PSYCHOLOGICAL SCIENCE Research Article Doing Better but Feeling Worse Looking for the "Best" Job Undermines Satisfaction Sheena S. Iyengar,' Rachael E. Wells, ' and Barry Schwartz Management Division, Graduate School of Business, Columbia University, and Psychology Department, Swarthmore College ABSTRACT-Expanding upon Simon's (1955) seminal the- spending so much time deciding on a show that little time would ory, this investigation compared the choice-making strat- be left for viewing. The satisficer would most likely channel-surf egies of maximizers and satisficers, finding that maximiz until he or she encountered the first acceptable show, put down ing tendencies, although positively correlated with object the remote control, and actually watch the show. Simon based his tively better decision outcomes, are also associated with distinction on the idea that the limited information-processing more negative subjective evaluations of these decision capacities of organisms make maximizing impossible. In the outcomes. Specifically, in the fall of their final year in modern world of almost unimaginable choice, this distinction is school, students were administered a scale that measured even more pertinent (see Iyengar & Lepper, 2000; Schwartz, maximizing tendencies and were then followed over the 2004a, 2004b). course of the year as they searched for jobs. Students with Expanding on Simon's classic theory, Schwartz et al. (2002) high maximizing tendencies secured jobs with 20% higher recently compared the decision-making processes of maxi- starting salaries than did students with low maximizing mizers and satisficers, finding that people who exhibit maxi- tendencies. However, maximizers were less satisfied than mizing tendencies, like the channel surfer just described, were satisficers with the jobs they obtained, and experienced less satisfied with their decision outcomes than their satisficing more negative affect throughout the job-search process. counterparts. The researchers asked participants about recent These effects were mediated by maximizers' greater reli- purchasing decisions and used a "maximization scale" to ance on external sources of information and their fixation measure individual differences in maximizing tendencies. Their on realized and unrealized options during the search and findings suggested that the experiences of maximizers differed selection process from those of satisficers during the decision-making process and also later, when they evaluated their final decision outcome. Specifically, compared with satisficers, maximizers were more likely to engage in an exhaustive search of all available options Success is getting what you want. Happiness is wanting what you and to compare their decisions with those of other people. Even get. -American proverb though maximizers invested more time and effort during the decision process and explored more options than satisficers- Half a century ago, Simon (1955, 1956, 1957) introduced an presumably in order to achieve greater satisfaction-they none- important distinction between maximizing and satisficing as theless felt worse about the outcomes that they achieved. Re- choice-making strategies. To maximize is to seek the best and sults showed that maximizing tendencies were positively requires an exhaustive search of all possibilities. To satisfice is correlated with regret, depression, and decision difficulty, and to seek "good enough," searching until encountering an option negatively correlated with happiness, life satisfaction, opti- that crosses the threshold of acceptability. For example, com- mism, and satisfaction with decision outcomes. pare the strategies of a maximizer versus a satisficer selecting a Such differences in the subjective choice-making experi- television show from choices available on 400 cable channels. ences of maximizers and satisficers are attributed to the fact that The maximizer would channel-surf, exploring all the channels, maximizers create a more onerous choice-making process for themselves. Initially, maximizers focus on increasing their Address correspondence to Sheena S. Iyengar, Management Division, choice sets by exploring multiple options, presumably because Columbia University Business School, 3022 Broadway, Room 714, expanded choice sets allow for greater possibilities to seek Uris Hall, New York, NY 10027, e-mail: ss957@columbia.edu. out and find the elusive "best." Yet, as the number of options Volume 17-Number 2 Copyright 2006 Association for Psychological Science 143Looking for the "Best" Job Undermines Satisfaction Sheena S. Iyengar, Rachael E. Wells, and Barry Schwartz proliferates, cognitive limitations prevent decision makers from plan to apply for more jobs, rely more on social comparison and individual items were averaged to create a composite maxi- participants were required to work and converted this infor- evaluating and comparing all options (Iyengar & Jiang, 2004; other external sources of information, and obtain jobs with mizing score. Overall, men (n = 166, M = 5.48) and women (n = mation into an estimated annual salary