Not Having What You Want versus Having What You ...

C O G N IT IO N AN D E M O TIO N , 1999, 13 (2), 129±148

Not Having What You Want versus Having What You Do Not Want: The Impact of Type of Negative Outcome on the Experience of Disappointment and Related Emotions Wilco W. van D ijk University of Amsterdam, the Netherlands

M arcel Z eelenb erg Tilburg University, the Netherlands

Jo op van der Pligt University of Amsterdam, the Netherlands

T he presen t resea rch fo cu ses o n the em otion al co nseq uences of negative o utco mes. Two types of n egat ive o u tco m es are d istin gu ish ed : T he ab sen ce o f a po sit ive outco me and the p resen ce o f a negative ou tcom e. It is argu ed that disapp o int ment , b eca u se o f its close link with ho p e, desire, and pro mise, is m ore asso ciat ed with the ab sen ce of a p ositive ou tcom e than with the presen ce o f a negat ive o utco m e. D isapp ointment is also expected to be m o re associated with the absen ce o f a p ositive o utco me than related n egat ive emo tio n s, such as sa dn ess, anger, frustr atio n , and regret . Th e result s o f fo u r stud ies, using differen t m et ho d ologies, co n® rm these pred ict io ns. In Study 1 and Study 2 participants recalled an autob iogr aphica l em otion al ep iso d e, and appraisals co ncern in g two d iffer ent typ es o f n egat ive ou tcom es were assessed. In Study 3 a scena rio m et ho d ology was used in which the type of n egative ou tco me was exp er iment ally m anip ula ted and ratin gs for differen t em otio ns were assessed . F ina lly, in Study 4 o n-line em o tio nal reaction s to the two d iffer ent types o f negat ive ou tco m es were assessed in an exp erim en t in which real mo ney cou ld be won o r lo st . Im p lication s for the st udy o f d isap po in tm en t are brie¯ y d iscu ssed . R eq uest s fo r reprint s sho u ld be sen t to Wilco W. va n D ijk, D ep a rtm ent of Socia l Psych o logy, F ree U n iver sit y of A m sterd am , va n d er Bo ech o rststra at 1, N L±1081 BT, Am st erd a m, Th e N et herla n d s; e-m a il: W W.va n.D ijk @p sy.vu .n l. We t ha n k Jan e Beatt ie for her help with design in g Study 3, a n d we a lso th a nk C a rst en F ried rich , K a thelijn e G o defrooij, M in ke M eth o rst, a n d C o rn eÂ lise Pa sto r fo r helping with t he d ata co llectio n o f Stud y 1. We t ha n k C ra ig Sm ith , To ny M an st ea d , an d two a no nym ou s reviewers fo r help ful co m ment s o n a ea rlier versio n of th is m a nu scrip t.

q

1999 P sych olo gy P ress Ltd

130

van DIJK, ZEELENBERG, van der PLIGT

INTRODUCTION N o child hoo d passes witho ut disapp ointm ent about a birth day present , no ad olescence seems t o be co mp lete witho ut a disapp ointing love affair, and h ardly anyo ne is a str anger to the un pleasa nt feeling th at stems from buying an expensive co nsu mer pro du ct th at tur ns o ut to be less th an expected. All in a ll, a life witho ut d isapp ointm ent seems rare. T h is int rospective view is su pp ort ed by m ore systematic research showin g t hat d isapp oint ment is o ne of th e m o st frequ en tly exp erienced emo tions following failu re (Weiner, R ussell, & L erm an , 1979). D isappoint ment has received som e attention from researchers in th e ® eld of behavio ur al decision ma king (Bell, 1985; L oom es & Sugd en, 1986). T h ey assum e t hat p eo ple a nticipate disappoin tm en t and take this int o accou nt when m ak in g decision s. For in sta nce, Shepp erd , Ou ellette, an d F ern and ez (1996) sh owed th at in divid uals tend to aban do n t heir o ptimism and may even becom e p essimistic in ant icipatio n o f self-relevant feedba ck. T hey a rgu e th at peop le a nticipate t he d isappoint ment th ey wo uld feel if th eir p erfor ma nce were t o fall shor t of th eir expectatio ns. T h us, peop le red uce their perfo r man ce estimates t o minim ise th e po ssibility of perfor ming wo rse th an expected a nd t o avoid d isappoin tment arising as a co nsequ ence. Ou r own research showed t hat th e probabilit y of a n ou tcom e an d t he effor t invested in attaining an ou tco me h ave a n im pact o n t he intensity of d isapp oint ment . T h e m ore pro bable a po sitive o ut co me was, the m ore int en se d isapp oint ment a person feels if th e ou tcom e is n ot attained (van D ijk & va n der Pligt, 1997). D isappo in tm en t is also m ore intense aft er h aving invested m ore effort in va in to attain the d esired o utcom e (van D ijk, va n der P ligt, & Z eelenb erg, 1998a ). T he way in which disapp oin tm en t is experienced ha s also been sub ject t o em pirical investigation . Z eelenberg, va n D ijk, M anstea d, an d va n der P ligt (1998a ) showed th at t he exp erience o f d isapp oint ment (as com pared to regret) involves feeling p owerlessness, a tend en cy to do no thing and t o get away from the situatio n, an d wa nting to d o no thing. A questio n th at was no t add ressed in o ur earlier research con cern s t he causes o f d isapp oint ment . In th e p resent art icle we aim to shed so me mo re light on this issue. We also investiga te whether disapp ointm en t, on the basis of these causes, ca n be distingu ished from related n egative em otion s such as sadn ess, an ger, fru str ation , an d regret. Disappointment and Desire, Hope and Promise. T h e va riou s d efinitio ns of disappo in tm en t seem to sh are one centr al featu re, t hat is, t he n onful® lment of an exp ectatio n. D isapp ointm ent h as b een d e® ned as ``n on achievemen t of an exp ected ou tcome’ ’ (F rijd a, 1986, p. 280), o r as ``a psycho logica l reaction to a n ou tco m e that do es not m atch u p to

DISAPPOINTMENT AND TYPE OF NEGATIVE OUTCOME

131

exp ectatio ns’’ (Bell, 1985, p. 1). Several aut hor s explicitly link disappo in tm ent to t he no nfu l® llm en t of a positive expectation . Shand (1914) d e® nes disappo in tm en t as an emo tion t hat is closely linked t o desire. O rto ny, C lore, an d C o llins (1988, p. 110) a lso stress the im por tan ce o f bo th d esire and no nfu l® lment of a n expected out co me in t heir de® nition of disappo in tm ent : ``t o be d ispleased abou t t he d iscon® rm atio n of th e pro spect of a desirable event’ ’ . T hey state th at th e intensity o f disapp ointm ent is affected by hope; t hat is, high h opes give rise to m ore int en se d isappoint ment if these ho pes are d ashed . M owrer (1960, p. 169) also relates disapp ointm ent to h ope: ``W hen a hop e signal app ea rs an d then disapp ea rs t he reaction is on e o f disapp ointm en t’ ’ . F rijda (1986, p. 287) link s d isappoint ment with promise: ``P rom ises gen erally t ur n in to d isapp oint ments when not ful® lled’ ’ . F inally, Or to ny et al. (1988, p. 110) make an exp licit distinctio n between being ``d isp leased about th e d iscon® r matio n o f th e pro spect o f a desirable event’ ’ and b eing ``d isplea sed abou t th e con® r mation o f the pro sp ect of a und esirable event’ ’ . T hey labelled t he fo r mer emo tiona l reactio n ``d isappoint ment’ ’ , t he latter as ``fears-con ® r med’ ’ . T hese d e® nitions su ggest that disappoint ment is p rima rily exp erienced in a situ ation in which so meth ing p ositive was exp ected but d id n ot occur. It seems to be closely linked with h ope, desire, a nd prom ise. Not Having What You Want vs. Having What You Do Not Want. Several au tho rs (F rijd a, 1986; H iggin s, 1989; M owrer, 1960; Roseman , 1984; R oseman, An ton io u, & Jo se, 1996; Roseman, Sp indel, & Jose, 1990) argued that negative em ot io ns can b e the result of t wo different negative situation s, which we refer to as type of negative outcome. N egative emot io ns can b e the result of either the absence o f a positive outcome (``n ot havin g wh at yo u want ’ ’ ) or th e presence of a negative o utcom e (``h aving what yo u do n ot want ’ ’ ). Appra isal th eo ry (see e.g. Ar nold , 1960; F r ijd a, 1986; Roseman , 1984; Scherer, 1984; Sm it h & E llswor th, 1985) states that evaluations an d in terp retation s of events d eterm in e which em otion is exp erienced. On e appraisal dimension that can lead t o d ifferen t emo tions is t ype of negative ou tco m e. F r ijd a refers to the absence o f a p ositive va lence o r t he presence of a negative va lence, t hat is, the absence o f so meth in g in tr in sically appetitive or th e presence of som ething in tr in sically aversive. Roseman refers to the absence of a reward o r t he presence of a p unish m en t. 1 T hu s, int erpretin g or eva lu ating a situ ation a s either one type of negative ou tco m e or th e 1 Th e ab sence of a reward an d th e p resen ce o f a pu nish men t is a co mb in a tion of a n ap pra isal of mo tivatio n al state (wheth er th e d o min a n t o perative mo tive is ap pet itive o r aversive; a reward wantin g to atta in o r a pu n ishm en t wantin g to avo id ) a nd a n a pp ra isa l o f sit uat io na l state (wheth er th e m o tivatio n al state is p erceived to b e present o r ab sent ).

