Punishment-related memory-guided attention: Neural dynamics of perceptual modulation

PUNISHMENT-RELATED MEMORY-GUIDED ATTENTION: NEURAL DYNAMICS OF PERCEPTUAL MODULATION Au ho s: Samuel Suá ez-Suá ez, Soco o Rod íguez Holguín, Fe nando Cada ei a, Anna Ch is ina Nob e, Sonia Doallo This is he pee e iewed e sion o he ollowing a icle: Suá ez-Suá ez S; Rod íguez Holguín S; Cada ei a F; Nob e AC; Doallo S (2019). Punishmen - ela ed memo y- guided a en ion: Neu al dynamics o pe cep ual modula ion. Co ex, 115, 231-245. doi: 10.1016/j.co ex.2019.01.029. This a icle may be used o non-comme cial pu poses in acco dance wi h Else ie Te ms and Condi ions o Use o Sel -A chi ed Ve sions. 2 Punishmen - ela ed memo y-guided a en ion: Neu al dynamics o pe cep ual modula ion Samuel Suá ez-Suá eza, Soco o Rod íguez Holguína, Fe nando Cada ei aa, Anna Ch is ina Nob eb, Sonia Doalloa a Depa men o Clinical Psychology and Psychobiology, Uni e si y o San iago de Compos ela, San iago de Compos ela, Spain b Depa men o Expe imen al Psychology and Ox o d Cen e o Human B ain Ac i i y, Depa men o Psychia y, Wellcome Cen e o In eg a i e Neu oimaging, Uni e si y o Ox o d, Ox o d, Uni ed Kingdom Co esponding au ho : Sonia Doallo Depa men o Clinical Psychology and Psychobiology Uni e si y o San iago de Compos ela Campus Vida, s/n, 15782 San iago de Compos ela Galicia-Spain Telephone: +34 881813916 E-mail: [email p o ec ed] Decla a ions o in e es : none. Pos -p in ( inal d a pos - e e eeing) 3 ABSTRACT Remembe ing he ou comes o pas expe iences allows us o gene a e u u e expec a ions and shape selec ion in he long- e m. A g owing numbe o s udies has shown ha lea ned posi i e ewa d alues impac spa ial memo y-based a en ional biases on pe cep ion. Howe e , whe he memo y-d i en a en ional biases ex end o punishmen - ela ed alues has ecei ed compa a i ely less a en ion. He e, we manipula ed whe he ecen spa ial con ex ual memo ies became associa ed wi h success ul a oidance o punishmen (po en ial mone a y loss). Beha iou al and elec ophysiological measu es we e collec ed om 27 pa icipan s du ing a subsequen memo y-based a en ion ask, in which we es ed o he e ec o punishmen a oidance associa ions. Punishmen a oidance signi ican ly ampli ied e ec s o spa ial con ex ual memo ies on isual sea ch p ocesses wi hin na u al scenes. Compa ed o non-associa ed scenes, con ex ual memo ies pai ed wi h punishmen a oidance lead o as e esponses o a ge s p esen ed a emembe ed loca ions. E en - ela ed po en ials elici ed by a ge s imuli e ealed ha acqui ed mo i a ional alue o speci ic spa ial loca ions, by i ue o hei associa ion wi h pas a oidance o punishmen , dynamically a ec ed neu al signa u es o ea ly isual p ocessing (indexed by la ge P1 and ea lie N1 po en ials) and a ge selec ion (as indica ed by educed N2pc po en ials). The p esen esul s ex end ou unde s anding o how memo y, a en ion, and punishmen - ela ed mechanisms in e ac o op imize pe cep ual decision in eal wo ld en i onmen s. Keywo ds: Long- e m memo y, isuospa ial a en ion, punishmen , e en - ela ed po en ials Pos -p in ( inal d a pos - e e eeing) 4 1. In oduc ion P e ious esea ch has demons a ed he abili y o spa ial con ex ual long- e m memo y (LTM) o guide a en ion wi hin scenes, and o modula e neu al signa u es o ea ly pe cep ual analysis (Summe ield, Rao, Ga side, & Nob e, 2011) and selec ion (Pa ai, Doallo, & Nob e, 2012) o ele an objec s p esen ed a memo ized loca ions. This memo y-d i en modula ion o isual p ocessing engages ac i i y in he on al- pa ie al ne wo k o isual-spa ial o ien ing in conce wi h ac i a ion in o he b ain egions implica ed in e ie al o objec loca ions wi hin speci ic con ex s (e.g., hippocampus) (S okes, A he on, Pa ai, & Nob e, 2012; Summe ield, Lepsien, Gi elman, Mesulam, & Nob e, 2006). Mo e ecen wo k by Rosen and colleagues shows ha LTM-guided isuospa ial a en ion ec ui s a ne wo k spanning pa ie al co ical a eas (la e al in apa ie al sulcus, pos e io p ecuneus and pos e io callosal sulcus) and subco ical egions ( he cauda e head, mediodo sal halamus and lobule VI/C us I o ce ebellum) (Rosen, S e n, De aney, & Some s, 2017; Rosen, S e n, Michalka, De aney, & Some s, 2016). A g owing numbe o s udies ha e also s a ed add essing he impo an ques ion o whe he and how lea ned ewa d alues impac spa ial memo y-based a en ional biases on pe cep ion. Recen wo k has shown ha ewa d- ou come associa ions boos memo ies and a en ional p io i y o speci ic spa ial loca ions (Chelazzi e al., 2014; Hickey, Chelazzi, & Theeuwes, 2014; see also Ande son, 2015; Pollmann, Es ocino a, Somme , Chelazzi, & Zinke, 2016), and dynamically impac di e en le els o isual co ical p ocessing o a ge s p esen ed a p e iously ewa ded loca ions (Doallo, Pa ai, & Nob e, 2013). These e ec s o ewa d his o y on a en ional p io i y o space add o accumula ing e idence indica ing ha ewa d-associa ions modula e bo om-up and op-down a en ional p io i y o s imuli ea u es and objec ca ego ies (e.g. Ande son, Lau en , & Yan is, 2011; Della Libe a & Pos -p in ( inal d a pos - e e eeing) 5 Chelazzi, 2006, 2009; Donohue e al., 2016; Ha is e al., 2016; Hickey, Chelazzi, & Theeuwes, 2010; Hickey, Kaise , & Peelen, 2015; Hickey & Peelen, 2015; Kiss, D i e , & Eime , 2009; o e iews see Ande son, 2013; Chelazzi, Pe la o, San and ea, & Della Libe a, 2013; Failing & Theeuwes, 2018; Pessoa, 2015; Vuilleumie , 2015). An impo an emaining ques ion is whe he punishmen - ela ed associa ions also in luence memo y-d i en spa ial-a en ional biases. In con as o he inc easing numbe o s udies examining posi i e ewa d- ela ed e ec s on hippocampus-dependen memo y and isuospa ial a en ion, less is known abou whe he simila e ec s a e obse ed when spa ial loca ions acqui e mo i a ional alue h ough he associa ion wi h an a e si e ou come (o i s a oidance). The complexi y o his issue inc eases i we conside ha he ques ion o whe he ewa d and punishmen p ocessing ely on a common o on dis inc neu al sys ems is s ill deba ed (Bissone e, Gen y, Padmala, Pessoa, & Roesch, 2014; B ooks & Be ns, 2013; Liu, Hai s on, Sch ie , & Fan, 2011). S imuli ha unde go a e si e condi