A Theoretical and Empirical Review of the Death-Thought Accessibility Concept in Terror Management Research

Page 1

A Theoretical and Empirical Review of the Death-Thought AccessibilityConcept in Terror Management Research

Joseph Hayes and Jeff SchimelUniversity of Alberta

Jamie ArndtUniversity of Missouri—Columbia

Erik H. FaucherUniversity of Alberta

Terror management theory (TMT) highlights the motivational impact of thoughts of death in variousaspects of everyday life. Since its inception in 1986, research on TMT has undergone a slight butsignificant shift from an almost exclusive focus on the manipulation of thoughts of death to a markedincrease in studies that measure the accessibility of death-related cognition. Indeed, the number ofdeath-thought accessibility (DTA) studies in the published literature has grown substantially in recentyears. In light of this increasing reliance on the DTA concept, the present article is meant to provide acomprehensive theoretical and empirical review of the literature employing this concept. After discussingthe roots of DTA, the authors outline the theoretical refinements to TMT that have accompaniedsignificant research findings associated with the DTA concept. Four distinct categories (mortalitysalience, death association, anxiety-buffer threat, and dispositional) are derived to organize the reviewedDTA studies, and the theoretical implications of each category are discussed. Finally, a number oflingering empirical and theoretical issues in the DTA literature are discussed with the aim of stimulatingand focusing future research on DTA specifically and TMT in general.

Keywords: death-thought accessibility, terror management theory, implicit social cognition, existentialpsychology

According to terror management theory (TMT; Greenberg,Pyszczynski, & Solomon, 1986), much of our daily behavior ismotivated by ongoing unconscious concerns about death. Becausethe conscious experience of such thoughts triggers the potential foranxiety, the human psyche responds with motivated avoidance. Inother words, the mind contains mechanisms that generally keepthoughts of death from becoming conscious and remove suchthoughts from focal attention when they do. The workings of thesemechanisms can be observed through indirect testing proceduresdesigned to enable inferences about the level of activation of deaththoughts, referred to as death-thought accessibility (DTA).

Over the years, the DTA concept has proved instrumental in thecontinued development of TMT. It has provided key insights intoissues emanating from existing literature, prompted new researchdirections, and in so doing provided novel insights into diverse

aspects of human social behavior. Indeed, the introduction of theDTA concept, along with a paradigm to measure it, facilitatedinvestigations into the precise cognitive processes involved inmanaging the awareness of death (Greenberg, Pyszczynski,Solomon, Simon, & Breus, 1994). This innovation ultimately ledto a refinement of the theory to distinguish between proximal(conscious, threat-focused) and distal (unconscious, symbolic) de-fenses against the awareness of mortality (Pyszczynski, Green-berg, & Solomon, 1999), which in turn has inspired additionalapplications and insights across diverse areas. In short, the DTAconcept is becoming an increasingly important tool for examiningthe role that death plays in a variety of life domains.

Due to the importance of the DTA concept, empirical studies ofTMT have undergone a slight but significant shift in recent yearsfrom use of thoughts of death almost exclusively as an independentvariable (the mortality salience manipulation) to a marked increasein the measurement of DTA under various conditions (e.g., fol-lowing threat). Given this increased reliance on DTA, a compre-hensive theoretical and empirical review is needed to organizeexisting knowledge concerning the DTA concept. As such, wecollected what we believe to be all of the extant DTA studies byperforming a PsycINFO search using various combinations of thekeywords death-thought accessibility, terror management theory,death, accessibility, and mortality salience and by gathering inpress and unpublished/forthcoming manuscripts from our own labsand from other researchers currently investigating terror manage-ment processes. The DTA concept has now been employed in oversixty published papers, totaling more than ninety empirical studies,

Joseph Hayes, Jeff Schimel, and Erik H. Faucher, Department of Psy-chology, University of Alberta, Edmonton, Alberta, Canada; Jamie Arndt,Department of Psychological Sciences, University of Missouri—Columbia.

This research was supported in part by a J. Armand Bombardier CanadaGraduate Scholarship awarded to Joseph Hayes and by Social Sciences andHumanities Research Council Standard Research Grant G124130386awarded to Jeff Schimel.

Correspondence concerning this article should be addressed to JosephHayes, University of Alberta, Department of Psychology, P-217 BiologicalSciences Building, Edmonton, Alberta T6G 2E9, Canada. E-mail:[email protected]

Psychological Bulletin © 2010 American Psychological Association2010, Vol. 136, No. 5, 699–739 0033-2909/10/$12.00 DOI: 10.1037/a0020524

699

Page 2

and these numbers are growing rapidly (see Figure 1). Indeed, thenumber of DTA studies published in just the last 3 years (2007–2009) now exceeds the total number published in the 13 yearsfollowing the initial DTA study (i.e., Greenberg et al., 1994). Thespeed with which such publications are emerging only underscoresthe importance of gathering and integrating relevant studies so asto arrive at an understanding of the concept’s contributions andpotential. In this way, unresolved issues from this recent wave ofresearch can be identified, such that direction for future researchcan be provided with the aim of enhancing our understanding ofthe role of existential cognition in social behavior.

In the pages that follow, we review and synthesize the extantDTA studies into a coherent and integrated body of research. Webegin with a brief overview of TMT and the postulates that gaverise to the DTA concept. In so doing, we chronologically trace thedevelopment of DTA as a measureable construct in studies usingthe mortality salience paradigm. Next, we review the extant DTAstudies, highlighting the contributions they have made to under-standing the role of existential concerns in a wide variety of socialdomains. On the basis of this review, we sum up the theoreticaland empirical significance of the DTA concept and provide a muchneeded update to the process model proposed by Pyszczynski et al.(1999; see also Arndt, Cook, & Routledge, 2004; Arndt, Green-berg, & Cook, 2002). Finally, we discuss lingering issues in theDTA literature in hope of stimulating future research on DTA andits relation to other psychological processes and phenomena.

TMT

TMT was derived largely from the writings of Ernest Becker(1971, 1973, 1975), who emphasized the important influence that theknowledge of death has on the way in which people live. He arguedthat although human beings are no different from other animals inpossessing a biological drive to preserve our existence, we differ withregard to our complex mental abilities, through which we are capableof becoming explicitly conscious of ourselves, the world around us,

and future possibilities (cf. Fingarette, 1969/2000). From this perspec-tive, awareness of the self as an object among objects, and of the factthat like other living organisms, all human beings must inevitably die,has been particularly important to the psychological and social devel-opment of the species. Becker argued that the knowledge of death,coupled with a basic desire for continued life, creates an existentialdilemma capable of producing potentially paralyzing anxiety, or ter-ror. However, death anxiety is not experienced by most people on adaily basis, nor is the conscious thought of death particularly preva-lent. Rather, thoughts of death and the attendant anxiety remainunconscious, because the mind employs defenses to cover them upand to keep such concerns out of explicit consciousness. The defen-sive structures employed to manage thoughts of death can be broadlyconceptualized as cultural worldviews and self-esteem.

A cultural worldview is a large-scale system of beliefs regardingthe nature of reality that quells anxiety about death in a number ofways. First, it structures the environment by implying the universeis orderly and stable. This has a tranquilizing effect by renderingthe universe intelligible and by fostering an implicit belief thatdeath is not an immediate possibility and is therefore not a causefor concern. A cultural mode of existence also pulls the individualin and occupies his or her time and thought, providing littleopportunity for explicit death contemplation. But perhaps mostimportant, a cultural worldview manages death anxiety by provid-ing meaning and purpose to life by delineating how life should belived. By prescribing a set of standards for evaluating the manypossible ways of being, it provides a basis for the feeling that onehas personal significance and value (i.e., self-esteem). Finally, forthe times in which thoughts of death cannot be avoided, culturalworldviews also provide a place for death, often suggesting that itis not the ultimate end of existence. Religious worldviews, forexample, dictate that life continues in a hereafter, and theyoutline the way in which one must live to qualify for theafterlife. Secular worldviews can also provide death transcen-dence, albeit of a symbolic form. Whereas the life of an indi-

18

20

s

14

16

DTA

stud

ie

8

10

12

ublis

hed

D

4

6

8

umbe

r of

pu

0

2

4

Nu

Year of Publication

Figure 1. Number of published DTA studies by year of publication (up to end of 2009). Studies included inthis figure are detailed in Appendices A–D. DTA � death-thought accessibility.

700 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 3

vidual may be short and fleeting, cultures and nations canendure.

In summary, according to TMT, given our biological propensity tolive, the knowledge of our impending death creates an existentialdilemma characterized by an ever looming potential for paralyzinganxiety. This anxiety is managed and kept largely unconscious by adual-component buffering mechanism consisting of (a) a culturalworldview and (b) self-esteem. Due to their death-anxiety bufferingproperties, these components are vigorously defended against threats,especially when people have been reminded of death.

The Birth of the DTA Concept: From ManipulatingConscious Thoughts of Death to Measuring the

Nonconscious Accessibility of Death-Related Cognition

Despite its diverse roots within psychology and widespreadapplication across fields of study from nursing to literature tocybernetics, TMT has relied heavily on experimental methodolo-gies to assess the hypotheses derived from it. This is one of TMT’smost significant offerings to the theoretical synthesis proposed byBecker, as the use of such procedures has led to conceptualadvances that may not have been possible from theoretical con-templation alone. Among the contributions TMT has made to theexistential–psychodynamic tradition, one of the most importanthas come from the ability to measure and thus illuminate thepotential role that nonconscious concerns about death have ineveryday life domains.

Manipulating Conscious Thoughts of Death:Mortality Salience

Beginning with the first empirical investigation of TMT (Rosen-blatt, Greenberg, Solomon, Pyszczynski, & Lyon, 1989), most ofthe research on this topic has manipulated conscious thoughts ofdeath. This is often done by having participants answer two open-ended questions asking them to (a) describe the emotions that thethought of their own death arouses in them and (b) imagine whatwill happen to them physically as they die. This procedure, re-ferred to as the mortality salience (MS) induction,1 is typicallyfollowed by exercises that introduce some kind of delay beforedependent measures are assessed. MS studies employ controlconditions in which participants answer two similarly wordedquestions about a topic other than death. This typically involvesthoughts of aversive topics, such as experiencing various kinds ofphysical pain, uncertainty, expectancy violations, or social exclu-sion or imagining a stressful future event (such as taking animportant exam), but has also included neutral, nonemotive topics,such as watching television or eating food. Although there havebeen some demonstrations of other aversive topics or events show-ing effects similar to MS (e.g., Proulx & Heine, 2008; van den Bos,Poortvliet, Maas, Miedema, & van den Ham, 2005), the prepon-derance of evidence indicates that the effects of MS cannot beexplained by elicitations of other aversive experiences (or negativeaffect/arousal, priming of cultural values, and a number of otheralternatives; see, e.g., Greenberg, Solomon, & Arndt, 2008). Infact, recent meta-analytic data suggest that although other types ofmeaning threats can evoke defensive responses similar to thoseobserved for MS, the effects produced by MS follow a unique,signature time course, consistent with the process model outlined

by Pyszczynski et al. (1999) and discussed in the following pages,that is not parallel to that of the effects produced by meaning-related threats (Martens, Burke, Schimel & Faucher, in press). Thisevidence suggests that conscious death reminders constitute aunique psychological threat.

Terror management studies that manipulate conscious thoughtsof death have typically been guided by the MS hypothesis. The MShypothesis states that if a psychological structure functions tobuffer awareness of death, inducing people to think of their deathshould increase their need for this psychological structure. In otherwords, reminding people of their mortality should increase theirneed for self-esteem and their faith in the cultural worldview.Accordingly, such reminders have been found to produce a varietyof reactions directed toward securing self-esteem and faith in thecultural worldview (for reviews, see, e.g., Greenberg et al., 2008;Pyszczynski, Greenberg, Solomon, Arndt, & Schimel, 2004). Forexample, Rosenblatt et al. (1989) found that having Americanparticipants think about death subsequently led to more negativereactions toward someone who was disparaging toward the UnitedStates and more positive reactions toward someone who hadupheld American values. This effect (and those like it) has beenobserved in a variety of countries and is now commonly referredto as worldview defense.

Suppression of Conscious Thoughts of Death

Although the effects of MS on cultural worldview defense werewell documented in the early days of the empirical development ofTMT, the theory did not offer a precise cognitive model of howMS ultimately produces such defenses until the mid-1990s. Some-what puzzled by the consistent lack of affective impact in responseto MS, TMT researchers began to explore the microlevel cognitiveeffects of consciously inducing thoughts of death. They reasonedthat if thoughts of death are really so troubling, participants whohave finished writing about the topic of death must attempt tobanish death-related thoughts from conscious awareness to preventdeath anxiety from taking hold. In other words, people engage inefforts to suppress thoughts of death, either by concentrating onother things (thereby providing distraction) or by minimizing theimportance of the issue so that it can be dropped from conscious-ness altogether.

