Context-dependent cheating: Experimental evidence from 16 countries

1 This is a p ep in o he ollowing pape : Con ex -dependen chea ing: Expe imen al e idence om 16 coun ies Fo hcoming in: Jou nal o Economics Beha io and O ganiza ion Da id Pascual‐Ezama Uni e sidad Complu ense de Mad id (UCM) Toke R. Fosgaa d Uni e si y o Copenhagen - Depa men o Food and Resou ce Economics Juan-Camilo Cá denas Uni e sidad de Los Andes P a een Kujal Middlesex Uni e si y Robe Fe ec Vesz eg Uni e sidad del País Vasco Bea iz Gil-Gómez de Liaño Depa men o Social Psychology and Me hodology, Uni e sidad Au ónoma de Mad id B ian Gunia Johns Hopkins Uni e si y - Ca ey Business School Do is Weichselbaume Johannes Keple Uni e si y Linz - Depa men o Economics; Ins i u e o he S udy o Labo (IZA) Ka ha ina Hilken Depa men o Applied Economics, F ee Uni e si y o B ussels (VUB) A menak An inyan Uni e si y o E langen-Nu embe g, Depa men o Economics Joyce Delnoij U ech Uni e si y - U ech Uni e si y School o Economics An onios P oes akis Join Resea ch Cen e o he Eu opean Commission Michael D. Ti a Uni e si y o Padua - Dipa imen o di Psicologia Gene ale 2 Yulius P a omo Sa ya Wacana Ch is ian Uni e si y Ta ek Jabe -López Jaume I Uni e si y Pablo B anas-Ga za Middlesex Uni e si y 3 Con ex -dependen chea ing: Expe imen al e idence om 16 coun ies Abs ac : Policy make s use se e al in e na ional indices ha cha ac e ize coun ies acco ding o he quali y o hei ins i u ions. Howe e , no e o has been made o s udy how he hones y o ci izens a ies ac oss coun ies. This pape explo es he hones y among ci izens ac oss six een coun ies wi h 1440 pa icipan s. We employ a e y simple ask whe e pa icipan s ace a ade-o be ween he joy o ea ing a ine chocola e and he disu ili y o ha ing a h ea ened sel -concep because o lying. Despi e he incen i es o chea , we ind ha indi iduals a e mos ly hones . Fu he , in e na ional indices ha a e indica i e o ins i u ional hones y a e comple ely unco ela ed wi h ci izens' hones y o ou sample coun ies. Keywo ds: Hones y, co up ion, cul u al di e ences 1. In oduc ion Imagine you sel on a uni e si y campus. You see he ollowing announcemen : “Is his you lucky day? Flip a coin and win a Chocola e”. You app oach he desk and ecei e he oppo uni y o win a delicious and beau i ully w apped Lind Lindo chocola e u le by lipping a coin wi h a black and a whi e side. You lip he coin in a box a a nea by able; he box p o ec s you anonymi y. You win a chocola e i you epo ha he whi e side came up and no hing i you epo black. I you ac ually olled black, he only hing keeping you om enjoying he u le is you mo al compass. You ace a adeo be ween he joy o ea ing a ine chocola e and he indi idual disu ili y o ha ing a h ea ened sel -concep because o lying. The e is no social shaming o os acism. We an his expe imen ( h ee ea men s) in six een coun ies o es how some egula ci izens a ound he wo ld beha e in such a si ua ion. In e es ingly, we did no ind any signi ican di e ences ac oss an o he wise e y he e ogeneous se o coun ies. Mos s udies abou cul u al di e ences ega ding dishones beha io ha e used he su ey me hodology. T anspa ency In e na ional epo s la ge di e ences in co up ion a ound he wo ld and he Wo ld Value Su ey documen s c oss-cul u al di e ences in opinions ega ding how “jus i iable” i is o chea on axes o public anspo a ion a es. Howe e , co up ion indices measu e pe cep ions and no ac ual beha io s. Fu he , a ia ions ac oss coun ies in academic dishones y (Rawwas, Al-Kah ib and 4 Vi ell, 2004) and ax e asion (Alm, Sanchez and de Juan, 1995; Cumming, Ma inez, McKee and To gle , 2009) ein o ce he imp ession ha chea ing