Memory performance as a function of modality and context: revealing the cognitive mechanisms underlying enhanced memory in virtual reality

ORIGINAL ARTICLE Vi ual Reali y (2025) 29:128 h ps://doi.o g/10.1007/s10055-025-01205-1 labo a o y, VR en i onmen s a e highly imme si e and ealis ic, hence elici ing mo e li elike mnemonic p ocess- ing (Kiske e al. 2021b; Schöne e al. 2019). In e es ingly, se e al s udies p o ide e idence o a memo y supe io i y e ec in VR, i.e., enhanced memo y pe o mance o VR expe iences as opposed o wo-dimensional p esen a ions commonly used in con en ional labo a o y se ups. This e ec has been ound o wo king memo y pe o mance (Gabana e al. 2017), memo y ecall pe o mance (K okos e al. 2019), and ecogni ion memo y pe o mance (Schöne e al. 2019). P e ious esea ch a ibu ed he enhanced memo y pe o mance o he in ol emen o he au obio- g aphical memo y sys em (Schöne e al. 2019) cha ac e - ized by pe sonal ele ance and he in eg a ion o newly made expe iences in o he na a i e o an indi idual’s li e expe ience (Conway 2005; Schöne e al. 2018). The i id 1 In oduc ion The applica ion o Vi ual Reali y (VR) as a psychological esea ch me hod o in es iga e eal-li e cogni i e p ocessing has led o a e-e alua ion o mnemonic p ocesses p e iously in es iga ed in he con en ional labo a o y (e.g., Co i eau Leca alie e al. 2020; Kiske e al. 2021c; Rod íguez e al. 2022). In con as o expe iences in he con en ional Ma ike Johnsdo [email p o ec ed] 1 Expe imen al Psychology I, Ins i u e o Psychology, Osnab ück Uni e si y, Osnab ück, Ge many 2 Depa men o Psychology, No wegian Uni e si y o Science and Technology, T ondheim, No way Abs ac Al hough ecen psychological esea ch p o ides e idence o enhanced memo y e ie al o Vi ual Reali y (VR) expe i- ences compa ed o con en ional labo a o y expe iences, he unc ional cha ac e is ics o he unde lying mnemonic p o- cesses emain unclea . Ini ial s udies sugges ha encoding and e ie al in ol e dis inc cogni i e p ocesses, ye he possible echnological a i ac s induced by VR head-moun ed displays (VR-HMDs) ha e no been accoun ed o . To de e mine he ac o s in luencing modali y-speci ic memo y p ocessing, 122 pa icipan s pe o med an old/new ecogni ion memo y ask wea ing a VR-HMD. S imuli we e i s ei he p esen ed wo-dimensionally on a sc een wi hin a VR en i on- men (media ed VR; mVR) o expe ienced immedia ely in a h ee-dimensional imme si e VR en i onmen (iVR). In he subsequen old/new ecogni ion memo y ask, pa icipan s had o ecall he s imuli ei he in he same o he o he modali y, p o iding a c oss-modali y compa ison. As a esul , he canonical con ex e ec , i.e., an ad an age o memo y pe o - mance in cong uen con ex s, was ound o bo h mVR and iVR modali ies. Mos impo an ly, he s udy u he p o ides e idence o a memo y supe io i y e ec ollowing iVR expe iences, as an equally high e ie al success was obse ed e en when he subsequen ecall was es ed in mVR. In conclusion, he imp o ed ecall obse ed sugges s ha imme si e en i onmen s play a c ucial ole in enhancing he encoding p ocess, whe eas echnological a i ac s, pa icula ly con ex ual ac o s o he VR-HMD, can be uled ou . Fu he mo e, he ans e abili y o in o ma ion om wo-dimensional o h ee- dimensional en i onmen s seems limi ed. Ou s udy hus ad oca es o in eg a ing imme si e VR in o lea ning con ex s o ex end he applicabili y o lea ned con en o a ious se ings. Keywo ds Vi ual eali y · Implici lea ning · Memo y · Encoding · Lea ning con ex Recei ed: 27 June 2024 / Accep ed: 15 July 2025 / Published online: 7 Augus 2025 © The Au ho (s) 2025 Memo y pe o mance as a unc ion o modali y and con ex : e ealing he cogni i e mechanisms unde lying enhanced memo y in i ual eali y Ma ikeJohnsdo 1· JoannaKiske 1· Me leSageho n1· ThomasG ube 1· BenjaminSchöne1,2 1 3 Vi ual Reali y (2025) 29:128 and expe ien ial na u e o VR closely mi o s eal-li e in e - ac ions, he eby embedding memo ies mo e deeply wi hin he au obiog aphical memo y amewo k (see Schöne e al. 2019). Addi ionally, in o ma ion which is encoded in a VR en i onmen and acco dingly s o ed wi hin au obio- g aphical memo y aces, is mo e e o lessly e ie able han memo ies o med unde con en ional labo a o y con- di ions, ega dless o memo y pe o mance (Kiske e al. 2021c). Recen elec ophysiological indings sugges ha hese quali a i ely dis inc mnemonic p ocesses in VR en i- onmen s (Kiske e al. 2021c) al eady occu du ing ea ly p ocessing s ages as e idenced by ex ensi e encoding p o- cesses in ol ing mul imodal neu al in eg a ion (Johnsdo e al. 2023a, b). The e o e, VR p o ides a unique app oach o in es iga e mnemonic p ocesses in ol ing au obiog aph- ical memo y unde mo e ealis ic condi ions. Howe e , se e al ac o s po en