Wegner’s (1994) seminal research on thought suppression sug-gests that such efforts may lead to some interesting and ironicconsequences, as actively attempting to remove a thought fromconsciousness can increase the likelihood that the thought willreturn once the suppressive activity ceases. Thus, although sup-pression may succeed in temporarily eliminating a thought fromconsciousness, the process ultimately makes the thought hyperac-cessible, meaning that it is highly active, albeit unconsciously so,

1 Examples of other operationalizations of MS include having partici-pants complete the Death-Anxiety Scale (Conte, Weiner, & Plutchik,1982), which assesses attitudes toward death and dying and thus inducesconscious contemplation of death (e.g., Arndt, Greenberg, Simon, Pyszc-zynski, & Solomon, 1998; Greenberg et al., 1995); having participantscomplete word search puzzles in which they search for death-related words(e.g., Jonas et al., 2008; Maxfield et al., 2007); and interviewing people inclose proximity to a funeral home or cemetery (e.g., Pyszczynski et al.,1996).

701DEATH-THOUGHT ACCESSIBILITY

Page 4

and capable of popping back into consciousness with relative ease(see also Wegner & Smart, 1997).

In keeping with Wegner’s theorizing and research, Greenberg etal. (1994) reasoned that if conscious contemplation of deathprompts suppression of the death-related material, thoughts ofdeath may become highly accessible in the unconscious after adelay, once suppressive activity relaxes, and therefore easily ca-pable of reentering consciousness. This logic led them to hypoth-esize that worldview defenses might result from these uncon-scious, yet highly accessible, thoughts of death.

Greenberg et al. (1994, Study 1) obtained initial support for thisreasoning by demonstrating that worldview defense was most pro-nounced in response to a subtle and brief encounter with deaththoughts. When the death-thought induction was more blatant, how-ever, requiring participants to imagine a more vivid encounter withdeath, worldview defense was absent. Presumably, participants in thisblatant death condition would still have been consciously attending tothoughts of death, whereas those in the subtle death condition wouldlikely have removed them from consciousness. If thoughts of deathinduce worldview defense only when they are unconscious but highlyaccessible, inducing people to continue thinking about death after thetypical MS induction should eliminate the effect of MS on worldviewdefense. This is precisely what was observed in Studies 2 and 3.Relative to participants who were distracted following the MS ma-nipulation, those who were required to keep thoughts of death inconsciousness evidenced low levels of worldview defense. Thus,when participants are able to suppress thoughts of death followingMS, such thoughts should become hyperaccessible in the unconsciousmind after a brief delay, at which point worldview defenses areinitiated.

Measuring Unconscious Thoughts of Death: DTA

To more directly confirm the notion that death thoughts first aresuppressed and then become hyperaccessible, Greenberg and col-leagues needed a way to measure unconscious DTA. Fortunately,projective testing procedures for measuring unconscious thought ac-tivation already existed. Although classic measures, such as the Ror-schach inkblot technique (Rorschach, 1921/1942) and the ThematicApperception Test (Murray, 1938/2008), had been in use for over halfa century, a number of newer techniques were emerging that mea-sured unconscious cognitive activation in terms of perceptual readi-ness or accessibility to consciousness (cf. Bruner, 1957; Higgins &King, 1981). These techniques, like the classic measures, are based onthe idea that a person’s underlying mental contents can be indirectlyinferred on the basis of responses to an unrelated task. In other words,when a particular construct is highly active in the unconscious mind,it can influence the individual’s interpretation of ambiguous events orspeed of responding on a judgment task, even if the individual has noconscious access to this material (for reviews, see, e.g., Bargh &Chartrand, 2000; Wegner & Smart, 1997).

Greenberg et al. (1994) opted for the use of ambiguous wordfragments, a technique that was successfully used in research onpriming effects (e.g., Tulving, Schacter, & Stark, 1982). Given thata fragment can be completed as different words, the way in whichparticipants complete word fragments can shed light on thethoughts that are on the edge of consciousness and likely to cometo mind. For example, in the present context, the fragment SK_ _Lcan be completed as either skill or skull. By observing the number

of fragments that participants complete in the death-related man-ner, one can infer whether thoughts of death are more or lessaccessible. To enhance the subtlety of the measure, Greenberg etal. embedded six possible death fragments among a group ofnumerous other word fragments that could be completed only in anon-death-related manner. Research participants are instructed tocomplete the word fragments as quickly as possible with the firstword that pops into their minds.

The Effect of Mortality Salience on Death-ThoughtAccessibility

Having appropriated a method of measuring DTA, Greenberg etal. (1994, Study 4) assessed the hypothesis that MS instigatesimmediate death-thought suppression, followed by a delayed in-crease in DTA. As predicted, when mortality was made salient,participants completed more word fragments as death words (in-dicating high DTA) only after a short delay. When measuredimmediately after the MS induction, DTA remained relatively low,presumably because participants were still engaged in active sup-pression of the death-related content.

The Continuing Contribution of the DTAConcept to TMT

Given that unconscious thoughts of death are central to theconceptual framework posited by TMT, having a way to measuresuch thoughts was an important step in the theory’s empiricaldevelopment. The acquisition of a measurable DTA concept sub-stantially increased the types of studies that could be conducted totest the basic tenets of the theory. Not surprisingly, this develop-ment has aided in a number of important theoretical innovationsand has vastly improved our understanding of how awareness ofdeath affects social judgment and behavior.

In the subsections that follow, we review the lion’s share of extantDTA studies. We divide them into four distinct categories for ease ofpresentation. A complete list and detailed summary of the publishedstudies falling into each category are presented in Appendices A–Dfor easy reference.2 We begin with the mortality salience DTA stud-ies, which provide insight into the psychological processes stemmingfrom the MS manipulation, followed by the death-association DTAstudies, which show that topics that are associated with death caninitiate terror management processes by virtue of either makingthoughts of death explicit or implicitly increasing DTA. Next, wereview the anxiety-buffer threat DTA studies, which demonstrate thatsignificant threats to the anxiety buffer (i.e., cultural worldviews andself-esteem) can bring mortality concerns closer to consciousness.Finally, we review the dispositional DTA studies, which show thatindividual differences in threat-buffering capability can influencebaseline levels of DTA, which can in turn predict worldview andself-esteem defensiveness.

It is important to note that these categories are not meant to bemutually exclusive. Some DTA studies could potentially fit into morethan one category, and in some cases, the category a particular study

2 Several studies that are mentioned more specifically in the method-ological and theoretical issues sections at the end of this paper are notincluded in this part of the review to avoid redundancy.

702 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 5

(or set of studies) falls into can be relatively ambiguous. With thissaid, however, we believe that the categories are nevertheless theo-retically meaningful in that each group provides a unique perspectiveon terror management processes. In presenting the categories below,we highlight the specific theoretical implications that these groups ofstudies provide at the end of each subsection.

The Mortality Salience DTA Studies

Like Greenberg et al.’s (1994) initial study, a large number ofDTA studies are also MS studies. For the most part, the inclusionof a DTA measure in these studies is designed to explore questionsabout the psychological processes involved in going from con-scious thoughts of death to worldview and self-esteem defenses.As mentioned above, conscious contemplation of death is believedto instigate efforts to remove thoughts of death from focal aware-ness, producing initially low levels of DTA while suppression orother such efforts are ongoing but ultimately resulting in highlevels of DTA once these efforts subside. Moreover, worldviewand self-esteem defenses are believed to result from these highlevels of unconscious DTA, rather than conscious thoughts ofdeath per se. Although Greenberg et al. provided initial support forthis sequence, more empirical work was needed to validate thishypothesized model.

Death-thought suppression. Arndt, Greenberg, Solomon,Pyszczynski, and Simon (1997) reasoned that if MS does indeedprompt suppression of death thoughts, denying people the cogni-tive resources required for suppressive activity should lead to animmediate increase in DTA (cf. Wegner, 1994). As expected, intheir Study 1, when MS was coupled with a cognitive load ma-nipulation, effectively tying up controlled processing resourcesand thereby hindering suppression, DTA was high immediatelyafter MS—without any need for the delay observed by Greenberget al. (1994). Arndt, Greenberg, Pyszczynski, and Solomon (1997)demonstrated that subliminal exposure to a death prime also in-creased DTA without delay. Because conscious attention is nec-essary to instigate suppressive activity, presenting the death stim-ulus below the threshold of consciousness precluded suppression.

Thinking about the self dying necessarily entails thinking about theself. Thus, focusing attention away from the self after contemplatingdeath can be an effective strategy of removing death thoughts fromconsciousness. In support of this idea, Arndt, Greenberg, Simon,Pyszczynski, and Solomon (1998) found that when participants wroteabout death in a cubicle that contained a large mirror, they spent lesstime writing than when there was no mirror present. Moreover,inducing participants to maintain self-focus after thinking about deathextended suppressive activity and resulted in lower levels of DTA,even after a delay. By contrast, fostering an external focus of attentionafter MS facilitated the removal of death thoughts from conscious-ness, leading to higher levels of DTA.

Moderating the effect of MS on DTA. Given the convergencebetween the effects of MS on increased DTA and investment insecurity providing domains of meaning and value, it follows thatindices of such psychological security should moderate the MS–DTAeffect. A range of studies is consistent with this prediction.

Fortified self-esteem and belief in the cultural worldview.According to TMT, faith in a cultural worldview and high levels ofself-esteem are two central means by which people acquire psy-chological protection against thoughts of death. Thus, fortifying

these structures should reduce the impact of MS on DTA. Insupport of this reasoning, Harmon-Jones et al. (1997) found thatbolstering participants’ self-esteem by administering positive per-sonality feedback prior to the MS manipulation subsequently elim-inated the usual delayed increase in DTA. The effect of MS onDTA is also eliminated when participants affirm a valued aspect ofthe cultural worldview prior to the MS induction (Jonas & Fischer,2006; Schmeichel & Martens, 2005). More recently, in their workon nostalgia, Routledge, Arndt, Sedikides, and Wildschut (2008)argued that remembering the past can provide meaning to life andhelp manage existential concerns. Accordingly, participants whowere high in nostalgia proneness (Study 2), or those who had beeninduced to reminisce about the past prior to the MS manipulation(Study 3), did not show the usual delayed increase in DTA fol-lowing MS. Finally, in arguing that the notion of progress is aculturally constructed yet vital defensive illusion that reducesmortality concerns, Rutjens, van der Pligt, and van Harreveld(2009, Study 3) demonstrated that strengthening faith in progress,by having people read an essay supporting the notion that civili-zation provides continuous improvement to the quality of life, alsobuffers the effects of MS on DTA.

Close relationships and the role of attachment style. In recentyears, some researchers have maintained that close relationshipsare an additional anxiety-buffering mechanism (see Hart, Shaver,& Goldenberg, 2005; Mikulincer, Florian, & Hirschberger, 2003).Indeed, according to Bowlby (1969), the attachment mechanism isthe primary means through which the child derives security fromthe caregiver. Furthermore, Becker (1971) argued that self-esteemoriginates from feelings of security that one acquires by living upto the parental standards for good behavior. This basis of securityis later transferred to the broader culture and to significant rela-tionship partners throughout life. Finally, from the parental per-spective, children are a source of symbolic immortality becausethey are representations of the self that will live on after one dies(Lifton, 1979).

In support of these assertions, MS has been found to promotegreater investment in close relationships (Florian, Mikulincer, &Hirschberger, 2002), an increased desire to have children (Fritscheet al., 2007; Zhou, Lei, Marley, & Chen, 2009), and an increaseddesire to affiliate with others (Taubman–Ben-Ari, Findler, &Mikulincer, 2002).3 For the purposes of the present analysis,activating significant relationship bonds prior to thinking aboutdeath can eliminate subsequent increases in DTA. In particular,inducing thoughts of positive support from one’s mother (Cox etal., 2008, Study 1) or viewing images of babies (Zhou et al., 2009,Study 2) eliminated the effect of MS on DTA.

Given evidence that attachment style moderates the effect of MSon worldview defenses (see Mikulincer et al., 2003), it should notbe surprising that it also moderates the effect of MS on DTA.

3 The effect of MS on relationship striving is shown to be moderated byindividual differences in attachment style. People who possess a secureattachment style appear more likely to manage existential concerns byseeking romantic attachment partners. Those who possess an anxious-ambivalent attachment style seem to favor parental attachments and cul-tural worldview defense. Finally, people with an avoidant attachment styleseem to engage only in worldview defense, preferring to avoid relation-ships altogether (Cox et al., 2008; Mikulincer & Florian, 2000).

703DEATH-THOUGHT ACCESSIBILITY

Page 6

Mikulincer and Florian (2000, Study 2) found immediate increasesin DTA following MS among anxious-ambivalent participants butonly after the usual delay for secure and avoidant participants.They reasoned that this pattern likely reflects anxious-ambivalentpeople’s difficulty suppressing troubling thoughts and tendency toruminate on anxiety-provoking stimuli such as death (see Miku-lincer, Florian, & Tolmacz, 1990), which may amount to beingconstantly burdened by cognitive load.

Rational vs. experiential processing. Another factor thatmoderates the effect of MS on DTA involves the way in whichmortality is processed when it is in consciousness. Recall that thedelayed increase in DTA and the consequent increase in world-view defense reactions is hypothesized to occur in a nonconscious,experiential processing system. As such, thinking about death in anintellectualized and rational manner (e.g., like a coroner conduct-ing an autopsy or a scientist investigating the effects of deathcontemplation) can insulate the thinker from the emotional impactthat such thoughts might otherwise produce (Epstein, 1980; Freud,1966). In support of this reasoning, Simon et al. (1997) found thatinducing people to think about death rationally (vs. experientially;see Epstein, 1994) eliminated the effect of MS on DTA. Moreover,participants who thought about death in a rational mode of pro-cessing did not show increased worldview defense after the usualdelay.