Iyengar & Lepper, 2000; Miller, 1956). Identifying the best higher expected returns (i.e., salary), but also experience greater 382, M = 5.10) from our sample population showed significantly becomes increasingly difficult, compelling maximizers to rely on negative affect and less outcome satisfaction throughout the higher maximizing tendencies than respondents in a recent Negative Affect external (often social) rather than internal standards to evaluate process and at the conclusion of their job search. national adult sample (Kliger & Schwartz, 2005; men: n = Participants' negative affect associated with the job-search and select outcomes (Lyubomirsky & Ross, 1997). In addition, 3,261, M = 4.9; women: n = 4,692, M = 4.77), t(165) = 7.03, process was measured at all three assessments. At TI and T2, the inevitability of trade-offs among attractive options intensi- METHOD <.0001 for men and t p .0001 women. participants were asked what extent does each of the fol- fies sting passing up one attractive alternative when these differences may be at least partly attributable to age lowing describe how you are generally feeling about job choosing a more increases expectations he quality utility chosen alternative. graduating students undergraduate seniors difference between two samples as maximization tendencies search process seven emotions listed have been found negatively correlated with but do very strategies that render maximizers less happy recruited from colleges universities varied in schwartz our sample maximizing rated emotion on scale than satisficers their decision outcomes also enable them geographical region university rank school size. significantly positively top-15 all t2 t3 achieve objectively better perhaps was female. median r .10 .05 male gender .17 same question repeated however three here is associated strategic pursuit real caucasian asian other racial-ethnic backgrounds. par- not any demographic or control variable added imagined options careful observation gathered. trated. addition who had accepted of- people choice-making experiences-utility re- ticipants majored social sciences arts hu- fers modified read: flected form effective deliberations object manities engineering natural math sessment option fixation following offer tively outcomes. unlike prior investigations business first responded response rates we used measures examine fixation. ti your upcoming new composite mea- elation sures constructed. current investigation examined effects second third assessments respectively. five measured number pursued: approximately many jobs anticipate applying both objective subjective ex- prizes raffled off among com- outcome satisfaction perience maker throughout process. pleted surveys. provided responses numerical form. note anticipated applications ranged items ac- thus expanding this nascent literature present study exhibiting extreme right skewness kurtosis cepted offers: satisfied allowed us test hypotheses. procedure therefore log-transformed. confident made searching choices deciding which select. november career services participating unrealized options: often fantasize choice where work next year hypothesized compared invest institutions directed just beginning quite different actual i am heavily gathering information external sources searches september pursuing. .88 score obtained participant by averaging incurring costs prioritizing exter- through survey web site. via e-mail no- disagree agree questions. nally valued criteria fixate realized tified follow-up on-line surveys regret size set: wish pursued opportunity hy- february completing applica- my demographics variables pothesis involved tions interviewing getting offers large gathered sex ethnicity family income hese decision-making pro cess contribute they accepting offers. although it difficult calculate level affiliation u.s. news successful percentage chose participate reliance influences world report geographic location academic yet result experiencing upon encountering advertisement consultations major ti. collected overall grade greater negative affect reduced well-being. career-services staff estimates created single measure .70 point average job- predictions within consequential total utilized influences. domain college job-search processes given year. asked: much using offered related activities stage influence ten- available only those office during dencies perceived affiliated months consulted experts ranking such results afforded reactions calculated companies growing fields etc. decisions being made. regardless finite sponse across sampling seek advice regarding in- preliminary analysis seekers un- institutions. put suggestions compare table reports means standard deviations function dertaking face raised own peers status dependent maxi- contemplating g an almost limitless peer comparison t2. mizers separated split. attrition set employment possibilities. determine little analyses demonstrated initial differed demo- orientation affects affective experiences ob- graphically samples: east asians children jective completed drawn foreign-born parents older