is bo h abundan and di e se a ound he wo ld. In he same line, ecen esea ch in beha io al and expe imen al economics has shown ha a la ge ac ion o indi iduals a e p one o chea ing (A iely, 2012; Gneezy, 2005). An excep ion o his end is Abele e al. (2014) who epo a phone-based incen i ized expe imen wi h a ep esen a i e sample in Ge many. They ind ha mos people epo hones ly a e lipping a coin in absolu e p i acy wi h a 50% chance o ge ing a payo o 15 eu os in cash o h ough an Amazon gi ce i ica e. Abele and colleagues also complemen ed hei s udy wi h a labo a o y expe imen and ind ha he e is a sligh ly highe le el o dishones epo ing in his se ing. The dominan iew in he li e a u e is ha indi iduals engage in dishones beha io as long as hey can main ain a posi i e sel -image while ob aining he maximum payo s om hei dishones y (Maza , Ami and A iely, 2008). Fu he , esea ch has shown ha c ea ing a jus i ica ion o a posi i e sel -image while beha ing dishones ly is con ex speci ic (Fosgaa d, Hansen and Pio esan, 2013; Gino and Galinsky, 2012; Pascual-Ezama, P elec and Dun ield, 2013). In his pape , we examine p ecisely his. Tha is, whe he chea ing pe se di e s ac oss di e en coun ies and whe he he con ex 1 in luences his beha io . We conduc ed an expe imen in which pa icipan s epo ed he ou come o a coin oss o win a p ize. We eplica ed he expe imen in a di e se se o 16 coun ies a ound he wo ld wi h 1440 subjec s, 90 in each coun y (30 pe ea men ). We used a simple chea ing ask based on Bucciol and Pio esan (2011). We ook g ea ca e o un each session unde simila condi ions. The loca ion o he expe imen was always a busy common a eas on uni e si y campuses (See Table 1 o an o e iew). Pa icipan s had o lip a black/whi e coin; i he ou come was epo ed whi e, hey ob ained a ed Lind Lindo T u le; i epo ed black, hey ob ained no hing. As ou sample includes subs an ial cul u al a ia ion—including pa icipan s om Anglo-Saxon, La in, Ge manic, No dic and Asia ic coun ies—one would expec subs an ial he e ogenei y in chea ing. In addi ion, and consis en wi h a social cons uc ionis iew, he e ec s o cul u e depend on he speci ics o he choice con ex (Gel and, 2013; K ame and Messick, 1995). In he i s ea men ( he Sel -Repo ed T ea men , SRT) each pa icipan lipped a coin in a p i a e a ea wi hou he p esence o he esea che s o o he pa icipan s. A e wa ds he subjec illed a epo shee – indica ing sex and he whi e/black coin- oss esul . I was clea ly indica ed ha he illed o m should be le in he box nea by. No in e ac ion wi h he expe imen e s occu ed in his ask. Taking in o accoun ha he e a e he e ogenous ypes o dishones people (Gneezy, Rockenbach and Se a-Ga cía, 2013), his ask is a good measu e o dishones y o se e al easons. Fi s ly, as a g owing body o esea ch sugges s ha in comple ely anonymous se ings, whe e he isk o being caugh 1 The con ex de ines he s eng h o he mo al compass in ou expe imen . 5 is inexis en , inding 100% chea ing is no unusual (Pascual-Ezama, P elec and Dun ield, 2013; Shu, Maza , Gino, A iely and Baze man, 2012), howe e , people may also es ic he amoun o chea ing (Ayal and Gino, 2011; Gneezy, 2005). In ou ask he e is no possibili y o es ic he amoun o chea ing, he decision is simply o be hones o no ( epo black o whi e). Fu he , decision making is immedia e and in ui i e, ins ead o delibe a i e and medi a i e in a cogni i e dual sys em (Baze man and Tenb unsel, 2011; Kahneman, 2011). In ou expe imen , chea ing is an au oma ic esponse and he need o jus i ica ion ma e s only when people ha e enough ime o delibe a e (Shal i, Elda and Be eby-Meye , 2012). Finally, he ewa d is a simple chocola e ha