Reducing MS-induced DTA through distal defense. In thestudies mentioned so far, distal worldview defense was found toresult from the same conditions that led to increases in DTA (e.g.,when MS is followed by a delay/distractor task, coupled withcognitive load, or is primed subliminally). Moreover, when MS ispaired with trait characteristics (e.g., nostalgia; Routledge et al.,2008) or experimental procedures (e.g., self-affirmation;Schmeichel & Martens, 2005) that activate protective beliefs andfortify the anxiety buffer, both DTA and worldview defense areabsent. These studies thus suggest that distal defense in response toMS results specifically from an increase in DTA. With this inmind, Arndt and colleagues reasoned that if distal defenses servethe ultimate purpose of defending against unconscious DTA, ini-tiating such defenses should eliminate the source of the threat. Inother words, distal defense should reduce the DTA aroused by MS.In support of this reasoning, studies show that, following MS,having the opportunity to defend the cultural worldview by dis-paraging a worldview detractor (Arndt, Greenberg, Solomon, etal., 1997, Study 3), or to enhance self-esteem by making self-serving attributions (Mikulincer & Florian, 2002), results in de-creased levels of DTA relative to conditions in which participantsare not given this opportunity. In fact, among participants whodefended their worldview or enhanced their self-esteem, DTA wasno higher than among participants who were not reminded of deathat all.

Greenberg, Arndt, Schimel, Pyszczynski, and Solomon (2001)clarified an important explanatory ambiguity in such effects, dem-onstrating that the reduction in DTA after worldview defense is notthe result of renewed suppression but rather of the dissipation ofunconscious thoughts of death. When participants were given theopportunity to defend their worldview after MS, DTA was reducedregardless of whether or not they were under cognitive load.Because cognitive load taxes cognitive resources and hinderssuppressive capability, DTA should have remained high undercognitive load even after defense, if lower DTA after worldview

defense was the result of renewed suppression. The absence of thiseffect suggests that defending one’s worldview reduced the un-conscious accessibility of death thoughts.

Mikulincer and Florian (2000) found similar results in theirresearch on the anxiety-buffering function of close relationships.Having the opportunity to defend the cultural worldview after MSreduced DTA only for participants with an avoidant attachmentstyle. Secure participants did not defend the cultural worldviewwhen given the opportunity (because they prefer relationship de-fenses), and DTA therefore remained high among these partici-pants. Anxious-ambivalent participants, on the other hand, didshow increased worldview defense, but this did not seem to haveany effect on DTA, which remained just as high as when they didnot have the opportunity for defense. Again, this is probablybecause of anxious-ambivalent people’s overall difficulty manag-ing anxiety and troubling thoughts.

Implications. Taken as a whole, the above studies suggestthat MS-induced worldview and self-esteem defense results fromhigh levels of DTA, which are reduced when such defenses areinitiated. This insight was important to the development of TMTand ultimately led to the first formal extension of the theory todelineate the precise cognitive processes involved in managingthoughts of death. Pyszczynski et al. (1999; see also Arndt, Cook,& Routledge, 2004) proposed that death thoughts are managedwith two types of defenses, each corresponding to the perceptualsystem in which these thoughts are active. Thoughts of death thatenter consciousness are managed via proximal defenses that arerational in nature and typically involve suppressing the thoughtsthrough distraction, reducing self-focused attention, or trivializingone’s vulnerability to death in the near future. More recent re-search indicates that such responses can also include proactiveefforts to reduce vulnerability by engaging in healthy decisions(Goldenberg & Arndt, 2008). Once these defenses removethoughts of death from consciousness, death thoughts becomenonconsciously accessible. Or alternatively, thoughts of death canbecome activated in the unconscious directly via subliminal per-ception. Either way, when death thoughts are activated outside ofconscious awareness (i.e., when DTA is high), they are managedthrough distal defenses that include efforts to shore up faith in thecultural worldview and attain self-esteem. The DTA concept waspivotal both in instigating the research designed to address ques-tions about process and in validating the dual-process model thatultimately resulted from this research. This, in turn, led to a morecomprehensive understanding of the differential effects of aware-ness of mortality on various forms of social behavior. Such appli-cations include explicating the multifaceted consequences of deathawareness for health decision making (i.e., the terror managementhealth model; Goldenberg & Arndt, 2008), legal decision making(Arndt, Lieberman, Cook, & Solomon, 2005), and consumer be-havior (Arndt, Solomon, Kasser, & Sheldon, 2004). Such insightswould not have been possible without the advent of theory andmethodology facilitating the study of DTA.

In addition to being the catalyst to the dual-process and othersuch models of TMT, the initial DTA studies provided the ground-work for a new paradigm investigating hypotheses regarding spe-cific anxiety-buffering beliefs. Prior to the introduction of thisconcept, the key to providing evidence that certain beliefs functionto allay concerns specifically relating to death resided in observingan increased need for such beliefs after death contemplation. The

704 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 7

DTA concept, by contrast, provided the opportunity to test whetheractivating the proposed protective belief before (or after) thinkingabout death eliminates the usual MS-induced increase in DTA.This reasoning continues to be used among researchers proposingnew (or specified) ways of buffering death anxiety (e.g., Rutjens etal., 2009, on progressive hope).

The Death-Association DTA Studies

The next set of DTA studies is distinct from the previous indemonstrating the provocation of DTA in the absence of a mor-tality salience manipulation. Rather, DTA is aroused in thesestudies due to the association between certain topics and death.According to cognitive conceptions of semantic memory, activat-ing (or priming) a certain thought can spread to activate concep-tually related thoughts (e.g., Collins & Loftus, 1975). As we shallsee, there are various stimuli that are, to one degree or another,related to death. These topics can therefore spread to activatedeath-related cognitions and, as such, can lead to both proximaland distal terror management defenses.

Direct death associations. For some topics, the associationwith death is so strong that thinking about these topics frequentlyentails consciously confronting the prospect of death.

Cancer and death-thought suppression. Most people cannotthink of health-related topics, such as cancer, without also thinkingabout the possibility that people are vulnerable to cancer, whichcan lead to death. Indeed, using a nationally representative samplein the United States, Moser et al. (2010) found that over 60% ofpeople see cancer as a possible death sentence. Thinking aboutcancer and other death-related issues can lead to proximal defensesaimed at removing death thoughts from consciousness. In fact,Arndt, Cook, Goldenberg, and Cox (2007) found the associationbetween cancer and death to be so strong that asking participantsto think about cancer prompted immediate and sustained suppres-sion of death-related thoughts. It is somewhat ironic that thoughtsof cancer induced even more death-thought suppression that didMS. The effect of cancer salience on DTA, unlike MS, could notbe seen after the usual delay (Study 1). Rather, this effect wasrevealed only when cancer salience was coupled with a cognitiveload manipulation, rendering suppression more difficult (Study 2),or when cancer was primed subliminally (Study 3). Suppressionpresumably was more vigorous in the cancer salience conditionbecause cancer represents a more specific threat of death than theabstract threat produced by MS. If so, this would suggest thatsuppression (and thus low DTA) should be especially likely whenperceptions of vulnerability are high. Study 4 supported this rea-soning; women who read an article highlighting their vulnerabilityto breast cancer had lower DTA than those who were not made tofeel vulnerable. Mediational analyses suggested that vulnerabilityincreased perceived threat, which in turn decreased DTA.

Along similar lines, Goldenberg and colleagues showed thatgetting a health checkup in the form of a self-administered breastexamination can also arouse death thoughts. In a pilot study,Goldenberg, Arndt, Hart, and Routledge (2008) measured DTAeither before or after participants performed the self-exam, withhigher levels of DTA shown after the exam. Subsequent studiesrevealed conditions under which such procedures can then influ-ence breast exam intentions and duration as both proximal anddistal defenses against death thoughts. For example, Cooper, Gold-

enberg, and Arndt (2010) interviewed women who had just re-ceived a screening mammogram, an experience likely to variablyincrease DTA. Among those with high DTA, framing of screeningbehavior (e.g., breast self-exams) as empowering to sense of selfand identity increased future breast self-exam intentions. Althoughthis study begins to reveal how naturally occurring DTA caninterface with identity-relevant motivations, the point to note hereis that measurement of DTA has allowed for novel insights intohow certain health contexts impact health behavior.

Life-threatening risks and self-esteem enhancement. The ca-pacity to index DTA has also provided insights on a range of issuespertaining to risky behavior. Highlighting the death-related conse-quences of certain types of risky behavior, such as unprotected sex(Taubman–Ben-Ari, 2004, Study 2), binge drinking (Jessop &Wade, 2008, Study 1), unsafe driving (Jessop, Albery, Rutter, &Garrod, 2008; Shehryar & Hunt, 2005), and tobacco use (Hansen,Winzeler, & Topolinski, 2010), has been shown to increase DTAafter the usual delay. Moreover, these studies have shown thatpriming these topics can produce distal defenses aimed at securingself-esteem. For example, in an ironic twist, priming the death-related consequences of a risky behavior (binge drinking or smok-ing) was found to increase intentions to engage in the behavioramong participants who believed that this behavior would conveyself-esteem (Hansen et al., 2010; Jessop & Wade, 2008).

Activating protective beliefs following exposure to these issueswill also eliminate the subsequent increase in DTA. Jessop,Albery, et al. (2008) found that priming the death-related conse-quences of risky driving increased DTA predominantly amongmen. In Study 2, the results indicated that women automaticallyactivated safe driving constructs, which buffered the effect of themanipulation on DTA. Consistent with this interpretation, in Study4, experimentally priming men with the idea of safe driving alsoserved to attenuate the link between risky driving and DTA.

Terrorism, violent conflict, and prejudice. In much the sameway that certain risky behavior can increase DTA and distaldefenses, other topics that are associated with death can alsoproduce these effects. One such linkage is found with exposure toterrorism and war-relevant information. The September 11 terroristattacks served to remind not only the people of the United Statesbut people all over the Western world that we are not imperviousto threats from abroad and that death can come at the mostunexpected times (see Pyszczynski, Solomon, & Greenberg,2003). Not surprisingly, reminding people of terrorism, war, orassociated topics produces effects parallel to those elicited by MS.In the first such demonstration, Landau, Solomon, et al. (2004)found distal defenses with explicit reminders of 9/11 and alsofound that subliminally priming participants with the letters WTC(for World Trade Center) or the numbers 911 increased DTA.

These effects can be provoked not just by laboratory-orientedstimuli but also by the images and information to which people areexposed as they navigate through their daily affairs. For example,Das, Bushman, Bezemer, Kerkhof, and Vermeulen (2009) foundthat exposing participants to a news story about Islamic terrorismincreased both DTA and distal worldview defense in the form ofprejudice toward Arabs. Following Lifton’s (1979) idea that deg-radation of the architectural symbols of humanity constitutes “im-agery of extinction,” Vail, Arndt, Pyszczynksi, and Motyl (2010)found similar DTA and worldview defense effects after exposure

705DEATH-THOUGHT ACCESSIBILITY

Page 8

to pictures of buildings that had been destroyed through bombingor in ways otherwise unconnected to violent action.

Death-related insurance logos and consumer behavior. Arecent study by Fransen, Fennis, Pruyn, and Das (2008) showedthat even fairly mundane stimuli, such as corporate logos, cansometimes elicit DTA. Fransen et al. exposed participants to thelogo of a well-known insurance company for five minutes. Toensure a thorough confrontation with the brand, they asked par-ticipants to report all the thoughts that came to mind during thisperiod. This procedure prompted thoughts pertaining to death, ashigh levels of DTA were subsequently observed following theusual delay. Moreover, given the value placed on consumer goodsamong people in Western culture (see Arndt, Solomon, et al.,2004), the extent to which the insurance prime increased DTA thenpredicted participants’ intentions to buy luxury food and entertain-ment items. This finding is particularly notable, given that itprovides direct meditational evidence that the extent of DTA canexplain the degree of response elicited by the priming stimulus.Although a number of studies have explored the role of terrormanagement processes in consumer behavior (see Arndt,Solomon, et al., 2004; Kasser & Sheldon, 2000), the Fransen et al.(2008) study illustrates how measurement of DTA can be espe-cially useful in understanding the role of existential fear in suchdomains.

Indirect death-associations. The topics noted above are oftenmore intuitively linked to death, but for other topics the deathassociation is more subtle. When topics do not induce consciousthoughts of death, proximal defenses such as suppression wouldnot generally be expected. Nonetheless, when DTA is increasedoutside of conscious awareness, it should still produce distal de-fenses. And given that thoughts of death never enter consciousnessin these cases, the need for delay in the production of thesedefenses is eliminated.

Creatureliness and the belief in human uniqueness. Accord-ing to TMT, the human body is a constant source of potential deathanxiety, because it is ultimately the body that is subject to deathand decay (Goldenberg, Pyszczynski, Greenberg, & Solomon,2000). Highlighting the frailty of the human body by observing itin old age (Martens, Greenberg, Schimel, & Landau, 2004) or bythinking about a person with a physical disability (Hirschberger,Florian, & Mikulincer, 2005), for example, has been found toincrease DTA in the absence of any explicit death reminders.