132


o ther, th at is, as either th e absence of a po sitive ou tcome o r as the presence o f a negative ou tcom e, can lea d to d ifferen t negative emo tions. Because o f its close lin k with hop e, desire, and pro m ise, we expect d isapp oint ment to be an em ot io n cau sed by a situation which is app ra ised as an absence of a p ositive ou tcome. F u rt herm ore, we exp ect that disappo in tm en t is m o re associated with t he absence of a p ositive o utcom e th an several o ther related negative em o tions, such as sadn ess, anger, frustr ation , an d regret. Disappointment and its Relation to Sadness, Anger, Frustration, and Regret. D isappo intm ent is hardly ever experienced in isolatio n. I ts experience is closely linked t o ot her n egative em o tion s. For instance, it h as been argued t hat b oth sadness and anger can be the result o f disappo in tm en t (L evine, 1996; M owrer, 1960). D isappo in tment about not attainin g a n exp ectatio n or a goal co uld result in sad ness or an ger, depending on b eliefs abou t whether th e origina l expectation or goa l can be reinstated. Sad ness is associated wit h the b elief th at go als can no t be reinstated, wherea s anger is associated with th e b elief th at so meth in g can b e d one t o reinstate a goal (L evine, 1996). Sad ness is assumed t o result from evaluatin g a situ ation as the absence o f a po sitive ou tcome (F rijd a, 1986; H iggin s, 1989; Roseman, 1984; R oseman et al., 1990, 1996). Anger, o n t he oth er ha nd, is oft en assum ed to be cau sed by b oth types o f n egative o utcom es (F rijda, 1986; Ro seman , 1984; Roseman et al., 1990, 1996). Frustration is also related to disappo in tment. T h e t erm ``d isapp ointm en t’ ’ is d e® n ed in Webster’s Third New International Dictionary, (1981) as fo llows: ``fa ilu re of expectation o r h op e: fru stratio n’ ’ . F r ustratio n, like an ger, is a ssum ed to be ca used by bo th typ es of negative o utcom es (F rijd a, 1986; Roseman, 1984; Ro sema n et a l., 1990, 1996). T h e relation between disappo in tm en t an d regret is at least twofo ld. F irst, b oth a re related to decision m ak in g and ch oice (Bell, 1982, 1985; L o om es & Su gden, 1982, 1986). Second , bo th a re considered t o be co unt erfa ctu al em o tions arisin g from t hou ght s about ``what migh t have b een’’ (Inm an, D yer, & Jia, 1997; L o om es & Sugden , 1984; O rt ony et a l., 1988; Z eelenb erg et al., 1998b ). R egret is generally associated with bo th types of n egative ou tco m es, that is, b ot h th e absence o f a positive ou tco me an d t he p resence of a negative outcome ca n give r ise to regret (F rijda, 1986; Ro sema n, 1984; Roseman et al., 1990, 1996; Z eelenb erg & Beattie, 1997; Z eelenb erg, Beattie, va n d er Pligt, & de Vr ies, 1996). To date, we d o n ot know o f any em p ir ical study th at h as exp licitly investigated t he in¯ uence o f typ e of negative ou tco me on t he int ensity of d isapp oint ment . T he m o st relevan t stu dy is the o ne repo rt ed by Roseman (1991). H is p articipant s read brief stor ies abo ut event s th at h app en ed to va r io us p rotagon ists. I n th ese stor ies, infor m atio n relevan t t o ® ve appr aisals


133

was systematically va r ied, and par ticipan ts r ated th e intensity of the em otio ns th at they believed t he p rotagon ists felt in response to th e events. H alf of th ese stor ies concern ed n egative em ot io ns an d are of in terest for o ur present pu rpo ses. T hese stor ies were concerned with th e absence o f a po sitive out co me or the presence o f a negative o utcom e. Alth ough disappo in tment was no t the fo cus of Roseman’s study (it dealt with sor row, 2 anger, fru str ation , regret, and several oth er em otion s), in tensity r atings o f d isappoin tment were obt ain ed. Roseman fou nd th at bo th th e absence of a positive ou tcom e a nd t he p resence o f a negative ou tco m e resulted in in creased ratin gs fo r all negative emotion s. F urt herm ore, h e con clud ed that th e differences in negative emotion s d ue t o th e typ e of negative o utcom e were co m par atively sma ll. In Table 1 we su mm ar ise th e m ea ns for sor row, anger, frustration, and regret as repo rted by Roseman (1991), an d also give t he m ean d isapp oint ment r atings. 3 Table 1 shows that the differences between the two different types of negative outcomes are indeed comparatively small for sorrow, anger, and regret. T he difference, however, is larger for frustration, and particularly large for disappointment. T his provides some preliminary support to our reasoning that disappointm ent is more closely associated with the absence of a positive outcome than with the presence of negative outcome. In the present series of studies we look explicitly at the relation between the two types of negative outcomes and disappointm ent. In the ® rst two studies participants recalled an autobiographical emotional episode and appraisals concerning typ e of negative outcome were assessed. In Study 3 we experim entally manipulated type of negative outcome and assessed ratings for different emotions, using a scenario methodology. In Study 4 we aga in manipulated typ e of negative outcome, but this tim e we assessed on-line emotio nal reactions by using a ga me-like task in which real money could be lost or won. In all studies we compare the relation between typ e of negative outcom e and disappointment with the relation between type of negative outcom e and sadness, anger, frustration, a nd regret.

STUDY 1 In Study 1 par ticipants were asked to recall a speci® c event in which they exp erienced either disappo intm en t, sad ness, anger, fru stratio n, or regret. F ur therm ore, they were a sked to in dicate to what exten t th e situatio n 2 In m o re recen t work Rosem a n u ses th e emo tio n term ``sa dn ess’ in stead o f so rrow. O ne rea so n fo r th is was th a t sadn ess is mo re co mm o nly used t ha n sorrow in th e En glish la n gu a ge (I.J. Ro sem an , p erso na l co mm un icatio n , 22 Ap ril 1997). 3 Th ese mea n s were no t repo rt ed by Ro sema n (1991). H owever, m ea n in t ensit y ratings o f d isap po in tm en t o n th e d ifferen t stories were report ed. O n th e b a sis o f th ese in t ensit y ratin gs we ca lcu lat ed t he m ean s o f d isap p oin t ment rat in gs fo r th e two typ es of negat ive o ut co m es.