ioning ha e been shown o exhibi enhanced p ocessing unde challenging condi ions (Padmala & Pessoa, 2008), o cap u e a en ion au oma ically du ing isual sea ch (Schmid , Belopolsky, & Theeuwes, 2015), and o coun e ac he a en ional blink (Lim, Padmala, & Pessoa, 2009). G owing e idence sugges s ha s imulus ea u es and objec s associa ed wi h a mone a y loss h ough associa i e lea ning also a ec a en ional and pe cep ual p ocesses. Face s imuli associa ed o he eceip o a mone a y loss a e mo e likely o be ecognized han o he aces in a subsequen a en ional blink ask (Raymond & O’B ien, 2009) and ecei e enhanced isual p ocessing in a masked ecogni ion ask (O’B ien & Raymond, 2012). S onge eo ien ing p ocesses o ab up -onse (exogenous) colo ed cues linked o a mone a y punishmen ha e also been epo ed (Bucke & Theeuwes, 2016; bu see Ru he o d, O'B ien, & Raymond, 2010), as well as inc eased a en ional cap u e by Pos -p in ( inal d a pos - e e eeing) 6 colo ed single on dis ac o s associa ed o e asion o a mone a y loss in a isual sea ch ask (Wen u a, Mülle , & Ro he mund, 2014). Recen da a e eal ha pe cep ual ea u es associa ed o eceip o punishmen boos p ima y isual co ex esponses (as e lec ed by he highe ampli ude o he C1 po en ial), ela i e o ewa d-associa ed and neu al ones (Rossi e al., 2017), al hough e idence o an enhanced ep esen a ion in isual en al a eas o gain- e sus loss-associa ed s imuli has been ound o symbol cues du ing an economic decision-making ask (San Ma ín, Appelbaum, Hue el, & Woldo , 2014) and o objec ca ego ies p esen ed in images o na u al scenes (Ba ba o, Peelen, & Hickey, 2017). Though e idence o modula o y e ec s o punishmen associa ions in isual pe cep ual and a en ional asks, whe he i can be obse ed o speci ic loca ions in clu e ed na u alis ic con ex s emains an open ques ion; u he mo e, i is unclea whe he i has a las ing e ec ha can modula e LTM-based a en ion. He e, we aimed o shed ligh on whe he and how punishmen - ela ed mechanisms in luence he abili y o LTM o guide pe cep ion in na u alis ic con ex s. We used a simila expe imen al app oach as in ou p e ious s udy (Doallo e al., 2013) o disen angle whe he spa ial expec a ions om LTM inco po a e alue signals speci ically ela ed o pas a oidance o an a e si e ou come o magni y expe ience- based biases upon pe cep ual decisions on ele an objec s in clu e ed scenes. E en - ela ed po en ials (ERPs) we e eco ded o e eal he ime cou se o e which punishmen -associa ed memo y-guided a en ion exe s i s e ec s on he ongoing neu al ac i i y in isual co ical a eas. We analysed he P1, N1 and N2pc componen s. These po en ials e lec isual p ocessing (P1 and N1) and a ge selec ion (N2pc). The P1 po en ial, a posi i e de lec ion peaking abou 100 ms a e s imulus p esen a ion, is modula ed by isuospa ial a en ion and hough o e lec a senso y gain con ol Pos -p in ( inal d a pos - e e eeing) 7 mechanism wi hin ex as ia e isual co ex (Heinze e al., 1994; Hillya d, Vogel, & Luck, 1998; Ma ínez e al., 1999). The N1 po en ial is a nega i e de lec ion, subsequen o he P1, which is also modula ed by a en ion. The N1 a en ion e ec is belie ed o index a op-down modula ion o disc imina i e p ocessing in a eas o he en al isual s eam (Hop , Vogel, Woodman, Heinze, & Luck, 2002; Luck, 1995; Vogel & Luck, 2000). The N2pc is an enhanced nega i i y a pos e io elec ode si es con ala e al o he loca ion o an a ended a ge , ypically eme ging 200-300 ms a e a ge onse . I is hough o e lec a en ional selec ion o a a ge among dis ac e s (Eime , 1996; Luck & Hillya d, 1994) and is gene a ed in pa ie al and in e io occipi al- empo al co ical a eas (Hop e al., 2000). These ERP componen s ha e been p e iously shown o be modula ed by memo y-guided o ien ing (Doallo e al., 2013; Pa ai e al., 2012; Summe ield e al., 2011) and ha e also been used o in es iga e alue-based modula ion o pe cep ual p ocessing and a en ional selec ion in o he ypes o asks (Donohue e al., 2016; Ha is e al., 2016; Hickey e al., 2010; I hipu ipa , Cha, Rangsipa , & Se ences, 2015; McLean & Giesb ech , 2015; Qi, Zeng, Ding, & Li, 2013). In he p esen expe imen , pa icipan s i s pe o med a lea ning ask du ing which hey lea n he spa ial loca ion o a p ede ined a ge (a small key) embedded wi hin isual scenes. Punishmen - ela ed associa ions we e manipula ed by punishing bad pe o mance du ing he las block o he lea ning ask on a p opo ion o ials. Twen y- ou hou s la e , hey comple ed an LTM-cued isual sea ch ask in which hey had o disc imina e he p esence o absence o a ge key s imuli wi hin he p e iously s udied scenes while ERPs we e eco ded. The ini ial p esen a ion o he scene (wi hou he a ge ) se ed as a con ex ual memo y-based cue o o ien spa ial a en ion in each ial. Con ex ual scene cues could ei he ha e been associa ed wi h a oidance o an Pos -p in ( inal d a pos - e e eeing) 8 a e si e ou come (i.e. po en ial mone a y loss) o ha e had no ou come associa ion. Based on ou p io wo k (Doallo e al., 2013), we hypo hesized ha i expe ience- dependen a en ional biases on pe cep ion a e in luenced by punishmen a oidance associa ions, his would lead, speci ically, o (i) enhanced beha iou al pe o mance, as e ealed by imp o ed eac ion imes and accu acy, and (ii) modula ion o neu al signa u es o ea ly isual p ocessing (exp essed as enhanced ampli udes o he P1 po en ial and ea lie la encies o N1) and a ge selec ion (indexed by educed ampli udes o N2pc). 2. Ma e ials and Me hods 2.1. Pa icipan s Thi y- wo heal hy s uden s om he Uni e si y o San iago de Compos ela (Galicia, Spain) pa icipa ed in his s udy o mone a y compensa ion. Pa icipan s comple ed a ques ionnai e ega ding pe sonal his o y o neu ological and/o psychia ic disease, exis ing ch onic disease and/o cu en pha macological ea men . All pa icipan s ga e w i en consen . Da a om i e pa icipan s we e disca ded du ing EEG p ep ocessing because o excessi e a i ac s in hei EEG eco ding. The emaining 27 pa icipan s (18 women, age ange 19-27, mean age 20.89 ± 2.02) we e all igh -handed and had no mal o co ec ed- o-no mal ision. The s udy was app o ed by he Bioe hics Commi ee o he Uni e si y o San iago de Compos ela. 2.2. Expe imen al p ocedu e The ask used in his s udy was a modi ied e sion o he expe imen al design employed in Doallo e al. (2013). The e we e wo phases o he expe imen . Pa icipan s i s pe o med a lea ning ask (see Figu e 1), ollowed on he second day by a memo y-cued Pos -p in ( inal d a pos - e e eeing) 9 o ien ing ask (in which EEG ac i i y was eco ded) (Figu e 2A) and a spa ial memo y ecall ask (Figu e 2B). 