You are on page 1of 11

Fading Optimism in Products: Temporal Changes in Expectations about Performance Author(s): Ashwani Monga and Michael J.

Houston Reviewed work(s): Source: Journal of Marketing Research, Vol. 43, No. 4 (Nov., 2006), pp. 654-663 Published by: American Marketing Association Stable URL: http://www.jstor.org/stable/30162438 . Accessed: 03/02/2013 07:54
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

American Marketing Association is collaborating with JSTOR to digitize, preserve and extend access to Journal of Marketing Research.

http://www.jstor.org

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

ASHWANI MONGA and MICHAEL J. HOUSTON*


from set of a that The authorsdemonstrate choosingone product about the chosen alternatives can change expectations competing abouttheproduct's consumers becomeoptimistic can such that product fadeaway.Infive can and experiments, performance, thisoptimism then in is of this showthat phenomenon fading theauthors optimism products and and across different robust categories experimental settings product and the attitude toward product is moderated prior category ambiguity by oftheproduct's performance.

in Temporal Optimism Products: Fading About Changes inExpectations Performance


a about good are Product expectationspredictions how Oliver et will be performance (Bouldingal. 1993; product's choice satisfacboth 1997). can They influence product and to if A tion. product bechosen itis expectedperform may if to which lead satisfaction well 1997, (Oliver p. 69), may meets theperformance or exceeds (Oliver expectations how change 1980).Thisresearch explores expectations isrevealed. but after choice before performance a revealed is frequently Product during performance the For choice. example, performance after (e.g., period is over of fabric softness) a shirt purchased theInternet after after a revealedfew choice, itarrives mail. by days the quality) (e.g., Similarly, performance photograph ofa after is revealed camera not immediately choice disposable and takes the butafter consumer pictures has thefilm the herein we The processed. research report addresses folofa chosen the Before performance prodlowing question: about can is uct revealed, consumers' expectations product to the (from prechoicethe change time performance over the remains even product as knowledge stage), postchoice but of Two same? streamsliterature plausible consuggest a and this We resolve conflict present answers. first flicting We inproducts. then about present optimism theory fading to cameras establish that three experimentsusedisposable that effect the fading-optimismandtwoexperiments use The the drinks replicate effect. soft-drink to soft experiattiand also ments examine ambiguity prior performance and out as tude moderatorsrule analternative explanation.
CONFLICTINGPREDICTIONS ABOUT TEMPORAL CHANGESIN EXPECTATIONS

from a chooses product a setofcoma Afterconsumer about his how alternatives,might orher expectations peting reduction on Research dissonance the sugchange? product on but should that gests expectations increase, research decrease. that should suggests they management strategic on based a resolution onresearchfading We offer optimism.
ShouldIncrease DissonanceReduction: After Expectations Choice

choice an who a Imagine consumer makes irrevocable A andB; both two between competing products products, After and some have aspects. negative aspects some positive those Product theconsumer A, owning regrets choosing todissonance toB. are of aspects A that inferior According attitude -behavior 1956;Festinger 1957), (Brehm theory attithe occurs negative inconsistency because consumer's with A the toward inferior of isinconsistenthis tude aspects the motivatescontoward behavior A.This her ofMarketing, or favorable *Ashwani is anassistant Department professor, Monga is Because behavior ashwani. sumer resolve inconsistency. the at Antonio this of Texas San to ofBusiness, (e-mail: University College M.Grieve Chair isEcolab-Pierson MichaelHouston to for the monga@utsa.edu). J. irrevocable,consumer's preferenceA increases of Carlson ofMarketing, School inInternational Department Marketing, with line Consistent this inline behavior. attitude with bring mhouston@csom.umn. of Minnesota (e-mail: University Management, that demonstratesdesirability Brehm (1956) first ofreasoning, the dissertation on is This articlebased a doctoral by completed edu). Simithan after is ofa productgreater itis chosen before. thank The author. authors the of second under guidance the author the committee -Deborah the of members dissertation and other Rao, John, races of contexthorse (Knox Inkster inthe Akshay 1968) larly, on project. and their Hawkins -for feedback this and Eric Klinger, Douglas and (Rosenfeld, Kennedy, Giacalone games gumball ofOmNarasimhan, from suggestions has This article alsobenefited the are that demonstrated it opti1986), hasbeen people more JMR and anonymous reviewers. Chandy, the Rajesh than an after the mistic about outcome choosing option
Association American (c) 2006, Marketing 1547-7193 ISSN: 0022-2437 (electronic) (print), 654 Research Journal Marketing of XLIII Vol. 2006), (November 654-663

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

in Optimism Products Fading

655
show about H1:Expectations product performance an inverted U shape time arises after choice over suchthat optimism but subsequently fades away. More specifically, (a) > 1 (i.e.,when is distant)prechoice postchoice performance 2 and performance expectations, (b) postchoice(i.e.,when is imminent) 1 < postchoiceexpectations. Overview

of A before. similar dissonance-based bolsteringexpectafor tions been has 1997, 259). (Oliver p. suggestedproducts to research, therefore, expectations Accordingdissonance consumerstoreasincrease choice after because should try will that surethemselves thechosen product perform well-that that made right the decision. is, However, they on researchstrategic management the suggests opposite.
ShouldDecrease Management: Expectations Strategic Choice After