134

van DIJK, ZEELENBERG, van der PLIGT TABLE 1 Mean Intensity Ratings of Different Negative Emotions as a Function of Type of Negative Outcome (adapted from Roseman, 1991) Emotions Outcome PA NP

Disappointment

Sorrow

Anger

Frustration

Regret

7.16 6.27

5.42 5.26

6.33 6.23

7.19 6.62

5.57 5.52

Note: PA , p o sitive ab sen ce; N P, n egat ive p r esen ce. I n R o sem a n’s o rigin a l work is referred to PA a s th e absence o f a reward (M S+ SS2 ) a n d t o N P a s t he p resen ce of a p un ish men t (M S 2 SS+ ).

represent ed o ne of th e t wo typ es of negative o utcom es. We expect t hat events in which d isappointm ent was exp erienced are m ore likely t o b e associated with th e absence o f a po sitive o utcom e than with th e presence o f a negative out co me, an d that th ese events are associated mo re wit h t he absence o f a po sitive out come than sadn ess, anger, frustration, and regret.

Method Design and Participants. St udy 1 ha d a ® ve grou p between-subjects d esign (D isappo intm ent vs. Sad ness vs. F ru stratio n vs. A nger vs. R egret). 4 St ud ent s at the U niversity o f A msterd am (N = 100) p art icipated in t his stu dy. T here were 20 p art icipant s in each cond it io n. T his study was p art of a large app raisa l study, t hat wa s ad ministrated du ring a large-scale test session. Par ticipants were pa id 10 D ut ch guilders (approximately $5.00) for their par ticipation. Procedure. Q uestion naires were r ando mly d istr ibuted am on g t he p art icip ant s. D ependin g on th e con ditio n th ey were in , p articipant s were asked to describe a situatio n in which th ey felt either in tense disapp ointm en t, sadn ess, frustration, anger, or regret. 5 N ext, p ar ticipan ts were asked th e fo llowing two q uestion s: ``To what extent did t he situation concern som eth ing p ositive (som ething yo u want ed ) th at d id not o ccur ?’ ’ and ``To 4 I n th e p resen t research pa rticip an ts were exp licitly a sked ab o ut d isap p oin t men t con cernin g a n o ut com e. H owever, o ne ca n o f co urse a lso be d isapp o int ed in a p erson. A mo re det a iled a ccou nt o n th e d ifferences b etween t hese t wo kin d s o f disapp oin tm en t ca n b e fo u nd in va n D ijk, va n der P ligt, a nd Zeelenb erg (1998b ). 5 T he em ot io n word s in th e present stu dy were in D ut ch, a s t hey were in Stud y 2 a nd 4. Study 3 was co n du ct ed in E nglish , at a u niver sit y in t he U n ited K ingdo m. We h ave n o rea so n t o b elieve th at t here a re a ny su b stan tia l d ifferences bet ween D ut ch a n d En glish in th e d eno ta tive o r co nn o ta tive m ean ing of t hese wo rd s.


135

what extent did t he situatio n con cern so meth in g n egative (som ething yo u did no t wa nt) t hat did occur ?’ ’ 6 Par ticipa nts cou ld answer bo th qu estions o n a 9-po int scale wit h end-po in ts labelled not at all (1) a nd to a great extent (9), an d t he mid point labelled a s neutral (5). F ur thermo re, p art icipan ts were asked how intense t hey exp erienced d isapp oint ment, sadn ess, frustration , anger, an d regret in the described situation. T hese in ten sity ratings were do ne on a 9-p oint scale labelled not at all (1) a nd very much (9).

Results and Discussion Appr aisal scores fo r bo th m ain d ep en dent va r iables (type of negative ou tco m e) were en tered in to a n AN OVA wit h type of n egative ou tcom e as a within-subjects factor an d th e em otion rated as a b etween-sub jects factor. A sign i® cant ma in effect o f type of negative out co me was foun d [F(1,95) = 6.10, P < .02], an d a signi® can t t wo-way intera ction b etween typ e o f negative ou tcom e an d th e emo tion r ated wa s fou nd [F(4,95) = 5.56, P < .001]. T he mean app raisal scores fo r t he ® ve target emo tion s are shown in Table 2. 7 Plann ed com par ison s revealed t hat disapp ointm ent wa s m o re closely associated with the absence of som ething po sitive th an with th e presence of so mething negative, [t(19) = 2.37, P < .05]. Sa dness an d regret were bo th mo re closely associated with th e presence of som eth ing negative [t(19) = 3.46, P < .005] and [t(19) = 2.86, P < .01], respectively. N o differences were fou nd for a nger an d frustration (ts < 1). F ur thermo re, a co ntr ast analysis between th e appr aisal ratings of d isappoint ment an d t he appr aisal ratin gs of the oth er em ot io ns revealed th at disappo in tm en t was m ore stron gly associated with the absence of so methin g p ositive t han th e oth er em ot io ns [F(4,95) = 2.61, P < .05]. D isappoint ment wa s m ore associated with this t ype of n egative ou tcome tha n sad ness, fru str ation , an d regret (see Table 2, com pa riso n within up per row). C oncerning th e presence of a negative o utcom e no signi® cant difference was fo un d between th e app raisa l r atin gs o f d isapp oint ment on the on e han d an d th e r atin gs of the oth er em ot ion s o n th e o ther. A lt hou gh d isappoint ment was less stro ngly associated with t he presence of som ething negative t han sad ness. N o sign i® can t d ifferences con cern ing th is type of

6 Bo th q uest io ns were pa rt o f a large ap pra isa l stu dy a nd were in tersp ersed b etween 13 o th er q uest ion s abo ut a pp ra isa ls. T h e o rder o f t hese q uest io ns was ra n do mly deter min ed . 7 Th e mea n in ten sity ra tin gs fo r disapp o in tm en t, sadn ess, a nger, frustratio n, a nd regr et were 5.40, 6.15, 5.80, 5.30, a nd 5.25, respectively. Th e inst a nces o f sad ness were sligh tly m ore in ten se th a n t ho se o f d isa pp oint men t, fru st ra tio n, a nd regret (Ps < .05). T he r an ge of scores, h owever, is sm all in absolu te t er ms.

136

van DIJK, ZEELENBERG, van der PLIGT TABLE 2 Mean Appraisal Ratings on both Questions concerning Type of Negative Outcome for Each of the Five Emotions (Study 1) Emotions Outcome PA NP

Disappointment d

5.75 4.35 a

Sadness a

3.50 5.90 b

Anger c, d

4.80 5.10 a ,b

Frustration a ,b ,c

4.15 4.45 a

Regret 3.50 a ,b 5.45 a ,b

Note: PA , po sitive ab sence; N P, negat ive presen ce. Scores co uld ra n ge from 1 to 9. H igh er sco res re¯ ect th e fo llo wing: p ositive is ab sent t o a greater ext ent ; n egative is p resen t to a grea ter ext ent . M ea n s within a row n ot sharing a com m on su perscript differ sign i® ca n tly (P < .05).