One of the primary ways that people ward off the death-relatedthreat of the body’s physicality is by embedding the body intocultural meaning systems that take place on a symbolic level. Putbroadly, cultural worldviews help us to deny our creaturelinessthrough the belief that humans are fundamentally different fromother animals. Elevating humans above the status of animalsremoves us from the association between creatureliness and death,providing at least some psychological distance from thoughts ofpersonal mortality. Research supporting this notion shows that theattempt to deny creatureliness influences several aspects of every-day life. First, Goldenberg, Pyszczynski, McCoy, Greenberg, andSolomon (1999) reasoned that sex is often psychologically prob-lematic because it represents a particularly potent way of remind-ing us that we are animals, which can in turn remind us of ourmortality. Thus, Goldenberg et al. explain, people are motivated toview sex as a symbolic expression of love, rather than animalcopulation, as a way of dissociating sex from physicality and thus

from death. In support of this idea, Goldenberg et al. (2000) foundthat priming the physical (vs. romantic) aspects of sex increasedDTA among participants who were high in neuroticism. Theseresearchers reasoned that neurotics may have particular difficultyin sustaining a meaning system (e.g., one that perpetuates a dis-tinction between humans and animals), which makes them espe-cially sensitive to the sex–death association. This, in turn, suggeststhat if the psychological distance between humans and physicalityis decreased for nonneurotics, they too should show evidence ofphysicality’s link to death-related cognition. Goldenberg, Cox,Pyszczynski, Greenberg, and Solomon (2002, Study 1) found, inline with this reasoning, that having participants read abouthuman–animal similarities prior to the physical sex prime in-creased DTA irrespective of levels of neuroticism; conversely,having participants read about human–animal dissimilarities ame-liorated this effect. Buttressing the belief that humans are unique,and not merely physical creatures, provided protection under con-ditions that otherwise increased DTA.

The association between the body and death has also been foundin research on disgust. Many prototypical elicitors of disgust, suchas defecation and urination, are relegated to private quarters whereothers need not be concerned. According to TMT, this is at least inpart due to the association between the body and death (see alsoHaidt, McCauley, & Rozin, 1994). Typically, however, the ten-dency to dissociate disgust-eliciting stimuli, such as feces, fromimplicitly activating thoughts of death is quite strong. Indeed,Dunkel (2009) found that inducing thoughts of feces produced animmediate drop in DTA. Likewise, inducing thoughts of death ledto an immediate drop in the accessibility of thoughts related tofeces. After a delay, however, when the risk of spreading activa-tion had diminished, accessibility levels returned to normal.

The effort to keep disgusting body-related stimuli from spread-ing to activate DTA does not always prevail, however, especiallywhen human creatureliness is already salient. In support of thisidea, Cox, Goldenberg, Pyszczynski, and Weise (2007, Study 2)found that priming human–animal similarities and having peoplecomplete a questionnaire regarding disgusting bodily functions(e.g., defecation and vomiting) resulted in high levels of DTA.When human–animal dissimilarities were primed, on the otherhand, completing the disgust questionnaire had no effect on DTA.This converges with findings by Goldenberg et al. (2002) to furthersupport the notion that the belief in human uniqueness buffersthoughts of death. When the disgusting stimuli were quite extreme,however, such as viewing graphic images of feces, DTA wasaroused regardless of human–animal similarity/dissimilarityprimes (Cox et al., 2007, Study 1).

The threat of physicality and its link to death-related cognitioncan also be observed in studies comparing humans to robots.Researchers working on the development of humanlike androidshave documented what they refer to as an uncanny valley (Mori,1970), in which people’s impressions of androids improve as theylook more and more human, up to the point at which only subtleimperfections in the appearance of an android remain. The slightimperfections serve not only to belie the android as unreal but alsoto highlight the mechanical nature of the human body. This un-canny effect can be psychologically unsettling, and it leads to anegative aesthetic evaluation of the robot. MacDorman (2005)found, consistent with a TMT conceptualization of the associationbetween human physicality and thoughts of death, that exposing

706 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 9

participants to these uncanny androids elicits increased DTA andworldview defense.

Wilderness and death. Similar processes unfold when peopleconfront the natural world. Koole and Van den Berg (2005) arguedthat although wilderness can be rejuvenating and aestheticallypleasing, it also has a dark side because of its association withdeath and uncontrollability. Indeed, in Study 1, participants re-ported being more likely to think about death in the wild than incultivated nature or in the city. Koole and Vand den Berg thusproposed that an appreciation for the positive aspects of natureinvolves suppressing the association between wilderness anddeath, something at which people who are action-oriented (i.e.,who possess well-developed self-regulatory capabilities; see Kuhl,1981) are particularly adept. As such, action-oriented individualsnot only perceived more beauty in nature (Study 3) but alsoshowed decreased DTA when viewing pictures of untamed wil-derness, suggesting that they were actively suppressing the asso-ciation between wilderness and death (Study 4). Thus, evidenceshows that thoughts about being in the wild can spread to activatethoughts of death and that appreciating the beauty of nature in-volves suppressing this association.

The value and meaning of life. Recent research by King,Hicks, and Abdelkhalik (2009) suggests that thoughts of death canbe aroused by promoting the meaning and value of life. Theseresearchers argued that the scarcity principle (see, e.g., Cialdini,2001) suggests that because death highlights life’s scarcity, thisshould in turn increase its perceived value. Moreover, given thatthe association between scarcity and value is so overlearned (Dai,Wertenbroch, & Brendl, 2008), highlighting the value of some-thing tends to make people think that it must also be scarce.Accordingly, in two studies, King et al. (2009) found that manip-ulating the value of human life subsequently increased DTA,which they construed as reflecting cognitions about the scarcity oflife. Having participants read that the human body is worth a large(vs. small) sum of money, or having them write about how thestatement “Human life is purposeful and meaningful”is true (vs.untrue), led to higher levels of DTA. Similar research byTaubman–Ben-Ari (in press) found that inducing thoughts of themeaning of life (by having participants complete either a meaningsalience induction, similar to the MS manipulation, or the Purposein Life Test; Crumbaugh & Maholick, 1969) also increased DTA.In contrast to King et al.’s explanation for the effect, however,Taubman–Ben-Ari argued that thinking about life’s meaning re-quires understanding that life is limited by death. Although addi-tional research is likely needed to reach a precise explanation forthe association, it seems that life’s meaning and value are seman-tically connected to death, such that thinking of one spreads toactivate the other.

Objective self-awareness. One final topic that appears to beimplicitly associated with death is objective self-awareness. Al-though Arndt et al. (1998) demonstrated that self-awareness candisrupt efforts to suppress conscious thoughts of death, there isalso evidence that inducing heightened self-awareness can beenough to increase DTA. According to Silvia (2001), self-awareness entails thinking about the self as an existential object.From a dialectical perspective, thinking about the self as an object thatexists also implies thinking about the self as an object that could notexist. Consistent with this idea, Silvia demonstrated that inducingself-awareness, either by having participants sit in front of a mirror

(Study 1) or by having them ponder the self as a distinct object(Study 2), subsequently increased the accessibility of thoughtspertaining to life and existence as well as thoughts pertaining todeath. Given the large body of research showing that inducingself-awareness will lead people to behave more consistently withtheir standards of value (Duval & Wicklund, 1972), it may be thatsuch behavior also functions to secure self-esteem as a distaldefense against high levels of DTA. Additional research couldassess this possibility.

Implications. Taken together, the aforementioned studiesconverge on the idea that topics associated with death (directly orindirectly) can increase DTA and therefore lead to distal defensesaimed at reducing DTA. Moreover, when the topics are moreexplicitly related to death (e.g., cancer), proximal defenses canbecome activated to remove death thoughts from consciousness.Thus, the death-association studies have shown effects that areparallel to those produced by MS. The main difference is that insome cases DTA can become aroused implicitly, producing distaldefenses without first producing proximal defenses.

The death-association studies thus provide support for TMTpostulates that converge with support provided by the MS studies.As the early studies with subliminal death primes suggested, thekey ingredient for MS-induced defenses appears not to be open-ended questions requiring participants to contemplate death butrather any stimulus capable of arousing DTA. Although thesestudies suggest that cognition pertaining to death is thus a keycause of worldview and other such defenses (rather than otheraversive thoughts with which contemplating mortality may beassociated), this is not to say that all death-related thought iscreated equal in terms of its capacity to motivate the same formsof psychological defense (Cozzolino, 2006). We return to thisissue in the Current Issues in the DTA Literature section towardthe end of the paper. At present, we highlight the insights that havebeen revealed in varying domains (e.g., sexuality, reactions toterrorism, cancer, disgust, self-awareness, connections to nature)as a result of the ability to measure DTA in the absence ofexplicitly manipulating thoughts of death. Further, we see thepivotal role that the DTA concept has played in elucidating theubiquity of death-related concerns and terror management pro-cesses.

The Anxiety-Buffer Threat DTA Studies

The next set of studies is somewhat similar to the death-association DTA studies in that both categories demonstrate thearousal of DTA without a MS manipulation. The key differencebetween the previous set of studies and the anxiety-buffer threatstudies is that in these studies DTA is not aroused via association.Rather, as the name implies, DTA is aroused due to a significantthreat that momentarily compromises the ability of the anxietybuffer to keep thoughts of death out of consciousness. The theo-retical thrust of TMT is that many beliefs about reality are culturalconstructions aimed in part at reducing concern about death. Itfollows that when people are exposed to a picture of realitywithout these constructions, or to a line of thought that contradictstheir protective beliefs, the result should be a resurgence of thevery thoughts that these beliefs function to allay. In other words,threatening people’s protective beliefs should increase DTA. Thisreasoning is evident in the previously discussed studies regarding

707DEATH-THOUGHT ACCESSIBILITY

Page 10

human creatureliness and DTA, and it eventually led to the pro-nouncement of a broad-level DTA hypothesis (Hayes, Schimel,Faucher, & Williams, 2008; Schimel, Hayes, Williams, & Jahrig,2007) outlining the conditions under which DTA should bearoused following threat. According to this hypothesis, if a psy-chological structure provides protection against thoughts of death,(a) threatening that structure should increase DTA and (b) forti-fying that structure should reduce DTA or prevent an increase inDTA due to subsequent threats.

Threatening attachment bonds increases DTA. The firststudies showing the DTA-arousing effect of a threatened anxietybuffer pertained to the attachment bond that develops betweenimportant relationship dyads. As mentioned earlier, attachmentfigures are a significant source of protection from death anxietyand are therefore an important means of keeping thoughts of deathaway from consciousness. But when important attachment figuresare no longer available, or the attachment bond is threatened, thethreat of death can become more difficult to manage. Accordingly,Mikulincer, Florian, Birnbaum, and Malishkevish (2002) foundthat having participants contemplate separation from a romanticpartner increased DTA among people high in attachment anxiety.Given that anxious-ambivalent people are already on shaky groundwhen it comes to managing existential anxiety (Mikulincer &Florian, 2000), simply imagining separation from a security-providing attachment figure can be enough to inflame deeplyrooted fears associated with death. Indeed, those who evidencedthe most DTA also evidenced the highest levels of distress in theirmusings on separation.

But the importance of attachment figures in regulating DTA isnot limited to those who have insecure attachment styles. Havingparticipants who were involved in a serious romantic relationshipcontemplate real problems in that relationship subsequently in-creased DTA, regardless of attachment style (Florian et al., 2002).Along similar lines, Taubman–Ben-Ari (2004, Study 3) found thatexperimentally manipulating thoughts about relationship insecu-rity by having participants complete a Fear of Intimacy Scale(Descutner & Thelen, 1991) increased DTA regardless of howparticipants answered the questions. The mere contemplation ofrelationship insecurity and fear of intimacy was enough to bringthoughts of death closer to consciousness. Moreover, Bassett(2005) found that disparaging the institution of marriage (an im-portant social structure that is devoted toward the maintenance ofsignificant attachments) led to greater efforts to dissociate the selffrom concepts pertaining to death.

The effect of threatening attachment bonds on DTA is also notlimited to romantic attachments. This should not be surprising,given that the attachment system first develops between parent andchild. Although attachment bonds might function primarily toensure the safety of the child, children are also an important meansof death denial for the adult. The role of parent is a culturallyvalued social role that provides self-esteem. Also, as Lifton (1979)and others have argued, children are a potent symbol of deathtranscendence and symbolic immortality, as they are the ones whowill remember us and carry on our legacy after we die. In supportof these ideas, Taubman–Ben-Ari and Katz–Ben-Ami (2008)found that imagined separation from one’s infant increased DTAamong first-time mothers of children age 3–12 months. In sum,because attachment bonds help us feel that we are connected toothers and therefore part of an everlasting cultural fabric, separa-

tion from important others (real or imagined) can bring the prob-lem of death to the edges of consciousness.

Threatening cultural worldview beliefs increases DTA. Al-though studies showing the DTA-arousing effect of threatenedattachment bonds were the first to emerge in the literature, dem-onstrating that threats to the cultural worldview can also arouseDTA was an important step toward understanding how culturalbeliefs provide protection from the awareness of death and thusvalidating the DTA hypothesis. Although a wide variety of beliefslikely can be significant to an individual’s worldview, extantresearch has focused on a few components that appear to beimportant to a number of worldviews.

