Computers in Human Behavior 31 (2014) 94–99

Contents lists available at ScienceDirect

Computers in Human Behavior

journal homepage: www.elsevier .com/locate /comphumbeh

Research Report

Death of a salesman: Webpage-based manipulations of mortalitysalience

0747-5632/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.chb.2013.10.022

⇑ Corresponding author. Address: Department of Psychology, University ofMichigan, 530 Church Street, Ann Arbor, MI 48109, United States. Tel.: +1 (734)763 5146; fax: +1 (734) 647 9440.

E-mail address: chopik@umich.edu (W.J. Chopik).

William J. Chopik ⇑, Robin S. EdelsteinDepartment of Psychology, University of Michigan, Ann Arbor, United States

a r t i c l e i n f o

Article history:Available online 1 November 2013

Keywords:Terror managementAdvertisementConsumer behaviorInternet methods

a b s t r a c t

Most people are accustomed to ignoring the advertisements that they encounter while surfing the Inter-net, despite the profound effects such advertisements can have on behavior. We showed that webpageadvertisements can remind people of their mortality (Study 1) and lead them to invest in culturally val-ued behavior (Studies 2–4). Specifically, individuals in the ‘‘mortality salience’’ condition reported greaterworldview defense (Study 2) and spent more money on luxury items (Studies 3 and 4) than those in thecontrol condition, consistent with proposals set forth by terror management theory. In Study 4, death-related thoughts mediated the relationship between mortality salience and willingness to spend moneyon luxury items. Findings are discussed in the context of online consumer behavior.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Any individual surfing the web is likely to encounter advertise-ments for a wide variety of products, events, and services. Most ofus have been conditioned to ignore these pesky annoyances and togo on clicking and attending to the space between the ads on thewebpage. However, there is some evidence that individuals do infact process these advertisements and that such advertisementscan increase preferences for the content featured in them (Yoo,2009). These effects are found regardless of whether or not peopleactively attend to the advertisements (Yoo, 2008). With the transi-tion of much of our behavior online (e.g., communication, network-ing, purchasing goods and services), it is particularly important toexamine the characteristics of an online environment. The purposeof the current studies was to examine the effects that banneradvertisements have on consumers using predictions made by ter-ror management theory (TMT; Greenberg, Pyszczynski, & Solomon,1986).

TMT posits that individuals are motivated to alleviate thethoughts and concerns that arise from the knowledge that deathis inevitable (Greenberg et al., 1986). To deal with this anxiety,people often cling to cultural worldviews in an effort to connectwith something that extends beyond their physical death (Solo-mon, Greenberg, & Pyszczynski, 2004). As a result, inducingdeath-related thoughts (i.e., mortality salience) leads participantsto react more negatively to those who threaten and criticize their

cultural worldviews (this defensive reaction is termed ‘‘worldviewdefense’’; Greenberg et al., 1990). Also, people seek to enhancetheir self-esteem by achieving standards of value that are definedby a particular culture. For instance, research suggests that main-taining high status and engaging in materialistic consumer behav-ior can buffers mortality salience among United States citizens(Arndt, Solomon, Kasser, & Sheldon, 2004; Mandel & Heine,1999). Further, mortality salience has been shown to increase peo-ple’s financial expectations for themselves 15 years in the futureand to elevate greedy behavior in a resource consumption scenario(Kasser & Sheldon, 2000). In this study, participants were told toimagine that they were in competition with three other companiesharvesting timber in an ever-dwindling national forest. Partici-pants made bids (ranging from small to large) on how many acresthey would be willing to harvest in a given year (out of an available100 acres) and how much they would like to profit more than theother companies (a measure of greed). Participants in the mortalitysalience condition reported higher levels of greed and noted thatthey were willing to harvest 12–13 more acres of forest than par-ticipants in the control condition. The results from this study sug-gest that mortality salience can motivate individuals to alleviatemortality concerns by increasing greed and consumption behavior.However, the effects of mortality salience should be limited to theacquisition of possessions that are high-status and tied to self-es-teem, such as luxury items (Mandel & Heine, 1999). Luxury prod-ucts are considered expensive and reflect high status, sowillingness to spend more money on these items represents an ef-fort to obtain greater status through possessions or, put anotherway, materialism.