this were was Participants toldthat experiment being on conductedbehalf a company wastryingdeterof that to to and mine which three of cameras launch. parAll to According self-handicapping (Berglas theory disposable Jones sometimes impediments create toper- ticipants assured inthe they were that receive 1978), people end, would not for exam") reduce to formancewill study the three cameras seeitssample and ("I poten- oneofthe photographs. tial from a low Before used camera, photographs the would ("I scored disappointment badperformance participantsthe I In because didnotstudy"). linewith logic, give this a themsense its of performance. only and that strateTheinformation available form to was (2001) Kopalle Lehmann suggest consumers expectations a inanattempt reduce toreduce brand attribute that displayedthe ofthe x table was at top product expectations gically from the were It potential disappointment product. pagewhenever expectations measured.compared and Lehmann's research is not focus on three and (2001) Kopalle Click-n-Throw, cameras-Single-Clickle, Buy-nhow reduce choice onhow but three dimensions ofphotograph resoluexpectations after strategic Click-on quality: reduction with varies individual However, essence tion, traits. the color and were richness, color accuracy. Participants oftheir is research that are in that had the "expectationsdetermined, told anindependent service rated phototesting to future on 0 to To part, the by desire enhance satisfaction" (p.387). graphs a scalefrom ("worst") 100("best"). After a product, want consumers tobe satisfied ensure alternatives competing, camera that were each was choosing with In anattemptenhance it. to satisfaction, should rated on onedimension not theothers. but on The they high reduce expectations. prediction This is in for color and strategically their resolution, richness, color respective ratings direct conflict the with dissonance prediction. were 95 Single-Clickle; and theory 55,85, accuracy 55,50,and for 60for and 60 Buy-n-Click. Click-n-Throw; 50,and for 90, Resolution theConflict: in of FadingOptimism Products Onthe basis these of marked ratings, participants expectaOur resolution is predicatedthe that on idea though dis- tions times, filler and gaps to multiple with tasks time used sonance effects known occur after are to soon clear it their about The choice, is memory prior responses. purported notclear whether effects strategic the of reduction for will reason multiple measurements examining wasthat the occur after soon choice some later. or time Research fad- same on information sometimes additional provides insights. an Todemonstrate temporal that the in ing optimism provide answer. helps changesexpectations Confidence performance about oncertain drops were dueto some tasks as not unforeseen correlate time, of we the ofperforming time the or "moment oftruth," measured over for chosen and tasks, the expectations time both draws (Gilovich, andMedvec near Ifsome extraneous effects Kerr, 1993). Similarly, rejected products. temporal were as time exam of feedback students' all the should the show same approaches, optimism, operating, products pattern. fades and sen- However,the if hypothesized were effects operating, the near, away, as graduation draws college day but juniors, lower show with to inverted should for iors, not optimism respect U-shaped pattern emerge thechosen thesalaries willreceive graduation after but for rejected ones.2 they (Shepperd, product not the and leads Ouellette, Fernandez This 1996). literature toour about in Design We conceptualization fading optimismproducts. that reduction not does occur after soon suggest strategic The design a 3 (time: was 1, prechoice, postchoice choice rather when but is about later, performance tobe mixed 2) postchoicex 2 (focal chosen) product: rejected, revealed. other In ina fades words, optimism choice in away design which wasa within-subjectsand time factor focal when is imminent. experiments Three performance using product a between-subjects The dependent was factor. cameras this test core variable expectations photograph photograph of disposable quality was about For quality. idea.1 for we asked, your "In example, Single-Clickle, opinion, what overall of will quality photographs be produced by EXPERIMENT1 = "poor 9= (1 Single-Clickle?" quality," "excellent qualThis tests prediction the We about temporal ity"). adapted measure prior this from research on experiment the of there be et and sequence expectations. Specifically, should an expectations (Boulding al. 1993; Kopalle Lehmann increase expectations choice, in after followed a 2001). by decrease expectations performance tobe in when is about revealed.
1A11 assume such of conditions, as presence hypotheses dissonance and ofchoice alternatives, competing 1956). irrevocability (Brehm Given half century a of researchdissonance, not on wedo attempt toprovide further evidence itbut maintain for conditions inwhichis known it to only occur. then how evoked We study the fades optimism subsequently away. 2For the U shape. rejected products,expected isnot inverted pattern an From topostchoice should the (i.e., 1,expectations remain same prechoice = postchoice even decrease prechoice > postchoice prechoice 1)or (i.e., 1) ifparticipants disparage rejected to the try (Brehm From products 1956). 1to should (i.e., occur postchoice 1= 2,no postchoice postchoice change the of revelation chosen ofthe 2), postchoicebecause time performance isirrelevant toexpectations rejected ofthe product products.

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

656 Procedure

JOURNALOF MARKETING RESEARCH, NOVEMBER 2006 Figure1

businessstudents Forty-two undergraduate participated in exchange partial for coursecredit and a freedisposable camera.On Day 1, participants marked (i.e., expectations for cameras thebasis of theratings on prechoice) all three items from Then,as a filler task, provided. they completed an unrelated individual difference scale. Next,theychose thecamera that wouldliketo receive Day 3, knowon they wererandomly ing thatthechoice was final.Participants to the or cell.Halfofthe assigned either rejected thechosen marked the (i.e., 1) participants expectations postchoice for one cameratheychose,and halfmarked for expectations thetwo camerasthey rejected. Theythenleftfortheday. On Day 3, participants first were reminded attheendof that thestudy, wouldreceivethecamerathey had chosen they on Day 1, as well as its samplephotographs. marked They the (i.e., 2) expectations postchoice foreither chosenor the on werein. (Day rejected products, depending thecell they 1 questionnaires uniqueidentification had based numbers, on whichDay 3 questionnaires werecustomized each for At was debriefed participant.) the end, each participant aboutthestudy and givena Kodak disposable camerafor participation. Results Foreachparticipant to cell, assigned therejected product we averagedexpectations the two rejectedproducts. of identical We results.) conducted (Separate analyses yielded a repeatedmeasuresanalysisof variance(ANOVA) in whichexpectations thedependent was time(prevariable, choice,postchoice postchoice was thewithin-subjects 1, 2) and variable, focalproduct chosen) independent (rejected, was the between-subjects variable.Figure 1 independent the presents results. Notsurprisingly, for were expectations higher thechosen thanfortherejected as product products, indicated the by maineffect focal product of (F(1, 40) = 16.1,p < .001). Morepertinent H1,themultivariate revealed main to test a effect time(Wilks'X = .80; F(2, 39) = 4.9,p < .05) that of was driven thechosen-product rather thantherejectedby x as evidenced thetime focal-product condition, product by interaction (Wilks'X= .78; F(2, 39) = 5.7,p < .01). Planned contrasts the for supported inverted U-shaped prediction the chosenproduct; from increased to expectations prechoice 1 (M = 6.1 versus F(1, 40) = 17.0,p < .001) 7.2; postchoice andthen decreased from 1 2 postchoice to postchoice (M = 7.2 versus6.3; F(1, 40) = 13.9,p < .01). In contrast, the showedno expectation rejected products changesbetween and 1 5.0; F(1, 40) < prechoice postchoice (M = 5.1 versus 1 2 1,p > .90) or between postchoice andpostchoice (M = 5.0 versus.5.1; 40) < 1,p > .80). F(1, EXPERIMENT2 1 an Experiment presents expectations sequencethatis consistent ourtheory. with it is However, showsthat fading correlated theimminence performance, that is with of not it causedbyit.To provide evidence thecausalmechanism, for 2 the of revelation that Experimenttests effect performance is manipulated be imminent to BecauseH2 (versus distant). proposes an increase in expectationsfrom pre- to it in of postchoice, is framed terms postchoice optimism. Consistent withHia (postchoice1 [distant] prechoice >