ially con ibu ing o he memo y supe io i y e ec in VR emain o be e alua ed. In addi ion o he a o emen ioned di e ences obse ed be ween con en ional labo a o y and i ual condi ions, and con adic o y o he memo y supe io i y e ec in VR, se e al s udies ac ually indica e equal, o poo e memo y pe o mance in VR in compa ison o con en ional 2D expe- iences depending on he e ie al ask (e.g., Cade and Chainay 2020; Johnsdo e al. 2023a, b; Kiske e al. 2021c; Mona o e al. 2024; Schöne e al. 2023). In pa icula , he ee ecall o s imuli encoded in a i ual en i onmen ei he p esen ed in VR o on a con en ional 2D sc een did no esul in di e ences in memo y pe o mance (Cade and Chainay 2020). In a u he s udy, pa icipan s pe o med poo e in an applica ion ask o p e iously lea ned ules a e lea ning ia a 3D/360° ideo compa ed o a 2D ideo. Mo eo e , no di e ences in memo y pe o mance we e ob ained be ween bo h g oups in ei he a mul iple-choice es o in a ee ecall es in he same s udy (Johnsdo e al. 2023a, b). Ano he s udy ha used a ecogni ion memo y ask in ol ing a emembe /know pa adigm (Tul ing 1985; Ga dine and Ja a 1990; Raja am 1993; see Kiske e al. 2021b) ound ha explo ing a i ual illage ei he in VR o ia a 2D sc een did no esul in di e en memo y pe - o mance (Kiske e al. 2021b). Compa ing memo y pe - o mance o objec s encoded in a i ual oom o ia 2D pic u es o he oom e ealed no di e ences be ween he wo condi ions in a ee ecall ask, a isual ecogni ion ask, and a sugges ibili y ask wi h he excep ion o a non- sugges i e e bal ask in which memo y pe o mance was be e in VR (Mona o e al. 2024). A i h s udy compa ed memo y pe o mance o ideos o di e en scenes wa ched ei he in VR o in 2D on a moni o . Al hough he VR g oup pe o med be e in a ee ecall ask, he wo g oups did no show di e ences in memo y pe o mance in a cued ecall ask (Schöne e al. 2023). O e all, he a ying memo y asks do no necessa ily p o ide he same conclusions abou he di e ences in memo y pe o mance be ween VR and 2D, al hough no s ingen pa e n eme ges ega ding memo y pe o mance as a unc ion o he e ie al ask. Impo an ly, i emains unclea whe he he a ying modali y-speci ic mnemonic p ocessing cha ac e is ics a e a ibu able o he ac ual exposu e o h ee-dimensional imme si e expe iences in VR, o whe he hey a e only in luenced by possible a i ac s induced by wea ing a VR head-moun ed display (VR-HMD). Since mos s udies assessed memo y pe o mance a e conduc ing he encod- ing phase in VR ia an HMD and in 2D ia a con en ional moni o (Johnsdo e al. 2023a, b; Kiske e al. 2021b, 2021c; Schöne e al. 2019), he di e ing ou comes in mne- monic p ocessing migh a ise om he use o he VR-HMD in only one o he compa ed condi ions, c ea ing he alse imp ession ha memo y p ocesses in VR di e ma kedly om hose unde 2D condi ions. On he one hand, he no - el y o he echnology and he associa ed high cogni i e load can ha e a nega i e impac on lea ning (Miguel-Alonso e al. 2023; Wu e al. 2013). On he o he hand, he h ill o echnological awe o wea ing a VR-HMD can also imp o e mo i a ion and subjec i e usabili y (Koch e al. 2018) as well as memo y pe o mance by inc easing a en ion and e o (Huang e al. 2021). The e o e, he no el y o wea ing a VR-HMD migh be a con ibu ing ac o o he memo y supe io i y e ec in VR, and pa icula ly o he di e ging esul s in mnemonic esea ch compa ing imme si e o con- en ional condi ions. In e es ingly, apa om mnemonic p ocessing, mo i a ional and emo ional eac ions s ill exhibi di e en unc ional cha ac e is ics on a neu al le el when con olling o a po en ial in luence o he VR-HMD by p esen ing bo h 3D/360° s imuli and 2D s imuli ia an HMD (Schöne e al. 2023). These indings indica e ha he h ee-dimensionali y o s imuli is a ele an ac o in luenc- ing a ec i e eac ions in VR. Fu he p ope ies in insic o VR migh accoun o he he e ogenei y in memo y pe o mance ou comes. The disc epancies migh be a ibu able o me hodological di - e ences ac oss s udies, and hus esul om me e ech- nological a i ac s like he echnical ea u es o he VR sys ems a hand along wi h a ying deg ees o in e ac i - i y and p esence a o ded by he VR en i onmen s ( o e iew see Smi h 2019). These speci ic echnological p op- e ies o a VR sys em gene a e di e en le els o imme - sion, i.e., he capabili y o a VR echnology o ep esen a ealis ic en i onmen (Sla e and Wilbu 1997). Gene ally, a highe deg ee o imme sion inc eases he sense o p es- ence (Dieme e al. 2015), i.e., he sense o ac ually being in an in e ac i e and esponsi e en i onmen (Sanchez-Vi es and Sla e 2005). Al hough a di ec in luence o p esence on memo y pe o mance could no always be obse ed (e.g., 1 3 128 Page 2 o 12 Vi ual Reali y (2025) 29:128 Smi h and Mulligan 2021), o he s udies ha e