gi es ins an g a i ica ion. The wo o he ea men s we e he W i en and he Ve bal Repo ed T ea men s (WRT and VRT, espec i ely). In WRT, pa icipan s comple ed a epo shee in p i a e and submi ed i o he expe imen e . The expe imen e made a no e o he epo ed ou come, and, i due, handed he chocola e o he pa icipan . In VRT, pa icipan s we e no asked o ill any o m. They e bally epo ed he ou come o he coin lip o he expe imen e . I epo ed whi e, he expe imen alis handed a chocola e o hem. Ou di e en ea men s allow us o unde s and how he le el o chea ing is shaped by con ex , i.e. he di e ences in epo ing ac oss ea men s. Building on he heo y o sel -image main enance (Maza , Ami and A iely, 2008), we p edic ha ou h ee ea men s will ha e di e en implica ions wi h ega d o he mo al p ocesses o epo ing inco ec ou comes. We conjec u e ha mis epo ing in he SRT is he leas mo ally di icul decision since epo ing does no in ol e any kind o social in e ac ion and, subsequen ly, no social p essu e om he expe imen e . Consequen ly, he decision o be dishones is only a adeo be ween he in e nal mo al p essu e (and he consequen guil expe ienced) agains he pleasu e o enjoying a Lind chocola e. In WRT, an addi ional mo al p essu e o igina es om he ac ha he decision is documen ed on pape and handed o ano he pe son. This addi ional social p ocess is likely o in ol e shame (on op o guil ) in case o dishones y, and hence we expec less dishones y occu ing in his ea men . Rela i e o e bal epo s, we conjec u e ha w i en epo s a e mo e dis an and make i easie o ake away in e nal mo al con ol (Shu e al. 2012). The e o e, we expec he VRT o b ing e en s onge sel -image e osion in case o chea ing, since he un u h ul decision now mus be s a ed ace- o- ace. In his si ua ion we expec he mos mo al p essu e o be p esen and hence he leas dishones beha io . This is consis en wi h Bandu a, Ba ba anelli, Cap a a and Pas o elli (1996) who ind ha e bal epo s a e likely o be less p one o dishones beha io han w i en epo s. Gi en his we hypo hesize ha sel - epo ing wi h no in e ac ion (wi h he expe imen e ) inc ease he emp a ion o be dishones . 6 Table 1. Summa y o coun ies s udied. Coun y Ci y Uni e si y Aus ia Linz Johannes Keple Uni e si y Linz Belgium B ussels V ije Uni e si ei B ussel (VUB) Colombia Bogo á Uni e sidad de los Andes Denma k Copenhagen Copenhagen Uni e si y Finland Oulu Uni e si y o Oulu Ge many Nu embe g Uni e si y o E langen-Nu embe g G eece Re hymno Uni e si y o C e e India Delhi Hans aj College, Uni e si y o Delhi Indonesia Sala iga Uni e si as K is en Sa ya Wacana I aly Pado a Uni e si à di Pado a Japan Tokyo Waseda Uni e si y Ne he lands U ech U ech Uni e si y Spain Mad id Uni e sidad Au ónoma Mad id Tu key Is anbul Bahçeşehi Üni e si esi USA Bal imo e Johns Hopkins Uni e si y UK London Middlesex Uni e si y 2. Expe imen al design We eplica e a simple coin ask by Bucciol and Pio esan (2011). As no ed be o e, pa icipan s had o lip a black/whi e coin. I he epo ed oss was whi e hey ob ained a ed Lind Lindo chocola e and no hing o he wise. A i s glance, he choice o chocola e as an incen i e migh seem as unusual compa ed o he con en ional choice o money. We belie e ha chocola es a e a good solu ion o s udying decision- making ac oss coun ies, gi en ha ou s udy is in a ield se ing and no in he labo a o y. In ac , we belie e ha i is ad an ageous o ell s uden s you ha e a chance o winning a chocola e, a he han elling hem you ha e a chance o win a dolla (o any co esponding local cu ency). Consuming a small snack in such a si ua ion is simply much mo e o an e e yday e en , and hence eels mo e na u al. And a mo e na u al se ing