Culture, Media & Deception

Joey F. GeorgeFlorida State University

OverviewJustificationLiterature reviewStudy 1: Media selectionStudy 2: Deception detectionConclusions

HistoryMy interest in deceptive CMC goes

back to about 1993AFOSR grant 2001-2006Deception literature had largely left

unexplored issues dealing with CMC, groups & culture

Four studies investigating cultural differences, two of which were dissertations that will be reported on here

Justification for Cultural Studies

With the rapid spread of CMC, it is now possible for billions of people all over the world to make video calls with each other via Skype for free

With increased (and low cost) exposure to people from many different cultures, it wouldn’t hurt to expand our understanding of other cultures

In any communication event, the possibility of deception is always present – What do we know about deceptive practices and attitudes towards deception in cultures other than our own?

Overall Research QuestionDo espoused cultural values affect

deceptive behavior and deception detection accuracy within and between people of varying cultures using CMC?

Literature ReviewComputer-mediated

communication (CMC)DeceptionCultureCMC & CultureDeception & CMCDeception & Culture

Literature ReviewComputer-mediated

communication (CMC)◦Media Synchronicity Theory (Dennis,

et al, 2008)DeceptionCultureCMC & CultureDeception & CMCDeception & Culture

MST

Literature ReviewComputer-mediated

communication (CMC)Deception

◦IDT (Buller & Burgoon, 1996)CultureCMC & CultureDeception & CMCDeception & Culture

Behavioral Adaptation

ReceiverInterpretation

&Judgment

Perceived Success

Discern Truth/

Deception

CO

NTE

XT

& R

ELA

TIO

NS

HIP

InitialMessage

Behavioral Adaptation

Sender

Interpersonal Deception Theory

Literature ReviewComputer-mediated

communication (CMC)DeceptionCulture

◦Theory of Cultural Differences (Hofstede, 1980)

CMC & CultureDeception & CMCDeception & Culture

Hofstede & CultureFour dimensions of national

culture:◦Collectivism◦Power distance◦Uncertainty avoidance◦Masculinity

Literature ReviewComputer-mediated communication

(CMC)DeceptionCultureCMC & Culture

◦Media use varies by culture (e.g., Lee & Lee, 2003)

Deception & CMCDeception & Culture

Literature ReviewComputer-mediated

communication (CMC)DeceptionCultureCMC & CultureDeception & CMC

◦Differences in cues transmitted (see chart)

Deception & Culture

Deception & CMCBehavior Video Audio WrittenLess talking time Detectable DetectableFewer details Detectable Detectable DetectableMore pressed lips DetectableLess plausibility Detectable Detectable DetectableLess logical structure Detectable Detectable DetectableMore discrepancies and ambivalence Detectable Detectable DetectableLess verbal and vocal involvement Detectable DetectableFewer illustrators Detectable Detectable DetectableLess verbal immediacy (all categories) Detectable Detectable DetectableLess verbal and vocal immediacy (impressions) Detectable DetectableMore verbal and vocal uncertainty (impressions) Detectable DetectableMore chin raises DetectableMore word and phrase repetitions Detectable DetectableLess cooperative Detectable DetectableMore negative statements and complaints Detectable DetectableLess facial pleasantness DetectableMore nervous and tense (overall) Detectable DetectableMore vocal tension Detectable DetectableHigher frequency, pitch Detectable DetectableMore pupil dilation DetectableMore fidgeting DetectableFewer spontaneous corrections Detectable DetectableLess admitted lack of memory Detectable Detectable DetectableMore related external associations Detectable Detectable Detectable

Literature ReviewComputer-mediated

communication (CMC)DeceptionCultureCMC & CultureDeception & CMCDeception & Culture

◦Some cultural differences discovered (see chart)

Sample of Deception-Related Cultural DifferencesStudy Countries Select FindingsTriandis et al 2001

Korea, Hong Kong, Greece, Japan, US, Australia, Netherlands, Germany

Collectivist groups more apt to deceive in business negotiations than individualist groups

Fu et al 2001

Canada & Chinese Canadians considered lies concealing pro-social behavior to be lies, but Chinese did not & rated such behavior favorably

Cheng & Broadhurst 2005

Hong Kong Chinese Observers better able to identify deception in their second language than in native language

Al-Simadi 2000

Jordan & Malaysia Individuals detected 52% of lies within their own cultures & 57% between cultures

Bond & Atoum 2000

US, Jordan & India Individuals do not perceive those from other cultures as more deceptive than individuals from their own culture

Study 1: Media selectionDissertation by Chris Furner,

West Texas A&M UniversityRQ: How does espoused national

culture influence media choice in a deceptive context?