sugges ed p esence o be one o he mos p o ound ea u es o VR ha has been shown o be posi i ely ela ed o memo y pe o - mance (Cade e al. 2022; Makowski e al. 2017). Taking in o accoun he po en ial echnological a i ac s esul ing om he use o he VR-HMD, his s udy aimed o un a el modali y-speci ic cogni i e mechanisms in bo h imme si e VR and con en ional 2D se ings e lec ed in enhanced e ie al success. The use o a VR-HMD in all condi ions ensu es ha di e ences in memo y pe o mance can be a ibu ed o speci ic unc ional cha ac e is ics o mnemonic p ocessing a he han me e echnological ea- u es. Memo y pe o mance was assessed by ansla ing a con en ional old/new ecogni ion pa adigm o bo h a medi- a ed VR en i onmen (mVR), p esen ing 2D s imuli on i - ual moni o wi hin VR, and an imme si e VR en i onmen (iVR), p esen ing 3D/360° s imuli in VR. The subsequen old/new ecogni ion memo y ask was ei he pe o med in he same o he o he modali y, p o iding a c oss-modali y compa ison. Classical indings om p e ious memo y esea ch showed ha memo y pe o mance bene i s om con ex e ec s, i.e., om a cong uen en i onmen du ing encod- ing and e ie al (Godden and Baddeley 1975). Since he s imuli in he mVR en i onmen we e p esen ed wo- dimensionally, i.e., as in he con en ional labo a o y, we expec ed o obse e high memo y pe o mance sco es when pa icipan s encoded and ecalled he s imuli in he same mVR en i onmen . The classical con ex e ec could be eplica ed wi hin di e en VR en i onmen s (Shin e al. 2021), hence we expec ed simila ly high memo y pe o - mance sco es o cong uen iVR con ex s du ing encod- ing and e ie al. Assuming ha he memo y supe io i y e ec in VR obse ed in p e ious esea ch was only caused by echnological a i ac s o he VR-HMD, only con ex e ec s should in luence memo y pe o mance in he p es- en s udy, and high memo y pe o mance sco es should only occu when modali ies a e cong uen du ing encoding and e ie al. Howe e , gi en he e idence o a memo y supe io i y e ec in VR when expe iences become pa o an au obiog aphical memo y ne wo k (Schöne e al. 2019), we also expec ed high memo y pe o mance sco es o pa - icipan s encoding in iVR and ecalling in mVR. This esul would indica e ha imme si e VR does no only se e as a me e con ex ual ac o bu ha i enhances encoding p o- cesses beyond wea ing a VR-HMD. Las ly, his e ec was no expec ed o pa icipan s encoding in mVR and ecall- ing in iVR as memo y pe o mance should be educed by he change in con ex , wi hou he bene icial e ec s o he h ee-dimensional p esen a ion o he s imulus ma e ial du - ing encoding. Addi ionally, due o he inconsis en indings ega d- ing he impac o p esence on memo y pe o mance (see Cade e al. 2022; Makowski e al. 2017; Smi h and Mul- ligan 2021), we assessed p esence bo h a e encoding and e ie al o u he in es iga e i s ela ion o memo y pe o - mance when con olling o he in luence o he VR-HMD. Since a high le el o imme sion inc eases he sense o p es- ence ( o e iew see Dieme e al. 2015), we expec ed o obse e a highe sense o p esence in he iVR modali y compa ed o he mVR modali y bo h du ing encoding and e ie al. 2 Me hods 2.1 Pa icipan s 139 pa icipan s we e ec ui ed ia he Uni e si y’s s u- den s e-mail lis , he Uni e si y’s online bulle in boa d and by s uden s o he bachelo ’s deg ee p og am in psy- chology. Pa icipan s we e sc eened o psychological and neu ological diso de s, subs ance abuse and symp oms o a SARS-CoV-2 in ec ion. All had no mal o co ec ed o no mal ision. Eigh pa icipan s had o be excluded due o un ul illed inclusion c i e ia, h ee da a acquisi ions had o be e mina ed because o echnical issues, wo pa ici- pan s p ema u ely e mina ed he expe imen and ou pa - icipan s did no show up o he appoin men . The e o e, da a o 122 pa icipan s we e included in he analyses (29 male, 93 emale, MAge = 22.32, SDAge = 3.610; nmedia ed = 30, nc oss−up = 32, nc oss−down = 29, nimme si e = 31). Pa icipan s ecei ed pa ial cou se c edi s o 15€ o hei pa icipa ion. The s udy was conduc ed in acco dance wi h he hygiene guidelines o he Uni e si y o minimizing he sp ead o SARS-CoV-2 du ing he COVID pandemic. The s udy was app o ed by he local e hics commi ee and all pa icipan s ga e in o med w i en consen . 