RESULTS FOR EXPERIMENT 1

7.37.1 6.96.76.5Performance 6.36.1 6.1

7.2

6.3

Product 5.9About
5.75.55.35.1 4.74.5_ Rejected Focal Product Time Prechoice Postchoice 1 Postchoice 2 Chosen 5.0 5.0 4.9Expectations 5.1

expectations), we expect optimism to arise in the distant

and withHib (postchoice [immi2 condition, consistent < 1 we expect nent] postchoice [distant] expectations), optimism to be lower in the imminent thanin the distant condition. arises the of H2:Postchoice optimism when time performance revelation distant fades is but it awaywhen is imminent. More > 0, < (a) specifically,PCODistant and(b) PC0Imininent
PC oDistant .

Overview We changedthebrandratings used in Experiment to 1 excludethepossibility ourresults tiedto a specific that are The ratings pattern. newonesare70, 40, and60 forSingleand Clickle;45, 65, and 60 forClick-n-Throw; 55, 40, and 75 for Buy-n-Click. Design x The designwas a 2 (time:prechoice, postchoice) 2 revelation: mixeddesign distant, imminent) (performance in which first the factor within was and subjects thesecond was between We as subjects. measured expectations we did in Experiment 1. Procedure business students in Fiftyundergraduate participated exchangeforpartialcourse creditand a freedisposable

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

in Fading Optimism Products

657 Overview Design and

camera. Day 1,participants expectations On marked (i.e., for on basis the cameras the of ratings prechoice) allthree table. Day8, they On chose camera they the that would a receivemonth Then, read performance later. they the revelation Theperformance-distant condition manipulation. read follows: as
At thetimeyou receive camera, will also the you receive some taken the photographs using camera you chose For if today. example, youchose Single-Clickle, takenusing you will be givensome photographs Please that camera receive Single-Clickle. note the you next month benew unused. will and These photographs would taken be a camera. using similar

The were to experimental and setting stimuli similarthat which used Experiment design wasa 2 we in 2. The used x2 of (time: prechoice, postchoice) (focus concern: prior future mixed in the choice, performance) design which first factor within was and second between was subsubjects the We aswe in 1 jects. measured expectations did Experiments and 2.
Procedure

The condition read performance-imminent asfollows:


After aredonewith survey, willreceive this you you somephotographs using camera chose taken the you For if today. example, youchoseSingle-Clickle, you willbe given somephotographs using taken SingleClickle. Pleasenote thecamera receive that next you month be newand unused. will Thesephotographs were taken a camera. using similar

business students in Fifty undergraduate participated for course credit a free and exchange partial disposable camera. Day 1,participants expectations On marked (i.e., for cameras the on basis the of ratings prechoice) allthree table. Day8,they On chose camera wanted. the Then, they were to condithey randomly assigned a focus-of-concern tion. Consistent H3, designed manipulations with we the not toinduce orientations prior toward choice versus general future to but concerns performance evoke specific regarding the The two. prior-choice condition asfollows: read
Nowtry think to about choice made. the the On you lastpage,youchoseone camera from three among available. Now that made options youhavealready final are doubts about the decision, youhaving your camera chose? other In are words, youconcerned you that choice mademight be thebestone? the not you Pleasewrite down in detail as your thoughts as much possible.

then marked theirexpectations Participants (i.e., for chosen each postchoice)the product. Finally, participant wasdebriefed given Kodak and a camera for disposable participation.
Results

Weconducted a repeated measures ANOVA which in condition as future-performance read follows wasthe time variable, (prechoice, The expectations dependent was within-subjects Nowtry think to postchoice) the variable, about sample the that independent photographs you andperformance revelation was aregoing see.In a few to the (distant, imminent) the minutes, willknow you that variable. with photograph quality is actually between-subjects produced the independent Consistent H2, by camera have chosen. you Are feeling anxious about there a time performance-revelation x was interaction you howgoodthephotograph will X= .90;F(1,48)= 5.6, < .05).Planned quality be? In other contrasts (Wilks' p are that words, youconcerned thephotograph quality revealed when that revelation distant, was performance be Please writedownyour might disappointing? suchthatexpectations postchoice optimism emerged in detail possible. as thoughtsas much increased pre- postchoice = 5.9 versus from to (M 6.5; when reve- Participants given minutes write were two to down F(1, = 5.0, < .05).However, performance 48) their p lation imminent, was faded such there thoughts. they Then, marked optimism away that (i.e., expectations postchoice) wasnoincrease pre- postchoice rather from to but a non- forthechosen each was product. Finally, participant decrease = 6.1versus F(1, = 1.3, > debriefed givena Fuji disposable and 5.8; 48) camera for significant (M p .25). participation. Weconducted a repeated measures ANOVA which in established sequence expectationsthe the of and Having was dependent time variable, (prechoice, causalmechanism, 3 tests underlying expectations the the Experiment was within-subjects postchoice) the variable, independent The after is process. process underlying optimism choice and of focus concern choice, performance) future was (prior that are on about people focused concerns whether they the variable.3 between-subjects independent made right the and choice, therefore try reassure they to Consistent H3, with there a time focus-of-concern was x themselves they Theprocess that did. underlying of fading interaction X= .90;F(1, = 5.0, < .05). Planned (Wilks' 48) p in thefaceof imminent is that optimism performance revealed when that concerns prior about choice are focused concerns whether perform- contrasts on about the people made such salient, postchoice optimism emerged that ance bedisappointing, will and therefore try lower were to they from to expectations (M = their If is and optimism.this true, optimism itsfading versus increased <pre- postchoice 6.1 when 6.8;F(1, = 8.9, .01).However, future48) p should to emerge simply by leading people these processes focus without (i.e., manipulating ofconcern) by changing the ofperformance time revelation. 3We coded responses focus-of-concern also the tothe (i.e., manipulation
arises when focus concern the of is H3:Postchoice optimism choice fades but when is future it prior away performance. More (a) specifically,PCOprior > 0, and(b) PCOFuture Choice
< PCoPriorChoicePerformance

EXPERIMENT3

Results

choice versus performance) of much future intermshow prior participants were toconvince themselvesthe that product begood bad. would or trying In theinterest parsimony,omit discussion. of we this the However, or was with results we expressed optimismpessimism consistent the report.