The belief in a just and orderly world. One of the mostfundamental aspects of any cultural worldview is that it structuresthe universe and makes things comprehensible. For example, the24-hr clock, 7-day week, and 12-month calendar all serve tostructure time. Abstract concepts in general serve to group thingsinto meaningful and intelligible categories. Structuring things ren-ders them predictable, giving us a sense of control over them. Therepetitive and cyclical nature of these concepts also serves to makethings seem eternal.

A common structural component of any worldview is the con-cept of justice. The belief in a just world (Lerner, 1980) has longbeen recognized as a fundamental part of managing anxiety. Theassumption of a just world helps enable people to take their ownpersonal security for granted. If bad things happen only to badpeople, and one knows that one has been good, there is no need toworry about potential evils such as death. But when bad thingsbefall good people, our sense of justice is compromised and theanxiety buffer along with it.

In support of this reasoning, Landau, Johns, et al. (2004) foundthat for people especially motivated to maintain epistemologicalstructure in their view of life (i.e., people high in personal need forstructure; Neuberg & Newsom, 1993), reading about an innocentvictim subsequently led to higher DTA. In contrast, when thevictim was portrayed in a negative manner, suggesting that he orshe may have deserved the negative event, DTA remained low.Similarly, Hirschberger (2006) found that having participants reada vignette about an innocent victim who was severely injuredelicited more DTA than did having them read a vignette about avictim who was responsible for his or her condition or who wasonly mildly injured. Given that the belief in a just world is animportant part of most cultural worldviews, information that con-tradicts this belief threatens the worldview and renders thoughts ofdeath more accessible to consciousness.

Nationalism. Given that much of the foundational work sup-porting TMT showed that MS leads to increased worldview de-fense in the form of nationalistic bias, pivotal support for the DTAhypothesis would be to show that threatened nationalism arouseshigher levels of DTA. Accordingly, Schimel et al. (2007) demon-strated that exposing participants who were heavily invested inCanadian nationalism to material that trivialized Canadianachievements and derogated Canadian culture was sufficient toincrease DTA, even in the absence of any explicit or implicitreminder of mortality. Similar material that criticized anothernation (e.g., Australia) did not produce the same effect amongthese participants.

Schimel et al. (2007) also examined several important questionsregarding the psychological processes involved in the transition

708 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 11

from worldview threat to heightened DTA. They found that theeffect of threat on DTA was not the result of a general increase inthe accessibility of negative constructs (Study 3), nor was it theresult of self-reported increases in anger or anxiety (Study 4).Rather, the degree to which participants reported being personallyoffended or offended as a Canadian (i.e., the extent to which theirnationalistic worldview was threatened) was responsible for theDTA effect. Moreover, the DTA effect does not seem to require adelay between the threat and the DTA measurement (Study 1).Recall that MS-induced DTA is observed only after a short delayor distraction that allows for death-thought suppression to subside.In much the same way that subliminal death primes and implicitdeath associations eliminate the need for delay by precludingsuppression, DTA aroused by worldview threat appears to occurimmediately because thoughts of death never enter consciousness.

Religious beliefs. Several additional studies have shown theDTA effect in response to threatened religious beliefs. In Study 5,Schimel et al. (2007) showed that participants who believe increationism (but not those who believe in evolution) evidencedhigh levels of DTA after being exposed to anticreationist argu-ments. Similarly, Friedman and Rholes (2007) found that amongself-identified Christians, reading about inconsistencies and con-tradictions in the Bible increased DTA but only among those whoscored high on religious fundamentalism (Altemeyer &Hunsberger, 1992). For fundamentalists who believe in the infal-libility of the Bible, hearing about textual inconsistency constituteda worldview threat and thereby aroused DTA. Furthermore, Hayes,Schimel, and Williams (2008) showed that threatening the ultimatevalidity of Christian participants’ worldview with the idea thatIslam is growing in acceptance and is gaining dominance intraditionally Christian cities led to increased DTA among thoseparticipants. Finally, recent research suggests that parallel effectson DTA can manifest for nonbelievers, if the appropriate belief istargeted. Hayes, Schimel, Howard, Webber, and Faucher (2010)reported that having nonbelievers (i.e., atheists) read compellingevidence for intelligent design also increased DTA.

Faith in progress. Recent work by Rutjens et al. (2009) hasdemonstrated the psychological benefit of believing that history isprogressive. According to Gray (2004), the widespread belief thatsecular society is progressive, in the sense that it provides contin-ual improvements in the standard of living, is a mere illusion towhich we desperately cling as a way of shielding ourselves fromthe fear of death. From this perspective, people take comfort in theidea that things will get better, even if they may not be optimal atthe moment. Indeed, the idea of a blissful afterlife, free from thetrials and tribulations of daily living, also contains the notion thathuman existence will ultimately improve. In three studies, Rutjenset al. supported these assertions. Consistent with the DTA hypoth-esis, their Study 2 demonstrated that exposure to the argument thatprogress is merely illusory increased DTA in participants relativeto controls.

Threatening self-esteem increases DTA. Since its inception,TMT has posited that faith in cultural worldviews and self-esteemare the two primary shields that insulate individuals from theawareness of death. Thus, although the above studies indicate thatcompromising faith in the worldview increases DTA, TMT alsopredicts parallel effects with threats to self-worth. The availableresearch supports this hypothesis. Ogilvie, Cohen, and Solomon(2008, Study 2) found that highlighting aspects of participants’

undesired self (Ogilvie, 1987), in terms of having participantsrecall when they felt they were at their worst, increased DTA.Recalling a time when one failed to meet current standards ofvalue highlights the fallibility of the self and compromises overallself-integrity.

Further support for the idea that self-esteem threat increasesDTA was obtained by Hayes, Schimel, et al. (2008), who, ratherthan having participants imagine past failures, experimentally ma-nipulated a failure experience. In Study 1, participants whoseself-esteem was invested in the belief that they were quite intelli-gent were subjected to information suggesting they possessedbelow average intelligence. In Study 2, participants were informedthat their personality was ill suited for their career aspirations.Study 3 more explicitly drew from the notion that people’s self-evaluations are largely dependent on the perceptions of others (cf.Cooley, 1902; Mead, 1934). Participants were told, with no time toprepare, that they would need to give an impromptu speech in frontof a small group of peers. Across all studies, these different formsof self-esteem threat provoked increased levels of DTA relative toconditions that did not threaten self-esteem. Moreover, consistentwith Schimel et al. (2007), the effect was not found to be the resultof a general increase in the accessibility of negative constructs.Thus, threatening either component of the cultural anxiety buffer(cultural worldviews or self-esteem) led to the same effect ofincreased activation of thoughts pertaining to death.

Reducing threat-induced DTA by defending the worldviewor bolstering self-esteem. In much the same way that worldviewdefense and self-esteem maintenance reduce DTA in response toMS and associated topics, there is evidence that such defenses canreduce DTA in response to threats to the anxiety buffer. Hayes,Schimel, et al. (2008, Study 3) found that having participantsaffirm their most important value reversed the effect of self-esteemthreat on DTA. Recall that participants’ self-esteem was threatenedby a looming public evaluation. When participants affirmed animportant value prior to completing the DTA measure, however,the effect of threat on DTA was eliminated. Thus, not only doesthreatening self-esteem increase DTA but fortifying self-esteemreduces DTA to baseline levels.

Likewise, research has shown that defending against worldviewthreats reduces the impact of those threats on DTA. Schimel et al.(2007, Study 2) showed that arming participants with informationthat allowed them to defensively dismiss the worldview threatreduced DTA relative to that for participants who were not armedwith the dismissive information. Similarly, Hayes, Schimel, andWilliams (2008) found that informing participants that some mem-bers of a worldview-threatening group had been killed in anaccident (i.e., the annihilation defense; see Berger & Luckmann,1967) also reduced the effect of worldview threat on DTA. Pre-sumably, when members of an opposing worldview die, this sug-gests that the opposing belief system is inferior or that it at leastprovides insufficient protection from the ultimate threat of death.

Implications. A growing number of studies now demonstrateincreased DTA in response to potent threats to the anxiety buffer,even though no death-related stimuli are presented. In conjunctionwith the MS and death-association DTA studies, the anxiety-bufferthreat DTA studies provide further support for the functionality ofspecific beliefs. This has enabled critical insights into not only howcultural worldview and self-esteem beliefs provide protectionwhen thoughts of death are activated but also how they function on

709DEATH-THOUGHT ACCESSIBILITY

Page 12

a continual basis to keep thoughts of death at bay. When thesebeliefs are threatened, they momentarily lose some of their func-tionality, and thoughts of death become more accessible.

Moreover, the anxiety-buffer threat DTA studies provide addi-tional evidence for the specificity of death as a critical motivatorbehind maintaining important relationships, asserting the validityof one’s cultural worldviews, and believing that one is living up tostandards of value. When these psychological structures are weak-ened, thoughts specifically pertaining to death become more ac-cessible to consciousness. The finding that threats to the culturalworldview and self-esteem increase death-related thoughts but notnegative thoughts in general suggests that these psychologicalentities perform specific functions tethered to managing thoughtsand concerns about death. Finally, the anxiety-buffer breakdownDTA studies provide yet another paradigm for testing TMT orother hypotheses concerning existential affairs pertaining to death,one that does not involve MS or any semantically death-associatedthreats.

The Dispositional DTA Studies

A new crop of DTA studies has emerged very recently. Thesestudies are distinct in that they do not manipulate threat or con-scious thoughts of death. Rather, these studies treat individualdifferences in baseline levels of DTA as a predictor variable forother terror-management-related defenses. The evidence suggeststhat DTA is ultimately responsible for distal defenses, so it may bethat at least in certain situations the way in which DTA is arousedis largely irrelevant. Whether DTA is aroused by a MS manipu-lation, a stimulus with death associations, or an anxiety bufferthreat, the initiation of distal defenses to reduce DTA may ensue.As such, naturally occurring (rather than experimentally manipu-lated) variations in levels of DTA may predict the degree to whichdistal defenses will be enacted.

Determinants of dispositional DTA. From a theoretical po-sition, the determinants of dispositional DTA should be muchthe same as those identified in other TMT studies involving DTA,such as having a recent encounter with death or a death-relatedstimulus. Furthermore, following from the DTA hypothesis, arelatively weak or unstable anxiety buffer, or one that is repeatedlyunder attack, should also increase one’s disposition for DTA.Recent research by Friedman and Rholes (2009) supports thisnotion, finding that people with highly fundamentalist religiousbeliefs have low levels of dispositional DTA. Because fundamen-talists have a rigid and unquestioned belief system that promisesliteral immortality for their fervent faith, thoughts of death can beeffectively managed on a regular basis, resulting in low levels ofdispositional DTA. Friedman and Rholes also found that for thosewho are not fundamentalist, having an interdependent self-construal is likewise associated with low levels of dispositionalDTA. Enmeshing the self within closely knit personal relationshipscan provide meaning to life and a sense that the self will beremembered by significant others and will thus continue to livesymbolically, even after death. Thus, perceiving the self as inter-dependent with others can be an alternative way of reducingunconscious DTA.

Recent research by Taubman–Ben-Ari and Noy (in press) showsthat one’s general level of self-consciousness is also associatedwith dispositional DTA. According to TMT, self-consciousness is

a key element in the human existential dilemma, as knowing thatone exists entails also knowing that one will eventually cease toexist. However, research by Trapnell and Campbell (1999) sug-gested that not all types of self-consciousness are equal and dis-tinguished between neurotic self-attentiveness (i.e., rumination)and intellectual self-attentiveness (i.e., reflection). Whereas reflec-tion involves a sort of self-consciousness motivated by curiosityand interest in self-discovery, rumination is motivated by per-ceived threat or injustice toward the self. Consistent with thisanalysis, Taubman–Ben-Ari and Noy (in press, Study 1) found thatparticipants who score high on the rumination (but not the reflec-tion) dimension of self-consciousness are more apt to have highlevels of DTA.

Another factor that has been shown to predict dispositional DTAis the level of ego resources that one has available at any giventime. The process of managing thoughts of death involves uncon-scious mechanisms that are continually working to keep troublingthoughts of death from entering consciousness. Implicit in thisnotion is that the mind must devote energy to this activity. And asis the case with all types of energy, the energy directed towardregulating the contents of the mind is not boundless (Baumeister,Bratslavsky, Muraven, & Tice, 1998). In this vein, Gailliot,Schmeichel, and Baumeister (2006) found that people who possessan increased ability to control the contents of consciousness (asassessed by the Self-Control Scale; Tangney, Baumeister, &Boone, 2004) are better able to keep thoughts of death undercontrol and are therefore less prone to experiencing high levels ofdispositional DTA. Conversely, when their ego resources are de-pleted, people are also susceptible to increased levels of DTA(Gailliot et al., 2006; Gailliot, Schmeichel, & Maner, 2007).