2.2 S imuli & expe imen al condi ions 120 3D/360° ideos ilmed wi h a esolu ion o 4 K a s an- da d eye le el we e selec ed om he lib a y o uni e sal Vi ual Reali y expe imen s (luVRe; Schöne e al. 2023). The ideos showed a a ie y o scenes, including o exam- ple di e en loca ions, animals o human in e ac ions ( o a comp ehensi e lis o he used scenes om he da abase see Supplemen a y Ma e ial S1). All scenes we e eco ded wi h sound. Each ideo was cu o en seconds and was ecoded o 2D o he p esen a ion on he i ual moni o wi hin a h ee-dimensional VR en i onmen in he mVR modali y. Du ing he encoding session, he ideos we e ei he p e- sen ed in mVR on in iVR, and he modali y in he e ie al 1 3 Page 3 o 12 128 Vi ual Reali y (2025) 29:128 2.3 P ocedu e The s udy had o be pa ly adap ed o he es ic ions due o he COVID pandemic. The e o e, he anamneses we e ini ially conduc ed online p io o he expe imen o mini- mize physical con ac , whe eas hey we e scheduled di ec ly be o e he expe imen a a la e s age o he da a acquisi ion. The expe imen was conduc ed on wo consecu i e days. On he i s day, pa icipan s we e andomly assigned o one o he ou condi ions and we e ins uc ed o pu on and adjus he VR-HMD co ec ly. They we e equipped wi h wo con olle s and headphones ha we e used o p o ide he p e iously eco ded ins uc ion. The encoding session was conduc ed ei he in mVR o in iVR (see Fig. 1a). Pa - icipan s in he mVR modali y we e loca ed in on o a 2D moni o on a able in a i ual h ee-dimensional labo a o y, while pa icipan s in he iVR condi ion we e loca ed in a neu al g ey oom. All pa icipan s we e ins uc ed ha hey would be p esen ed wi h ideos o a ious scenes, which hey we e asked o wa ch ca e ully. A e wa ds, hey could session was ei he consis en wi h he encoding session (cong uen con ex ) o was ca ied ou in he al e na i e modali y (incong uen con ex ). In he mVR modali y, he ideos we e p esen ed in 2D on he i ual moni o , while hey we e p esen ed in 3D/360° in he iVR modali y, i.e., he pa icipan s we e su ounded by he ideos and wa ched hem om a i s -pe son pe spec i e (see Fig. 1). This design esul ed in ou condi ions p o iding a c oss-modali y com- pa ison: Encoding in mVR and e ie al in mVR (media ed), encoding in mVR and e ie al in iVR (c oss-up), encoding in iVR and e ie al in mVR (c oss-down), and encoding in iVR and e ie al in iVR (imme si e). Bo h he mVR and iVR en i onmen s we e buil using Uni y ( e sion 2020, Uni y Technologies, San F ancisco, Uni ed S a es). They we e p esen ed using a s andalone HTC Vi e Focus Plus HMD (wi h a esolu ion o 1440 × 1600 pixels pe eye). Addi ionally, pa icipan s we e equipped wi h wo Vi e Focus Plus con olle s and Sony o e -ea headphones p o iding s e eoscopic audio. Fig.1 Se up o he encoding phase (a) and he e ie al phase (b) in media ed VR (mVR) and imme si e VR (iVR). The c oss-modali y compa ison esul ed in ou condi ions: Encoding in mVR and e ie al in mVR (media ed), encoding in mVR and e ie al in iVR (c oss-up), encoding in iVR and e ie al in mVR (c oss-down), and encoding in iVR and e ie al in iVR (imme si e) 1 3 128 Page 4 o 12 Vi ual Reali y (2025) 29:128 sepa a ely o each symp om (headache, nausea, and diz- ziness) and each expe imen al phase (encoding, e ie al). 2.4.2 Memo y pe o mance The sensi i i y index d-P ime (d’) was used o in es iga e memo y pe o mance in each condi ion du ing he e ie al phase. I is calcula ed by ela ing he hi s, i.e., he co ec ly ecognized old s imuli o he alse posi i es, i.e., he new s imuli inco ec ly a ed as old (d’ = zHi –zFalsePosi i e; Haa - ei e al. 2010; Swe s e al. 1961). To compa e memo y pe o mance be ween condi ions du ing he e ie al session on he second day o he expe imen , a 2 × 2 ANOVA wi h he ac o Encoding Modali y (mVR, iVR) and Re ie al Modali y (mVR, iVR) esul ing in he condi ion combi- na ions (mVR + mVR = media ed, mVR + iVR = c oss-up, iVR + mVR = c oss-down, iVR + iVR = imme si e) was pe - o med ollowed by pos -hoc - es s o independen sam- ples wi h a Bon e oni-co ec ed signi icance le el. SPSS (Ve sion 28) was used o s a is ical analyses. To u he cla i y whe he non-signi ican g oup di e - ences ega ding memo y pe o mance (d’) would indica e he absence o a meaning ul e ec , he wo one-sided - es s p ocedu e (TOST) was pe o med using he TOSTER pack- age (Caldwell 2022; Lakens 2017) in RS udio (Ve sion 2023.3.1.446). In compa ison o non-signi ican esul s o con en ional - es s, his equi alence es allows o s a is i- cally ejec ing he exis ence o e ec s ha a e la ge enough o be conside ed meaning ul, and he e o e o assume he absence o an e ec . Two s a is ically signi ican esul s in he TOST p ocedu e sugges s a is ical equi alence be ween he compa ed g oups. In his case, he es wi h he smalle es s a is ic is epo ed. The uppe and lowe equi alence bounds ha need o be speci ied o he TOST a e based on he smalles e ec size o in e es (SESOI). The SESOI used in ou analyses was de e mined based on p e ious esea ch ega ding memo y p ocessing in VR. Se e al VR s udies ha e epo ed d’ as an indica o o memo y pe o mance in di e en pa adigms (Kiske e al. 2021b, c; Schöne e al. 2019, 2023). The smalles e ec size o a signi ican e ec ega ding memo y pe o mance documen ed in hese s ud- ies was Cohen’s d = 0.5 (Schöne e al. 2023) which co e- sponds o a medium e ec (Cohen 1988). The e o e, based on he smalles e ec size obse ed o da e in simila s ud- ies, he uppe and lowe bounds we e se o Cohen’s d = 0.5 and Cohen’s d = − 0.5. 