Mortality salience is most often induced by asking participantsto respond to open-ended questions about the thought of one’s

W.J. Chopik, R.S. Edelstein / Computers in Human Behavior 31 (2014) 94–99 95

own death (Rosenblatt, Greenberg, Solomon, Pyszczynski, & Lyon,1989). However, other effective manipulations have also beenused: exposure to subliminal death-related stimuli (i.e., the word‘‘dead’’; Arndt, Greenberg, Pyszczynski, & Solomon, 1997), readingdeath-related facts (i.e., ‘‘In 1997, 3,599 people were killed. . .inroad traffic accidents.’’; Jessop, Albery, Rutter, & Garrod, 2008),completing a fear of death survey (Mikulincer & Florian, 2000),handing out fliers with reminders of death to pedestrians (Hirsch-berger, Ein-Dor, & Almakias, 2008), and even being interviewedoutside of a funeral home (Pyszczynski, Wicklund, Floresku, &Koch, 1996). The wide breadth of manipulations demonstrates thatboth overt and incidental reminders of mortality affect humanbehavior.

Despite the variety of past manipulations, researchers have notyet investigated how mortality salience can affect human behaviorin an online environment. The current studies were implementedto test the ability of webpage advertisements to elicit mortality sal-ience. We examined whether this manipulation led to higheraccessibility of death-related thoughts in one’s consciousness(i.e., ‘‘death-thought accessibility’’; Study 1) and greater culturalworldview defense (Study 2). Because webpage advertisementshave been shown to influence people unknowingly (Yoo, 2008),and mortality salience has been shown to increase materialisticbehavior (Arndt et al., 2004), we also examined whether banneradvertisements influence consumer behavior (Study 3) andwhether this relationship was mediated by death-related thoughts(Study 4). With the online purchasing of products becoming morecommon, there are important implications of advertisements thatunobtrusively affect consumer behavior.

2. Study 1

In this first study, we developed a subtle mortality saliencemanipulation administered through webpage advertisements andtested whether it increased the accessibility of death-relatedthoughts. Death-thought accessibility was operationalized as thenumber of word-fragments that were completed with a death-re-lated word. If participants complete a possible word-fragment(i.e., de_ _) with a death-related word (i.e., dead) instead of a neu-tral word (i.e., deed, debt, deal), it is hypothesized that death-re-lated thoughts are more consciously accessible (Greenberg,Pyszczynski, Solomon, Simon, & Breus, 1994).

2.1. Method

2.1.1. ParticipantsParticipants were 1864 online respondents (74.5% female,

Mage = 35.95, SD = 13.65) recruited through Amazon’s MechanicalTurk (MTurk). Mturk is a web service that matches requesters (of-ten companies or researchers) with members willing to completeweb-based tasks for a specified reward. Several evaluations ofMturk have shown that it is a reliable and valid source of psycho-logical data (Buhrmester, Kwang, & Gosling, 2011; Paolacci, Chan-dler, & Ipeirotis, 2010). Participation was limited to those in theUnited States, and participants were compensated $.05, which issimilar to other short studies implemented on MTurk (Buhrmesteret al., 2011). The majority of participants were recruited from Maythrough November 2010. All respondents finished in less than5 min.

2.1.2. Procedure and materialsParticipants responded to a request to participate in a study on

word-fragment completions. They were first directed to an in-formed consent page, which provided the cover story that the Uni-versity of Michigan was sponsoring the webpage. The banner

advertisements at the top and bottom of the first page read, ‘‘UMEvents & Services 2010’’ and had a University of Michigan logo.These banner advertisements were included to reduce suspicionabout the nature of the study by introducing advertisements earlyon in the study (on the consent page). On the next page, partici-pants were randomly assigned to one of two conditions. The mor-tality salience condition featured banner ads that read, ‘‘Join Us forUM’s Day of the Dead Celebration on Nov. 2nd!!!’’ with a picture ofa skull; the control condition featured banner ads that read, ‘‘JoinUs for UM’s Cherry Picking Outing on Aug. 5th!!!’’ with a pictureof cherries. The advertisements were created in consultation withresearch suggesting that exposure to certain words and imagesmakes that content more accessible to influence future behaviorand judgments (Higgins, Rholes, & Jones, 1977; Hirschbergeret al., 2008). No references to the content of the advertisementswere made in the survey. See the Appendix for illustrations ofthese banner ads.