n egative outcom e were fou nd b etween disappo in tm ent and a nger, fru str atio n, a nd regret (see Table 2, co m par ison wit hin lower row). Par ticipants who recalled a situ ation lin ked with a pa rticular emotion also rated th e intensity o f th e o ther t arget emotion s th ey experienced d ur ing th is situation. 8 To exa mine the extent to which d isappoint ment shares a un ique relation to the absence of so methin g po sitive a par tial cor relatio n analysis between th e appr aisal questio ns an d the in tensity of th e emotion s wa s cond ucted . A cross all d escrib ed situatio ns t he in tensity of d isapp oint ment wa s signi® cantly cor related with the absence of a po sitive o utcom e, when cor rected fo r th e o ther appr aisal qu estion (r = .18, P < .05). N on e of t he in tensity r atings of sadn ess, anger, frustr ation , an d regret were signi® cantly co rrelated wit h this t ype of negative ou t co me (r = 2 .07, n.s., r = .08, n.s., r = .10, n.s., r = .07, n.s.), respectively. 9 T he intensity of sadn ess was signi® cant ly correlated with the presence of a n egative o utcom e (r = .32, P < .005). T h e intensity ratings of disapp ointm ent, anger, frustr ation , and regret were n ot sign i® ca nt ly cor related wit h t his typ e of n egative o utcom e (r = 2 .06, n.s., r = .04, n.s., r = .03, n .s., r = .06, n.s.), respectively. 1 0 8 D isap po int men t sit u atio n s received in tensity rat ings o f 4.95, 4.75, 6.20, 3.60, for sadn ess, a n ger, fru st rat ion , a n d regret , resp ect ively. Sadn ess situ at io ns received in ten sity ra tin gs o f 4.05, 4.55, 4.50, 2.90, fo r disa pp o in tm en t, a nger, fru strat io n, a nd regr et , respect ively. A nger situ at ion s received in ten sity rat ings o f 5.20, 3.95, 5.40, 3.60, fo r disa pp o int men t, sad ness, frustrat ion , a nd regret, resp ectively. F ru st ratio n sit u atio n s received in ten sity ratings o f 5.20, 4.60, 5.90, 2.90, fo r d isap p oin tm ent , sa dn ess, a n ger, a n d regret, respect ively. R egr et sit uat io ns received in t en sit y ratings o f 5.15, 4.95, 4.10, 4.85, fo r disap po in tm en t, sad ness, a nger, a n d frust ration , resp ect ively. 9 Th e zero-order co rrelation s between th e int ensit y o f th e em ot io ns a n d th e absence o f a p o sitive o ut co m e were a lm o st id ent ical t o th e p a rtia l co rrelatio n s. Th ere was a sign i® can t n ega tive co rrelation b et ween th e two ap p raisa l q uest ion s (r = 2 .21, P < .05). 10 Th e zero-ord er correla tio ns b etween th e in ten sit y o f th e emo tio n s a nd th e p resence of a n ega tive o u tco me were a lm ost id ent ica l to th e pa rt ia l co rrelation s.


137

T hu s, o ur results su ppo rt the n otion th at disappo in tm ent is better chara cterised by th e absence o f a p ositive ou tcom e than by th e presence of a n egative o utcom e. F ur thermo re, results sh owed t hat disapp ointm ent can be distin guish ed on t he ba sis o f type of n egative ou tcome from sadn ess, anger, frustr ation , a nd regret. T hat is, disapp ointm en t is m ore associated with the absence of a p ositive o ut com e th an sa dness, anger, frustration , and regret. In the present stu dy we asked par ticipants to give ratin gs o f t he extent to which th e situ ation involved th e absence of som ething po sitive o r th e presence of something n egative. T h is meth odo logy m ight suffer from at least two p roblems, a s Roseman et al. (1996) recently po inted ou t. F ir st, ask ing for ratings char acterising th e co ntent cou ld b e different from asking about th e cau se of an exp erienced emotion (see also F r ijda , 1993; Park in son & M an stead, 1992; Roseman et al., 1990). T h is cou ld lea d to a less correct id ent i® cation of th e cau ses of em o tio ns. Second , em o tio n episod es d escribed by par ticipan ts cou ld encom pass several emo tions, each with their own appr aisal d eterm in ant s (see also Scherer, 1993; Sm ith & E llswort h, 1987). Roseman et al. (1996, p. 245) stated that ``U nless th e su bject is in str ucted t o specify the appra isals t hat are relevan t to th e pr im ar y emo tion u nd er investigatio n, appra isals relevant t o ot her emot io ns m ay be rep or ted, ob scur in g tr ue appr aisal-emo tion relatio nsh ip s’ ’ . Ro sema n et al. (1996) recom m ended cor recting these problem s by: (a ) a sking pa rticipants t o rate the cau se of an emotion rath er th an the t hou ght s t hat they had on ce th e em ot io n h ad begun: a nd (b) asking par ticipa nts about the appraisals th at led to their emotio ns, r ath er tha n by ask ing th em about t he event t hat led to their emotion .

STUDY 2 In Study 2 we tr ied to replicate the ® nd in gs of St udy 1 usin g the m etho dology recom m en ded by Roseman et al. (1996). A s in St udy 1, p art icipant s were asked t o recall an in tense situation in which t hey experienced one of the ® ve target emot io ns. Part icipan ts were asked to ind icate t o what extent the target emotion was a ssociated with on e of th e two types o f negative ou tco m es. T h e question th at captu red t he type o f n egative o utcom e wa s adapted from R oseman et al. 1 1

11 In th e Ro sem an et a l. (1996) t heo ry, th is qu estio n is inten d ed t o cap tu re th e a ppr aisal d im en sio n o f m o tivatio n al st ate. We did n o t in clud e a mea su re of sit u atio n al st ate beca use an n ega tive ou tco me is a ssum ed t o b e ap p ra ised a s m ot ive inco nsistent. T his st ud y was in t en d ed to exa min e whet her th e emo tion s un der invest igation s a re eit her in con sisten t with an ap petit ive mo tive (in con sistent with att a in ing a reward ) or in co nsistent with a n aver sive mo tive (incon sist ent with avo id ing a pu nish men t).

138


Method Design and Participants. St udy 2 ha d a ® ve grou p between-subjects d esign (D isapp oin tm en t vs. Sad ness vs. F r ustratio n vs. An ger vs. R egret). Stud en ts at the U niversity of Am sterda m (N = 110) par ticipated in this study. T h ere were 22 par ticipants in each con dition . T his stu dy wa s pa rt of a large appr aisal study, that was adm inistrated du ring a la rge-scale exp erim en tal session . Pa rticipant s were pa id 10 D ut ch gu ild ers (approximately $5.00) for their par ticipation. Procedure. Q uestion naires were r ando mly d istr ibuted am on g t he p art icip ant s. D ependin g on th e con ditio n th ey were in , p articipant s were asked to describe a situatio n in which th ey felt either in tense disapp ointm en t, sad ness, frustration, an ger, or regret. Pa rticipants were a sked to an swer the qu estion : ``M y [emo tion term ] was cau sed by: want in g to get r id o f or avoid som ething painfu l or wan tin g t o get or keep som ething p leasurable’ ’ . P articipan ts cou ld an swer on a 9-p oint scale labelled avoiding something painful (1) t o wanting something pleasurable (9).

Results and Discussion Scores on th e ma in depend en t va r iable (typ e of negative outcome) were entered int o an A N OVA, the single fa cto r b eing emotion . A nalysis revealed a signi® cant ma in effect d ue to em otio n [F(4,105) = 4.98, P < .001]. T he m ea n appra isal scores of t he ® ve target emo tions are shown in Table 3. 1 2 A con trast an alysis, in which disapp ointm ent was cont ra sted aga inst t he o ther four emo tions, revealed t hat disapp ointm ent is more strongly associated with wan ting so methin g p leasu rable t han th e oth er em ot io ns [t(100) = 4.03, P < .001]. Appr aisal r atings fo r disapp ointm ent were high er th an th e ratin gs for t he o ther target emotion s (see Table 3). T h is im plies th at disappo intm ent was m or e closely associated with wan ting som ething p leasurable t han were sadn ess, an ger, frustration, a nd regret. M oreover, o nly disappoint ment app raisal ratings differed sign i® can tly from t he n eu tra l m idpo in t of t he scale, imp lyin g th at d isapp oint ment was associated mo re strongly with o ne type o f n egative o utcom e, th at is, t he absence o f a po sitive o utcom e [t(21) = 5.40, P < .001]. Sa dn ess, anger,

12 Th e mea n in t en sity ratin gs fo r d isap p oin t men t, sad ness, a nger, frustratio n, a n d regr et were 6.81, 7.73, 7.64, 7.50, 6.48, respect ively. Th e o nly sign i® ca n t d ifferences were th o se b etween sad ness a nd regret , a nd a n ger a nd regret (Ps < .05). N ot e, h owever, th at desp ite it s st atist ica l sign i® ca n ce, th e difference b etween the lea st in ten se a n d th e m o st int ense em o tio n is sm a ll in ab so lu te t erms.