constituted smaller multiple et al. e.g. car listening job-market performance proportion fall final subsequently followed radio check stations see if something interviews did identify themselves specific measuring well playing even relatively received. larger tl. tually thought did. methodology shopping hard time finding received annual salary dollars per differ critical predictions: clothes really love item hour accepted. case hypotheses log pared would desire scores hourly wages determined hours week volume undermines sheena s. iyengar rachael e. wells barry mediated logged b=".21," regression models predicting mediator .001 prep=".99," fully salary: alized model female fact relation .29 .06 .11 .13 .03 .02 .18 science .08 tl experience included note. parentheses. refer .321 similar emerged. reported medians rather humanities ranked top education . cumulative discussion interacted tendencies. .20 financially proposed feel worse. quest placement after .27 further revealed already full-model r2 .54 graduation dv arz vs. .09 pursue likely majors measurement .21 ar previous degree rely infor- ma s younger f ratio mation satisficing seekers. efforts .23 higher gpas came wealthier socioeconomic back- degrees freedom payoffs: earn starting salaries x grounds relied satisficers. despite relative suc- controlled uni- .14 .28 variable. versity gpa ar2 .01 .10. pessimistic stressed tired anxious worried whether tables overwhelmed depressed why full including variables. merely high achievers killeen reporting poisson regression: x2="1.43," n.s. past successes superior credentials rightly learned provide probabilities repli- o expect themselves. no matter cating .05. worse because fail match values. expectations. certainly there evidence suggest histories success sig- wished still t2: main t3: nificantly top-ranked shown every one-unit schools. increased increase equating capability oversimplifies improved measure. additionally reliant story. find significant experiences. satisfac another marker gpa. furthermore assumed tion market simply effect attenuated attending high-ranked proxies qualifications versities indeed decrease mediate performance. though indices whereas lower-ranked counterparts predictive salary. mediators whatever fantasized achievement itself seems inade- pursuing tained based split quate explain affect. then showed mean combination start acted partial suggests unaccounted dispositionally statistical yielded sub- positive correlational decision. e amples characteristics request. prove predictor either observed partially accounting obser benefits best underestimate involves evaluating possible fixating affect: satisfaction: nonexistent. get want controlling always get. individual de- cision makers policymakers confronted dilemma: well-being value odds should prioritized .04 makes .00 acknowledgments-this research supported na- tional foundation young investigator award. majo .07 _02 references j2 national universities. p. .53 botti s.s. psychological pleasure pain ..28 choosing: prefer cost subsequent .25 .16 satisfaction. journal personality psychol ogy frederick loewenstein g. hedonic adaptation. d. kahneman diener n. schwarz well-being: foundations psychology york: russell sage. t1 .59 .31 jiang w. choose: retirement savings decisions. manuscript submitted .35 .50 publication. lepper m.r. rethinking choice: cultural perspective intrinsic motivation. person- ality .85 demotivating: assessment. can too good thing d happiness. die- ner behaviors expected satis- p.r. null-hypothesis significance t3. find- designed specifically faction impending assess jok tests. ings support earlier tasks necessarily tasks. resulting employment. kliger m. b. tendencies: unpublished raw data. contribution mediation differ- psychologists economists alike provi- schkade wouldn nice evaluations independent dispositional happi- ences explained sion beneficial allows uture feelings. ness. informa- opportunities preference matching en- instead tion. nonetheless global ables maximization. seem cast long shadow part body positing lyubomirsky ross l. consequences obtain elusive induces appraisals comparison: contrast unhappy people. consider possibilities thereby course findings limited ability preference-match miller g.a. magic plus minus two: some creasing potential engen- values limits capacity processing information. dering unrealistically creating mounting allow mispredicted review costs. integral identifying lesser stems paradox less. goal detract working conditions wilson gilbert ecco. ultimately derive professional atmosphere interaction colleagues organi- experienced april tyranny choice. scientific treated zational commitment advancement. ad- epitomize ward a. monterosso j. wh dividual ditionally type overestimate lehman d.r. versus satisficing: haj t.d. d.t. forecasting. m.p. advances experimental simon h.a. behavioral rational quarterly pp. san diego ca: press. economics structure environ- ment. man rational: mathematical essays human behavior. wiley. strangers ourselves: discovering adaptive revision unconscious. cambridge ma: harvard materials>