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

658

JOURNAL MARKETING OF 2006 RESEARCH,NOVEMBER


shouldhave no influence theproposed if holds.In theory twovariables attitude and 5, we employ Experiment (prior that the performance ambiguity) shouldmoderate Experiment results theproposed 3 if holdsbutthatshould theory haveno influence thealternative if holds. explanation EXPERIMENT4 Whenpeoplesimultaneously tasksthat engagein several consume are to resources, cognitive they known be cognior a load (Gilbert, and tively Pelham, busy, under cognitive Krull1988). If ourfading-optimism results due to difare ferences betweenjudgment and choice, theyshould be attenuated when cognitiveload is high because prior research has shownthatjudgments involvebroaderand more elaborate processingthan choices (Billings and Scherer increas1988; Shivand Huber2000). Specifically, load shouldreducethe moreelaborate ing cognitive proreducedifferences and, cessingofjudgments consequently, between results and choice-based results. judgment-based if are of Conversely, ourresults theoutcome theexplanation we propose, shouldnotbe affected cognitive load, they by becauseourtheory based noton theextent cognitive is of but focusof concern. Furtherprocessing on thecognitive that more, priorresearchhas demonstrated dissonance effects a free-choice in are automatic paradigm relatively andnotaffected cognitive load (Lieberman al. 2001). et by Given thatour effects are predicated a dissonanceon evoked that about optimism fadesawaybecauseofconcerns an imminent load not performance, cognitive should reduce theeffects we observed that we Therefore, prespreviously. enta nullhypothesis is related thefading-optimism that to effects that found Experiment we in 3. (H3) load the H4:Cognitive does not influence fading-optimism effect stated H3. in

concerns weremade salient, faded performance optimism away such that there was no increase from pre- to (M postchoice = 6.4 versus6.4; F(1, 48) < 1, p > .8). In other whendoubts werecreated abouthaving made words, theright themselves converted and choice, peoplereassured that concern intooptimism, whendoubtswerecreated but abouttheperformance this beingdisappointing, optimism was absent. 1-3 for Although Experiments findconverging support the proposedtheory about fadingoptimism products, in raisetwoconcerns. first The concern that experiis all they mentsused disposablecameras.Therefore, is not clear it whether results the are regarding fading optimism somehow of cameras moregenor uniqueto thecategory disposable eral.The secondand moreimportant concern thatthere is for mightbe an alternative explanation the resultswe observed.4 ALTERNATIVE EXPLANATION Preferences known changedepending howthey are to on are measured. For example,preferences from may differ choicetasksto ratings tasks(Nowlisand Simonson 1997); from evaluations jointevaluations to (Hsee 1996; separate Hsee et al. 1999); and from choicetaskstojudgment tasks, suchas matching and Sattath, Slovic 1988). The (Tversky, mostpertinent thealternative to is finding explanation that or dimensions important prominent might play a greater role in choice than in judgment(Tversky, and Sattath, Slovic 1988). To illustrate explanation, discussExperiment this we 3. At the prechoicestage,participants were in a judgment mode as theymarked for expectations each camerabased on somecombination of (e.g.,average) theavailable ratings of resolution, color richness, and color accuracy. Subsewhen participants found it difficult choose to quently, between three the chosetheone that had the cameras, they on thedimension believed be importo highest rating they for resolution. the rated tant, example, Resolution, highly now dimension, becamemoreprominent Sattath, (Tversky, andSlovic 1988).Therefore, in participants theprior-choice who were in a choice mind-set, condition, overweighted resolution when theywere asked for theirexpectations aboutthechosenproduct. Consequently, showedoptithey mism (i.e., postchoice> prechoiceexpectations). Conin condition versely, participants the future-performance were in a judgment mind-set because theywere thinking about the performance the one chosenproduct of rather than a choice between alternatives. Therefore,they the dimension as theydid at the weighted prominent just To prechoice stage,whichled to an absenceof optimism. addressthisalternative and explanation theissue of using in we just one product category our experiments, present twoadditional experiments. In Experiments and 5, we testwhether results 4 the that we found disposable for cameras Experiment emerge (in 3) forsoft drinks well.In addition, evaluate alternaas we the tiveexplanation twoways.In Experiment we employ in a 4, variable(cognitive the load) thatshouldmoderate Experiment3 results thealternative if holdsbutthat explanation
4Wethank anonymous an reviewer suggestions for this regarding alternative explanation.

Overview
In addition testing thisexperiment teststhe to also H4, robustness thefading-optimism (i.e., H3) in a difof effect ferent lemon-lime softdrinks. product category, namely, The ratings tablewas similar theone we used in Experito ments and 3, exceptthatthenamesused forsoftdrinks 2 wereBe-Cool, Cool-It,and Chill-Out, thedimensions and were those of taste-refreshingness, and flavorfulness, of carbonation. marked appropriateness Participants expectations that thinking theywere goingto findout theperformance in sessionat (i.e., thetasteof thedrink) a tasting theend.To makeparticipants a stakein theperformhave weretoldthat one ance,they consuming fullcan ofthesoft drink a mandatory ofthetasting was session.(Because part of this,3 of 80 participants refused participate the to in entire study.)