Death-thought accessibility was measured by a word-fragmentcompletion task similar to those used in previous terror manage-ment research (Greenberg et al., 1994). Out of the 25 word frag-ments, 6 could be completed with a death-related word (e.g., coff_ _; which could be completed as coffin or coffee). Death-thoughtaccessibility was operationalized as the average number ofdeath-related word completions. This completion task was placedin between the banner ads on the main part of the webpage. On thefollowing page, the cover story University advertisements re-sumed, and demographic characteristics were collected. Partici-pants were also given the opportunity to provide feedback orcomments about the experiment, which were later coded for anysuspicion that the advertisements had an effect on their responses.

2.2. Results and discussion

In Study 1, we tested whether mortality-based webpage adver-tisements increased the accessibility of death-related thoughts. Asseen in Fig. 1, participants in the mortality salience condition com-pleted more death-related words (M = 2.16, SD = 1.08) than partic-ipants in the control condition (M = 1.94, SD = .98), t(1862) = �4.75,d = .22, p < .001. That is, death-thought accessibility was higher inthe mortality salience condition than in the control condition.The pattern of findings for Study 1 remained significant after con-trolling for age and gender. No participants reported any suspicionabout the advertisements. In sum, this study demonstrated thatmortality-related banner ads lead to higher death-thoughtaccessibility.

3. Study 2

The purpose of Study 2 was to test whether our mortality sal-ience manipulation elicited increases in worldview defense, a com-mon consequence of mortality salience in the TMT literature. Wepresented participants with two essays about the United Statesostensibly written by two foreigners (one pro-U.S., one anti-U.S.).Worldview defense is operationalized as the difference betweenapproval of the pro-U.S. author and essay and disapproval of theanti-U.S. author and essay. A larger discrepancy between howmuch people like someone who validates their worldview andhow much people dislike someone who challenges their worldviewserves as a proxy for the defense of their cultural worldview.

3.1. Method

3.1.1. ParticipantsTwo hundred and ninety-eight participants (68.8% female,

Mage = 33.81, SD = 13.38) were recruited through Amazon’s Mturk

1.2

1.4

1.6

1.8

2

2.2

2.4

Mortality Salience Control

Com

plet

ed d

eath

-rel

ated

wor

ds

Condition

Fig. 1. Study 1: The effects of mortality salience on death thought accessibility.Error bars represent standard errors of the mean.

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

Mortality Salience Control

Wor

ldvi

ew D

efen

se

Condition

Fig. 2. Study 2: The effects of mortality salience on worldview defense. Error barsrepresent standard errors of the mean.

Table 1Mean favorability ratings of the target essay/authors by experimental condition(Study 2).

Type of target Mortality salience Control

M SD M SD

Pro-U.S. 5.13 1.11 4.84 1.15Anti-U.S. 3.59 1.45 4.02 1.45

96 W.J. Chopik, R.S. Edelstein / Computers in Human Behavior 31 (2014) 94–99

and compensated $.10. Participation was limited to those in theUnited States. All respondents finished in less than 10 min.

3.1.2. Procedure and materialsThe manipulation of mortality salience was identical to Study 1.

Participants responded to a request to participate in a study onworldview beliefs. They read two essays about the United States(one pro-U.S., one anti-U.S.) ostensibly written by two foreigners,adopted from previous research on world defense (Greenberg, Si-mon, Pyszczynski, Solomon, & Chatel, 1992). The order of the twoessays was counterbalanced across participants. Participants eval-uated the truth of each essay, their agreement with each essay, andhow likeable, intelligent, and knowledgeable they thought eachauthor was on a 9-point scale. Higher scores indicated more posi-tive evaluations. The evaluations for each essay were averaged andserved as measures of worldview-consistent (in the case of thepro-U.S. essay) and worldview-inconsistent (in the case of theanti-U.S. essay) opinions. Items showed a high internal reliabilityfor both the worldview-consistent questions (a = .89) and world-view-inconsistent questions (a = .93). Worldview defense was de-fined as the difference between these measures, so a largerdifference indicates a greater worldview defense (Schmeichelet al., 2009).