2.2.1. S imuli Two hund ed and ou digi al images o scenes we e ob ained om he image se used in ou p e ious s udy (Doallo e al., 2013). A se o 192 scenes was used in expe imen al ials, and addi ional 12 scenes we e used o amilia iza ion and p ac ice ials. Each scene was p epa ed, using Ma lab (Ma hwo ks, Na ick, MA), in wo di e en o ma s, one o he lea ning ask and one o he o ien ing ask. Fo coun e balancing pu poses, ou lea ning ask e sions we e p epa ed o each scene wi h he key (15 x 29 pixels, equi alen o 0.3º x 0.7º) placed in one o each o he ou isual quad an s, p e e ably in a hidden loca ion. The assignmen o scenes o di e en expe imen al condi ions, key p esence o absence (in he o ien ing ask) and key loca ion wi hin scenes we e coun e balanced ac oss pa icipan s. Fo he o ien ing ask, he scenes wi h keys we e emade o include a la ge and b igh e key (25 x 49 pixels; 0.6º x 1.1º) in he loca ion o he o iginal key o make i isible wi hin he b ie ly displayed a ge scene. Scene s imuli we e p esen ed using P esen a ion (Neu obeha io al Sys ems, Albany, CA) and sub ended 22º x 17º o isual angle a a iewing dis ance o 100 cm. 2.2.2. Lea ning ask Du ing he lea ning ask, pa icipan s iewed 192 na u alis ic scenes epea ed in andom o de o e i e blocks (Figu e 1A). Pa icipan s explo ed he scenes o e ly o sea ch o a small gold key a ge (p esen in all he scenes). Once loca ed, hey ac i a ed he mouse cu so wi h a le -side mouse click and indica ed he loca ion o he key by posi ioning he cu so on he loca ion o he key and making a second le -sided mouse click. A e a esponse o a e he a ailable sea ch ime expi ed, isual w i en eedback Pos -p in ( inal d a pos - e e eeing) 16 we e hus co ec ed o mul iple compa isons. Addi ional analyses should be conside ed as explo a o y and hypo hesis-gene a ing. Fo all p alues epo ed h oughou he pape , we p o ide nominal p alues and, whe e app op ia e, co ec ed p alues using FDR, as well as e ec sizes calcula ed as pa ial e a-squa ed (ŋ2p) alues. 2.5. Media ion analysis Media ion analyses we e conduc ed o examine whe he modula ion o wa e o m po en ials media ed he e ec o condi ion on pe o mance (i.e. whe he he bene i s in a oidance s. sa e ials we e media ed by modula ions o isual co ical p ocessing). We pe o med sepa a e media ion analyses o epea ed-measu es designs using he MEMORE mac o o SPSS (Mon oya & Hayes, 2017), wi h condi ion (a oidance s. sa e) as he independen a iable, ERPs (ampli ude and/o la ency pa ame e s) as he media o s, and beha iou (RT, accu acy and d’) as he dependen a iables. The p ocedu e desc ibed by Mon oya and Hayes (2017) concep ualizes media ion analysis as a pa h-analy ic amewo k in which media ion is assessed by a single es o he indi ec e ec (a*b; i.e. he conjunc ion o he e ec o condi ion on he po en ial media o [pa h a] and he e ec o he po en ial media o on beha io al pe o mance [pa h b]). The di ec e ec o condi ion on beha io al pe o mance ha does no ope a e h ough he media o is also calcula ed (pa h c’). Media ion analyses we e conduc ed using he pe cen ile boo s ap me hod wi h 10,000 i e a ions. The indi ec e ec was conside ed s a is ically signi ican i he con idence in e al (CI 95%) excluded ze o. 3. Resul s 3.1. Fo ma ion o obus LTMs o a ge loca ions wi hin na u al scenes Pos -p in ( inal d a pos - e e eeing) 17 Beha iou al analysis o he lea ning ask con i med ha pa icipan s we e able o es ablish obus memo ies o he spa ial loca ions a which a ge s imuli we e p esen ed. O e he cou se o he lea ning blocks, pa icipan s loca ed an inc easing numbe o a ge s, wi h inc easing speed (Block 1: mean accu acy ± SEM = 92.4 ± 1.1%, mean sea ch imes ± SEM = 4.2 ± 0.2 sec; Block 5: 97.8 ± 0.4%, 1.2 ± 0.1 sec) (Figu e 1B). ANOVAs es ing o linea dec eases in sea ch imes e ealed a signi ican linea con as o e he lea ning blocks, F(1, 26) = 386.64, p < .001, ŋ2p = .94. A signi ican linea inc ease in accu acy o e he lea ning blocks was simila ly e ealed, F(1, 26) = 48.53, p < .001, ŋ2p = .65. No di e ence was ound be ween a oidance and sa e scenes in he inal (six h) block ei he in sea ch ime (a oidance: 1.1 ± 0.05; sa e: 1.1 ± 0.05; F(1, 26) = .07, p = .79, ŋ2p = .003) o in accu acy (a oidance: 97.7 ± 0.5; sa e: 96.9 ± 0.5; F(1, 26) = 3.34, p = .08, ŋ2p = .11). The ecall ask pe o med immedia ely a e he o ien ing ask con i med ha pa icipan s e ained s ong memo ies o he key loca ions wi hin he lea ned scenes on he day a e he lea ning ask. On a e age, pa icipan s could explici ly ecall he co ec loca ions o a ge s on 75% o scenes (± 2.7 SEM). In addi ion, subjec s' con idence a ings a ied sys ema ically wi h he esponse dis ance om ac ual a ge loca ion, F(1.6, 41.6) = 64.96, p < .001, ɛ = .8, ŋ2p = .71, e ealing ha highe con idence a ings we e associa ed wi h mo e accu a e memo ies [mean dis ance in pixels ± SEM; Ra e 1: 81.4 ± 3.9; Ra e 2: 60.7 ± 3.5; Ra e 3: 36.9 ± 1.6] (see Figu e 2B). Recall did no di e be ween a oidance-associa ed and sa e key loca ions (accu acy: 75.9 ± 2.6% s. 73.8 ± 3%, F(1, 26) = 1.93, p = .18, ŋ2p = .07; o e all dis ance in pixels: 58.4 ± 3 s. 60.9 ± 2.9, F(1, 26) = .39, p = .54, ŋ2p = .02; con idence a ings: 2.67 ± .04 s. 2.68 ± .04, F(1, 26) = .01, p = .92, ŋ2p < .001). Pos -p in ( inal d a pos - e e eeing) 18 3.2. One single punishmen a oidance- ela ed associa ion enhances he abili y o spa ial memo ies o d i e isual sea ch in na u al scenes RT and accu acy le els o a ge -p esen and a ge -absen ials in a oidance and sa e condi ions in he o ien ing ask a e summa ized in Table 1. To es ou main hypo heses ega ding beha iou al pe o mance (imp o ed RT and accu acy), we examined he main e ec o condi ion on hese wo dependen a iables. The e was a signi ican e ec