enables g ea e expe imen e con ol and ex e nal alidi y. Fu he mo e, a chocola e 7 is a chocola e anywhe e, and subjec s ne e ha e o calcula e he mone a y alue o he good, jus ecei e a small delicious snack. Clea ly, he e a e many o he po en ial incen i e app oaches one could apply. A mone a y p ize could be scaled based on Big Mac p izes, GDP, o any pu chasing powe index. Al hough he e migh be coun y di e ences in chocola e p e e ence (as indica ed by Messe li, 2012), he choice o chocola e as paymen cons i u es jus as many o jus as ew conce ns as any o he c oss-coun y paymen ehicle, including coun y-adjus ed mone a y compensa ions. Ano he possible conce n is he e y small size o he incen i e. We acknowledge he small size, ye we wan o emphasize ha conside ing he ex emely sho du a ion o he expe imen (pe haps 1 o 2 minu es), ou paymen ansla es in o an hou ly paymen simila o wha is used in mos labo a o y s udies. Ou pa icipan s a e uni e si y s uden s2. The coin was lipped only once and by one pe son a a ime. The ask was pe o med in a p i a e a ea a a sa e dis ance om he expe imen e s o o he pa icipan s. The loca ion o he expe imen was wi hin he college/uni e si y campus, o example, in a hallway in he ca e e ia a ea, hence ensu ing a s eady low o po en ial pa icipan s. To ensu e ha we do s udy coun y di e ences, we ensu ed ha only na i e people pa icipa ed in he a ious loca ions. The impo ance o ensu ing his is suppo ed by e idence o he e ec o immig an s’ o iginal coun y social no ms in e hical beha io (Koun o is and Remoundou, 2013). The expe imen was ca ied ou in he local wo king language o bo h he e bal and he w i en pa .3 The whole s udy comp ises h ee di e en ea men s acco ding o he le el o p oximi y be ween he pa icipan and he expe imen e : • Sel -Repo ed T ea men (SRT): A s uden app oached he expe imen e (a able A, see Figu e A1 in he appendix o de ails), he/she was handed a epo shee , and was guided o able B ( a enough away om he esea che a ea) o lip he black/whi e coin p i a ely (see B in Figu e A1). The coin lip was ca ied ou inside a small ca dboa d box o make i uly p i a e. The s uden no ed gende and he esul o he coin lip on he epo shee a able B. A e illing ou he epo shee , he s uden walked o e o a hi d able ( able C in ano he p i a e a ea a a dis ance om he expe imen e , see C in Figu e A1) o deposi he epo shee in a box, and o ake he chocola e him/he sel in case o a whi e oss. • W i en Repo ed T ea men (WRT): Exac ly he same as SRT, bu now wi h he illed ou epo shee in hand, he s uden wen back o able A, ga e he epo shee o he 2 Recen pape s show ha ypical expe imen al subjec s a e no eally di e en om o dina y people (see o ins ance Exadak ylos, Espin and B anas-Ga za, 2013 o social dilemmas.) 3 A ew s uden s asked why he coin-chocola e e en was ca ied ou . Ou p epa ed explana ion was ha he pu pose was o unde s and he s uden s’ in e es o chocola es. This s a emen was in en ionally made imp ecise, such ha s uden s did no hink his was a es bu a he saw i as a chocola e p omo ion (we ne e said i was a Lind p omo ion). 8 expe imen e and ecei ed he chocola e i a whi e oss was epo ed. Table C is he e o e no used (see igu e A1, appendix). • Ve bal Repo ed T ea men (VRT): Exac ly he same as WRT, bu now he s uden did no ha e o ill ou he epo shee (only epo gende ) and had o in o m he expe imen e e bally abou he ou come o he coin oss. We an he expe imen wi h 90 pa icipan s in each coun y. The e we e 10 s uden s o each ea men and his was epea ed h ee imes: 10 (WRT), 10 (VRT), 10 (SRT), 10 (WRT), 10 (VRT), 10 (SRT), 10 (WRT), 10 (VRT), 10 (SRT). We s a ed each ea men wi h en chocola es on a