Consequences of dispositional DTA. Being particularly dis-posed to thoughts of death can have important consequences. Oneof the central tenets of TMT is that we avoid thoughts of deathbecause they can potentially produce burdensome anxiety thatwould bog down our daily pursuits. Indeed, a growing number ofrecent studies have demonstrated that dispositional DTA is nega-tively associated with indicators of psychological well-being. Forexample, Cox, Reid-Arndt, Arndt, and Moser (2010) found thatwomen who had recently received surgery for a cancerous breastmass reported lower levels of well-being than did women who hadsurgery for a noncancerous mass. This is a well-documented effectin the cancer survivorship literature. Perhaps most interesting,these women also showed higher DTA but did not report greaterexplicit worries about death. Elevated DTA, in turn, was nega-tively associated with well-being and mediated the effect of diag-nosis on well-being. This study thus showcases the potential forDTA to provide novel insights beyond what might be obtainedwith reports of explicit death-related worry.

Additional studies are documenting that the association betweendispositional DTA and well-being is moderated by certain person-ality characteristics. Routledge, Ostafin, Juhl, and Sedikides(2010) found that dispositional DTA interacted with self-esteem topredict perceived meaning in life, such that low levels of self-esteem coupled with high levels of DTA was associated withdecreased perceptions of meaning.

In another line of work, Vess, Routledge, Landau, and Arndt(2009) found that dispositional DTA was negatively associatedwith perceived meaning in life only among participants with lowlevels of personal need for structure (PNS; Neuberg & Newsom,

710 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 13

1993). For those high in PNS, DTA is somewhat positively asso-ciated with meaning in life. Vess et al. suggested that this occursbecause DTA signals the need to engage preexisting meaning-providing resources. Experimental studies manipulating MS con-verge to suggest that whereas people high in PNS have thesestructures readily in place, individuals low in PNS are more proneto search out novel experience for bolstering meaning-providingresources. Consistent with this hypothesis, at low levels of PNS,DTA was positively associated with the desire for novelty andexploration. Thus, the DTA construct provided critical insightsinto the different ways in which death is connected to meaning:When DTA is high, people with high PNS find meaning instructure, and people with low PNS find meaning in novelty.

Finally, recent research has also shown a relationship betweendispositional DTA and distal defensiveness. Webber (2009) foundthat people who possess high levels of dispositional DTA alsoshowed increased defensive self-handicapping on a subsequenttask. Thus, those who may be particularly predisposed to thoughtsof death may also be likely to behave defensively, even in theabsence of threat.

Implications. The use of dispositional DTA as a predictor ofterror management defensiveness is a relatively new researchstrategy, and it is therefore difficult to ascertain exactly where thisline of research will lead. However, the bulk of the evidencesuggests that higher baseline levels of DTA may be a vulnerabilityfactor for psychological problems such as anxiety and depression,increased defensiveness, and poor adjustment to life difficulties.Given this potential link, further research should continue to in-vestigate this possibility, but it should also distinguish betweenhigh dispositional DTA due to acute life circumstances (e.g.,serious illness either for oneself or a loved one, exposure totrauma) and that due simply to watching the news on a daily basisand insidious life circumstances (e.g., a general lack of meaningand self-worth). Indeed, all of these events may make deaththoughts more accessible to awareness on a chronic basis, albeit indifferent ways and in different contexts and because of differentcircumstances. Further research should assess possible differencesamong these sources of high dispositional DTA and how theyinteract with other individual differences. We return to variants ofthis idea later.

Interim Summary: The Theoretical and EmpiricalSignificance of the DTA Concept

The foregoing review of the DTA literature converges to high-light the importance of understanding the role of death awarenessin social behavior. The DTA concept has been a driving forcebehind a number of significant advances in TMT. Indeed, the veryfirst DTA study led to a wave of follow-up investigations prompt-ing the first major theoretical refinement of the theory, proposinga microlevel cognitive model of the processes involved in manag-ing thoughts of death (Pyszczynski et al., 1999).

The conceptual and empirical work that was inspired by inves-tigations of DTA has since gone far beyond this initial theoreticalrefinement, extending insights in ways that could not have beenanticipated by the cognitive model of dual defenses alone. This, inpart, is why it is critical to take stock of these contributions. Notonly have the foundational insights of the role of DTA in manag-ing awareness of death led to theoretical advances in varying

domains, but the DTA concept has expanded our understanding ofhow people use their beliefs about the self and their social worldto come to grips with the human existential predicament. Forexample, although the MS paradigm provided initial support forthe function of worldview and self-esteem beliefs, evidence show-ing that activating protective beliefs before or after MS thereafterreduces DTA added further support for this proposition. Moreover,studies showing increased DTA in response to anxiety-bufferthreats have provided even more evidence for the function of theseimportant beliefs.

The anxiety-buffer threat DTA studies also lend strong supportfor the notion that thoughts specifically relating to death areresponsible for MS-induced worldview and self-esteem defensive-ness. A number of theorists have criticized the theory over theyears, arguing that it is not thoughts of death per se but someby-product of those thoughts that produce MS effects. Most re-cently, some have argued that the uncertainty and lack of controlassociated with death are most threatening (e.g., Fritsche, Jonas, &Fankhanel, 2008; McGregor, Zanna, Holmes, & Spencer, 2001;van den Bos, 2001) and that thoughts of death produce defensiveresponding only to the extent that they arouse thoughts of uncer-tainty regarding what will happen to the self after the body dies(van den Bos et al., 2005). Others argue that death poses a threatto human meaning systems, which is more problematic than deathper se (Heine, Proulx, & Vohs, 2006). From this perspective,humans have an innate need to create meaning, and the idea ofdeath is problematic because it threatens to remove meaning fromlife. Still others suggest that death is an adaptive problem and thatadaptive problems of all types motivate coalitional support seeking(Navarrete & Fessler, 2005). Thus, thoughts of death induceworldview defense not to shore up faith in the ultimate validity ofanxiety-buffering beliefs but to seek support to help defend againsta realistic threat. Although death does indeed suggest all of thesethreats and more (for a full review, see Pyszczynski, Greenberg,Solomon, & Maxfield, 2006), research on DTA points specificallyto the problem of death as a driving force in worldview andself-esteem defensiveness. As presented in the above review, agrowing number of empirical studies are demonstrating the DTA-arousing effect of threatening important components of people’scultural anxiety buffer. Why would thoughts of death becomeaccessible to consciousness if these thoughts were not a pivotalaspect of the problem?

Furthermore, measuring DTA, rather than being restricted tomanipulating thoughts of death, has provided researchers withadditional paradigms for investigating TMT and other propositionsconcerning the role of death in life. By allowing for methodolog-ical convergence and triangulation among different strands ofresearch, measuring DTA enables the larger existential picture tobe seen from a diversity of perspective. In so doing, this researchhas been critical in moving TMT beyond a conceptual frameworkfor explaining laboratory effects derived from reminding people oftheir mortality. Such research showcases the relevance of theseprocesses for experiences that people encounter as they navigatethe daily affairs of life.

Toward an updated TMT process model. It has been morethan 10 years since Pyszczynski et al. (1999) introduced thedual-process model. Since then, as mentioned above, research onDTA has led to a number of insights that have expanded ourknowledge of terror management processes. As such, we feel, a

711DEATH-THOUGHT ACCESSIBILITY

Page 14

modest update to this model is now in order (see Figure 2). For thesake of clarity and efficiency in depicting the processes leading toand from DTA, moderating variables have been included in theform of questions that can be seen in the spherical objects of themodel. For each question, “yes” and “no” responses lead to dif-ferent paths/processes in the model.

In keeping with the original model, Figure 2 shows MS to firstproduce proximal defenses but only to the extent that controlled-processing resources are available to initiate such defenses. If theseresources are not available (e.g., cognitive load is high), proximaldefenses are circumvented along with the need for delay to observedownstream processes. Both paths ultimately lead toward highDTA but only to the extent that anxiety-buffering capabilities arerelatively weak (e.g., low self-esteem) or have not been preemp-tively called into action. Possessing high levels of self-esteem(e.g., Harmon-Jones et al., 1997) or bolstering the anxiety bufferby activating protective beliefs (e.g., affirming important values,Schmeichel & Martens, 2005; taking the opportunity to defendone’s worldview, Arndt, Greenberg, Solomon, et al., 1997; orenhancing one’s self-esteem, Mikulincer & Florian, 2002) willpreclude death thoughts from becoming highly accessible. Whenself-esteem is low or protective beliefs are not preemptively acti-vated, however, the result will be high levels of DTA, which willmotivate distal defenses retroactively in order to reduce DTA.Although this basic process should be similar for all individuals,the precise nature of the distal defenses enacted to reduce DTA canvary depending on dispositional and situational constraints ondefensive style (see also Arndt, Cook, & Routledge, 2004). People

with a secure attachment style, for instance, are more likely toengage in relationship-type defenses, whereas those with moreinsecure attachment styles seem to prefer cultural worldview de-fenses (e.g., Cox et al., 2008; Mikulincer & Florian, 2000).

The main additions to the current model from the original arethe paths that begin either with a death-associated stimulus or withan anxiety-buffer threat. Death-associated stimuli will produce thesame array of processes as the MS manipulation, provided that theassociation leads thoughts of death to become conscious. If not,however, activation can spread directly to death-related cognition,but only to the extent that dispositional buffering capabilities areweak or defensive beliefs have not been preemptively activated.Highlighting the uniqueness of humans relative to other animals,for example, can effectively inhibit the spread of activation fromsex- to death-related cognition, keeping DTA low. But whendefensive beliefs are not activated or are dispositionally weak (e.g.,recently primed human–animal similarities; high levels of neurot-icism), the death-associated stimulus will succeed in spreading toactivate thoughts of death nonconsciously, and this activation willin turn promote distal defenses to reduce DTA (e.g., Goldenberg etal., 1999, 2002).

Finally, anxiety-buffer threats will more or less directly lead tohigh DTA, unless defensive beliefs have been activated preemp-tively or immediately in response to the threat (e.g., self-affirmation; Hayes, Schimel, et al., 2008). When a threat leads toa momentary breakdown in the anxiety buffer, however, DTAincreases and distal defenses will ensue in much the same way aswhen DTA is aroused by other means (e.g., Hayes, Schimel, &Williams, 2008).

Current Issues in the DTA Literature

Despite the emergence of a relatively coherent body of studiesfrom the measurement of the DTA concept, several methodolog-ical and theoretical issues remain. We outline a number of theseissues with the goal of clearing up ambiguities and stimulatingfuture research.

Methodological Issues

Why is DTA a better method of assessing concern aboutdeath than self-reports? One question involves why a measureof DTA is a better indicator of concern regarding death than is ameasure of self-report. The answer here depends on what one istrying to understand. If one is interested in conscious reports offears of death, self-report measures are certainly appropriate. How-ever, interest often lies in the influence of death-related thoughtoutside of what people can report. The rationale behind the use ofan indirect, rather than a direct, measure is based on the TMTproposition that much of the influence of concerns about death isunconscious. People do not always have direct introspective accessto the psychological processes that are thought to be operational.Thus, using self-report measurement is likely to generate specula-tion on the part of participants as to how they feel about death (cf.Nisbett & Wilson, 1977). Moreover, these introspective specula-tions could be influenced by motivated defensiveness. Even whenpeople might actually be concerned about dying, they may notalways admit it, even to themselves. This was evident, for exam-ple, in the recent findings of Cox et al. (2010), wherein DTA

Mo

Death associated s�mulusDeath-associated s�mulus

Do death-thoughts become conscious?

NO

Anxiety-buffer threat

Is the anxor have def

a

High death-t

Dis

Disposi�onal and situa�onal constraints

on defensive styleon defensive style

rtality salience

Are controlled-processing resources available?

YES

NO

YES

Proximal defense

Delay

xiety-buffer strong, fensive beliefs been c�vated?

YESNO

thought accessibility

stal defense

Low death-thought accessibility

Figure 2. Theoretical model depicting the processes leading to and fromdeath-thought accessibility.

712 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 15

predicted well-being subsequent to cancer diagnosis but explicitworries about death did not. Given that TMT researchers are ofteninterested in an implicit cognitive process of which people areunaware, and which could potentially be distorted by psycholog-ical defense, indirect measurement is essential (Greenwald &Banaji, 1995).

Multiple operational definitions of DTA: Are they all thesame? Across the over ninety published studies, DTA has beenmeasured with a number of techniques. In the vast majority ofthese studies, following Greenberg et al.’s (1994) original DTAstudy, a word-fragment completion task has been used to assessDTA. As previously mentioned, the idea behind this measure isthat a respondent should complete more fragments in the death-related manner if thoughts of death are highly accessible than ifsuch thoughts are not accessible. Within studies using this proce-dure, there have been a number of variations. As shown in Ap-pendices A–D, some studies have embedded six death fragments(e.g., Greenberg et al., 1994; Schimel et al., 2007) and others haveused only four (e.g., Arndt, Greenberg, Solomon, et al., 1997). Thenumber of neutral fragments within which these death fragmentsare embedded has also varied: Some studies included 14 neutralfragments (e.g., Greenberg et al., 1994), whereas others included20 (e.g., Arndt, Greenberg, Pyszczynski, & Solomon, 1997). Thespecific death words and the letters omitted to make the fragmentshave also been variable.4 Moreover, the fragment-completion taskhas been employed in several languages, including English (e.g.,Greenberg et al., 1994), Hebrew (e.g., Mikulincer & Florian,2000), Dutch (e.g., Das et al., 2009), and Chinese (Zhou et al.,2009). The fact that these variations have yielded similar resultsacross studies attests to the overall construct validity of this mea-surement procedure.