3.2. Results and discussion

In Study 2, we tested whether mortality-based webpage adver-tisements increased worldview defense. As seen in Fig. 2, partici-pants in the mortality salience condition reported higherfavorability for worldview-consistent compared to worldview-inconsistent opinions (M = 1.54, SD = 1.85) than those in the con-trol condition (M = .82, SD = 1.88), t(296) = �3.33, d = .39, p = .001.That is, worldview defense was higher in the mortality saliencecondition than in the control condition. Favorability ratings foreach target essay (pro- and anti-U.S.) are presented in Table 1.Favorability was highest for the pro-U.S. author/essay and lowestfor the anti-U.S. author/essay in the mortality salience condition.The pattern of findings for Study 2 remained significant after con-trolling for age and gender. No participants reported any suspicionabout the advertisements.

Study 2 replicated a common finding in the terror managementliterature, namely that mortality salience leads to greater world-view defense. As expected, mortality-based webpage advertise-ments increased favorability of worldview-consistent comparedto worldview-inconsistent opinions.

4. Study 3

The purpose of Study 3 was to determine whether the effects ofbanner advertisements translate to consumer behavior—spending

money on luxury products. People spend a large portion of theirtime online buying products, and online purchases are a largesource of profits for companies and advertisers. Previous researchhas linked death anxiety with increased materialism (Arndt et al.,2004; Mandel & Heine, 1999), so Study 3 can uncover some ofthe motivations behind individuals’ online buying intentions.

4.1. Method

4.1.1. ParticipantsFour hundred and sixty-seven participants (66.2% female,

Mage = 33.58, SD = 13.46) were recruited through Amazon’s Mturkand compensated $.10. Participation was limited to those in theUnited States. All respondents finished in less than 10 min.

4.1.2. Procedure and materialsThe manipulation of mortality salience was identical to the first

two studies. Participants responded to a request to participate in astudy on consumer behavior. Materialism was measured by askingan open-ended question about how much participants would bewilling to spend on each of four luxury products (a TV, watch, cellphone, and camera; adopted from past research on consumerbehavior; Kesebir, Chiu, & Pyszczynski, 2013). Luxury productsare considered expensive and reflect high status, so a willingnessto spend more money on these items represents an effort to obtaingreater status through possessions. Pictures and model descrip-tions for each of the products were provided for the participants.Two non-luxury products (a ballpoint pen and a package of toiletpaper) were also included to examine whether mortality salienceincreased willingness to spend money overall. Because the effectsof mortality salience on materialism are thought to be specific onlyto high status (luxury) items, we would expect that the willingnessto spend money on low status (non-luxury) items should not differbetween conditions. Total willingness to spend money for eachgroup of products (luxury and non-luxury) was computed by aver-aging the dollar amount across the items.

4.2. Results and discussion

Study 3 tested whether our mortality salience manipulation in-creased individuals’ intentions to buy luxury products. We also

$0

$100

$200

$300

$400

$500

$600

Mortality Salience ControlMon

ey S

pent

on

Pro

duct

s

Condition

Luxury Products

$0

$10

$20

$30

$40

$50

$60

ControlMortality Salience

Money Spent on P

roducts

Condition

Non-Luxury Products

Fig. 3. Study 3: The effects of mortality salience on materialism. Values represent US dollars. Error bars represent standard errors of the mean.