pla e on able B (VRT and WRT) o able C (SRT). When we changed he ea men we e illed he chocola es aken by he pa icipan s and egis e ed he exac numbe o chocola es eplaced4. A he beginning o each ea men we s a ed wi h en chocola es in he pla e. A he end o he expe imen we compa ed he numbe s in ou eco ds wi h he epo s by he pa icipan s. In WRT and SRT pa icipan s w o e he colo on he epo shee . In VRT pa icipan s only w o e he gende . In his case, when pa icipan s ga e us he epo shee and old us he colo we o e a li le co ne o he epo shee when hey old us black so a he end we could know how many blacks and whi es we e decla ed. No e ha , as we kep no names o indi idual-speci ic eco ds, we had o be e y ca e ul abou whe he someone e u ned. Expe imen e s we e ins uc ed o in o m anyone who ied o e u n o lip he coin again ha i was only possible o pa icipa e once. In case, acul y membe s showed up o pa icipa ion, i was emphasized ha i was in ended o be a s uden e en . Comple ing he expe imen ook abou a minu e o wo. 3. Resul s 3.a. Di e ences by ea men s To ou g ea su p ise, o e all only 57% o he pa icipan s epo ed whi e ac oss all h ee ea men s and coun ies. Since he p obabili y o ob aining whi e is 50%, his esul indica es ha 86% esis ed he emp a ion o lie [100% - 2(57%-50%)]. We conclude ha li le dishones y is obse ed ac oss all coun ies and ea men s. Looking a he di e en ea men s, he deg ee o dishones beha io , lis ed as he pe cen age o whi e ou comes, con i ms ou ini ial conjec u es. We obse e a g ea e amoun o lying in he SRT, wi h WRT in he middle, and wi h VRT ha ing he smalles amoun o lying. 62% (Sel RT) > 57% (W i enRT) >53% (Ve balRT) 4 No e ha ha ing mo e o less chocola es on he pla es is no an issue o pa icipan s as hey ne e ge o know he o al. Wha hey see (on he pla e) is he only e e ence hey ha e. 9 The di e ences be ween SRT and WRT a e no s a is ically signi ican (χ2=.316; p- alue=.574), al hough he a e age ac oss he six een socie ies we e 5 pe cen age poin s highe o he o me . Howe e , using a con ingency able (Pea son χ2) he di e ence be ween SRT and VRT was la ge (9 pe cen age poin s) and signi ican a he 5% le el (χ2=4.479; p- alue=0.034). Fu he , we do no ind any signi ican di e ences be ween WRT and VRT (χ2=2.422; p- alue=0.120). Thus, compa ing he di e en ea men s we ind ha he e bal epo s a e likely o be less p one o dishones beha io han w i en epo s in line wi h p e ious li e a u e (Bandu a, Ba ba anelli, Cap a a and Pas o elli, 1996; Fes inge and Ca lsmi h, 1959; Maza , Ami and A iely, 2008) and we u he ind ha sel - epo ing he answe s (and no ha ing o deli e i o he expe imen e ) inc eases his e ec . 3.b. Di e ences by coun ies A second impo an esul is ha we ind no s a is ically signi ican di e ences ac oss he 16 coun ies ac oss any ea men : SRT (χ2 = 16.953; p- alue=0.322); VRT (χ2 = 15.691; p- alue=0.403) and WRT (χ2 = 22.674; p- alue=0.091). This sugges s ha he e is a high le el o homogenei y in beha io ac oss ou sample coun ies. Addi ional in e es ing esul s show up om a close examina ion o he da a. Since ossing a coin ollows a binomial dis ibu ion, we looked o hose coun ies whe e he e is an anomalous numbe o whi e (awa ded) coins epo ed. Table 2 epo s de ailed in o ma ion by coun y and ea men s5. A he coun y le el, he sha e o whi es epo ed in excess o 50% (ac oss all ea men s), is signi ican (a a 5% signi icance le el) in h ee ou o six een coun ies. Ye , he o e all coun y le el e ec s can mask unde lying e ec s a he ea men le el. We he e o e ocus on he ea men e ec s a he coun y le el. Ob iously a sample size o 30 pe ea men allows us only o make ough assessmen o he deg ee o chea ing, bu we s ill belie e his ough measu e p o ides a use ul compa ison ac oss coun ies and ea men s6. 