2.4.3 P esence All p esence da a we e analyzed using SPSS (Ve sion 28). To analyze p esence, mean alues o all IPQ subscales on bo h s a he expe imen by p essing a bu on in he i ual en i- onmen . Du ing he encoding phase, a o al o 60 ideos we e andomly selec ed om all 120 ideos and p esen ed in andom o de . The e o e, each o he 120 ideos could be andomly p esen ed in all ou condi ions, wi h no ideo being p esen ed epea edly. A e he encoding phase, pa icipan s illed in he Ge - man e sion o he iG oup P esence Ques ionnai e (IPQ; Schube e al. 2001). The IPQ assesses he sense o p es- ence and consis s o he subscales Gene al P esence (one i em), Spa ial P esence ( i e i ems), In ol emen ( ou i ems) and Realness ( ou i ems). A Like -Scale om one o se en was used o he IPQ a ings. Addi ionally, pa - icipan s we e asked o symp oms o mo ion sickness, i.e., headache, nausea, and dizziness. Du ing he e ie al phase, pa icipan s had o pe o m an unannounced old/new ecogni ion ask exac ly 24 h a e he encoding phase. Again, hey we e ins uc ed o pu on and adjus he VR-HMD co ec ly. Depending on he condi ion hey we e assigned o, pa icipan s expe ienced he e ie al phase in he mVR o he iVR modali y (see Fig. 1b). In bo h modali ies, he p e iously eco ded ins uc- ion was p esen ed h ough he headphones. Du ing he old/ new ecogni ion ask, he pa icipan s we e p esen ed wi h all 120 ideos in andom o de , esul ing in 60 old and 60 new ideos. E e y ideo was played a leas o one sec- ond and a maximum o en seconds, and none o he id- eos was epea edly p esen ed. The pa icipan s’ ask was o indica e whe he o no hey emembe ed he scene om he encoding session by p essing he co esponding bu on as as as possible (see Fig. 1b). Immedia ely a e he bu - on p ess, he cu en ideo was skipped and he nex ideo was p esen ed, esul ing in a ying p esen a ion imes o he espec i e ideos du ing e ie al. When pa icipan s did no answe wi hin en seconds, he ideo was paused un il hey ga e an answe . Following he old/new ecogni ion ask, pa icipan s again had o ill in he Ge man e sion o he IPQ and we e asked o symp oms o mo ion sickness, i.e., headache, nausea, and dizziness. 2.4 Da a analysis 2.4.1 Di e ences be ween condi ions To accoun o po en ial he e ogenei y be ween pa icipan s in he ou expe imen al condi ions, we analyzed whe he age and gende di e ed be ween condi ions. Age was in es- iga ed by means o a one-way ANOVA wi h he ac o Condi ion (media ed, c oss-up, c oss-down, imme si e), whe eas gende was analyzed using a χ2- es . Addi ion- ally, po en ial di e ences in symp oms o mo ion sickness be ween condi ions we e in es iga ed by means o χ2- es s 1 3 Page 5 o 12 128 Vi ual Reali y (2025) 29:128 condi ions (χ2(3) = 0.32, p =.955, V = 0.052). Addi ionally, he equency o symp oms o mo ion sickness did no di - e be ween condi ions (all χ2s ≤ 2.01, all ps ≥ 0.571, see Supplemen a y Ma e ial Table S1), excep o he headache symp om du ing e ie al (χ2(3) = 9.35, p =.025, V = 0.277, see Supplemen a y Ma e ial Table S2 o he equencies o he mo ion sickness symp oms in each condi ion). 3.2 Memo y pe o mance Fo memo y pe o mance, a signi ican main e ec o he ac o Encoding Modali y (FEncoding Modali y(1, 118) = 15.85, p <.001; η2 = 0.118), a signi ican main e ec o he ac o Re ie al Modali y (FRe ie al Modali y(1, 118) = 7.21, p =.008; η2 = 0.058), and a signi ican in e ac ion o bo h ac o s (FEncoding Modali y × Re ie al Modali y(1, 118) = 24.35, p <.001; η2 = 0.171) we e ob ained. Pos -hoc - es s u he speci ying he in e ac ion e ec e ealed be e memo y pe o mance o pa icipan s in condi ions media ed, imme si e and c oss-down in com- pa ison o condi ion c oss-up (all s ≥|4.50|, all ps < 0.001, see Table 1; Fig. 2). Addi ionally, he equi alence es (TOST) e ealed sig- ni ican esul s o all non-signi ican g oup di e ences ega ding memo y pe o mance (d’) epo ed in Table 1 (all s ≥|3.28|, all ps < 0.001, see Table 2), indica ing ha mem- o y pe o mance can be conside ed equal compa ing condi- ions media ed, c oss-down and imme si e. 3.3 P esence The independen - es s compa ing p esence be ween he mVR and iVR modali ies on bo h days sugges a highe sense o p esence in iVR compa ed o mVR indica ed by he IPQ subscales. Excep o Spa ial P esence in he encoding phase days we e compa ed be ween he mVR and he iVR modali- ies using independen samples - es s. To assess a po en ial modi ica ion o he sense o p esence om he i s measu emen o he second measu emen as a unc