Design
The designused was a 2 (time:prechoice, x postchoice) 2 (focusof concern: x choice,future prior performance) 2 load: low,high)mixeddesignin whichthefirst (cognitive factor within was and werebetween subsubjects theothers used formeasuring was jects.The question expectations as follows: youropinion, "In whatwill be theoveralltasteof Be-Cool?" (1 = "poortaste," = "excellent 9 taste").

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

in Fading Optimism Products Procedure

659

ANOVA in which cessful, we conducteda univariate was variable cognitive and reported busyness thedependent businessstudents Seventy-seven undergraduate particiload (low,high)and focusof concern choice,future (prior for and a freesoft coursecredit patedin exchange partial were the between-subjects performance) independent drink. Day 1,participants On marked (i.e., expectations prevariables. was for Indeed, reported busyness higher thosein forall three softdrinks thebasis of theratings on choice) thehigh-cognitive-load forthosein thelow-cognitivethan table.On Day 8, participants weretoldthat wouldbe they load condition = 4.1 versus 2.9; F(1, 73) = 7.4,p < .01). (M in participating theremaining of the soft-drink part study All other effects werenonsignificant> .7). (p andin a different on memorizing numbers. study Theywere Forthemainanalysis, conducted repeated we a measures witha brief abouttheimportance of provided description ANOVAin which was thedependent variable, expectations in numbers consumption situations credit card (e.g.,prices, time(prechoice, was thewithin-subjects indepostchoice) and thenwere given 20 seconds to memorize numbers) and load (low,high)and focus variable, cognitive pendent eithera two-digit number (31; low cognitive load) or a of concern(priorchoice, future were the performance) number (6475031; highcognitive load). They seven-digit variables. between-subjects independent were not permitted writethe number to This anywhere. Consistent withH3 and theresults observed diswe for was based on an established of manipulation procedure the effect for posable cameras, fading-optimism emerged numbers inducecognitive to load (Roch et al. 2000; using soft drinks theform thetime focus-of-concern in x of interShivandFedorikhin weretoldthat 1999).Respondents they = .85; action(Wilks'21/4, F(1, 73) = 13.0,p < .01). Planned wouldbe askedto recallthenumber minutes ten laterand contrasts revealedthatwhenconcerns aboutpriorchoice that wouldcomplete soft-drink the in they study themeanwere made salient, PCO emerged such thatexpectations time. drink they that wanted consume to Theychosethesoft in the tasting sessionand thenread the focus-of-concern increasedfrompre- to postchoice(M = 5.8 versus6.4; F(1, 73) = 14.2, p < .001). However, when futureThe condition as follows: read manipulation. prior-choice concerns were made salient, faded performance optimism Nowtry think to about choice made. the the On you away such that there was no increase from pre- to lastpage,youchoseone soft for drink tasting, from but a decrease(M = 5.9 postchoice rather nonsignificant three available. Now thatyou have among options versus5.7; F(1, 73) = 1.7,p > .2). Although none of the made yourfinaldecision, you having are already effects cognitive of load weresignificant ps > .3), the (all doubts about soft the drink chose? other In words, you effect thatdeservesspecial mention the is nonsignificant areyouconcerned the that choice made not you might time x cognitive-loadx focus-of-concern interaction be the one?Pleasewrite best down in your thoughtsas much detail possible. as (Wilks'X = .99; F(1, 73) = .42,p > .5). Thiswas consistent with nullprediction the basedon ourtheory andincon(H4) The future-performance condition readas follows: sistent with alternative the The was explanation. interaction not onlynonsignificant also directionally but Nowtry think to about soft the drink youaregoing that oppositeto that which alternative the toconsume. a few In that the minutes, willknow taste explanation predicted; is, fadyou ofthesoft drink youchose. youfeeling that Are anxwas stronger in the highing optimism (not weaker) iousabout goodthe how taste be?In other will words, conditionthan in the low-cognitive-load cognitive-load areyouconcerned thetaste that be might disappointcondition. Please write down in detail ing? your thoughtsas much In Experiment we examineda variablethatshould 4, as possible. have had an influence the alternative if were explanation valid.However, had no effect. nowexamine it We variables Afterparticipants wrote down their thoughts, they thatshouldhave a moderating if influence thevalidexplamarked for (i.e., expectations postchoice) thechosenprodnation is fading optimismratherthan the alternative uct.Then,they tried recallthenumber to had memothey rized. a manipulation As checkfor were explanation. load,they cognitive also asked to indicatethe extentof mentaleffort they EXPERIMENT5 exerted keepthenumber memory to in the during previous = fewminutes = "did notexert mental effort all," 9 at (1 The variables consider priorattitude we are toward the "exerted lot of mental a and mind effort") how busytheir and of If product category ambiguity product performance. was whiletrying keepthenumber memory to in the ourresults due to thedifferential are of during weighting a promifewminutes = "mymindwas notbusyat all," (1 nent dimension previous there no is (i.e.,thealternative explanation), 9 = "my mind was extremely reasonforthesevariables have anyinfluence. to busy"). Because the two However, measures werehighly correlated = .82,a = .90), we averif our results due to theproposedtheory (r are (i.e., fading to a of aged them form measure reported busyness. Finally, optimism),these variables should act as moderators. were debriefed, freecans of Coke, Diet and because thesevariables related percepparticipants are to Specifically, weredistributed. Coke,and Sprite tions performance of rather to thechoicedecision, than they shouldnotaffect optimism occursbecauseof conthe that Results cerns about choice (prior-choice condition)but should For the memory for influence fading the thatoccursbecause of contask,recall was correct almostall optimism 39 in aboutperformance as participants-all participants the low-cognitive-load cerns condition), (future-performance conditionand 36 of the 38 participants the highin we discussnext. condition. was a mistake one of Priorattitudes predispositions are toward (One error cognitive-load objects.Such and another one was a mistakeof two digits.)To notions can guide interpretationnew inforof digit, preexisting checkwhether manipulation cognitive the of load was sucmation(Petty and Cacioppo 1986) and affect information