5 We ha e also conduc ed a Logi analysis o explain he ou come o indi idual coin osses. In he eg ession, among o he ac o s, we con olled o di e en income le els ac oss coun ies (wi h he help o pe -capi a GNI) and o coun y-le el pe -capi a cocoa consump ion. While income does no seem o ha e a signi ican impac , pe -capi a cocoa consump ion has a nega i e e ec on chea ing, i.e. he highe he pe -capi a consump ion he lowe is chea ing. 6 We also conduc he analysis by using Ingleha -Welzel cul u al ca ego ies (based on he Wo ld Value Su eys). Again we do no ind any obus e ec o cul u e on chea ing beha io . 16 Bucciol, A. and M. Pio esan. 2011. Luck o chea ing? A ield expe imen on hones y wi h child en. Jou nal o Economic Psychology, 32 (1), 73-78. C oson, R. and U. Gneezy. 2009. Gende di e ences in p e e ences. Jou nal o Economic Li e a u e, 47(2), 1-27. Cummings, R.G., J. Ma inez-Vazquez, M. McKee and B. To gle . 2009. Tax Mo ale E ec s Tax Compliance: E idence F om Su eys And An A e ac ual Field Expe imen . Jou nal o Economic Beha io and O ganiza ion, 70, 447-457. D ebe , A., Johannesson, M., 2008. Gende di e ences in decep ion. Economic Le e s, 99 (1), 197–199. Exadak ylos, F., A.M. Espín and P. B añas-Ga za. 2013. Expe imen al subjec s a e no di e en . Scien i ic Repo s 3(1231). DOI: 10.1038/s ep01213 Falk, A. and J. Heckman. 2009. Lab Expe imen s A e a Majo Sou ce o Knowledge in he Social Sciences. Science, 326 (5952), 535-538. Fes inge , L. and J.M. Ca lsmi h. 1959. Cogni i e consequences o o ced compliance. Jou nal o Abno mal and Social Psychology, 58, 203–210. Fosgaa d, T., L.G. Hansen and M. Pio esan. 2013. Sepa a ing Will om G ace: An Expe imen on Con o mi y and Awa eness in Chea ing. Jou nal o Economic Beha io and O ganiza ion, 93, 279- 284. Gel and, M.J. e al. 2013. Towa d a Cul u e-by-Con ex Pe spec i e on Nego ia ion: Nego ia ion Teams in he Uni ed S a es and Taiwan. Jou nal o Applied Psychology, 98 (3), 504-513. Gino, F. and A.D. Galinsky. 2012. Vica ious Dishones y: When Psychological Closeness C ea es Dis ance om One's Mo al Compass. O ganiza ional Beha io and Human Decision P ocesses,119 (1), 15-26. Gneezy, U. 2005. Decep ion: The ole o consequences. Ame ican Economic Re iew, 95 (1), 384-394. Gneezy, U., B. Rockenbach and M. Se a-Ga cía. 2013. Measu ing lying a e sion. Jou nal o Economic Beha io & O ganiza ion, 93, 293-300. Kahneman, D. 2011. Thinking: Fas and Slow. Fa a S aus and Gi oux, New Yo k. Koun ou is, Y. and K. Remoundou. 2013. Is he e a cul u al componen in ax mo al? E idence om immig an s in Eu ope. Jou nal o Economic Beha io and O ganiza ion, 96,104-119. K ame , R.M. and D. Messick. 1995. Nego ia ion as a social p ocess. Thousand Oaks, CA: Sage. Magnus, J.R., V.M. Pol e o ich, D.L. Danilo and A.V. Sa a ee . 2002. Tole ance o chea ing: an analysis ac oss cul u es. The Jou nal o Economic Educa ion, 33, 125-135. Maza , N., O. Ami and D. A iely. 2008. The Dishones y o Hones People: A Theo y o Sel -Concep Main enance. Jou nal o Ma ke ing Resea ch, 45 (6), 633-644. 17 Messe li, F.H. 2012. Chocola e consump ion, cogni i e unc ion, and nobel lau ea e. New England Jou nal o Medicine, 367, 1562–1564Pascual-Ezama, D., D. P elec and D. Dun ield. Mo i a ion, Money, P es ige and Chea s. Jou nal o Economic Beha io and O ganiza ion, 93, 367-373. Rawwas, M.Y.A., J.A. Al-Kha ib and S.J. Vi ell. 2004. Academic dishones y: A c oss-cul u al compa ison o U.S. and Chinese ma ke ing s uden s. Jou nal o Ma ke ing Educa ion, 26, 89-100. Shal i, S., O. Elda and Y. Be eby-Meye . 