on RT, F(1, 26) = 6.67, p = .016, p(FDR co ec ed) = .048, ŋ2p = .204, e ealing ha a ge disc imina ion was as e in a oidance (700.03 ± 31.13) e sus sa e (709.07 ± 31.22) ials. The main e ec o condi ion on accu acy was no signi ican (a oidance: 0.91 ± 0.01, sa e: 0.92 ± 0.01; F(1, 26) = .09, p = .77, p(FDR co ec ed) = .81, ŋ2p = .003). In addi ion, RT and accu acy ANOVAs showed he ollowing esul s. Ta ge disc imina ion was as e in a ge -p esen ials (p esen : 651.58 ± 32.3, absen : 757.53 ± 32.37; F(1, 26) = 36.59, p < .001, ŋ2p = .59), bu no signi ican in e ac ion was ound be ween condi ion and a ge p esence on RT, F(1, 26) = .58, p = .45, ŋ2p = .02. Analysis o accu acy e ealed ma ginally signi ican e ec s o a ge p esence (p esen : 0.91 ± 0.01, absen : 0.93 ± 0.01; F(1, 26) = 3.93, p = .058, ŋ2p = .13) and o he in e ac ion be ween condi ion and a ge p esence, F(1, 26) = 3.89, p = .059, ŋ2p = .13, which showed a end o highe accu acy o sa e s. a oidance condi ions in a ge absen ials. To es he possibili y ha RT e ec s e lec ed a speed-accu acy ade-o , Pea son’s co ela ion be ween RT and accu acy e ec s (calcula ed by sub ac ing sa e ials om a oidance ials) was pe o med. A signi ican nega i e co ela ion ( = -.48, p = .006) indica ed ha as e esponses in a oidance s. sa e scenes we e associa ed wi h highe le els o accu acy. Pos -p in ( inal d a pos - e e eeing) 19 Finally, he analysis o pe cep ual sensi i i y showed ha he d’ measu e was equi alen be ween a oidance and sa e ials (a oidance: 3.27 ± 0.63, sa e: 3.2 ± 0.61; (26) = .58, p = .57). 3.3. Punishmen a oidance-associa ed memo ies modula e a ge - ela ed neu al ac i i y 3.3.1. Ea ly isual p ocessing (P1 and N1 po en ials) Ta ge -p esen scenes elici ed he expec ed isual po en ials P1 and N1 o e pa ie o- occipi al scalp egions in all condi ions (Figu e 3). As explained in he In oduc ion sec ion, d awing on ou p e ious wo k (Doallo e al., 2013; see also Summe ield e al., 2011), we expec ed ha punishmen a oidance- associa ed memo y-based o ien ing would esul in enhanced ampli udes o he P1 po en ial and ea lie la encies o N1. As p edic ed, he e was a signi ican main e ec o condi ion on P1 ampli ude, F(1, 26) = 4.31, p = .048, ŋ2p = .14, e ealing ha P1 was la ge o a ge s appea ing a a oidance-associa ed emembe ed loca ions ela i e o sa e loca ions. A e ha ing iden i ied ha one ol age alue was ou side 3SD om he mean, we e i ied ha he main e ec o condi ion on P1 ampli ude emained signi ican when he ou lie was eplaced by he se ies’ mean, F(1, 26) = 4.68, p = .040, ŋ2p = .15. This esul , howe e , did no su i e he FDR co ec ion, p(FDR co ec ed) = .067. Rega ding he N1 la ency, he e was no a signi ican main e ec o condi ion, F(1, 26) = 2.45, p = .13, p(FDR co ec ed) = .81, ŋ2p = .09. The ANOVAs ca ied ou on he P1 ampli ude and N1 la ency also e ealed he ollowing esul s. A signi ican main e ec o elec ode on P1 ampli ude, F(2.03, 52.66) = 3.73, p = .03, ɛ = .338, ŋ2p = .13, indica ed ha i was maximal o e elec odes P5/6 and PO7/8, bu no o he main e ec s o in e ac ions eached signi icance. The analysis o N1 la ency showed ha i peaked Pos -p in ( inal d a pos - e e eeing) 20 ea lie o e he mo e pos e io si es (O1/2, PO3/4; main e ec o elec ode, F(3.09, 80.27) = 10.77, p < .001, ɛ = .515, ŋ2p = .29), and o e con ala e al (149 ± 2 msec) e sus ipsila e al (151 ± 2 msec) si es (main e ec o hemisphe e: F(1, 26) = 4.72, p = .039, ŋ2p = .15). The e also was a signi ican hemisphe e x condi ion in e ac ion on N1 la ency, F(1, 26) = 6.31, p = .019, ŋ2p = .195, which showed he la encies o be ea lies o a ge s p eceded by a oidance-associa ed memo y cues (149 ± 2 msec) han o hose p eceded by sa e memo y cues (153 ± 3 msec) o e he ipsila e al hemisphe e (p = .030). The pos -hoc compa isons also e ealed he e ec o hemisphe e being signi ican o sa e ials (con ala e al s. ipsila e al: 150 ± 3 s. 153 ± 3 msec; p = 0.009) bu no o a oidance ials (149 ± 2 s. 149 ± 2 msec; p = 0.708). The analysis o P1 la ency showed ea lie la encies a pos e io elec odes (O1/2, PO3/4; main e ec o elec ode: F(3.13, 81.29) = 10.28, p < .001, ɛ = .521, ŋ2p = .28), bu no signi ican e ec s o hemisphe e o condi ion; only subsidia y ANOVAs on a signi ican 3-way in e ac ion be ween elec ode, hemisphe e and condi ion, F(4, 104) = 2.48, p = .049, ɛ = .667, ŋ2p = .09, e ealed a signi ican hemisphe e x condi ion in e ac ion a PO9/10 elec odes, F(1, 26) = 5.43, p = .028, ŋ2p = .17, showing a end o ea lie P1 la encies in a oidance (95 ± 2 msec) e sus sa e (100 ± 3 msec) ials o e ipsila e al si es (p = 0.053). The analysis o N1 ampli ude showed ha i was la ge a pos e io elec odes (O1/2, PO3/4; main e ec o elec ode, F(3.33, 86.59) = 12.83, p < .001, ɛ = .555, ŋ2p = .33), bu nei he a main e ec no an in e ac ion in ol ing he ac o condi ion we e ound, indica ing ha he ampli ude o he N1 componen was una ec ed by ype o memo y cue. Pos -p in ( inal d a pos - e e eeing) 21 ___________________________________________ Inse Figu e 3 a ound he e ___________________________________________ 3.3.2. Ta ge selec ion (N2pc) The N2pc po en ial was elici ed by a ge -p esen scenes a pa ie o-occipi al elec odes con ala e al o he side o he a ge (Figu e 4). The eliabili y o he N2pc was con i med by a main e ec o hemisphe e, F(1,26) = 32.28, p < .001, ŋ2p = .55, on mean ampli udes 190-270 msec a e a ge onse . I as p edic ed, based on ou p e ious esea ch, N2pc ampli ude is modula ed by he mo i a ional alue o he p eceding cue, his should esul in an in e ac ion be ween hemisphe e and condi ion. The p edic ed wo-way in e ac ion was indeed signi ican , e en a e FDR co ec ion, F(1,26) = 6.23, p = .019, p(FDR co ec ed) = .048, ŋ2p = .19, indica ing ha , as expec ed, he ampli ude o N2pc became a enua ed by punishmen a oidance associa ed-memo y cues. Pos - hoc analysis indica ed ha a signi ican N2pc was elici ed in sa e ials (p < .001) and, al hough smalle , was also p esen in a oidance ials (p = .001). To con i m ha he N2pc was ela ed o he selec ion o he a ge in i s scene con ex and o ule ou ha i may ha e been d i en simply by he o ien ing o spa ial a en ion, we also es ed o he p esence o he