139

TABLE 3 Mean Appraisal Ratings for Each of the Five Emotions (Study 2) Emotions Appraisal Avoidin g p a inful vs. Wan ting p lea sur able

Disappointment

Sadness

Anger

Frustration

Regret

7.45 c

5.41 a , b

4.27 a

5.73 b

5.95 b

Note: Scores cou ld ra nge from (1) avoiding something painful t o (9) wanting something pleasurable. M ean s wit h in a row n o t sh a r in g a co m m o n sup erscr ip t d iffer sign i® ca n tly (P < .05).

frustratio n, and regret were not differentially a ssociated with one type of negative o utcom e (ts < 2.04, n.s.). T hu s, the result s of this study sup por t our p red iction s and replicate th e ® n dings of Stu dy 1. D isappo in tm en t appear s to be m ore caused by th e absence o f a p ositive ou tcom e t han by t he presence o f a n egative outcome. M o reover, results sh owed t hat disappo intm ent can b e distingu ished from the related emo tions o f sad ness, an ger, frustration, and regret on the ba sis of typ e o f n egative o utcom e. I n Study 3 we attempt to extend these ® nd in gs usin g a different approach.

STUDY 3 In t he t wo previou s studies par ticipants were asked t o recall an instan ce of a ta rget emo tion an d to r ate the extent to which this emo tion was ca used by a speci® c typ e of negative outcom e. In t he p resent stu dy we used a scenario m eth od , in which we m an ipu lated typ e of negative ou tcom e and asked fo r in tensity ratings of t he target emot io ns. Par ticipan ts were confronted with eith er t he absence of a po sitive out co me or th e p resence of a negative ou tco m e. We exp ect th at disappo in tm en t ratin gs are h igher in t he fo rm er situation tha n in t he latter situatio n. F ur therm ore, we expect that in th e for m er situatio n disappo in tm en t r atings are high er t han are sadn ess, anger, frustratio n, and regret ratings.

Method Design and Participants. St udy 3 ha d a two -gro up b etween-sub jects design (Positive-Ab sence vs. N egative-Presence). Stud en ts at the U niversity of Sussex, U K (N = 40) par ticipated in th is stu dy o n a volunt ary basis. T here were 20 p art icip ant s per co nditio n.

140


Procedure. Q uestion naires were r ando mly d istr ibuted am on g t he p art icip ant s. Part icipan ts in the Positive-Ab sence co ndition (PA ) rea d t he fo llowin g story: It is th e tim e o f yea r when yo u h ave to ® ll in yo u r tax fo rm. You h ave so rted o ut all yo u r ® n an cia l in s an d o u ts. You h ave ® lled in yo u r tax fo rm to th e b est o f yo u r ab ilit ies. You h ave d o n e th is b efo re, so it is n o t too b ig a hassle. A fter re-exa m in ing yo ur tax fo rm an d sign in g it , yo u exp ect that yo u get a tax reb ate o f £150. A fter a while yo u receive a let ter fro m th e tax o f® ce. D u e to a n ew ta x rule, so m e o f yo u r ca lcu lation s were n o t valid a n d so yo u will n ot receive th e exp ected reb ate o f £150. You d o n o t h ave to p ay any ext ra ta xes. H ow wou ld yo u feel ab o u t this o u tco m e?

Pa rticipants in th e N egative-Presence con dition (N P) read th e following stor y: It is th e tim e o f yea r when yo u h ave to ® ll in yo u r tax fo rm. You h ave so rted o ut all yo u r ® n an cia l in s an d o u ts. You h ave ® lled in yo u r tax fo rm to th e b est o f yo u r ab ilit ies. You h ave d o n e th is b efo re, so it is n o t too b ig a hassle. A fter re-exa m in ing yo u r tax fo rm an d signin g it , yo u exp ect that yo u won’t h ave to p ay any extr a tax, bu t yo u also exp ect that yo u won’t get a ny tax reb ate. A ft er a while yo u receive a let ter fro m th e tax o f® ce. D u e to a n ew tax ru le, so m e of yo ur ca lcu latio n s were n o t va lid an d n o w yo u h ave to p ay £150 in ext ra ta xes. H ow wou ld yo u feel ab o u t this o u tco m e?

After reading the story par ticipa nts were ® rst asked to in dicate h ow n egative their feelings would be in general after t he ou tco m e. T his r ating was d one on a 9-p oin t scale wit h end -p oints labelled not at all negative (1) to very negative (9). T his questio n was asked in ord er to give par ticipan ts an o ppor tun ity to give a gen eral evaluation of the situ atio n. N ext, p art icipant s were asked to give int en sity r atin gs of mo re speci® c emo tions, (i.e. d isapp oint ment , regret, fru str ation , sad ness, and anger). T hese questio ns h ad the following word ing: ``H ow much [em otion term ] wou ld yo u exp erience after t his ou tcom e?’ ’ R atings o f the speci® c em o tio ns were a ll m ade on 9-poin t scales with en dpo in ts labelled none (1) t o a lot (9).

Results and Discussion F irst, we exam ined negative feelings in genera l that wou ld b e exp erienced. T h e results showed that negative feelings were mo re intense in t he N P con dition (M = 7.05) t han in t he PA cond ition [(M = 6.10), t(38) = 2.71, P < .01]. T he m ean intensity ratings fo r t he ® ve target emo tions are repor ted in Table 4. I ntensity r atings o f th ese em ot io ns were en tered int o a n A N OVA , usin g cond ition as a between-subjects fa cto r and


141

emot io n as a with in-subjects factor. An alyses revealed a marginally sign i® can t m ain effect o f condit io n [F(1,38) = 3.52, P < .07 ], an d a sign i® can t two -way interactio n between co nd ition and em ot io n [F(4,35) = 5.58, P < .001]. U nivariate tests revealed t hat d isapp oint ment was m o re intense in th e PA co nd itio n th an in th e N P con dition [F(1,38) = 4.40, P < .05], whereas, anger and frustr ation were more intense in th e N P cond it io n th an in th e PA cond it io n (Fs > 7.00, Ps < .05). N o d ifferen ces between con dition s were fou nd for sad ness a nd regret (Fs < 2.14, n.s.). We also p redicted t hat disapp ointm ent wo uld b e the d om in ant em o tion in t he PA con dition . Plann ed com par ison s co n® r m ed t his p redictio n an d revealed th at disappo in tm en t is m ore int en se in the PA co nd ition th an th e oth er target emo tions (all Ps < .005). Alth ough disapp oin tm en t was mo re in tense in th e N P con ditio n than were sad ness and regret (Ps < .005), disappo in tm en t was th e o nly emot io n th at was mo re intense in th e PA cond it io n tha n in th e N P co nd it io n. T hu s, the results of Study 3 su pp ort our predictio ns an d replicate an d extend the ® n dings o f Stu dies 1 and 2. D isappoin tment is m o re intense aft er experiencing th e absence of a p ositive o utcom e tha n a fter experiencing the presence o f a n egative o ut com e. M oreover, d isappoint ment is mo re in tense t han sad ness, anger, fru str atio n, and regret aft er experiencing th e absence of a po sitive ou tco me. H owever, it sh ould be no ted th at t his study was a simulation (i.e. p art icipant s were inferr in g what t hey wo uld feel if they were in th ese situatio ns), a nd th ere is no 100% guar ant ee that th eir in feren ces were fully accurate. To overcom e t his prob lem we d esigned an add it io nal exp erim en t in which on-line em ot io nal reactions towa rd the two typ es of negative o ut com es were assessed.