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

660

JOURNAL MARKETING OF 2006 RESEARCH,NOVEMBER


effect H5:Thefading-optimism is moderated theproduct's by and performance ambiguity theconsumer's attitude prior toward product the Morespecifically, when category. (a) the of focus concern prior is is choice, postchoice optimism not influenced either or but attitude, (b) by ambiguityprior when the focus of concernis future performance, is when attitude unfais postchoice optimism lower prior vorable than when is favorable; it when moreover, prior attitudeunfavorable, is is when postchoice optimismlower is low when is high, when it and attiambiguity than prior tude favorable, is is postchoice optimismlessinfluenced by ambiguity. Overview Design and The experimental and stimuliwere similarto setting thosewe used in Experiment The designwas a 2 (time: 4. x choice, prechoice, postchoice) 2 (focusof concern: prior future x low, performance)2 (performance ambiguity: high) x 2 (prior attitude: mixeddesignin unfavorable, favorable) whichthe first factor was within subjectsand the others were betweensubjects. We measured as expectations we didin Experiment 4. We conceptualized as performance ambiguity thepotential formultiple (Ha interpretations and Hoch 1989; Hoch and Ha 1986),and we manipulated usingtheopinions it of a soft-drink readexcerpts from pura expert.5 Participants ported interviewwith a food scientist.In the highthe claimed that a condition, scientist ambiguity tasting soft drink liketasting is is but wine;thetaste notclear-cut rather is subjective and open to interpretation. if the soft Even drink objectively a person stillconvince is can himor bad, herself ittastes that in other lies in the words, good; beauty In the condition, eyes of thebeholder. the low-ambiguity scientist stressed tasting softdrink notliketasting that a is winebutrather morelike tasting is fruit juice; thetasteis clear-cut unambiguous. thesoftdrink objectively and If is thatit bad, a personcan neverconvincehim-or herself tastes good. Prior attitude thesumof twomeasures = .75, a = was (r measure askedparticipants muchthey how .86). An affect likedlemon-lime softdrinks = "don'tlike at all," 9 = (1 "likea lot"),and a behavior measure askedthem howoften consumedthe drinks(1 = "not at all," 9 = "very they often").6 Procedure One hundred business students sixty-three undergraduate in for coursecredit a free and participated exchange partial soft drink of 168 didnotparticipate becausethey not did (5 wantto consumethe softdrink). Day 1, participants On marked their for soft (i.e., expectations prechoice) all three drinks thebasis oftheratings on table.On Day 8, they first readtherandomly Then, assigned ambiguity manipulation. as in Experiment they chosethesoft drink wanted to 4, they consumeand verbally to responded the focus-of-concern
5Wepretested manipulation 40 participants. expert's this with The opinions influenced of of in participants' opinions tasteambiguity softdrinks theintended direction. 6Priorattitude was dichotomized versus (Mdn = 5.0) into favorable unfavorable. better For of and conexposition ouroverallresults planned we The results trasts, discussonlytheANOVA results. usingregression attitude continuous as wereidentical. (with variable) prior

processing(Russo, Meloy, and Medvec 1998). When with unfavorable an a peoplearepredisposed (versus favortoward product a able) attitude travel), category (e.g.,airline it could heighten their aboutproduct anxiety performance Consider twopassengers withdiffercomfort). (e.g., flight entprior attitudes areabouttotakethesameflight. who The whose priorattitude moreunfavorable is passenger (i.e., believesthatflights uncomfortable) morelikelyto are is dreadtheexperience a potentially of uncomfortable flight. In an effort braceagainst to this discomfort, paspotential is senger morelikelyto reduceexpectations (e.g., by conhim-or herself theflight goingto be horrithat is vincing for withan unfavorable ble). Therefore, thisperson (versus is favorable) priorattitude, optimism goingto fade away morestrongly. there Furthermore, will be a multiplicative effect to performance due ambiguity. In the fading-optimism literature (Gilovich,Kerr,and Medvec 1993),theimplicit is assumption that performance is clearly revealed themoment truth. clear-cut at of For performance such as thoseof gradesand salaries revelations, and Fernandez1996), the threat of Ouellette, (Shepperd, short expectations real,and thereof is performance falling fore makessenseto lowerexpectations. it However, product is and It performance notalwaysas objective clear-cut. can the it varywith waytheconsumer interprets(Ha and Hoch dimen1989; Hoch andHa 1986).Ambiguous performance sionswouldbe thecomfort a carortheelegance a fine of of restaurant (Oliver 1997, p. 89), whereasless ambiguous dimensions wouldbe thecar's gas mileageor an entree's size. serving For unambiguous because it is difficult to performance, inflate it performance perceptions, makessense to reduce expectations 2001). For strategically (KopalleandLehmann can however,performance be ambiguousperformance, to it there less incenis interpretedbe highafter is revealed; tiveto lowerexpectations beforehand. disconFurthermore, firmation expectations "much of is less likely ambiguous in environments because of the availability littledirectly of information" (Hoch and Ha 1986,p. 223). It contradictory is notnecessary worry to aboutexpectations beingexplicbecauseperformancenota is itlyshattered performance, by moment truth of (Gilovich,Kerr,and Medvec 1993) but rather truth a thatis subjectto multiple interpretations andHa 1986). (Hoch becauseprior attitude ambiguity related and are Overall, to performance, willhaveno effect whenpeoplefocus they on priorchoice butwill moderate fading optimism the of that occurs when people focus on future performance. Because peoplewithan unfavorable attitude more are prior aboutperformance people witha favorthan apprehensive able attitude, theywill show lowerpostchoice optimism. whenpriorattitude unfavorable, is Moreover, postchoice will be even lowerwhenambiguity low than is optimism whenit is highbecause a person who is already apprehensive aboutperformance be concerned will evenmoreifthe revelation goingto be explicit. is Whenprior performance attitude favorable, is willbe less influenced expectations by The will abouta disambiguity. consumer be less worried or at appointing performance maynotbe worried all. Thereis fore, the explicitnesswith which the performance revealed willmatter or notat all. less