N2pc in a ge -absen ials. In hese cases, he e was no signi ican main e ec o hemisphe e, F(1,26) = .73, p = .40, ŋ2p = .03, on mean ampli udes 190-270 a e scene onse , con i ming ha N2pc was no obse ed when no a ge was p esen . ___________________________________________ Inse Figu e 4 a ound he e ___________________________________________ Pos -p in ( inal d a pos - e e eeing) 22 Visual inspec ion o he wa e o ms also e ealed a la e la e alized e ec , ollowing he N2pc wi h opposi e pola i y (see Figu e 4), which has been also epo ed in ou p e ious s udies examining LTM-guided isual sea ch in na u alis ic scenes (labeled as pos e io con ala e al posi i i y, PCP; Doallo e al., 2013; Pa ai e al., 2012). This e ec was an enhanced posi i i y o e pos e io con ala e al ( ela i e o ipsila e al) scalp loca ions o he a ge side in he la ency window be ween 320 and 380 msec pos s imulus. An ANOVA analyzing ERP mean ampli udes h ough his la ency window o e PO9/10, PO7/8, PO3/4 and O1/2 con i med he p esence o his la e alized e ec (main e ec o hemisphe e: F(1,26) = 14.67, p < .001, ŋ2p = .36), bu i was no di e en ially modula ed by punishmen a oidance-associa ed con ex ual memo ies (hemisphe e x condi ion: F(1,26) = .45, p = .508, ŋ2p = .02). 3.4. Rela ionship be ween beha io al and elec ophysiological measu es Media ion analyses showed ha he P1 ampli ude media ed he e ec o condi ion on RT, ab = -4.227, boo SE = 2.739, CI95% [-10.499, -.102], which was he only indi ec e ec signi ican ly di e en om ze o (see also Supplemen a y Figu e 1). This esul sugges s ha as e esponses in a oidance ela i e o sa e condi ions may occu h ough changes in senso y p ocessing e lec ed by he P1 po en ial (i.e. P1 ampli ude seems o p edic a signi ican p opo ion o he co- a ia ion be ween condi ion and RT). The comple e pa e n o esul s om he media ion analyses is shown in Supplemen a y Tables 1, 2 and 3. 4. Discussion The indings o his expe imen demons a e he ole o pas a oidance o an a e si e ou come in magni ying memo y-d i en a en ional biases upon pe cep ual decisions on Pos -p in ( inal d a pos - e e eeing) 23 ele an objec s embedded in clu e ed na u al scenes. Impo an ly, ou indings e ealed he abili y o one single punishmen -a oidance associa ion in memo y o enhance a en ion and isual sea ch p ocesses in scenes and o modula e ongoing p ocessing in isual co ical a eas. Beha io al esul s showed ha spa ial expec a ions om LTM associa ed o success ul a oidance o punishmen con e ed beha io al bene i s, as e ealed by as e esponses o a ge s placed in punishmen a oidance- e sus sa e- ela ed emembe ed loca ions. These indings eplica e and ex end ou p e ious esul s (Doallo e al., 2013) by showing ha punishmen - ela ed LTM, simila ly o wha is obse ed when posi i e ewa d is in ol ed, allows a en ion o each he a ge loca ion mo e apidly compa ed wi h when memo ies no associa ed o mo i a ional alues guide spa ial o ien ing. The obse ed e ec s in beha io hus indica e ha punishmen a oidance can bias a en ion h ough associa ions in memo y bu in he absence o immedia e po en ially nega i e consequences (i.e. no punishmen was a s ake in he isual sea ch ask). The p esen da a expand upon he g owing beha io al e idence showing ha s imuli associa ed h ough lea ning o punishmen a oidance (Wen u a e al., 2014) o o an ac ual mone a y loss (Bucke & Theeuwes, 2016) can in luence a en ional deploymen and ecei e acili a ed p ocessing (O'B ien & Raymond, 2012). Unlike ou p io s udy manipula ing posi i e ewa ds, we did no ind signi ican e ec s on accu acy. Taking in o accoun ha he e was also a end o highe accu acy in sa e ela i e o a oidance ials o a ge absen scenes, i is ha d o say, om he p esen esul s, i his di e ence may be a ibu able o speci ically manipula e a oidance- ela ed alues, leading o di e en ial beha io al e ec s han when gains a e in ol ed. In his sense, RT e ec s in he absence o accu acy ones, ela ed o punishmen -associa ed s imuli, ha e been epo ed p e iously (Bucke & Theeuwes, 2016). Pos -p in ( inal d a pos - e e eeing) 24 The momen -by-momen eco d o a ge - ela ed ERPs showed ha he beha io al imp o emen d i en by a oidance- ela ed associa ions was accompanied by modula ion o mul iple s ages o isual p ocessing. Simila ly o he pa e n o esul s obse ed in ou p e ious s udy, he ea lies e ec was obse ed on he P1 po en ial, an index o a en ional gain in ex as ia e isual co ex, which showed la ge ampli udes a pa ie o-occipi al si es o a ge s appea ing a punishmen a oidance-associa ed emembe ed loca ions ( ela i e o sa e- ela ed loca ions), in he absence o e ec s on he P1 la ency. These indings p o ide e idence ha a oidance-associa ed memo y- d i en a en ion biases ea ly pe cep ual analysis o ele an objec s embedded in c owded eal-wo ld scenes. Al hough he main e ec o condi ion on P1 ampli ude, which passed a nominal alpha h eshold o .05, did no su i e FDR co ec ion, he inding ha he P1 ampli ude modula ion signi ican ly media ed he ela ionship be ween punishmen - ela ed associa ions in memo y and bene i s in he eac ion ime u he unde sco es he signi icance o his po en ial in he p io i ized p ocessing o s imuli a spa ial loca ions linked o e asion o punishmen in he pas . Ou indings seem o di e om some ecen epo s o ERP modula ions by acqui ed posi i e and nega i e mo i a ional salience. Baye e al. (2017) in es iga ed he in e play be ween ewa d p ospec (i.e. pe o mance-based mone a y incen i es) du ing a cued pa e n disc imina ion ask on he ac i i y o he p ima y and ex as ia e isual co ex. Mo i a ional ele ance con eyed by he cue ( ela ed o bo h ewa d app oach and punishmen a oidance) inc eased he ampli ude o he C1 po en ial ( e lec ing ea ly pe cep ual p ocessing in V1) bu had no e ec s a he P1 le el. On he o he hand, Hamme schmid , Sennhenn-Reulen, and Schach (2017) ound ha neu al aces p e iously associa ed wi h h ee di e en mone a y ou comes (gain, loss o ze o) impac ed di e en ly he ampli ude o he P1 po en ial: he acili a ed senso y p ocessing Pos -p in ( inal d a pos - e e eeing) 25 o s imuli wi h associa ed mo i a ional salience, as e lec ed by la ge P1 ampli udes, was con ined o ewa