2012. Hones y equi es ime (and lack o jus ifica ions). Psychological Science, 23, 1264–1270. doi: 10.1177/0956797612443835 Shu, L., N. Maza , F. Gino, D. A iely and M.H. Baze man. 2012. Signing a he beginning makes e hics salien and dec eases dishones sel - epo s in compa ison o signing a he end. P oceedings o he Na ional Academy o Sciences, 109 (38), 15197-15200. 18 Appendix: De ails on expe imen al design We used a pos e wi h he ollowing ex : “Is his you lucky day? Flip a coin and win a Chocola e”. When he pa icipan s a i ed, he esea che s ga e a epo shee o hem wi h di e en ex (depending on he ea men ). E e y pa icipan in he h ee di e en ea men s ma ked he gende in he epo shee ( ha was impo an because in he SRT ea men esea che s canno eco de he gende ). Ins uc ions we e he ollowing: • Non Repo ed T ea men : “I you wan o win a chocola e, go o able B and lip a coin inside he box. Then ma k on he pape he colo you had, black o whi e, and you gende . Go o he able C, and deposi he pape in he box. I you ma k whi e you can ake a chocola e, i you ma k black you jus lea e”. • W i en Repo ed T ea men : “I you wan o win a chocola e, go o able B and lip a coin inside he box. Then ma k on he pape he colo you had, black o whi e, and you gende . Come back and gi e us he pape . I you ma k whi e we will gi e you a chocola e, i you ma k black, no hing” • Ve bal Repo ed T ea men : “I you wan o win a chocola e, go o able B and lip a coin inside he box. Then ma k on he pape you gende , come back, and ell us he colo you had. I i was whi e we will gi e you a chocola e, i i is black, no hing” 19 Figu e A1. P ocedu e and ables’ dis ibu ion. Block A: WRT and VRT ables’ dis ibu ion. We used a able (Table A) wi h a Pos e (whe e he esea che s we e) in a isible si ua ion o s uden s ( o example in hallway in he ca e e ia a ea). Wi h a couple o me e s o dis ance in ano he able (Table B) he e was a box wi h a black/whi e coin inside whe e he pa icipan s lipped he coin (inside he box) wi hou nobody could see hem. Block B: SRT ables’ dis ibu ion. In SRT he e was ano he able (Table C) wi h was a small box whe e he pa icipan s ga e he epo shee and ook he chocola es. Block C: Pic u e example in Spain. Al hough he expe imen ules a e w i en in he epo s he esea che s p o ide o he pa icipan s a he beginning o he ask, all esea che s a e na i e in o de o a oid misunde s andings in he communica ion. CHOCOLATES REPORT BOX TABLE C COIN BOX TABLE B PEN RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A CHOCOLATES REPORT BOX TABLE C CHOCOLATES REPORT BOX CHOCOLATES REPORT BOX TABLE C COIN BOX TABLE B PEN COIN BOX TABLE B COIN BOXCOIN BOX TABLE B PENPEN RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e RESULTS BOX RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A COIN BOX TABLE B PEN CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e CHOCOLATES RESULTS BOX CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A COIN BOX TABLE B PEN COIN BOX TABLE B COIN BOXCOIN BOX TABLE B PENPEN A B C CHOCOLATES REPORT BOX TABLE C COIN BOX TABLE B PEN RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A CHOCOLATES REPORT BOX TABLE C CHOCOLATES REPORT BOX CHOCOLATES REPORT BOX TABLE C COIN BOX TABLE B PEN COIN BOX TABLE B COIN BOXCOIN BOX TABLE B PENPEN RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e RESULTS BOX RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A COIN BOX TABLE B PEN CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e CHOCOLATES RESULTS BOX CHOCOLATES RESULTS BOX Is you lucky day? Flip a coin and win a Chocola e TABLE A COIN BOX TABLE B PEN COIN BOX TABLE B COIN BOXCOIN BOX TABLE B PENPEN A B C