law and Human Behavior, 16, 663-676.Leo,R. A. (1996). Insidethe interrogation room.Journal o f CriminalLaw and Criminology,86,266-303.Leo, R. A., & Ofshe, R. (1998).The consequences of false confessions: Deprivations of l ibertyandmiscarriages of justice in the age of psychological interrogation. Journal of Criminal Law andCriminology, 88,429-496.Leo,R. A., &Thomas,G. C.(1998).TheM iranda deb ate: Law, justice,andpolicing. Boston: NortheasternUniversityPress.Macdonald, J. M., & Michaud, D. L . (1987). Th e confession: Interrogation and criminal profiles for policeofficers. Denver: Apache.Mirandav.Arizona, 384U.S.336 (1966).Munsterberg, H.(1908).On thewitness stand. Garden City,NY: Doubleday.Ofshe, R. J., & Leo, R. A. (1997).The social psychology of police interrogation: The theory andclassification of true and false confessions. S tudies in La w, Politics, and So ciety, 16,189-251.O'Hara,C. E., &O'Hara,G. L.(1981).Fundam entals o f crim inal investigation.Springfield,IL: CharlesC. Thomas.Penrod,S. D., & Cutler, B.(1995). Witness confidenceand witness accuracy: Assessing their forensicrelation. Psychology, Public Policy, an d Law, 1,817-845.Porter,S., &Y uil le,J. C.(1996).Thelanguageofdeceit:An investigationof theverbal cuesto deceptioninthe interrogation context. Law and Human Behavior,20,443-458.Radelet,M. L., Bedau,H. A., &Putnam,C. E. (1992). Inspiteof innocence: Erroneous convictions incapital cases.Boston: Northeastern UniversityPress.

Judgments of Truth and D eception 515


18/18

Schulhofer, S. J. (1996). Miranda s practical effect: Sub stantial benefits and vanishingly small socialcosts. Northwestern University La w Review, 90 ,500-564.Shuy,R. W.(1998).The language of confession, interrogation, and deception. Thousand Oaks, CA: Sage.Simon, D. (1991). Homicide: A year on the killing streets.N ew York: Iv y Books.Snyder, M.(1984).When belief creates reality.Advances in Experimental Social Psychology, 18,207-303.Snyder, M., & Swann, W. B. (1978). Hypothesis-testing processes in social interaction. Journal ofPersonality and Social Psychology, 36,1202-1212.Sporer, S. L.(1997).Th e less travelled road to truth: Verbal cues in deception detection in accountsof fabricated and self-experienced events. Applied Cognitive Psychology, 11,373-397.Sporer,S. L.,Penrod,S. D.,Read,J. D., &Cutler ,B. L. (1995). Choosing, confidence,and accuracy:Ameta-analysis of the confidence-accu racy relation in eyewitness identification studies. PsychologicalBulletin, 118,315-327.Vrij, A. (1994). Th e impact of inform ation and setting on detection of deception by police detectives.Journal of Nonverbal Behavior, 18,117-136.Walkley, J.(1987).P olice interrogation: Handbook fo r investigators. London: Police Review.Wells,G. L., &Murray,D. M.(1984).Eyewitness confidence.In G.Wells& E.Loftus(Eds.),Eyewitnesstestimony: Psychological perspectives. New York: Cambridge Universi tyPress.Wrightsman,L. S., & Kassin, S. M.(1993). Confessions in the courtroom. Newbury Park , CA: Sage.Zuckerman, M., DePaulo, B. M., & Rosenthal , R. (1981). Verbal and nonve rbal communicat ion ofdeception. Advances in Experimental Social Psychology, 14, 1-59.Zuckerman , M.,Knee,C. R., Hodgins, H. S., &Miyake,K .(1995).Hypothesis confirmation:Th ejointeffect of positive test strategy and acquiescence response set. Journal of Personality and SocialPsychology, 68,52-60.Zuckerman ,M .,Koestner,R ., &Alton,A. O. (1984). Learningtodetect deception.Journal o f Personalityan d Social Psychology, 46,519-528.Zulawski, D. E., & Wicklander, D. E. (1993). P ractical aspects of interview and interrogation. Oxford:Elsevier.

516 Kassin and Fong

adevarul si minciuna in camera de ascultare

Documents