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

in FadingOptimism Products (i.e., They thenmarkedtheirexpectations manipulation. attitudes. and reported their Finally, they prior postchoice) and were debriefed, freecans of Coke, Diet Coke, and weredistributed. Sprite Results measures We conducted repeated a ANOVA in which time(prechoice, was thedependent variable; expectations was the within-subjects variable; independent postchoice) and prior attitude favorable), (low, (unfavorable, ambiguity choice,future (prior performhigh),and focusof concern To variables. ance) werethebetween-subjects independent 2 the of ease presentation results, Figure presents meansfor (i.e., expectations). optimism postchoice prechoice There was a main effectof time (Wilks' = .96; consistent H3,was qualwith F(1, 155) = 6.9,p < .01) that, interaction ifiedby a timex focus-of-concern (Wilks' = the .90; F(1, 155) = 16.4, p < .001), indicating fadingwhen concerns effect yet again. Specifically, optimism about prior choice were salient,postchoice optimism increasedfrompre- to emergedsuch that expectations 6.2; (M postchoice = 5.5 versus F(1, 155) = 21.2,p < .001). were whenconcerns aboutfuture However, performance faded away such that therewas no salient,optimism but a increase from pre-to postchoice rather nonsignificant decrease = 5.9 versus 5.7; F(1, 155) < 1,p > .40). (M interaction The timex prior-attitude (Wilks' X = .97; F(1, 155) = 5.0,p < .05) was qualified thetimex focusby x prior-attitude x interof-concern performance-ambiguity action(Wilks'X = .96; F(1, 155) = 7.3, p < .01). Further ambicondition, analysis supported In theprior-choice H5. and prior attitude no effect ps > .15). In the had (all guity bothvariables had condition, however, future-performance Postchoice was higher whenprior an influence. optimism it attitude favorable when was unfavorable = .2 was than (M versus-.5; F(1, 155) = 4.4, p < .05). Whenpriorattitude did was favorable, performance ambiguity not have any effect > .5), but whenpriorattitude was unfavorable, (p was lowerwhenambiguity low was postchoice optimism thanwhenit was high(M = -1.1 versus.15; F(1, 155) = low6.7, p < .05). In the unfavorable-prior-attitude, that there was clearpessimism, is, a condition, ambiguity declinefrom to prechoice postchoice expectasignificant tions(M = -1.1; F(1, 155) = 12.3,p < .001). effect these results the Overall, replicate fading-optimism demonstrate moderating the effects they (H3). In addition, ofperformance and attitude ambiguity prior (H5),providing of conclusive evidence theproposed for optiprocess fading the of mism.Furthermore, alternative explanation choice versus cannot account these for results. judgment GENERAL DISCUSSION in is A proposed aboutfading theory optimism products fiveexperiments, threedealingwithphotosupported by 1 and two withsoft-drink Experiment taste. graph quality an U demonstrates inverted shapeovertimesuchthat optimism arisesafter choiceandfadesawaywhenperformance 2 evidence for is aboutto be revealed. Experimentprovides that fadesaway thecausalmechanism showing optimism by when is (versus distant). performanceimminent Experiment the thatopti3 supports hypothesized processby showing is mismfadesawaywhenthefocusof concern future per1.2 1.0 .8 .6 .4 Optimism .2 -.0 -.2 Postchoice -.4 -.6 -.8 -1.0 -1.2 Unfavorable Favorable .8 .4 .6

661 Figure2

RESULTS FOR EXPERIMENT 5 A: Focuson PriorChoice 1.1

Toward Product Category PriorAttitude Ambiguity Low ambiguity product of performance of I Highambiguity product performance

B: Focuson Future Performance 1.2 1.0 .8 .6 .4 Optimism .2 -.0 -.2 Postchoice -.4 -.6 -.8 -1.0 -1.2 -1.1 Unfavorable Favorable .2 .3 .1

PriorAttitude Toward Product Category Ambiguity of Low ambiguity product performance of performance Highambiguity product

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

662

JOURNAL MARKETING OF 2006 RESEARCH,NOVEMBER


that will expectations consumers haveat thetimeperformance is revealed. understanding phenomenon fadthe of By in and to attiing optimism products its relationship prior tude and performance ambiguity, managerscan better how expectations the predict might varyfrom timea product is beingconsidered choiceto thetimeits performfor anceis revealed. REFERENCES Steven Edward Jones and E. Choice a as Berglas, (1978),"Drug to SucSelf-Handicapping inResponse Noncontingent Strategy Journal Personality SocialPsychology,(April), and 36 cess," of 405-417. Robert andLisa L. Scherer S. of Billings, (1988),"TheEffect ModeandImportance Decision-Making on StrateResponse Versus Behavior and gies:Judgments Choice," Organizational Human Decision 41 1-19. Processes, (February), Richard and A. William, Staelin, Valarie Boulding, AjayKalra, Zeithaml "A Process ModelofService (1993), Dynamic Qualto Journal Intentions," ity:From Expectations Behavioral of 30 7-27. Research, (February), Marketing Jack "Post-Decision in DesirabilBrehm, W.(1956), Changes the Journal Abnormal SocialPsycholand ityofAlternatives," of 52 384-89. ogy, (May), Leon A Dissonance. StanFestinger, (1957), Theory Cognitive of CA: Press. ford, Stanford University Daniel Brett Pelham, Douglas Krull W. and S. Gilbert, T., (1988), "OnCognitive When Person Perceivers Persons Meet Busyness: Journal Personality SocialPsychology, and 54 Perceived," of 733-40. (May), Husted Medvec Gilovich, Thomas, Kerr, Margaret andVictoria of on Confi(1993),"Effect Temporal Perspective Subjective Journal Personality Social Psychology, and 64 dence," of 552-60. (April), and J. ProHa, Young-Won Stephen Hoch(1989),"Ambiguity, and JourInteractions," cessing Strategy, Advertising-Evidence nalofConsumer 16 354-60. Research, (December), J. "Product Is JourHoch, Stephen (2002), ExperienceSeductive," nalofConsumer 29 448-54. Research, (December), and Ha "Consumer AdverYoung-Won (1986), Learning: and of Journal tising theAmbiguity Product Experience," of Consumer 13 221-33. Research, (September), K. An Hsee,Christopher(1996),"TheEvaluability Hypothesis: for Reversals Between Joint Sepaand Explanation Preference rate Evaluations Alternatives," of Behavior and Organizational Human Decision 67 247-57. Processes, (September), F. and , George Loewenstein, Blount, MaxH. BazerSally man "Preference Reversals Between andSeparate Joint (1999), Evaluations Options: Review Theoretical of A and Analysis," 125 576-90. Bulletin, (September), Psychological Robert andJames Inkster E. A. DisKnox, (1968),"Postdecision