d- ela ed aces ( ela i e o he neu al, ze o-ou come, ones). When add essing he di e ences be ween ea lie s udies and ou s, i is impo an o ake in o accoun ha indings such as hose om Hamme schmid e al. (2017) indica e a spa ially unspeci ic e ec ( a he han limi ed o speci ic spa ial posi ions) o acqui ed mo i a ional salience. The di e ences be ween he esul s epo ed by Baye e al. (2017) and hose o he cu en s udy could e lec di e en ial e ec s o isual spa ial cues con eying in o ma ion abou pe o mance-based mone a y incen i es and con ex ual memo y cues which acqui ed alue in a p io lea ning expe ience. Ano he con ibu ing ac o o hese di e ences could be he challenging pe cep ual condi ions in ou ask. Ou a ge s imuli we e embedded wi hin an associa ed clu e ed scene, which may a o p io i iza ion e ec s (as e lec ed by highe ampli udes o P1) by which s imuli appea ing a spa ial loca ions associa ed wi h a oidance o an a e si e ou come in pas encoun e s win ep esen a ion a he expense o o he s imuli. Rega ding he N1 po en ial, an elec ophysiological ma ke o disc imina i e p ocesses in a eas o he en al isual s eam (Hop e al., 2002; Luck, 1995; Vogel & Luck, 2000), ou esul s showed a signi ican in e ac ion be ween condi ion and hemisphe e on i s la ency, indica ing ha o ien ing o spa ial a en ion based on loss a oidance end o sho ened he la ency o he N1 componen o e ipsila e al si es. Al hough ou p ima y hypo hesis es ing did no e eal a main e ec o condi ion on N1 la ency, his in e ac ion sugges s ha con ex ual memo ies associa ed o a oidance o punishmen a e also able o speed up he disc imina ion o a ge s appea ing a hese loca ions in subsequen encoun e s, in line wi h ou p io indings (Doallo e al., 2013). This esul , howe e , should be in e p e ed wi h cau ion because i was no subjec ed o mul iple es ing co ec ion. Pos -p in ( inal d a pos - e e eeing) 32 his issue, i is impo an o ake in o accoun ha he es ic ed s a is ical powe , mainly due o he cha ac e is ics o his ype o expe imen al design, makes necessa y o balance he likelihood o Type II e o s. We should no e ha he ecological alidi y achie ed by he use o memo ies o speci ic a ge loca ions in eal scenes comes a a cos . I limi s he numbe o scenes we can use, and p e en s us om epea ing scenes du ing he pe cep ual disc imina ion ask (wi hou possibly comp omising he s a e o he memo ies being in es iga ed). This consequen ly educes he numbe o ials a ailable and limi s he s a is ical powe ha can be achie ed in o he ypes o designs. To deal wi h his limi a ion and balance he likelihood o Type I and Type II e o s, we co ec ed ou p ima y hypo heses o mul iple compa isons, as epo ed h oughou his s udy. In conclusion, he p esen s udy p o ides new e idence ha memo y-dependen spa ial a en ional biases on pe cep ion a e in luenced by punishmen a oidance associa ions, leading o enhanced beha io al pe o mance and modula ion o neu al signa u es o a ge p ocessing. I also ex ends ou unde s anding o he ole o acqui ed mo i a ional alue in selec i ely p io i izing speci ic spa ial loca ions when sea ching o objec s in na u alis ic con ex s. Acknowledgmen s This esea ch was suppo ed by a P ojec G an o S.D. om he Conselle ía de Educación e O denación Uni e si a ia (Xun a de Galicia, Spain) (EM2012-017). S.S-S was suppo ed by a g an o p edoc o al con ac s om he Spanish Minis y o Economy and Compe i i eness (BES-2016-076298). A.C.N. is suppo ed by Wellcome Pos -p in ( inal d a pos - e e eeing) 33 T us Senio In es iga o Awa d (104571/Z/14/Z). We hank Kenia Co ea o assis ance in EEG da a collec ion and p ep ocessing. Da a s a emen We epo how we de e mined ou sample size, all da a exclusions, all inclusion/exclusion c i e ia, whe he inclusion/exclusion c i e ia we e es ablished p io o da a analysis, all manipula ions, and all measu es in he s udy. No pa o he s udy p ocedu es o analyses was p e egis e ed p io o he esea ch being unde aken. Due o policy o he esea ch g oup, public a chi ing o da a and digi al s udy ma e ials will be done a he ime ou ins i u ion makes a ailable an ins i u ional eposi o y o ha pu pose. Un il hen, eade s seeking access o he da a should con ac he co esponding au ho . REFERENCES Adcock, R. A., Thanga el, A., Whi ield-Gab ieli, S., Knu son, B., & Gab ieli, J. D. (2006). Rewa d-mo i a ed lea ning: Mesolimbic ac i a ion p ecedes memo y o ma ion. Neu on, 50(3), 507-517. doi: 10.1016/j.neu on.2006.03.036 Ande son, B. A. (2013). A alue-d i en mechanism o a en ional selec ion. Jou nal o Vision, 13(3), 7-7. doi:10.1167/13.3.7 Ande son, B. A. (2015). Value-d i en a en ional cap u e is modula ed by spa ial con ex . Visual Cogni ion, 23(1-2), 67-81. doi: 10.1080/13506285.2014.956851 Ande son, B. A. (2016). The a en ion habi : how ewa d lea ning shapes a en ional selec ion. Annals o he New Yo k Academy o Sciences, 1369(1), 24-39. doi: 10.1111/nyas.12957 Pos -p in ( inal d a pos - e e eeing) 34 Ande son, B. A., Kuwaba a, H., Wong, D. F., Gean, E. G., Rahmim, A., B ašic, J. R., … Yan is, S. (2016). The ole o dopamine in alue-based a en ional o ien ing. Cu en Biology, 26(4), 550-555. doi: 10.1016/j.cub.2015.12.062 Ande son, B. A., Kuwaba a, H., Wong, D. F., Robe s, J., Rahmim, A., B ašic, J. R., & Cou ney, S. M. (2017). Linking dopamine gic ewa d signals o he de elopmen o a en ional bias: A posi on emission omog aphic s udy. Neu oimage, 157, 27- 33. doi: 10.1016/j.neu oimage.2017.05.062 Ande son, B. A., Lau en , P. A., & Yan is, S. (2011). Value-d i en a en ional cap u e. P oceedings o he Na ional Academy o Sciences, U.S.A., 108, 10367-10371. doi: 10.1073/pnas.1104047108 Ande son, B. A., Lau en , P. A., & Yan is, S. (2014). Value-d i en a en ional p io i y signals in human basal ganglia and isual co ex. B ain Resea ch, 1587, 88-96. doi: 10.1016/j.b ain es.2014.08.062 Ba ba o, L., Peelen, M. V., & Hickey, C. (2017). Valence, no u ili y, unde lies ewa d- d i en p io i iza ion in human ision. Jou nal o Neu oscience, 37(43), 10438- 10450. doi: 10.1523/JNEUROSCI.1128-17.2017 Baye , M., Rossi, V., Vanlessen, N., G ass, A., Schach , A., & Pou ois, G. (2017). Independen e ec s o mo i a ion and spa ial a en ion in