ion o he encoding modali y, each IPQ subscale was analyzed using a mixed ANOVA wi h he wi hin-subjec s ac- o Timepoin (Day 1, Day 2) and he be ween-subjec s ac- o Condi ion (media ed, c oss-up, c oss-down, imme si e). Signi ican esul s we e u he analyzed by co esponding independen samples - es s o IPQ a ings on he second day o he expe imen , wi h a Bon e oni-co ec ed signi icance le el. To in es iga e whe he memo y pe o mance in he ou condi ions can be p edic ed by he le el o p esence du ing encoding o e ie al, eigh mul iple linea eg essions we e pe o med using all IPQ subscales on day one o on day wo as p edic o a iables and he d’ sco es sepa a ely by condi ion as dependen a iables. The ollowing eg ession equa ion was modelled o each o he ou expe imen al con- di ions o day one and day wo, espec i ely: d′=a+b Gene al P esence × x Gene al P esence +bSpa ial P esence ×x Spa ial P esence +bIn ol emen ×xIn ol emen +bRealness ×xRealness. 3 Resul s 3.1 Di e ences be ween condi ions No di e ences ega ding he pa icipan s’ age we e obse ed be ween he ou condi ions (FCondi ion(3, 118) = 0.19, p =.905; η2 = 0.005). Simila ly, he χ2- es did no e eal any di e ences ega ding gende balance be ween he Table 1 Tes s a is ics o pos -hoc independen samples - es s ega ding memo y pe o mance (d’) be ween all condi ions (media ed, c oss-up, c oss-down and imme si e). A Bon e oni-co ec ed signi icance le el was applied G oup s a is ics Independen samples es M SD d pCohens da Media ed 1.54 0.34 5.48 60 < 0.001 1.39 C oss-up 0.99 0.44 Media ed 1.54 0.34 0.72 57 0.475 0.19 C oss-down 1.47 0.39 Media ed 1.54 0.34 –0.97 59 0.338 − 0.25 Imme si e 1.63 0.41 C oss-up 0.99 0.44 − 4.50 59 < 0.001 − 1.15 C oss-down 1.47 0.39 C oss-up 0.99 0.44 − 6.01 61 < 0.001 − 1.51 Imme si e 1.63 0.41 C oss-down 1.47 0.39 − 1.57 58 0.123 − 0.40 Imme si e 1.63 0.41 aCohen’s d uses he pooled s anda d de ia ion. 1 3 128 Page 6 o 12 Vi ual Reali y (2025) 29:128 Pos hoc - es s e ealed ha he sense o Gene al P esence and Spa ial P esence du ing he e ie al phase we e highe in he c oss-up condi ion compa ed o he media ed, c oss- down and imme si e condi ions (all s ≥|3.06|, all ps < 0.001, see Supplemen a y Ma e ial Table S3). Addi ionally, he sense o Realness du ing he e ie al phase was highe in he c oss-up condi ion in compa ison o he c oss-down condi- ion ( (59) = 3.68, p <.001, o all pos -hoc compa isons see Supplemen a y Ma e ial Table S3). The mul iple linea eg essions de e mining whe he he d’ sco es can be p edic ed by he le el o p esence sepa a ely o bo h days in all ou condi ions yielded no signi ican esul s (all ps ≥ 0.214, see Supplemen a y Ma e ial Tables S4 and S5). 4 Discussion The aim o he s udy was o cla i y whe he expe iences in Vi ual Reali y (VR) in ol e modali y-speci ic mnemonic p ocesses ha a e dis inc om hose ypically obse ed unde 2D labo a o y condi ions, when me iculously ( (120) = − 1.73, p =.087), all u he IPQ a ings yielded sig- ni ican esul s (all s ≥|2.07|, all ps ≤ 0.041; see Table 3). The mixed ANOVA e ealed a signi ican in e ac ion be ween he ac o s Timepoin and Condi ion o all sub- scales o he IPQ (Gene al P esence: FTimepoin *Condi ion(3, 118) = 28.55, p < 0.001; η2 = 0.421; Spa ial P esence: FTimepoin *Condi ion(3,118) = 14.43, p < 0.001; η2 = 0.268; In ol emen : FTimepoin *Condi ion(3,118) = 6.28, p < 0.001; η2 = 0.138; Realness: FTimepoin *Condi ion(3,118) = 28.82, p < 0.001; η2 = 0.423). Table 2 Tes s a is ics o Welch wo sample - es (TOST) o non- signi ican compa isons o memo y pe o mance (d’) be ween he con- di ions G oup S a is ics TOST M SD d p Media ed 1.54 0.34 − 4.5 55.44 < 0.001 C oss-down 1.47 0.39 Media ed 1.54 0.34 4.23 57.75 < 0.001 Imme si e 1.63 0.41 C oss-down 1.47 0.39 3.28 57.98 < 0.001 Imme si e 1.63 0.41 Fig.2 Compa ison o memo y pe o mance (d’) be ween all ou con- di ions (encoding in mVR and e ie al in mVR (media ed), encoding in mVR and e ie al in iVR (c oss-up), encoding in iVR and e ie al in mVR (c oss-down), and encoding in iVR and e ie al in iVR (imme si e)). Memo y pe o mance (d’) was signi ican ly enhanced o pa icipan s in condi ions media ed, imme si e and c oss-down in compa ison o condi ion c oss-up (all ps <.001). S anda d e o s a e depic ed by e o ba s. 1 3 Page 7 o 12 128 Vi ual Reali y (2025) 29:128 ma kedly dis inc om hose p edominan ly s udied in labo- a o y se ings. The enhancemen in memo y pe o mance obse ed can he e o e be a ibu ed o he unique unc ional cha ac e is ics o mnemonic p ocesses in imme si e VR en i onmen s. F om a p ac ical s andpoin , ou esul s sugges ha o asks in ended o be ca ied ou in h ee-dimensional space, lea ning ou comes a e signi ican ly enhanced when aining o lea ning occu s in an iden