W.J. Chopik, R.S. Edelstein / Computers in Human Behavior 31 (2014) 94–99 97

hypothesized that this pattern would not be found among non-lux-ury items. We conducted a repeated measures ANOVA predictingwillingness to spend money on consumer goods (within-subjectscondition: luxury vs. non-luxury) from mortality salience condi-tion (between-subjects condition: mortality salience vs. control).A significant interaction emerged between mortality salience con-dition and type of consumer goods, F(1,465) = 5.42, p < .05. Partic-ipants in the mortality salience condition were willing to spendmore money on luxury products (M = $507.46, SD = 398.74) thanthose in the control condition (M = $419.42, SD = 430.46),F(1,465) = 5.24, d = .21, p < .05. Also as expected, willingness tospend money on non-luxury products did not differ between themortality salience (M = $5.07, SD = 6.56) and the control condition(M = $6.53, SD = 19.10), F(1,465) = 1.21, p = .27. This pattern of re-sults can be seen in Fig. 3. No participants reported any suspicionabout the advertisements. The pattern of findings for Study 3 re-mained significant after controlling for age and gender.

Study 3 demonstrated that mortality-related advertisementsled participants to spend more on luxury products, but not non-luxury products.

5. Study 4

The purpose of Study 4 was to show that the effects of banneradvertisements on consumer behavior were mediated by death-re-lated thoughts. Previous research shows that exposure to manipu-lations of mortality salience increases death-thought accessibility,which in turn increases personal spending intentions (Fransen,Fennis, Pruyn, & Das, 2008). Study 4 can ensure that the relation-ship between our mortality-based advertisements and consumerbehavior is indeed caused by the mechanism1 we propose—in-creased mortality salience (Das, Bushman, Bezemer, Kerkhof, & Ver-meulen, 2009; Pyszczynski et al., 1999).

5.1. Method

5.1.1. ParticipantsFive hundred and twenty-one participants (53.2% female,

Mage = 35.97, SD = 13.52) were recruited through Amazon’s Mturkand compensated $.15. Participation was limited to those in theUnited States. All respondents finished in less than 10 min.

5.1.2. Procedure and materialsThe manipulation of mortality salience was identical to the first

three studies. Death-thought accessibility was measured using the

1 One possible hypothesis is that our mortality salience manipulation may alsoincrease personal fear of death (Mikulincer & Florian, 2000; Pyszczynski, Greenberg, &Solomon, 1999). To determine whether mortality-based advertisements also increasepersonal fear of death, in a follow-up study (N = 204 volunteer Internet respondents)we manipulated mortality salience in an identical way to Studies 1–4 and measuredparticipants’ personal fear of death (Florian & Kravetz, 1983). Participants in themortality salience condition reported greater fear of death (M = 3.70, SD = .92) thanthose in the control condition (M = 3.37, SD = .89), t(202) = 2.66, d = .37, p < .01.

,

same word-fragment completion task from Study 1 (Greenberget al., 1994). Materialism was measured using the same designand products from Study 3 (Kesebir et al., 2013). Total willingnessto spend money for luxury items was computed by averaging thedollar amount across the items.

5.2. Results and discussion

Replicating the effects we observed in Study 1, participants inthe mortality salience condition completed more death-relatedwords (M = 2.25, SD = 1.14) than participants in the control condi-tion (M = 1.89, SD = 1.07), t(516) = 3.67, d = .32, p < .001. Replicat-ing Study 3, a significant condition � consumer goods interactionemerged from the repeated measures ANOVA, F(1,519) = 4.10,p < .05. Participants in the mortality salience condition were will-ing to spend more money on luxury products (M = $511.07,SD = 532.21) than those in the control condition (M = $417.48,SD = 495.74), F(1,519) = 4.32, d = .18, p < .05. Also as expected, will-ingness to spend money on non-luxury products did not differ be-tween the mortality salience (M = $9.76, SD = 36.75) and thecontrol condition (M = $7.28, SD = 24.38), F(1,519) = .82, p = .36.

To test whether death-related thoughts mediated the associa-tion between mortality salience and willingness to spend on luxuryitems, we conducted a regression-based mediation analysis (Baron& Kenny, 1986). As shown in Fig. 4, the effect of death-relatedthoughts on willingness to spend remained significant aftercontrolling for mortality salience. However, the effect of mortalitysalience was reduced to non-significance when controlling fordeath-related thoughts, Sobel test = �2.23, p < .05. Thus, ourfindings provide evidence that increases in death-related thoughtsmediated the effects of mortality salience on willingness to spendon luxury items. No participants reported any suspicion about theadvertisements. The pattern of findings for Study 4 remainedsignificant after controlling for age and gender.