he human isual co ex. Social Cogni i e and A ec i e Neu oscience, 12(1), 146-156. doi:10.1093/scan/nsw162 Benjamini, Y., & Hochbe g, Y. (1995). Con olling he alse disco e y a e: a p ac ical and powe ul app oach o mul iple es ing. Jou nal o he Royal S a is ical Socie y Se ies B, 57, 289–300. h p://www.js o .o g/s able/2346101. Bissone e, G. B., Gen y, R. N., Padmala, S., Pessoa, L., & Roesch, M. R. (2014). Impac o appe i i e and a e si e ou comes on b ain esponses: Linking he Pos -p in ( inal d a pos - e e eeing) 35 animal and human li e a u es. F on ie s in Sys ems Neu oscience, 8, 24. doi:10.3389/ nsys.2014.00024 B ignani, D., Lepsien, J., & Nob e, A. C. (2010). Pu ely endogenous cap u e o a en ion by ask-de ining ea u es p oceeds independen ly om spa ial a en ion. Neu oImage, 51(2), 859-866. doi: 10.1016/j.neu oimage.2010.03.029 B ooks, A. M., & Be ns, G. S. (2013). A e si e s imuli and loss in he mesoco icolimbic dopamine sys em. T ends in Cogni i e Sciences, 17(6), 281- 286. doi:10.1016/j. ics.2013.04.001 Bucke , B., & Theeuwes, J. (2016). Appe i i e and a e si e ou come associa ions modula e exogenous cueing. A en ion, Pe cep ion & Psychophysics, 78(7), 2253- 2265. doi:10.3758/s13414-016-1107-6 Chelazzi, L., Eš očino á, J., Calle i, R., Lo Ge o, E., Sani, I., Della Libe a, C., & San and ea, E. (2014). Al e ing spa ial p io i y maps ia ewa d-based lea ning. Jou nal o Neu oscience, 34(25), 8594-8604. doi:10.1523/JNEUROSCI.0277- 14.2014 Chelazzi, L., Pe la o, A., San and ea, E., & Della Libe a, C. (2013). Rewa ds each isual selec i e a en ion. Vision Resea ch, 85, 58-72. doi:10.1016/j. is es.2012.12.005 Della Libe a, C., & Chelazzi, L. (2006). Visual selec i e a en ion and he e ec s o mone a y ewa ds. Psychological Science, 17(3), 222-227. doi: 10.1111/j.1467- 9280.2006.01689.x Della Libe a, C., & Chelazzi, L. (2009). Lea ning o a end and o igno e is a ma e o gains and losses. Psychological Science, 20(6), 778-784. doi: 10.1111/j.1467- 9280.2009.02360.x Pos -p in ( inal d a pos - e e eeing) 36 Doallo, S., Pa ai, E. Z., & Nob e, A. C. (2013). Rewa d associa ions magni y memo y- based biases on pe cep ion. Jou nal o Cogni i e Neu oscience, 25(2), 245-257. doi:10.1162/jocn_a_00314 Donohue, S. E., Hop , J-M., Ba sch, M. V., Schoen eld, M. A., Heinze, H-J., & Woldo , M. G. (2016). The apid cap u e o a en ion by ewa ded objec s. Jou nal o Cogni i e Neu oscience, 28(4), 529-541. doi: 10.1162/jocn_a_00917 Eime , M. (1996). The N2pc componen as an indica o o a en ional selec i i y. Elec oencephalog aphy and Clinical Neu ophysiology, 99(3), 225-234. doi:S0921884X96957112 Engelmann, J. B., Dama aju, E., Padmala, S., & Pessoa, L. (2009). Combined e ec s o a en ion and mo i a ion on isual ask pe o mance: ansien and sus ained mo i a ional e ec s. F on ie s in Human Neu oscience, 3, 4. doi: 10.3389/neu o.09.004.2009 Failing, M., & Theeuwes, J. (2018). Selec ion his o y: How ewa d modula es selec i i y o isual a en ion. Psychonomic Bulle in & Re iew, 25, 514-538. doi: 10.3758/s13423-017-1380-y Ghazizadeh A., G iggs, W., Leopold, D. A., & Hikosaka, O. (2018). Tempo al- p e on al co ical ne wo k o disc imina ion o aluable objec s in long- e m memo y. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica, 115(9), 2135-2144. doi: 10.1073/pnas.1707695115 G iggs, W. S., Ami a, H., Gopal, A., & Hikosaka, O. (2018). Visual neu ons in he supe io colliculus disc imina e many objec s by hei his o ical alues. F on ie s in Neu oscience, 12, 396. doi: 10.3389/ nins.2018.00396 Pos -p in ( inal d a pos - e e eeing) 37 Hamme schmid , W., Sennhenn-Reulen, H., & Schach , A. (2017). Associa ed mo i a ional salience impac s ea ly senso y p ocessing o human aces. Neu oImage, 156, 466-474. doi:S1053-8119(17)30325-7 Ha is, J. A., Donohue, S. E., Schoen eld, M. A., Hop , J-M., Heinze, H-J., & Woldo , M. G. (2016). Rewa d-associa ed ea u es cap u e a en ion in he absence o awa eness: E idence om objec -subs i u ion masking. Neu oImage, 137, 116- 123. doi: 10.1016/j.neu oimage.2016.05.010 Heinze, H. J., Mangun, G. R., Bu che , W., Hin ichs, H., Scholz, M., Mün e, T. F., … Hillya d, S. A. (1994). Combined spa ial and empo al imaging o b ain ac i i y du ing isual selec i e a en ion in humans. Na u e, 372(6506), 543-546. doi: 10.1038/372543a0 Hickey, C., Chelazzi, L., & Theeuwes, J. (2010). Rewa d changes salience in human ision ia he an e io cingula e. Jou nal o Neu oscience, 30(33), 11096-11103. doi: 10.1523/JNEUROSCI.1026-10.2010 Hickey, C., Chelazzi, L., & Theeuwes, J. (2014). Rewa d-p iming o loca ion in isual sea ch. PLoS One, 9(7), e103372. doi: 10.1371/jou nal.pone.0103372 Hickey, C., Di Lollo, V., & McDonald, J. J. (2009). Elec ophysiological indices o a ge and dis ac o p ocessing in isual sea ch. Jou nal o Cogni i e Neu oscience, 21(4), 760-775. doi:10.1162/jocn.2009.21039 Hickey, C., Kaise , D., & Peelen, M. V. (2015). Rewa d guides a en ion o objec ca ego ies in eal-wo ld scenes. Jou nal o Expe imen al Psychology: Gene al, 144, 264-273. doi: 10.1037/a0038627 Hickey, C., & Peelen, M. (2015). Neu al mechanisms o incen i e salience in na u alis ic human ision. Neu on, 85, 512-518. doi: 10.1016/j.neu on.2014.12.049 Pos -p in ( inal d a pos - e e eeing) 38 Hickey, C., & Peelen, M. V. (2017). Rewa d selec i ely modula es he linge ing neu al ep esen a ion o ecen ly a ended objec s in na u al scenes. Jou nal o Neu oscience, 37(31), 7297-7304. doi: 10.1523/JNEUROSCI.0684-17.2017 Hilimi e, M. R., Moun s, J. R., Pa ks, N. A., & Co ballis, P. M. (2010). E en - ela ed po en ials dissocia e e ec s o salience and space in biased compe i ion o isual ep esen a ion. PloS One, 5(9), e12677. doi: 10.1371/jou nal.pone.0012677 Hillya d, S. A., Vogel, E. K., & Luck, S. J. (1998). Senso y gain con ol (ampli ica ion) as a mechanism o selec i e a en ion: elec ophysiological and neu oimaging e idence. Philosophical T ansac ions o he Royal Socie y o London, Se ies B, Biological Sciences, 353(1373), 1257-1270. doi: 10.1098/ s b.1998.0281 Hop , J. M., Luck, S. J., Gi elli, M., Hagne , T., Mangun, G. R., Scheich, H., & Heinze, H. J. (2000). Neu al sou ces o ocused a en ion in isual sea ch. Ce