Much like the word fragments, which can be completed in atleast two ways, ambiguous images were recently used to measureDTA (Gailliot et al., 2006). In one study (Study 1B), the image canbe seen either as a lady sitting at a vanity mirror or, if one focusesdifferently, as a picture of a skull. Similarly, in another study(Study 3) the ambiguous image can be seen either as a skull or astwo men sitting at a table. In both cases, Gailliot et al. (2006) hadparticipants view the images and write down the first 10 words thatpopped into their mind. Participants whose ego resources had beendepleted, or who dispositionally lacked such resources, were morelikely to report thoughts of death when viewing these images. It isinteresting that Gailliot et al. used the term explicit DTA to de-scribe the construct measured by the ambiguous images and theterm implicit DTA to describe that measured by the word-fragmenttask. The possibility that some measurement procedures mightassess conscious thoughts of death and others might measureunconscious ones seems appealing. However, in the absence offurther research, it is currently unclear whether there are any realdifferences between these two measures. Both of these approachesinvolve responding to an ambiguous stimulus, whether it is a wordfragment or a picture. With regard to word fragments, it is wellestablished that accessible content will increase the likelihood ofcompleting a fragment in a particular way (Fiske & Taylor, 1991;Srull & Wyer, 1980). As it pertains to ambiguous images, writingout more death words might simply represent the same process.Alternatively, it might mean that seeing the death-related perspec-tive causes more thoughts of death to enter consciousness, becausethey were already highly accessible. Either way, this would not be

any different than the process occurring in response to the wordfragments. Thus, the distinction between implicit and explicit DTAis tenuous and should be made with caution.

Although the majority of DTA studies have employed the word-fragment task, a number of other, more recent studies have mea-sured DTA in terms of reaction times, generally in the context ofa lexical decision task (Arndt et al., 2007, Study 3; Bassett, 2005;Fritsche et al., 2008, Study 4; Hayes, Schimel, et al., 2008, Studies1 and 2; Koole & Van den Berg, 2005, Study 5; Schimel et al.,2007, Study 3). According to De Houwer (2003), reaction time(RT) measures are best understood in terms of stimulus–responsecompatibility versus incompatibility (see also Kornblum, Has-broucq, & Osman, 1990). For a lexical decision task that containsdeath-related words, a high level of DTA is compatible with thecorrect response option for these words (i.e., that the letter stringdoes indeed form a word). As such, people should react morequickly and make fewer errors in response to the death words(relative to other words) when death thoughts are highly accessible(e.g., Schimel et al., 2007). In contrast to the lexical decision task,Bassett’s (2005) study employed an implicit-association test(Greenwald, McPhee, & Schwartz, 1998), which assesses therelative strength of association between concept pairs. In this case,the study examined the associative strength between self-relatedand death-related concepts. Bassett (2005) reasoned that peopleshould inhibit the association between self and death after world-view threat. Within the context of the implicit association measure,the inhibition of this association would be incompatible withaccurate responding when both self- and death-related words aregiven the same response option, thereby increasing RTs and in-ducing more errors—which is what was observed. The point tonote here is that different RT measures are based on different typesof processes (compatible vs. incompatible) and can therefore yieldvery different results (e.g., Gawronski & Bodenhausen, 2005).Researchers should keep these differences in mind when conduct-ing future research on DTA.

Even though all of these (i.e., reaction time, fragment type)measures are indirect (or implicit) assessment techniques, theymay differ on important dimensions, such as the degree to whichthey are deliberative versus spontaneous (see Vargas, Sekaqua-ptewa, & von Hippel, 2007). For example, partially structuredmeasures (e.g., the fragment completion task or the ambiguousimage task) afford some degree of deliberation and require moreexplicit resources from the respondent. Measures based on perfor-mance on an objective task (e.g., the lexical decision task or theimplicit association test), on the other hand, rely more heavily onthe respondent’s spontaneous information-processing abilities. Al-though both types of measures tap automatic responses to somedegree, deliberative implicit measures are more apt to be tainted bycontrolled processes, making them more sensitive to issues such asself-presentational concerns. A given respondent may have highDTA and may be automatically inclined to complete multiplefragments in the death-related manner, for example, but the re-

4 Presumably, more death fragments may enable a more sensitive as-sessment, but this comes with the cost of participants becoming tuned in towhat one is trying to assess. Use of more neutral fragments may help tooffset this concern, but such use needs to be balanced with practicalconcerns with participant time and questionnaire length.

713DEATH-THOUGHT ACCESSIBILITY

Page 16

spondent may also become aware of this inclination and choose tomodify his or her initial tendencies, completing the fragments innon-death-related terms instead. This possibility is less likely tooccur for spontaneous measures, for which automaticity is theprevailing process. With this said, however, one should not con-clude that spontaneous measures are more desirable or valid thandeliberative measures, as the latter may be more appropriate incertain circumstances. For example, deliberative measures afford agreater variety of responses and might therefore facilitate investi-gation into different types of accessible death thoughts. This ideahas yet to be thoroughly examined. Thus, despite the relativelyconsistent results that these different measures have yielded thusfar with respect to DTA, researchers should carefully consider thesubtle differences that exist among them and remain vigilant to thepossibility that these differences may in some cases produce dis-crepancies.

What are the effects of measuring DTA? Another method-ological issue for DTA research is the potential effect of measuringDTA. The distinguishing feature of DTA is that it representsthoughts of death that are not in focal awareness but are nonethe-less prone to entering awareness. The mere act of measuring DTAmight therefore make thoughts of death conscious; even if deathremains unconscious, the measurement procedure could primethoughts of death and thereby increase DTA. Given the importanceof language in making something explicitly conscious (Fingarette,1969/2000) and the fact that measurement procedures such as theword-fragment task involve having participants spell out the firstword that comes to mind, it seems possible that the mere admin-istration of a DTA measure can, at least in some cases, bringthoughts of death into focal awareness.5

Recent studies by Zhou, Liu, Chen, and Yu (2008) using aChinese version of the DTA word-fragment task may support thisreasoning. All participants completed 30 word fragments; in anMS condition 10 could be completed only with death-relatedwords, and in a control condition none of the fragments could becompleted with death words. After the usual delay, participantsread a description of the Chinese one-child birth control policy andrated their attitudes toward the policy. Given that children are animportant means of death transcendence (as noted earlier in thisreview), responses were taken to reflect a terror managementstrategy. Participants who were given death-word fragments re-ported significantly less support for the birth control policy thandid participants in the control condition.

It is important to note that the fragment task used by Zhou et al.(2008) is slightly different from the typical DTA measure. Partic-ipants had no choice but to complete the 10 death fragments withthe death words, whereas the typical word-fragment DTA measurecontains fragments that can be completed either as death-related oras neutral words. Moreover, Zhou et al.’s measure contained 10death-related words, which is considerably more than the usualfour to six possible death completions. And on average, evenparticipants who evidence high levels of DTA rarely complete allpossible death fragments in the death-related manner. Neverthe-less, it might be reasonable to assume that the measurement ofDTA can make thoughts of death conscious, provided that (a) theparticipant actually completes the possible death fragments withthe death-related word and (b) the participant completes enough ofsaid fragments. Exactly how many death-related fragment com-

pletions are necessary to activate thoughts of death is a matter forfuture research.

Thus, it remains to be determined whether and, if so, under whatconditions measuring DTA increases nonconscious DTA or in-duces conscious thoughts of death. In other words, measuringDTA might be subject to Heisenberg’s principle of indeterminacy,wherein measuring a phenomenon influences the object of thatmeasurement (see Hergenhahn, 2001). As we discuss below, thispossibility has created some issues with using DTA as a medita-tional variable, and given other potential ramifications of thisissue, future research should assess the effects of measuring DTAon DTA and related phenomena.

Theoretical Issues

Do all threats that produce defense also increase DTA?TMT does not posit that awareness of death is the only motivatorof self-esteem striving or identification with cultural beliefs. Thus,it is certainly plausible from this and many other theoreticalperspectives that threats unrelated to the prospect of mortalitywould engender defensiveness without increasing DTA. Indeed,Burris and Rempel (2004, Study 6) found that having participantsread a graphic description of dust mites (a microscopic bug thatfeeds on people’s dead skin) subsequently increased worldviewdefense but not DTA. Similarly, Proulx and Heine (2008, 2009)found that some types of meaning threats, such as surreptitiouslyreplacing the experimenter midstudy (Proulx & Heine, 2008,Study 1b), heightened worldview defense or general meaningseeking without also increasing DTA.

However, TMT and the research reviewed here (i.e., theanxiety-buffer threat studies) do implicate death as a core psycho-logical threat that is brought to the surface when the anxiety-bufferis compromised. Given this assumption of TMT, why have somestudies demonstrated that meaning threats can increase defensive-ness without increasing DTA? One possibility is that some mean-ing threats may be potent enough to trigger defenses aimed atmanaging the threatening thought or idea but not potent enough toweaken the anxiety buffer and elicit high DTA. Consider also thatan anxiety-buffer threat will increase DTA only when defensivebeliefs are not preemptively activated to maintain the integrity ofthe anxiety buffer (see Figure 2). As such, finding increaseddefense following a meaning threat without also finding increasedDTA does not rule out the possibility that concerns about death areat the root of the problem.

Finally, there may be methodological reasons why some studieshave not found increased DTA following meaning threats. Mea-sures of DTA are designed to assess implicit death thoughts, thatis, thoughts of death that are just outside of awareness. Experi-mental procedures that enhance the implicit properties of the DTAmeasure may therefore increase its effectiveness. For example, onthe word-fragment completion task, the most common method ofassessing DTA, participants are instructed to complete each frag-ment with their most natural gut-level response without spendingtoo much time on each item (Greenberg et al., 1994). If researchers

5 Of course, this may be more likely with some measures than others,and this possibility may in part underlie Gailliot et al.’s (2006) reference toimplicit and explicit measures.

714 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 17

do not take steps to emphasize these instructions, the DTA mea-sure could easily be disrupted by conscious thought processes.Moreover, in the majority of the studies showing heightened DTAin response to anxiety-buffer threats, a cognitive load componentwas built into the procedures such that participants were simulta-neously completing a memory task and/or were explicitly told torespond very quickly to the DTA measure (e.g., Schimel et al.,2007). It is conceivable that taxing mental resources in this waymay have precluded participants from engaging in other DTA-reducing defenses. Future work is certainly needed to better un-derstand the particular conditions under which meaning threatsincrease defensiveness because of unconscious death-thought ac-tivation.

Does DTA always entail defense? Is DTA always indicativeof psychological threat, and even if so, can one experience highlevels of DTA without clinging to worldview beliefs or forti-fying self-esteem? Evidence suggests some circumstances inwhich people show high levels of DTA but not increasedworldview defense. Florian, Mikulincer and Hirschberger(2001, Study 1) found that high levels of personal hardiness (ameasure of psychological security) was associated with lowerlevels of worldview defense in response to MS. In Study 2,however, hardiness did not reduce the effect of MS on DTA.Similarly, in their research on attachment mechanisms andTMT, Mikulincer and Florian (2000) found that high attach-ment security reduced worldview defense in response to MS.Once again, however, attachment security did not reduce theeffect of MS on DTA. More recently, Norenzayan, Dar-Nimrod,Hansen, and Proulx (2009) found that MS increased DTAequally among religious and nonreligious participants, but onlynonreligious participants responded to a religiously orientedworldview threat with increased defense. In investigating theeffect of implicit versus explicit self-esteem on reactions to MS,Schmeichel et al. (2009, Study 3) found that MS increased DTAregardless of self-esteem levels, but only those with low im-plicit self-esteem and high explicit self-esteem responded de-fensively. Finally, Fritsche et al. (2008) found that relative to agroup of participants who merely imagined death, participantswho imagined dying in a controlled and autonomous mannerdid not show increased worldview defense. Once again, how-ever, this manipulation did not reduce DTA. Thus, certainpersonality traits or experimental manipulations seem capableof reducing worldview defense, but they do not always appearto reduce DTA.

Although additional research is needed, there are severalexplanations for these findings. One possibility is that peoplewith different beliefs or personality characteristics defendagainst the threat of death in different ways. As noted byFlorian et al. (2001), the mere observation that certain peopledo not show increased defense, as measured in a particularstudy, even though they have increased DTA does not mean thatthese people would not defend if given a different opportunity.For example, whereas those high in attachment security do notshow more punitive judgments against moral transgressors afterMS (Mikulincer & Florian, 2000), they do increase investmentin close relationships (e.g., Cox et al., 2008). Further, in somecases these defenses might even occur at an implicit level (e.g.,Gurari, Strube, & Hetts, 2009).

In other cases, the manner in which death is contemplated may,itself, be defensive or may elicit a different set of psychologicalprocesses. Imagining an autonomous heroic death (e.g., Fritsche etal., 2008), or one that would be looked upon favorably by others,for example, may in some cases preclude the need for furtherdefensiveness. Although soldiers marching off to war are con-fronted with the prospect of dying, imagining that their death willserve a higher purpose and that the world will be a better place asa result of their sacrifice might provide solace and the fortitude toplace oneself at risk.