6. Discussion

Four experiments demonstrated the effects of mortality-relatedadvertisements on social and consumer behavior. In Studies 1 and2, mortality salience manipulations led to higher death-thoughtaccessibility and greater worldview defense. In Studies 3 and 4,participants in the mortality salience condition were willing tospend more money on luxury products but not non-luxury prod-ucts, and this relationship was mediated by death-thought accessi-bility. These findings are consistent with predictions made fromterror management theory. To our knowledge, our study is the firstto use webpage advertisements, which are commonly encounteredonline, to induce a psychological state that increases people’s will-ingness to pay for expensive items.

Our findings reveal that some of the most central effects of mor-tality salience were observed using webpage-based manipula-tions—higher death-thought accessibility (Greenberg et al., 1994),greater cultural worldview defense (Greenberg et al., 1992), and

-.09* (-.07)

.13** -.16**

Death-Related Thoughts

Mortality Salience Condition

Willingness to Pay for Luxury Items

Fig. 4. Study 4. Standardized regression coefficients representing the relationship between mortality-based advertisements, willingness to pay for luxury items, and death-related thoughts. The effect of mortality salience on willingness to pay for luxury items, controlling for death-related thoughts, is in parentheses. �p < .05; ��p < .01.

98 W.J. Chopik, R.S. Edelstein / Computers in Human Behavior 31 (2014) 94–99

increased willingness to pay for luxurious items (Arndt et al.,2004). Thus, researchers need not be confined to laboratory manip-ulations (Rosenblatt et al., 1989) and limited by small sample sizeswhen examining terror management effects. Future work will al-low researchers to test hypotheses about online behavior moreeasily and investigate how online behavior of all types can resem-ble that of the real world (Chandler, Konrath, & Schwarz, 2009). Forexample, previous research demonstrates that exposure to well-known insurance brands leads to many of the same consequencesas traditional manipulations of mortality salience: increased per-sonal spending intentions, higher charitable donations, and prefer-ence for domestic products over foreign products (Fransen et al.,2008). Follow-up studies will be able to test whether webpage orTV advertisements for insurance companies elicit the same effectsthat were observed in this study.

Of course, because our measure of consumer behavior involvedhypothetical products, our data cannot speak to how much peoplewould spend on products when actually confronted with the deci-sion to purchase these items. However, studies in other domains(such as close relationships) demonstrate that people respond insimilar ways when confronted with hypothetical and real events(Fincham & Beach, 1988), so self-reports of willingness to spendmoney on these products may still be very useful. Some research-ers also note that there are several scenarios in which intentions toperform a certain behavior do not predict actual behavior (e.g., Aj-zen, 1991), particularly with measures of willingness to pay(‘‘WTP’’; Kahneman, Ritov, Jacowitz, & Grant, 1993). Critics of theWTP measure suggest that it reflects an expression of an individ-ual’s attitude or liking towards the stimuli rather than an actualintention to pay a specific amount of money. Importantly, in Stud-ies 3 and 4, we were interested primarily in the value participantsplaced on luxurious goods rather than the absolute amount ofmoney they would be willing to spend on these goods. To thisend, willingness to pay is an appropriate measure in testing thehypothesis that, after a mortality salience manipulation, individu-als would place more value on the acquisition of luxury goods (ortheir ‘‘attitude’’ towards luxurious products). Also, measuring thevalue that individuals place on luxury goods is entirely consistentwith Kahneman et al.’s (1993) observation that willingness to paymay reflect more of a general liking or evaluation of a stimuli(Brough & Chernev, 2012). Nevertheless, future studies can addressthis limitation by providing participants with real money and ask-ing them to selectively spend it on luxury items, or better yet, col-lect actual online shopping behavior data.