ical o simila h ee- dimensional en i onmen . This unde sco es he limi ed ans e abili y o lea ned con en om wo-dimensional o h ee-dimensional con ex s and highligh s he unique cogni i e p ocesses in ol ed in lea ning in imme si e en i onmen s. The esul s a e in line wi h p e ious VR s udies show- ing ha in con as o con en ional 2D pa adigms, di e en mechanisms a e employed o encode (B éche e al. 2019; I iye and S . Jacques 2021; Johnsdo e al. 2023a, b; Kiske e al. 2024; Kou esis e al. 2021) and o e ie e con en in imme si e en i onmen s (Reggen e e al. 2018). In com- pa ison o he encoding o 2D s imuli, a b oade co ical ne wo k associa ed wi h mul imodal in eg a ion is in ol ed du ing he encoding o complex i ual s imuli (Johnsdo e al. 2023a, b). Mo eo e , he e ie al o imme si e VR expe iences in a classic old/new EEG pa adigm e ealed ha e en when he e ie al modali y ollows he con en- ional 2D p esen a ion, he mnemonic mechanisms di e o he ex en ha he usually cha ac e is ic di e ence in he he a band esponse o old and new s imuli was no obse ed a e VR-based encoding (Kiske e al. 2021c). This is pa - icula ly ema kable as he he a-based co ico-hippocampal con olling o he e ec s o he VR head-moun ed display (VR-HMD) on memo y pe o mance. D awing on p e ious esea ch, we hypo hesized ha hese unique mnemonic p o- cesses would be mani es ed h ough enhanced e ie al suc- cess. Acco dingly, we compa ed he e ie al success o he same s imulus ma e ial inciden ally encoded in imme si e VR (iVR) o in media ed VR (mVR), using a VR-HMD in all condi ions. This design allowed us o de e mine whe he VR me ely se es as a con ex o encoding o e ie al, o indeed acili a es a unique o m o mnemonic p ocessing. Gene ally, he canonical con ex e ec , i.e., an ad an- age o memo y pe o mance de i ed om he cong uency o encoding and e ie al con ex s (Godden and Bad- deley 1975), was eplica ed o bo h he media ed and he imme si e condi ion as pa icipan s exhibi ed equally high memo y pe o mance. Ye mos impo an ly, we ound ha expe iences ecalled in mVR esul ed in he same e ie al success when encoded in iVR as hose encoded in mVR. In o he wo ds, encoding in iVR no only enhances mnemonic p ocessing bu also achie es ecall success a he le el o con ex e ec s o s imuli encoded and ecalled in mVR. To ou bes knowledge, his is he i s s udy o igo ously con ol o he e ec s o he HMD and he gene al impac o being in VR. Ou indings p o ide solid e idence ha he enhanced memo y pe o mance o en obse ed in VR se ings is no me ely a echnological a i ac ha can be a ibu ed o he VR se ing i sel . Mos impo an ly, he da a e eals ha enhanced memo y e ie al o iVR expe i- ences s ems om he encoding s age in iVR and does no solely ely on he e ie al s age. Thus, expe iencing e en s in iVR acili a es a o m o mnemonic p ocessing ha is Table 3 Tes s a is ics o independen samples - es s ega ding p esence (all IPQ subscales) in he modali ies mVR and iVR on bo h days G oup S a is ics Independen Samples -Tes M SD d pCohens da Day 1 Gene al P esence mVR 4.58 1.56 − 2.28 120 0.025 − 0.41 Gene al P esence iVR 5.22 1.52 Spa ial P esence mVR 4.69 1.08 − 1.73 120 0.087 − 0.31 Spa ial P esence iVR 5.02 0.98 In ol emen mVR 4.27 1.40 − 3.09 120 0.002 − 0.56 In ol emen iVR 5.01 1.23 Realness mVR 3.41 0.97 − 4.02 120 < 0.001 − 0.73 Realness iVR 4.10 0.95 Day 2 Gene al P esence mVR 3.97 1.59 − 5.10 120 < 0.001 − 0.93 Gene al P esence iVR 5.37 1.44 Spa ial P esence mVR 4.32 1.20 − 4.33 120 < 0.001 − 0.78 Spa ial P esence iVR 5.20 1.05 In ol emen mVR 4.41 1.57 − 2.07 120 0.041 − 0.37 In ol emen iVR 4.95 1.32 Realness mVR 3.33 1.41 − 3.50 120 < 0.001 − 0.63 Realness iVR 4.02 1.05 aCohen’s d uses he pooled s anda d de ia ion. 1 3 128 Page 8 o 12 Vi ual Reali y (2025) 29:128 li elike au obiog aphical memo y p ocesses. Ou knowl- edge abou said p ocesses could u he mo e be ex ended by combining his unique app oach wi h elec ophysiologi- cal me hods, e.g., elec oencephalog aphy, o neu oimag- ing me hods, e.g., unc ional magne ic esonance imaging. U ilizing VR o c ea e con olled li elike au obiog aphical scena ios he e o e p o ides he po en ial o in es iga e mnemonic p ocessing no only on he beha io al bu also on he neu al le el. Ye de eloping i ual en i onmen s o psychological expe imen s equi es he app op ia e echnological and me hodological expe ise. In designing a s udy in ol ing VR, i is c ucial o ca e ully conside bo h echnological aspec s and cha ac e is ics o he i ual en i- onmen , as hese ac o s migh in luence he na u e o he mnemonic p ocesses exhibi ed ( o e iew see Smi h 2019). Whe eas echnological ad ancemen s migh educe hese es ain s by minimizing he e o o c ea e app op ia e i - ual en i onmen s in he u u e, i is impo