TABLE 4 Mean Intensity Ratings per Condition for Each of the Five Emotions (Study 3) Emotions Outcome PA NP

Disappointment c(a )

7.25 6.30 b (b )

Sadness a (a )

3.15 4.10 a (a )

Anger b (a )

4.95 6.60b (b )

Frustration b (a )

5.25 7.00b (b )

Regret 3.10 a (a ) 3.85 a (a )

Note: PA, p o sitive ab sen ce; N P, negative presence. A high er score ind ica tes a h igh er int en sit y o f th e emo tio n . M ea n s within t he sa me row n ot sharin g a co mm o n ® rst su perscrip t d iffer sign i® can tly (P < .005). M ea ns within t he same colu m n n ot sharin g a co mm on seco n d (between pa ren theses) superscript d iffer sign i® can tly (P < .005).

142


STUDY 4 I n t he ® rst t wo stu dies we used a retrospective meth od, in which pa rticip ants recalled actu al events in wh ich they experienced par ticula r em ot io ns an d m easu res concerning typ e of negative o utcom e were obt ain ed . In Study 3 we used a scenario m ethod , in which we ma nipulated type of n egative ou tcom e an d asked for int en sity ratin gs of th e target em ot io ns. I n the present study we fo cu s on on-line emotional reactions. Pa rticipants in th is experim en tal study were confron ted with eith er th e absence of a (rea l) p ositive ou tco m e or t he p resence of a (rea l) negative o utcom e. T his was d one by using a fra ming procedure in wh ich an identica l o utcom e was p resented either in win-term s or in loss-term s. T his is bo th a com m on and effective procedure u sed in t he ® eld of behavio ur al decision ma king (see e.g. K ahneman & T versky, 1982). We expect to replicate our previo us ® nding, th at is, we expect that disappo in tment r atin gs are h igher in t he fo rm er situatio n t han in th e latter. F u rth erm ore, we expect th at in th e for m er situ ation d isapp ointm ent ratings are high er th an a re sad ness, an ger, frustratio n, and regret ratings.

Method Design and Participants. Study 4 h ad a t wo-group between-subjects d esign (Positive-A bsence vs. N egative-Presence). St udent s at N ijm egen U niversity (N = 40) p art icipated in t his study. T h ere were 20 par ticipan ts in ea ch cond it io n. 1 3 T h is study wa s a dm in istr ated dur ing a large exp erim en tal session . Pa rticipant s were pa id 10 D ut ch gu ild ers (approximately $5.00) for their par ticipation in the large exp erimenta l session , an d, as p art o f the present stu dy, cou ld ga in an ad dition al 5 D ut ch guild ers. Procedure. A ll par ticipa nts were pro mised 10 D u tch guild ers for t heir p art icip ation . H owever, par ticipa nts in th e N egative-Presence con dition were end owed wit h 5 D utch gu ilders extr a. T hey were given 15 G uilders at th e sta rt o f t he exp erim ent a nd were t old th at th ey would p lay a ga me at th e end of th e session in which they co uld win o r lose 5 D utch gu ilders or th ey cou ld win o r lose n o m oney. A t th e end o f th e session qu estio nn aires were ran dom ly distribut ed am on g t he par ticipa nts. Par ticipa nts in th e N egative-Presence co ndition read t he followin g in str uction (T he PositiveA bsence co ndition is shown in b rackets): 13

In o rder to o bt a in 20 p a rticip a n ts in t he t wo releva nt co nd it ion s th e gam e was p layed with 84 p a rticip a n ts. O n ly t he da ta of th ose p ar ticip a n ts who lo st t he ga m e were in clu ded in t his stu dy.


143

We are ab o ut to play a ga m e in which m o n ey ca n b e lo st [wo n ]. T h is resea rch is co n cer n ed with yo u r ch oices an d yo u ca n lo se [win ] rea l m on ey. It is ther efo re d ifferen t fro m m o st o th er resea rch that is co n cer n ed with hyp o thet ica l ch oices an d co n seq u en ces. [T h e m o n ey yo u ca n win is in d ep en d en t fro m the 10 D u tch gu ild ers that yo u receive for yo u r particip atio n in the whole sessio n , th is m o n ey yo u get anyway.] H o w d o es th e ga m e work? In a m om en t yo u m ay th ro w a d ie an d if yo u th row a lo sing [win n in g] nu m b er yo u lo se [w in ] 5 D utch gu ild er s. If yo u th ro w o n e o f th e o ther nu mb er s yo u lo se [win ] n othin g. You may th row th e d ie yo u rself an d also ch oo se th ree nu m b er s with which yo u lo se [w in ] m o n ey an d th ree nu m b er s wit h which yo u lo se [win ] noth in g.

Aft er p art icipan ts read the in struction an d ® lled in th e three winnin g o r lo sing num bers the exp erim en ter ca me alon g with a die an d t he gam e was played. Par ticipa nts were p aid imm ediately according to the result of th eir throw. For p articipant s in th e Positive-Absence con dition th is m eant th at they did not win the extra 5 gu ilders, a nd for t hose in the N egative-Presence cond it io n it m ean t that t hey lost 5 guilders. (Because pa rticipant s in th e N egative-Presence condit io n h ad received 15 guild ers at t he b egin ning of the la rger experim en tal session, a ll p ar ticipa nts left with 10 gu ilders mo re tha n when th ey entered t he experiment.) N ext, par ticipan ts were ® rst asked to in dicate h ow negative their feelings are in genera l after the o utcom e. T his rating wa s do ne on a 9-poin t scale with end-poin ts labelled not at all (1) to very much (9). T his q uestion was asked in order to give par ticipants an op por tun it y t o give a gen eral affective evaluation of the situation. N ext, par ticipa nts were asked to give in tensity ratings of m or e speci® c em otio ns, nam ely, disapp ointm ent, regret, fru str atio n, sad ness, and anger. T hese qu estions ha d t he following word in g: ``H ow m uch [em otion term] are yo u exp eriencing aft er this ou tcom e?’ ’ R atings of th e speci® c em otion s were all m ade o n 9-po in t scales wit h en dpo in ts labelled none (1) to very much (9).

Results and Discussion F irst, we examined negative feelings in general that were experienced. Results showed that negative feelings in general were not m ore intense in the N P condition (M = 4.00) than in the PA condition [(M = 3.7), t(38) < 1, n.s.]. T he mean intensity ratings for the ® ve target emotions are reported in Table 5. Intensity ratings of these emotions were entered into an AN OVA, using condition as a between-subjects factor and emotion as a within-subjects factor. Analyses revealed only a signi® cant two-way interaction between condition and emotion [F(4,35) = 3.15, P < .05]. U nivariate tests revealed th at, as p red icted, disapp ointm ent wa s mo re in tense in th e PA cond it io n tha n in the N P co nditio n [F(1,38) = 6.05.,

144

van DIJK, ZEELENBERG, van der PLIGT TABLE 5 Mean Intensity Ratings per Condition for Each of the Five Emotions (Study 4) Emotions Outcome PA NP

Disappointment a (a )

4.75 3.40 a (b )

Sadness d (a )

1.35 1.70 c(a )

Anger d (a )

1.40 1.65c (a )

Frustration c(a )

2.25 2.60 a ,b (a )

Regret 3.10 b(a ) 1.95 b, c(a )

Note: PA , po sit ive ab sence; N P, negative presen ce. A h igher sco re in d icates a h igh er int ensity o f t he em ot ion . M ea n s within th e sa me row n o t sh a ring a co mm on ® rst superscript differ sign i® can t ly (P < .005). M ea n s within th e sam e co lu mn n ot sh a ring a co m mo n seco nd (between pa rent heses) sup erscrip t differ signi® ca nt ly (P < .05).