Another limitation of the present research is that we did not askany explicit questions about the believability of our webpageadvertisements or whether they considered our advertisementsto be legitimate advertisements for university activities. Also, thedegree to which individuals were actively or passively attendingto the advertisements was not measured or manipulated. It isimportant to note that, across four studies, participants did not re-port any suspicion about the webpage advertisement or the pur-pose of the studies. Nevertheless, future research can manipulatespecific aspects of the advertisements (e.g., the font, images) to

see if this increases or decreases participants’ conscious processingor attention to them. Future research can also utilize eye-trackingtechnology to pinpoint the exact elements (e.g., the word ‘‘dead’’ orthe image of a skull) of the advertisements that participants attendto and how attending to these elements yields the psychologicaloutcomes observed in the current studies.

Our results from the webpage-based manipulations are consis-tent with a small body of literature suggesting that the content ofadvertisements is influential even when individuals do not activelyattend to them (Yoo, 2009). It is important to note that, in each ofthe studies, no participants reported any indication that the adver-tisements were part of the study. Studies 3 and 4, in particular,have many important implications for how individuals buy and ex-change products online. The design of these studies was meant tomirror the context of online shopping, namely that it is done at anindividual’s leisure, online, and from a place of their choosing. Be-cause seeking out material gain is one of the many ways in whichindividuals alleviate the anxiety of death (Mandel & Heine, 1999),behavior in the online marketplace may be susceptible to influ-ences of which companies, advertisers, and consumers areunaware.

Taken together, the results of these four studies suggest thateven the subtlest stimuli we encounter can have profound effectson our behavior (buyers beware!). So the next time you decideto surf the web, be wary of more than just the space betweenthe advertisements!

Acknowledgement

The first author was supported by a National Science Founda-tion Graduate Research Fellowship.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.chb.2013.10.022.

References

Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and HumanDecision Processes, 50, 179–211.

Arndt, J., Greenberg, J., Pyszczynski, T., & Solomon, S. (1997). Subliminal exposure todeath-related stimuli increases defense of the cultural worldview. PsychologicalScience, 8, 379–385.

Arndt, J., Solomon, S., Kasser, T., & Sheldon, K. M. (2004). The urge to splurgerevisited: Further reflections on applying terror management theory tomaterialism and consumer behavior. Journal of Consumer Psychology, 14,225–229.

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction insocial psychological research: Conceptual, strategic, and statisticalconsiderations. Journal of Personality and Social Psychology, 51, 1173–1182.

Brough, A. R., & Chernev, A. (2012). When opposites detract: Categorical reasoningand subtractive valuations of product combinations. Journal of ConsumerResearch, 39, 399–414.

Buhrmester, M. D., Kwang, T., & Gosling, S. D. (2011). Amazon’s Mechanical Turk: Anew source of inexpensive, yet high-quality data? Perspectives on PsychologicalScience, 6, 3–5.

W.J. Chopik, R.S. Edelstein / Computers in Human Behavior 31 (2014) 94–99 99

Chandler, J., Konrath, S., & Schwarz, N. (2009). Online and on my mind: Temporaryand chronic accessibility moderate the influence of media figures. MediaPsychology, 12, 210–226.

Das, E., Bushman, B. J., Bezemer, M. D., Kerkhof, P., & Vermeulen, I. E. (2009). Howterrorism news reports increase prejudice against outgroups: A terrormanagement account. Journal of Experimental Social Psychology, 45, 453–459.

Fincham, F. D., & Beach, S. R. (1988). Attribution processes in distressed andnondistressed couples: V. Real versus hypothetical events. Cognitive Therapyand Research, 12, 505–514.

Florian, V., & Kravetz, S. (1983). Fear of personal death: Attribution, structure, andrelation to religious belief. Journal of Personality and Social Psychology, 44,600–607.

Fransen, M. L., Fennis, B. M., Pruyn, A. T. H., & Das, E. (2008). Rest in peace? Brand-induced mortality salience and consumer behavior. Journal of Business Research,61, 1053–1061.

Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The causes and consequences ofa need for self-esteem: A terror management theory. In R. Baumeister (Ed.),Public self and private self (pp. 189–192). New York/Berlin: Springer-Verlag.