an no o neglec e hical conside a ions ega ding VR esea ch. Since VR can elici li elike emo ional, physiological, and beha io al eac ions (e.g Kiske e al. 2021a; Kiske e al. 2021a, b, c, d), esea che s should be awa e o he po en ial de imen- al consequences o expe iencing ealis ic i ual scena ios, in pa icula hose o a s ong nega i e alence. The e o e, when de eloping VR esea ch, he po en ial cogni i e and beha io al consequences o VR expe iences (Mada y and Me zinge 2016) should be add essed in he ligh o e hical conside a ions. Al hough we main ained he a i icial se ial p esen a ion o he i ual scenes, ou s udy suppo s he idea ha iVR en i onmen s enhance he sense o p esence, wi h his e ec being pa ially mo e p onounced when a less imme si e en i- onmen has p e iously been encoun e ed. While he ela ion be ween an inc eased sense o p esence and memo y pe o - mance has been epea edly discussed in p e ious esea ch ( o e iew see Smi h 2019), a di ec s a is ical ela ionship be ween p esence and memo y pe o mance emains elusi e in ou esul s. Howe e , he absence o his ela ion in ou da a is consis en wi h p e ious indings showing ha while di - e en manipula ions o imme sion in VR en i onmen s ha e a ying e ec s on memo y and p esence, he e is no de ini- i e e idence ha he subjec i e sense o p esence media es he impac o imme sion on episodic memo y pe o mance (Smi h and Mulligan 2021). Po en ially, p esence ques ion- nai es like he IPQ (Schube e al. 2001) o he ITC-Sense o P esence In en o y (Lessi e e al. 2001) do no ully co e he c i ical aspec o p esence ha indeed acili a es he o ma- ion o au obiog aphical memo y. Al hough signi ican posi- i e co ela ions be ween di e en p esence ques ionnai es could be obse ed (Kobe and Neupe 2013), he in luence o a ious ac o s can cause inconsis encies in esul s ega ding p esence assessed wi h hose ques ionnai es (e.g., G a and in e play is hypo hesized o unde lie many di e en ypes o memo y, wi h episodic memo y being one o he mos ho oughly s udied ones (Tul ing 2002). Mos s udies in es iga ing episodic memo y assume ha he same cogni i e p ocesses a e elici ed by p esen - ing pa icipan s wi h images on a compu e sc een du ing encoding o e ie al ha occu unde eal-li e condi ions. Howe e , p e ious s udies ha e shown minimal o e lap in b ain egion ac i a ion be ween labo a o y-based and eal- li e memo y expe imen s (Monge e al. 2018). This disc ep- ancy is unde s andable, as labo a o y asks o en p esen a i icial e en s lacking pe sonal meaning, whe eas eal-li e e en s a e mul imodal and signi ican , occu ing in physical p oximi y. Labo a o y e en s usually only a ec he isual domain, se ing as eminde s o p e iously expe ienced e en s (Ma co i and Jacques 2018). S udying hese labo a o y e en s has signi ican ly ad anced ou unde s anding o mnemonic p ocesses and b ain unc ioning. The pa adigm o sho se ial s imuli p e- sen a ion is cen al o psychological science, especially wi h psychophysiological measu emen s ha ely on epea ed p esen a ion o s imuli. Unlike eal-li e expe imen s, s an- da d pa adigms o e unpa alleled expe imen al con ol and ep oducibili y, which a e undamen al o psychological science. Howe e , we should also ecognize his con en- ional app oach’s limi a ions. S imuli o en se e me ely as eminde s, which can signi ican ly obscu e he ue na u e o p ocessing hem (Snow and Culham 2021), he eby se e ely dec easing ou abili y o gene alize om labo a- o y o eal-li e se ings. Howe e , ha does no imply ha we canno gene alize o ce ain eal-li e condi ions bu ha we should be mo e conscious o wha ou pa adigms a e in es iga ing, and exe cise cau ion in o e -gene alizing. Fo example, wa ching pic u es o ideos on a compu e sc een is a signi ican pa o ou li es as we cons an ly engage wi h moni o s. Ye , i is c ucial o esea ch o acknowledge ha i is speci ically his aspec o ou daily li es ha is being in es iga ed wi h moni o -based pa adigms. Towa ds a mo e imme si e app oach in psychological science enhancing ecological alidi y, we c ea ed a pa a- digm ha b idges he gap be ween classic o con en ional pa adigms and eal-li e expe iences. On he one hand, we main ained he se ial p esen a ion o s imuli, on he o he hand we used pho o ealis ic, i.e., imme si e s imuli o c e- a e a ealis ic en i onmen . The main ained se ial p esen- a ion o he i ual en i onmen s mos ly con adic s he p obabilis ic dis ibu ion o ou daily li e expe iences as scenes do no change ha d as ically wi hin a ac ion o a second. Ne e heless, his combina ion o con en ional and imme si e aspec s o he pa adigm seems o achie e an op imal balance whe e expe imen al con ol is s ill high bu he mnemonic p ocesses in es iga ed di e and esemble 1 3 Page 9 o 12 128