P < .05]. C oncern in g the oth er emotion s n o signi® cant differences between th e t wo co nd ition s were fou nd (Fs < 3.60, n.s.). We a lso p redicted that disapp ointm ent is th e d om in ant em o tion in th e PA cond it io n. P lann ed com p ariso ns con® r med th is p redictio n and revealed t hat disapp ointm ent wa s mo re intense in t he PA co nditio n t han were th e ot her target em otio ns (Ps < .005). Altho ugh d isappoint ment was mo re in ten se in th e N P co nd it io n th an were sad ness, anger and regret (Ps < .05), d isapp oint ment was the on ly emotion t hat wa s m ore int en se in th e PA con dition th an in t he N P cond it io n. A possible reason why disapp ointm ent ratin gs were also relatively h igh in the N P conditio n co uld be th at so me par ticipa nts in t his co nd it ion did no t un eq uivocally perceive t he situation (i.e. losin g mo ney) a s th e p resence of a n egative ou tco m e. T hese p art icipant s m ay h ave been p art ly respo nding to th e absence o f a positive ou tco me, t hat is, n ot h aving mo ney. 1 4 T his issue may b e resolved by in clud in g in fut ure research qu estion s concerning h ow par ticipants appr aise t he situatio n with which th ey are con fron ted. T h us, th e result s of Stu dy 4 sup por t ou r predictions and replicate and extend th e ® nd ings o f Stu dies 1, 2, a nd 3. D isappo intm ent was m ore int en se aft er experiencing th e absence o f a positive ou tco m e t han aft er experiencing the p resence of a negative out com e. M o reover, disapp ointm en t was m ore in ten se tha n sad ness, a nger, fru stratio n, a nd regret aft er experiencing t he absence of a p ositive ou tcom e.

14 Th is cou ld a lso a cco un t for th e relative h igh d isap po in tm ent rat ings in t he N P co n dit io n o f Study 3.


145

GENERAL DISCUSSION In t he present p aper we distingu ished between two t ypes of negative ou tcomes, th e absence of a p ositive ou tco m e and the presence o f a negative ou tco m e. I t is a rgu ed that disappoint ment , because of its close lin k with ho pe, desire, an d pro mise, is m ore ca used by t he fo r mer typ e of ou tcom e tha n th e latt er. M o reover, disappo intm ent can be distin guish ed on t he ba sis of type o f n egative out co me from related emo tions, such as sadness, anger, frustratio n, an d regret. D isappo in tm en t is caused m ore by t he absence o f a po sitive out com e th an t hese oth er em otion s. R esults of four stu dies u sing different m etho dologies p rovide co nvergen t su pp ort fo r th ese p red iction s. T hese results also su pp ort th e explicit distinctio n between being disp leased abou t t he d iscon® rm atio n of the pro spect o f a d esirable event (i.e. disappoin tm en t) an d b eing displeased abo ut the co n® rm atio n o f th e pro sp ect of a und esirable event (i.e. fear s-con ® r m ed ) ma de by O rto ny et al. (1988). T he d e® nition s of disapp ointm ent describ ed ear lier in t his paper suggest that disappo intm ent is p rima rily experienced in a situation in which so methin g p ositive wa s expected but d id n ot o ccur. T h is seems closely linked with h ope, desire, a nd p rom ise. It sho uld be not ed , h owever, that, a lt hou gh exp ectatio ns, hop e, d esire, an d prom ise are related, th ey are n ot syno nym ous. Fo r examp le, one m ay ho pe for som ething with out expecting it to happen. I n o ur view d isapp ointm ent is the result of expectatio ns that were un ful® lled, and were in it ially desired o r hop ed for. A s Shan d (1914, p. 487) stated ``D isappoint ment . . . im plies that we h ave hitherto been ho peful of the issue, if no t con® dent’ ’ . Altho ugh disappo intm ent wa s t he ma in emo tion u nd er investigation , we also investigated th e relatio n between sa dness, a nger, fru stratio n, an d regret and t ype o f negative o utcom e. T his enabled us to com pa re ou r results con cern in g this relation wit h the predictio ns o f several em o tio n research ers. According to ma ny th eo rists, sa dn ess is mo re oft en associated with the absence o f a positive o utcom e, whereas anger, frustration , an d regret are m ore oft en a ssociated with bot h types of n egative o ut com es (F rijda, 1986; H iggin s, 1989; Roseman , 1984; Roseman et al., 1990, 1996). I n ou r stu dies we fo und that a nger, fru stration , an d regret a re associated with bo th t ypes o f negative o utcom es. C o ncerning sadn ess we did not ® n d evid ence th at t his emotio n is stro ngly associated with the absence of so m eth ing positive. A possible exp lan ation for the la ck of a strong relatio n between sad ness and th e absence o f a po sitive ou tcom e could be a pro cess o f psycho logica l ``r epa ckaging’ ’ . 1 5 Peop le ca n tra nsfo r m an appr aisa l o f the absence of p ositive ou tcom e int o an app raisa l o f th e 15

We th an k an a n onym ou s reviewer for brin ging th is p o ssib ilit y to o ur a tten tio n .

146


p resence o f a n egative out co me. Fo r exa mp le, one co uld feel sad because of th e d eath of a loved on e. D epend in g on on e’s focus, t his experience cou ld b e app raised as th e absence of so meth ing p ositive (e.g. when t he focus is on th e absence of a loved one) o r as the p resence of a n egative outcome (e.g. when t he fo cus is on the presence of death ). T he focus on eith er the absence o f a po sitive out co me or t he presence of a negative out com e m ay chan ge in tim e. T h e result o f th ese different focuses co uld help to exp lain why sadn ess, in ou r studies, was n ot associated with o ne typ e of n egative o utcom e. Perhaps fur ther re® nemen ts in th e m ea su rem en t of appr aisals cou ld h elp to clar ify this issue. T h e results o f t he present stu dies, however, im ply th at d isappoint ment is less affected by the p rocess o f psychological ``rep ack agin g’ ’ . In ou r fou r stu dies disapp ointm en t was clearly a ssociated with o ne typ e of negative o utcom e, that is, the absence o f a positive o utcom e. U ntil recently very litt le em otion research h as focused o n disapp ointm en t. H owever, in t he ® eld of decision ma king disappo in tm en t is con sidered t o be a n im po rta nt emo tion. D ecision research ers (e.g. Bell, 1985; L oo mes & Sugden, 1986) have stressed th e n otio n th at d ecision m akers a nticipate d isapp oint ment an d take it into a ccoun t when m aking d ecisions. If peo ple d o an ticipate disappo in tm ent , research o n disapp ointm en t co uld be h elpful in un derstan din g how this emot io n affects d ecision p rocesses. F or in stan ce, th e fa ct that disapp ointm en t is p rim arily associated wit h the absence of positive o utcom es, together with p eo ple’s tend en cy to avoid d isapp oint ment, cou ld help t o explain why p eo ple tend to b e cau tious in their decision m aking when d ealing with po sitive o utcom es. K a hneman and T versky (1982), for instance, sh owed t hat p eo ple are gen erally r isk-averse an d that this tend en cy is stron ger in th e dom ain of ga ins (where t ype of negative ou tco me is the absence of som eth ing p ositive) t han in the do ma in of losses. O ne reason for this tend en cy co uld be that tak ing r isks in t he d om ain o f gain s is a ssociated with m ore ant icipated disappoin tm en t. T his in creased anticipation of d isapp oint ment could m ake p eo ple m ore risk-averse. W h en p eop le are r isk-averse they are likely to get what they expect and therefore do not r un th e risk o f becom ing disapp ointed. H owever, disappoin tm en t is n ot o nly a n u np leasant emo tion t hat is an ticipated o r avoided. T he experience of disappo in tm en t cou ld also h ave a b right sid e. Shan d (1914, p. 489) stressed t he u seful fun ction of d isapp oint ment in desire: ``D isappoint ment , in it s after-effect o n desire, always tend s to counteract the excesses of h ope, to evoke anxious questio nings, to sup press all co n® d en ce th at is no t well-found ed; so far as it is checked and bala nced by ho pe itself’ ’ . Or as St anley (cited in Shan d, 1914, p. 488) stated, ``D isapp oint ment tu rns life from false dream s t o stern realit ies: I t prom p ts a n investigation of causes and arou ses cognition to a