Greenberg, J., Pyszczynski, T., Solomon, S., Rosenblatt, A., Veeder, M., Kirkland, S.,et al. (1990). Evidence for terror management theory II: The effects of mortalitysalience on reactions to those who threaten or bolster the cultural worldview.Journal of Personality and Social Psychology, 58, 308–318.

Greenberg, J., Pyszczynski, T., Solomon, S., Simon, L., & Breus, M. (1994). Role ofconsciousness and accessibility of death-related thoughts in mortality salienceeffects. Journal of Personality and Social Psychology, 67, 627–637.

Greenberg, J., Simon, L., Pyszczynski, T., Solomon, S., & Chatel, D. (1992). Terrormanagement and tolerance: Does mortality salience always intensify negativereactions to others who threaten one’s worldview? Journal of Personality andSocial Psychology, 63, 212–220.

Higgins, E. T., Rholes, W. S., & Jones, C. R. (1977). Category accessibility andimpression formation. Journal of Experimental Social Psychology, 13, 141–154.

Hirschberger, G., Ein-Dor, T., & Almakias, S. (2008). The self-protective altruist:Terror management and the ambivalent nature of prosocial behavior.Personality and Social Psychology Bulletin, 34, 666–678.

Jessop, D. C., Albery, I. P., Rutter, J., & Garrod, H. (2008). Understanding the impact ofmortality-related health-risk information: A terror management theoryperspective. Personality and Social Psychology Bulletin, 34, 951–964.

Kahneman, D., Ritov, I., Jacowitz, K. E., & Grant, P. (1993). Stated willingness to payfor public goods: A psychological perspective. Psychological Science, 4, 310–315.

Kasser, T., & Sheldon, K. M. (2000). Of wealth and death: Materialism, mortalitysalience, and consumption behavior. Psychological Science, 11, 348–351.

Kesebir, P., Chiu, C. -Y., & Pyszczynski, T. (submitted for publication). The sacred: Anexistential anxiety buffer.

Mandel, N., & Heine, S. J. (1999). Terror management and marketing: He who dieswith the most toys wins. Advances in Consumer Research, 26, 527–532.

Mikulincer, M., & Florian, V. (2000). Exploring individual differences in reactions tomortality salience: Does attachment style regulate terror managementmechanisms? Journal of Personality and Social Psychology, 79, 260–273.

Paolacci, G., Chandler, J., & Ipeirotis, P. G. (2010). Running experiments on AmazonMechanical Turk. Judgment and Decision Making, 5, 411–419.

Pyszczynski, T., Greenberg, J., & Solomon, S. (1999). A dual-process model of defenseagainst conscious and unconscious death-related thoughts: An extension ofterror management theory. Psychological Review, 106, 835–845.

Pyszczynski, T., Wicklund, R. A., Floresku, S., & Koch, H. (1996). Whistling in thedark: Exaggerated consensus estimates in response to incidental reminders ofmortality. Psychological Science, 7, 332–336.

Rosenblatt, A., Greenberg, J., Solomon, S., Pyszczynski, T., & Lyon, D. (1989).Evidence for terror management theory: I. The effects of mortality salience onreactions to those who violate or uphold cultural values. Journal of Personalityand Social Psychology, 57, 681–690.

Schmeichel, B. J., Gailliot, M. T., Filardo, E.-A., McGregor, I., Gitter, S., & Baumeister,R. F. (2009). Terror management theory and self-esteem revisited: The roles ofimplicit and explicit self-esteem in mortality salience effects. Journal ofPersonality and Social Psychology, 96, 1077–1087.

Solomon, S., Greenberg, J., & Pyszczynski, T. (2004). The cultural animal: Twentyyears of terror management theory and research. In J. Greenberg, S. L. Koole, & T.Pyszczynski (Eds.), Handbook of experimental existential psychology (pp. 13–34).New York, NY: Guilford Press.

Yoo, C. Y. (2008). Unconscious processing of web advertising: Effects on implicitmemory, attitude toward the brand, and consideration set. Journal of InteractiveMarketing, 22, 2–18.

Yoo, C. Y. (2009). Effects beyond click-through: Incidental exposure to webadvertising. Journal of Marketing Communications, 15, 227–246.