formance 4 (versus choice).Experiments and 5 repliprior cate the fading-optimism rule out an alternative effect, and that is explanation, demonstrate theeffect moderated by attitude performance and Postchoice prior ambiguity. optimism lowest-in fact, is arises-whenprior attipessimism tudeis unfavorable ambiguity low. and is Theoretical Contributions This researchpresentsa dynamicnatureto product and a aboutwhether expectations resolves conflict expectationsshouldincrease 1957) or decrease(Kopalle (Festinger and Lehmann choice.It introduces 2001) after fading optimism(Gilovich,Kerr, and Medvec 1993) to therealmof and two of products identifies moderators this phenomenon: attitude performance and prior ambiguity. In addition, research this addsto KopalleandLehmann's on of (2001) research how strategic management expectationsvarieswithindividual traits showing evenfor that by thesameperson, reduction happenwhenpercan strategic formance abouttobe revealed. is With to respect postchoice research within whichhas focusedmainlyon marketing, are postperformance processes,such as how expectations either confirmed disconfirmed performance or (Oliver by demonstrates interesting that 1980), thisresearch expectationprocessescan occurafter choice butbefore performance. Moreover, expectations if changeovertime,predictionsaboutsatisfaction as might vary well. the Finally,this researchreinforces notionthat consumersare "adept at adaptingto unchangeable circumstances"(Hoch 2002, p. 452). We show thatwhen consumersare concernedabout productperformance, they lowertheirexpectations because theycannotchangethe itself. this when Moreover, effect performance strengthens is that of performanceambiguous, is, whenperceptions performance cannotbe changedbecause the evidenceis not to subject multiple interpretations.

ManagerialImplications Clearimplications after-sales for communication emerge. LosciutoandPerloff (1967,p. 289) notethat "[m]anufacturers and retailers in interested maintaining satisfied customers benefit providing purchaser the with might by ways of reducingdissonance,"such as givingliterature that assures"thepurchaser his choicewas a wise one."Our that researchsuggeststhatsuch reassurance shouldbe given after is If choice performance revealed. it is givenbetween and performance it revelation, will increaseexpectations the for and,thus, potential disappointment. sonanceat PostTime,"Journal Personality Social Psyand of also arise forwarranties. After consumers 8 319-23. Implications chology,(April), Praveen andDonaldR. Lehmann K. offeroptionalwar(2001),"Strategic Kopalle, buy products, companiesfrequently of The ranties. Because warranties Management Expectations: Role of Disconfirmation help reducetheriskof a bad and Journal Research, Sensitivity Perfectionism," ofMarketing consumers shouldbe less likelyto buythem performance, 38 (August), 386-94. soon after are optimistic aboutperpurchase(whenthey Matthew Kevin Ochsner, N. Daniel Gilbert, T. and D., and whenperformance Lieberman, formance) morelikelyto buythem DanielL. Schacter Exhibit (2001),"Do Amnesiacs Cognitive is imminent. example, BestBuy.com, a customer For at after Dissonance Reduction? Role of Explicit The and Memory adds a television theshopping to a cart, warranty (performin Attention Attitude Science,12 Change," Psychological ance service for Best Buy might plan) is offered purchase. 135-40. (March), have a better chanceof sellingthewarranty it is offered Losciuto, if Leonard andRobert A. Perloff of (1967),"Influence for on is Product PreferenceDissonance on Journal Marlater, example, telephone thedaythetelevision Reduction," by of 4 286-90. Research,(August), keting goingto be delivered. M. Simonson "Attribute-Task Nowlis, to withperformance meets (1997), that Stephen andItamar Finally, designproducts as a ofConsumer Preference ReverCompatibility Determinant or exceedsexpectations, first need to gauge the managers Journal Marketing 39 205-219. sals," Research, (May), of

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

in Optimism Products Fading


Richard (1980), Cognitive L. "A of Antecedents Model the Oliver, andConsequences Satisfaction of Journal MarDecisions," of 17 460-69. Research, (November), keting A on (1997),Satisfaction: Behavioral Perspective the New Consumer. York: McGraw-Hill. T. Richard andJohn Cacioppo E. (1986),Communication Petty, and Routes Attitude to and Persuasion: Central Peripheral New Change. York: Verlag. A.S. D. Scott Roch, G., Sylvia John Lane,Charles Samuelson, T. L. and Load Allison, JenniferDent(2000),"Cognitive andthe A ModelofResource Heuristic: Two-Stage OverconEquality in Behavior and Organizational sumption Small Groups," Human Decision 83 185-212. Processes, (November), A. G. and Rosenfeld, Paul,John Kennedy, Robert Giacalone A of the (1986),"Decision Making: Demonstration PostdecisionDissonance Journal Social Psychology, 126 Effect," of 663-65. (October),

663
J. G. and H. Russo, Edward, Margaret Meloy, Victoria Medvec of Distortion Product (1998), "Predecisional Information," Journal Marketing 35 438-52. Research, (November), of James Judith Ouellette, Julie Fernandez A. and K. A., Shepperd Unrealistic Performance Esti(1996),"Abandoning Optimism: mates the and Temporal of Feedback," ProximitySelf-Relevant Journal Personality Social Psychology, (April), and 70 of 844-55. BabaandAlexander Fedorikhin in "Heart Mind and Shiv, (1999), Conflict: Interplay Affect Cognition Consumer The of in and Decision Journal Consumer 26 Research, (DecemMaking," of 278-92. ber), and Joel Huber "TheImpact Anticipating of Satis(2000), faction Consumer on Journal Consumer Choice," Research, of 27 (September), 202-216. Shmuel and "ContinAmos, Sattath, PaulSlovic (1988), Tversky, in and gentWeighting Judgment Choice,"Psychological 95 371-84. Review, (July),

This content downloaded on Sun, 3 Feb 2013 07:54:10 AM All use subject to JSTOR Terms and Conditions

You might also like