147

full u nderstan ding o f t he situ ation . H o pe th ereby, becom es m ore rationa l and rea lisable’ ’ . M a nuscrip t received 18 Ju ly 1997 R evised ma nuscrip t received 19 Ju ly 1998

REFERENCES Ar no ld , M .B. (1960). Emotion and personality. N ew Yo rk: Co lu mb ia. Bell, D .E . (1982). R egret in d ecisio n m a king un der un cer ta in ty. Operations Research, 30, 961±981. Bell, D .E . (1985). D isap p o in t m en t in d ecisio n m a k in g u n d er u n cer t a in t y. Operations Research, 33, 1±27. F rijda , N.H . (1986). The emotions, C am b rid ge, U K : C a m bridge U n iver sity P ress. F rijda , N.H . (1993). T he p la ce of app ra isa l in emo tio n. Cognition and Emotion, 7, 357±387. H iggins, E.T. (1989). Self-discrepa ncy t heo ry: what p att er ns of self-belief ca u se p eop le to su ffer? Advances in Experimental Social Psychology, 22, 93±136. Inm a n , J.J., D yer, J.S., & Jia , J. (1997). A gen era lized ut ility mo d el of d isa pp o int men t a nd regret effects o n p o st-cho ice va lu atio n. Marketing Science, 16, 97±111. K a h nem an , D., & T ver sky, A . (1982). T he p sych o logy of p references. Scienti® c American, 246, 136±142. Levin e, L .J. (1996). T he a n ato my o f d isa pp oin t ment : A n atu ra list ic test o f a ppr aisal m od els of sadn ess, a nger, a n d ho p e. Cognition and Emotion, 10, 337±359. Lo om es, G ., & Sugd en , R . (1982) R egret T heo ry: A n a lt erna tive th eo ry o f r atio n al cho ice un der u n cert ain ty. Economic Journal, 92, 805±824. Lo om es, G., & Sugden, R . (1984). T he im p ort a nce o f what migh t h ave b een. In O. H a gen a nd F. Wen st rop (Ed s.), Progress in utility and risk theory (pp. 219±235). D o rdrecht: D . R eid el. Lo o mes, G ., & Sugden , R . (1986). D isap p oint m ent a n d dyn a m ic in co nsistency in cho ice un der u n cert ain ty. Review of Economic Studies, 53, 271±282. M owrer, O.H . (1960). Learning theory and behavior. N ew Yo rk: W iley. Or to ny, A ., C lore, G .L., & C ollins, A . (1988). The cognitive structure of emotions. N ew York : C a mb rid ge U n iver sit y Press. Pa rkin so n, B., & M a nstead, A.S.R . (1992). App ra isa l a s a ca use of emo tio n . In M .S. C la rk (Ed .), Review of personality and social psychology (Vol. 13, p p. 122±149). N ewbury Park , C A : Sage. Ro sema n , I.J. (1984). C ognit ive d eterm ina n ts o f em ot io ns: A stru ctu ra l t heo ry. In P. Shaver (Ed .), Review of personality and social psychology (Vol, 5, pp. 11±36). Bever ly H ills, C A : Sage. Ro sema n , I.J. (1991). Ap pra isa l d eterm in an t s o f d iscrete em ot io ns. Cognition and Emotion, 5, 161±200. Ro sema n , I.J., A n ton io u , A .A., & Jose, P.E . (1996). A pp ra isa l d eter min a n ts of emo tion s: C on str ucting a mo re a ccu rate a nd com p rehensive t heo ry. Cognition and Em otion, 10, 241±277. Ro sem a n , I.J., Spind el, M .S., & Jose, P.E . (1990). A pp ra isa l of em ot io n-elicitin g even ts: Testin g a th eo ry of discrete emo tio n s. Journal of Personality and Social Psychology, 59, 899±915. Scherer, K .R . (1984). Em o tio n a s a mu lt i-co mp on ent process. In P. Shaver (Ed .), Review of personality and social psychology: Vol: 5. Emotions, relationships, and health (pp. 37±63). N ewbury Park, C A : Sage.

148


Scherer, K .R . (1993). Studying th e em ot ion -an teced en t app ra isa l p rocess: A n exp ert syst em ap proa ch. Cognition and Emotion, 7, 325±355. Shan d , A .F. (1914). The foundations of character. L on d on : M a cm illa n . Shep p erd , J.A ., O u ellett e, J.A ., & F ern an d ez, J.K . (1996). A ba ndo n in g un realistic op tim ism : Perfo rm a nce estim at es a n d th e t em p ora l p roximity of self-releva n t feed b a ck . Journal of Personality and Social Psychology, 70, 844±855. Sm it h , C.A ., & E llswo rth , P.C . (1985). Patter ns o f cogn itive a pp ra isa l in em o tio n . Journal of Personality and Social Psychology, 48, 813±838. Sm it h , C .A . & E llswort h, P.C . (1987). Pat terns o f ap p ra isa l a n d emo tio n rela ted to t ak in g a n exa m . Journal of Personality and Social Psychology, 52, 475±488. va n D ijk , W.W., & va n d er Pligt , J. (1997). T he im pa ct o f p robab ilit y a n d ma gn it ud e o f o ut co m e on d isa pp oint men t a n d elation . Organizational Behavior and Human Decision Processes, 69, 277±284. va n D ijk , W.W., va n der Pligt, J., & Zeelen berg, M . (1998a ). Effort invested in vain: The impact of effort on the intensity of disappointment and regret. M a nu script submit ted for p ublication . va n D ijk , W.W., va n d er P ligt , J., & Z eelen b erg, M . (1998b ). Investigating the appraisal patterns of outcome-related and person-related disappointment. M a nu scrip t su bm it ted for p ublication . Webster’ s Third New International Dictionary (3rd ed .). (1981). Sprin g® eld, M A : M erriam Webster. Wein er, B., Ru ssel, D ., & Lerm a n , D . (1979). Th e cognit io n-em o tio n p rocess in a ch ievem en trelated co nt exts. Journal of Personality and Social Psychology, 37, 1211±1220. Z eelen b erg, M ., & Bea ttie, J. (1997). C o n sequ ences o f regret aver sion : 2. A dd itio n a l eviden ce fo r effects o f feedba ck o n d ecisio n ma k in g. Organizational Behavior and Human Decision Processes, 72, 63±78. Z eelen b erg, M ., Bea ttie, J., van d er Pligt , J., & d e Vries, N.K . (1996). Co n sequ ences of regr et aver sio n : Effects of exp ected feedba ck on risk y d ecision m a kin g. Organizational Behavior and Human Decision Processes, 65, 148±158. Z eelen b erg, M ., va n D ijk , W.W., M a n st ea d, A .S.R ., & van der P ligt, J. (1998a). Th e experien ce o f regret an d disapp o in tm en t. Cognition and Emotion, 12, 221±230. Z eelen b erg, M ., van D ijk , W.W., va n d er Pligt , J., M a n st ead , A .S.R ., va n Em pelen , P., & R ein d er ma n , D. (1998b ). E mo tio n al rea ctio n s t o th e o ut com es o f decisio n s: Th e role o f cou n terfactu a l t ho u gh t in t he experien ce o f regret a n d disa pp o in tm en t. Organizational Behavior and Human Decision Processes, 75, 117±141.