Research DesignCreated 4 scenarios, which varied by:

◦Familiarity (stranger or friend)◦Severity of the situation (trivial or serious)

Embedded scenarios in questionnaires, which also included demographic and other items

Questionnaire translated into Mandarin & back to English; discrepancies addressed

Distributed to 261 American students and 194 Chinese students (PRC)

Research ProceduresQuestionnaires distributed to

students at universities in US & PRCEach questionnaire contained 1 of the

4 scenariosIn each scenario, boss asks employee

to lieRespondent asked to choose one

medium for the deceptive taskRespondent asked to give a reason

for the choice

Overall Choice Frequencies

Option Choice PercentFace-to-face 185 40.7Telephone 93 20.5E-mail 54 11.9Refuse 50 11.0Memo 32 7.0Letter 20 4.4Videoconferencing 13 2.9Voice-mail 4 0.9IM 3 0.7

Choice by GroupOption US Percent PRC PercentRefuse 44 16.9 6 3.1Telephone 64 24.5 29 15.9Memo 29 11.1 3 1.6E-mail 28 10.7 26 13.5Face-to-face 80 30.7 105 54.4Letter 12 4.6 8 4.1Videoconferencing

3 1.1 10 5.2

Voice-mail 1 0.4 3 1.6IM 0 0.0 3 1.6Totals 261 100 193 100

Ranked Choices by Groups

Option US Percent PRC PercentFace-to-face 80 30.7 105 54.4Telephone 64 24.5 29 15.9Refuse 44 16.9Memo 29 11.1E-mail 28 10.7 26 13.5Videoconferencing

10 5.2

Letter 12 4.6 8 4.1Refuse 6 3.1Videoconferencing

3 1.1

Voice-mail 1 0.4 3 1.6IM 0 0.0 3 1.6Memo 3 1.6Totals 261 100 193 100

Edited Choice FrequenciesOption US PRC TotalFace-to-face 80 105 185Telephone 64 29 93E-mail 28 26 54Memo 29 3 32Letter 12 8 20Totals 213 171 384

* Chi-square test is significant at the p < .000 level

Findings by Cultural Characteristic

Individuals who scored highly on espoused collectivism preferred to lie using text-based media (F (3, 370) = 2.811, p=0.039)

Individuals who scored highly on espoused power distance preferred to lie using voice-based media (F (3, 370) = 3.01, p=0.030)

Individuals who scored highly on espoused masculinity preferred to use visual media when lying (F (3, 370) = 7.683, p < 0.001)

Study 2: Deception detection

Dissertation by Carmen Lewis, now at Troy University

Work supported by Gabe Giordano, who was at IESE in Barcelona at the time data were collected, & who is now at Miami University

RQ1: To what extent does CMC affect deceptive behavior and deception detection?

RQ2: How do espoused cultural values affect deceptive behavior and deception detection accuracy within and between people of varying cultures using CMC?

Experimental Design

Experimental Procedures

Phase 1Conduct CMC Résumé Interviews

Subjects: Students

Honest and dishonest communication took place during the questioning of the résumé-based interview

The interviewee was videotaped:20 American, 20 Spanish

Phase 2Edit Tapes

The interview tapes were edited to separate honest and dishonest exchanges

2 stimulus tapes32 snippets per tape: 16 honest, 16 dishonest8 audio/video, 8 audio, 8 video, 8 text

Phase 3Test Deception Detection Ability

Third-party observers watched the stimulus via a computer:106 American, 104 Spanish

Each observer was asked to document where the lying occurred and what cues indicated that the interviewee was being dishonest

Observer: Interviewee:

The Stimulus “Reel”Part of what the participants sawExamples to show you:

◦One audio◦One text◦One video only◦4 audio/visual examples:

2 American: one honest, one not 2 Spanish: one honest, one not

Part of the questionnaire itself

Audio

Text• Interviewer: How would this scholarship help

you in any way?• Interviewee: Umm, the scholarship would

really help me out with umm … Well I am actually a student completely umm financially independent from my parents. So, the scholarship would help me with uh finishing up paying my tuition, my books, and my living expenses here on campus.

• Interviewer: And what’s your year in college?• Interviewee: I’m a senior.

Video Only

4Full A/V Examples

The Questionnairehttp://www.surveymonkey.com/s.

aspx?sm=21uW_2f2xgGMHctt7u3JBDZw_3d_3d

Some Preliminary Findings:Veracity Judgment SuccessCulture of the Judge Culture of the Interviewee

U.S. Spain

U.S. 15.15 (47%) 19.23 (60%)

Spain 16.37 (51%) 18.92 (59%)

Veracity Judgment Success (cont’d)

Culture Veracity Judgment SuccessTruths Deceptions

U.S. Judge9.83 (61%) 5.37 (34%)U.S. Snippet

U.S. Judge10.69 (67%) 8.56 (54%)Spain Snippet

Spain Judge10.08 (63%) 8.85 (55%)Spain Snippet

Spain Judge9.02 (56%) 7.18 (45%)U.S. Snippet

Veracity Judgment Success (cont’d)

Condition Mean SD % Correct

Audio and Video 4.58 1.38 57%

Audio Only 4.48 1.32 56%

Text-Based 4.35 1.40 54%

Video Only 4.00 1.48 50%

Preliminary Findings Regarding Reliable Indicators of Deception

Both groups, visual cues:◦Adaptors (excessive hand movements,

fidgeting)Spanish interview participants, visual

cues:◦Smiling◦Swallowing more strongly than usual◦Pressed lips

American interview participants, visual cues:◦Less facial pleasantness

Reliable Indicators of Deception (con’t)

Both groups, verbal cues:◦Changes in vocal pitch◦Repetition◦Illogical sentence structure◦Brief replies◦Pauses & hesitations

Reliable Indicators of Deception (con’t)

Easy cues for all judges to detect:◦Pauses & hesitations◦Changes in vocal pitch

One incorrect cue commonly cited:◦Gaze aversion

Concluding RemarksThere are differences in

deceptive behavior and these differences do seem to have some impact on deception detection

However, there is still much to learn about these differences, especially at the intersection of culture, deception & CMC