eb al Co ex, 10(12), 1233-1241. Hop , J. M., Vogel, E., Woodman, G., Heinze, H. J., & Luck, S. J. (2002). Localizing isual disc imina ion p ocesses in ime and space. Jou nal o Neu ophysiology, 88(4), 2088-2095. doi:10.1152/jn.2002.88.4.2088 I hipu ipa , S., Cha, K., Rangsipa , N., & Se ences, J. T. (2015). Value-based a en ional cap u e in luences con ex -dependen decision-making. Jou nal o Neu ophysiology, 114(1), 560-569. doi: 10.1152/jn.00343.2015 Je de, T. A., Me iam, E. P., Riggall, A. C., Hedges, J. H., & Cu is, C. E. (2012). P io i ized maps o space in human on opa ie al co ex. Jou nal o Neu oscience, 32(48), 17382-17390. doi: 10.1523/JNEUROSCI.3810-12.2012 Kahn , T., Pa k, S. Q., Haynes, J. D., & Toble , P. N. (2014). Disen angling neu al ep esen a ions o alue and salience in he human b ain. P oceedings o he Pos -p in ( inal d a pos - e e eeing) 39 Na ional Academy o Sciences o he Uni ed S a es o Ame ica, 111(13), 5000- 5005. doi: 10.1073/pnas.1320189111 Kim, H. F., Ghazizadeh, A., & Hikosaka, O. (2015). Dopamine neu ons encoding long- e m memo y o objec alue o habi ual beha io . Cell, 163(5), 1165-1175. doi: 10.1016/j.cell.2015.10.063 Kiss, M., D i e , J., & Eime , M. (2009). Rewa d p io i y o isual a ge single ons modula es ERP signa u es o a en ional selec ion. Psychological Science, 20(2), 245-251. doi: 10.1111/j.1467-9280.2009.02281.x Kuo, B. C., Rao, A., Lepsien, J., & Nob e, A. C. (2009). Sea ching o a ge s wi hin he spa ial layou o isual sho - e m memo y. Jou nal o Neu oscience, 29(25), 8032-8038. doi:10.1523/JNEUROSCI.0952-09.2009 LaBa , K. S., & Cabeza, R. (2006). Cogni i e neu oscience o emo ional memo y. Na u e Re iews Neu oscience, 7(1), 54-64. doi: 10.1038/n n1825 Lim, S. L., Padmala, S., & Pessoa, L. (2009). Seg ega ing he signi ican om he mundane on a momen - o-momen basis ia di ec and indi ec amygdala con ibu ions. P oceedings o he Na ional Academy o Sciences o he Uni ed S a es o Ame ica, 106(39), 16841-16846. doi:10.1073/pnas.0904551106 Liu, X., Hai s on, J., Sch ie , M., & Fan, J. (2011). Common and dis inc ne wo ks unde lying ewa d alence and p ocessing s ages: A me a-analysis o unc ional neu oimaging s udies. Neu oscience & Biobeha io al Re iews, 35(5), 1219-1236. doi:10.1016/j.neubio e .2010.12.012 Luck, S. J. (1995). Mul iple mechanisms o isual-spa ial a en ion: Recen e idence om human elec ophysiology. Beha iou al B ain Resea ch, 71(1-2), 113-123. Pos -p in ( inal d a pos - e e eeing) 40 Luck, S. J., & Hillya d, S. A. (1994). Spa ial il e ing du ing isual sea ch: E idence om human elec ophysiology. Jou nal o Expe imen al Psychology. Human Pe cep ion and Pe o mance, 20(5), 1000-1014. MacLean, M. H., & Giesb ech , B. (2015). Neu al e idence e eals he apid e ec s o ewa d his o y on selec i e a en ion. B ain Resea ch, 1606, 86-94. doi: 10.1016/j.b ain es.2015.02.016 Ma ínez, A., Anllo-Ven o, L., Se eno, M. I., F ank, L. R., Bux on, R. B., Dubowi z, D. J., … Hillya d, S. A. (1999). In ol emen o s ia e and ex as ia e isual co ical a eas in spa ial a en ion. Na u e Neu oscience, 2(4), 364-369. doi: 10.1038/7274 Mohan y, A., Gi elman, D. R., Small, D. M., & Mesulam, M. M. (2008). The spa ial a en ion ne wo k in e ac s wi h limbic and monoamine gic sys ems o modula e mo i a ion-induced a en ion shi s. Ce eb al Co ex, 18(11), 2604-2613. doi: 10.1093/ce co /bhn021 Mon oya, A. K., & Hayes, A. F. (2017). Two-condi ion wi hin-pa icipan s a is ical media ion analysis: A pa h-analy ic amewo k. Psychological Me hods, 22(1), 6- 27. doi: 10.1037/me 0000086 Mu y, V. P., LaBa , K. S., & Adcock, R. A. (2012). Th ea o punishmen mo i a es memo y encoding ia amygdala, no midb ain, in e ac ions wi h he medial empo al lobe. Jou nal o Neu oscience, 32(26), 8969-8976. doi: 10.1523/JNEUROSCI.0094-12.2012 Mu y, V. P., Ri chey, M., Adcock, R. A., & LaBa , K. S. (2010). MRI s udies o success ul emo ional memo y encoding: A quan i a i e me a-analysis. Neu opsychologia, 48(12), 3459-3469. doi: 10.1016/ j.neu opsychologia.2010.07.030 Pos -p in ( inal d a pos - e e eeing) 41 O'B ien, J. L., & Raymond, J. E. (2012). Lea ned p edic i eness speeds isual p ocessing. Psychological Science, 23(4), 359-363. doi:10.1177/0956797611429800 Padmala, S., & Pessoa, L. (2008). A ec i e lea ning enhances isual de ec ion and esponses in p ima y isual co ex. Jou nal o Neu oscience, 28(24), 6202-6210. doi:10.1523/JNEUROSCI.1233-08.2008 Pa ai, E. Z., Doallo, S., & Nob e, A. C. (2012). Long- e m memo ies bias sensi i i y and a ge selec ion in complex scenes. Jou nal o Cogni i e Neu oscience, 24(12), 2281-2291. doi:10.1162/jocn_a_00294 Pessoa, L. (2015). Mul iple in luences o ewa d on pe cep ion and a en ion. Visual Cogni ion, 23(1-2), 272-290. doi:10.1080/13506285.2014.974729 Pollmann, S., Es ocino a, J., Somme , S., Chelazzi, L., & Zinke, W. (2016). Neu al s uc u es in ol ed in isual sea ch guidance by ewa d-enhanced con ex ual cueing o he a ge loca ion. Neu oImage, 124(P A), 887-897. doi:10.1016/j.neu oimage.2015.09.040 P ak, R. (2012). The on opa ie al a en ional ne wo k o he human b ain: ac ion, saliency, and a p io i y map o he en i onmen . The Neu oscien is , 18(5), 502- 515. doi: 10.1177/1073858411409051 Qi, S., Zeng, Q., Ding, C., & Li, H. (2013). Neu al co ela es o ewa d-d i en a en ional cap u e in isual sea ch. B ain Resea ch, 1532, 32-43. doi: 10.1016/j.b ain es.2013.07.044 Raymond, J. E., & O'B ien, J. L. (2009). Selec i e isual a en ion and mo i a ion: The consequences o alue lea ning in an a en ional blink ask. Psychological Science, 20(8), 981-988. doi:10.1111/j.1467-9280.2009.02391.x Pos -p in ( inal d a pos - e e eeing) Figu e1 Click he e o download high esolu ion image Pos -p in ( inal d a pos - e e eeing) Figu e 2 Click he e o download high esolu ion image Pos -p in ( inal d a pos - e e eeing) Figu e 3 Click he e o download high esolu ion image Pos -p in ( inal d a pos - e e eeing) Figu e 4 Click he e o download high esolu ion image Pos -p in ( inal d a pos - e e eeing) Suppl. ma e ial o online publica ion only Click he e o download Suppl. ma e ial o online publica ion only: Supplemen a y da a.docx Pos -p in ( inal d a pos - e e eeing)