This leaves the question of whether there are legitimate casesin which DTA can be aroused without necessarily leading todefense. The existential perspective from which TMT waspartly derived suggests this is theoretically possible. For exam-ple, Heidegger (1927/1962) asserted that the authentic lifeinvolves owning up to and acknowledging death rather thantrying to cover it up and deny it by escaping into the sharedcultural mode of existence. The distinction between authenticand inauthentic modes of being toward death is a significantpart of nearly all existential theories. For Jean-Paul Sartre, itappears in the notion of bad faith; for Søren Kierkegaard, indouble-mindedness (Fingarette, 1969/2000). Thus, respondingto thoughts of death with worldview or self-esteem defense, oras Heidegger put it, falling into the world, may not be the onlyanswer. An authentic mode of existence involves living accord-ing to values that are self-determined rather than dictated byculture. It involves preparing for death by living truthfullyrather than hiding from it in the comfortable illusions affordedby a cultural worldview. Becker (1973) made a similar argu-ment in the conclusion of his book The Denial of Death, arguingthat we should remove our heads from the sand and acknowl-edge our finitude. He nevertheless appeared ambivalent towardthis possibility and doubted people’s ability to live with theanxiety of death. Becker seemed resigned to the idea thathumans will always need to use defenses to quell death anxiety.He believed we should live according to a higher ethical stan-dard, choosing to defend against our death anxiety in ways thatdo not increase suffering for others or escalate conflict amongpeople with divergent worldviews.

Given the possibility of authentic and inauthentic modes ofengaging death, an important question would be whether or notauthentic ways of being will also provide psychological reliefagainst death anxiety. If so, it would be difficult to distinguishauthentic ways of being from those that are defensive (and inau-thentic), given that both provide tranquilization. Future TMT re-search on the impact of thoughts of death should neverthelessexplore the possibility of authentic versus inauthentic responsesand the role that DTA might play in these responses. One encour-aging and complementary trajectory is to consider whether allthoughts of death are, in a sense, created equal in their capacity toprovoke defensiveness. Cozzolino (2006), for example, has arguedthat whereas most MS manipulations conjure up an abstract spec-ter of death, more specific and detailed self-focused experienceswith the prospect of death may engage a different existentialprocess that is oriented more toward psychological growth. Ofcourse, whether these confrontations elicit the same implicit DTAthat has been the focus of this review, or whether they serve to holddeath ideation in consciousness and thus represent a different

715DEATH-THOUGHT ACCESSIBILITY

Page 18

psychological state (perhaps more akin to TMT’s proximal mode),remains to be investigated.6

What is the role of DTA in the production of worldview andself-esteem defense? An implied assumption of TMT is that inthe face of mortality awareness DTA mediates, or at least partiallymediates, worldview and self-esteem defensiveness. However, fewstudies have directly tested this assumption by statistical inference.To our knowledge, only five studies have used the classic Baronand Kenny (1986) analysis to verify whether or not DTA mediatesdefense. In the first, conducted by Fransen et al. (2008), exposureto an insurance brand logo increased DTA and worldview-consistent spending intentions. In this case, the results indicatedthat DTA fully mediated the effect of brand exposure on spendingintentions.

Another series of studies found that exposure to images ofdeteriorating infrastructure (i.e., rubble) increased DTA and world-view defense (e.g., dogmatism, endorsement of militaristic re-sponses against terrorism) and that DTA did in fact mediate theeffect of destruction images on worldview defense (Vail et al.,2010). In the study establishing mediation, DTA was measured viareaction times in a lexical decision task that also assessed responselatencies for nation-oriented and war-oriented words. DTA wasfound to be a unique mediator, as the influence was not reproducedby the accessibility of nation or war words.

In two other studies conducted by Das et al. (2009), terrorismnews was found to increase DTA and prejudice toward Arabs. Inboth of these cases, however, statistical mediation was not sup-ported, because there was no overall relationship between DTAand prejudice across conditions. Instead, the relationship betweenDTA and prejudice was present in the threat condition (in whichparticipants read the terrorism news article) but not in the controlcondition (in which participants read an article about the Olym-pics). Although Das et al. (2009) described the observed relation-ship as moderation, Harackiewicz, Abrahams, and Wageman(1987) referred to it as interactional mediation (cf. Judd & Kenny1981), wherein the independent variable interacts with the processvariable to predict the outcome variable. The basic differencebetween simple mediation and interactional mediation is thatthe latter denotes a relationship between the mediator and theoutcome in one experimental condition but not in another, whereassimple mediation involves the same mediated process regardless ofcondition.

Finally, a recent study by Hayes et al. (2010) investigating theextent to which people will change their worldview beliefs inresponse to a worldview-threatening argument found a similarpattern. In this case, the relationship between DTA and defensiveauthor derogation was found only in the threat condition. How-ever, this relationship in the threat condition was further mediatedby the degree to which participants were willing to change theirworldview beliefs in line with the threatening communication,indicating moderated mediation.

Thus, despite the discrepancy in the relationship between DTAand defense in the control conditions of these studies, all five casessupport the notion that DTA mediates defense in the threat con-ditions. One can only speculate as to why the hypothesized processwas present in the control conditions of Fransen et al. (2008) andVail et al. (2010) but not of Das et al. (2009) and Hayes et al.(2010). As Spencer, Zanna, and Fong (2005) noted, when inves-tigating psychological processes that are proposed to mediate a

certain effect, one must consider whether or not the same mediatedprocess would be expected in the treatment and control conditions.Perhaps the nature of the defense measures in Fransen et al. andVail et al., which involved spending intensions or endorsement ofmilitaristic responses, was more amenable to capturing subtledifferences in DTA in the control condition than was the prejudicedefense measure used by Das et al. or the author-derogationmeasure used by Hayes et al. Although our previous discussion ofdispositional DTA suggests that all types of defense might occurdue to baseline variations in DTA, there may also be reason tobelieve that dispositional DTA would not predict certain antisocialdefenses, such as prejudice, unless the potential target of prejudiceis presented as a source of threat. This is ultimately a matter forfuture research.

Setting aside the differences in processes occurring in threatversus control conditions, Spencer et al. (2005) have argued,moreover, that a series of experiments may in some cases be moreappropriate for establishing theoretical mediation. They noted thatone concern with seeking to establish statistical mediation is thatmeasurement of the process variable must “not interfere with howthe process leads to the effect of interest” (pp. 847–848). Recallthat the measurement of DTA may in some cases bring thoughts ofdeath into consciousness or prime the concept for all participants.Thus, in certain cases, designs dependent on statistical medita-tional analysis may be inappropriate for investigating whether ornot DTA mediates defense. At the least, it may behoove research-ers to consider DTA measures that are more spontaneous and lessdeliberative.

In lieu of this method, Spencer et al. (2005) proposed theexperimental-causal-chain design, which they argued is ideal forcases in which both measurement and manipulation of the processare relatively easy. In this design, a series of experiments isconducted to show that manipulation of the independent variable(in our case, threat) affects the proposed process (DTA) and thatmanipulation of the process (DTA) in turn affects the outcome(worldview and self-esteem defense). As the above empiricalreview shows, manipulating threat does indeed affect DTA, andmanipulating DTA affects defense. And we can go one step furtherin noting that the employment of worldview and self-esteemdefense has been found to reduce DTA. Thus, based on Spencer etal.’s (2005) experimental-causal-chain design, most of the avail-able evidence supports the proposed process that DTA mediatesthe impact of threat on defense. Although the majority of extantevidence suggests mediation, researchers should continue to mon-itor this issue as additional research is conducted.

6 Another variant of the question of whether all thoughts of death arecreated equal is whether thoughts of death need to be self-referent to elicitworldview and other such defenses. Although some of the death-association studies suggest this might not be the case, other studies havefound stronger worldview defense effects when thoughts of death are tiedto the self versus focused on the death of others (Greenberg et al., 1994;Nelson, Moore, Olivetti, & Scott, 1997). This suggests that some degree ofself-reference when death thoughts are active may in fact prompt moredefense.

716 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 19

Conclusion

‘Tis after death that we measure men.

—James Baron Hope, “Our Heroic Dead”

Terror management theory and the hundreds of studies it hasinspired over the last 20 years have helped to move existentialtheorizing about fundamental aspects of human social behavior outof the closet and into the common parlance of academic psychol-ogy. Still, developments must be made on both conceptual andmethodological fronts if any theory is to progress and remainactive. The DTA concept is a vital development to this effect, as ithas simultaneously contributed to the growth and refinement of thetheory, enabled exploration of previously uncharted territory, and,perhaps most important, provided a generative foundation forfuture research. Although James Baron Hope may be right aboutthe measurement of men (and women) after death, it appears thatDTA may provide a means of measuring death, or at least theextent to which it occupies our minds while we live.

References

Altemeyer, B., & Hunsberger, B. (1992). Authoritarianism, religious fun-damentalism, quest, and prejudice. International Journal for the Psy-chology of Religion, 2, 113–133. doi:10.1207/s15327582ijpr0202_5

Arndt, J., Cook, A., Goldenberg, J. L., & Cox, C. R. (2007). Cancer and thethreat of death: The cognitive dynamics of death thought suppressionand its impact on behavioral health intentions. Journal of Personalityand Social Psychology, 92, 12–29. doi:10.1037/0022-3514.92.1.12

Arndt, J., Cook, A., & Routledge, C. (2004). The blueprint of terrormanagement: Understanding the cognitive architecture of psychologicaldefense against the awareness of death. In J. Greenberg, S. L. Koole, &T. Pyszczynski (Eds.), Handbook of experimental existential psychology(pp. 35–53). New York, NY: Guilford Press.

Arndt, J., Greenberg, J., & Cook, A. (2002). Mortality salience and thespreading activation of worldview-relevant constructs: Exploring thecognitive architecture of terror management. Journal of ExperimentalPsychology: General, 131, 307–324.

Arndt, J., Greenberg, J., Pyszczynski, T., & Solomon, S. (1997). Sublim-inal exposure to death-related stimuli increases defense of the culturalworldview. Psychological Science, 8, 379–385. doi:10.1111/j.1467-9280.1997.tb00429.x

Arndt, J., Greenberg, J., Simon, L., Pyszczynski, T., & Solomon, S. (1998).Terror management and self-awareness: Evidence that mortality salienceprovokes avoidance of the self-focused state. Personality and SocialPsychology Bulletin, 24, 1216–1227. doi:10.1177/01461672982411008

Arndt, J., Greenberg, J., Solomon, S., Pyszczynski, T., & Simon, L. (1997).Suppression, accessibility of death-related thoughts, and cultural world-view defense: Exploring the psychodynamics of terror management.Journal of Personality and Social Psychology, 73, 5–18. doi:10.1037/0022-3514.73.1.5

Arndt, J., Lieberman, J. D., Cook, A., & Solomon, S. (2005). Terrormanagement in the courtroom: Exploring the effects of mortality sa-lience on legal decision-making. Psychology, Public Policy, and Law,11, 407–438. doi:10.1037/1076-8971.11.3.407

Arndt, J., Solomon, S., Kasser, T., & Sheldon, K. M. (2004). The urge tosplurge: A terror management account of materialism and consumerbehavior. Journal of Consumer Psychology, 14, 198–212. doi:10.1207/s15327663jcp1403_2

Bargh, J. A., & Chartrand, T. L. (2000). The mind in the middle: Apractical guide to priming and automaticity research. In H. T. Reis &C. M. Judd (Eds.), Handbook of research methods in social and per-

sonality psychology (pp. 253–285). Cambridge, England: CambridgeUniversity Press.

Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variabledistinction in social psychological research: Conceptual, strategic, andstatistical considerations. Journal of Personality and Social Psychology,51, 1173–1182. doi:10.1037/0022-3514.51.6.1173

Bassett, J. F. (2005). Does threatening valued components of culturalworldview alter explicit and implicit attitudes about death? IndividualDifferences Research, 3, 260–268.

Baumeister, R. F., Bratslavsky, E., Muraven, M., & Tice, D. M. (1998).Ego depletion: Is the active self a limited resource? Journal of Person-ality and Social Psychology, 74, 1252–1265. doi:10.1037/0022-3514.74.5.1252

Becker, E. (1971). The birth and death of meaning: An interdisciplinaryperspective on the problem of man (2nd ed.). New York, NY: Free Press.

Becker, E. (1973). The denial of death. New York, NY: Free Press.Becker, E. (1975). Escape from evil. New York, NY: Free Press.Berger, P. L., & Luckmann, T. (1967). The social construction of reality:

A treatise in the sociology of knowledge. New York, NY: Anchor Books.Bowlby, J. (1969). Attachment. New York: Basic Books.Bruner, J. S. (1957). On perceptual readiness. Psychological Review, 64,

123–152. doi:10.1037/h0043805Burris, C. T., & Rempel, J. K. (2004). “It’s the end of the world as we

know it”: Threat and the spatial-symbolic self. Journal of Personalityand Social Psychology, 86, 19–42. doi:10.1037/0022-3514.86.1.19

Cialdini, R. B. (2001). Influence: Science and practice (4th ed.). Toronto,Ontario, Canada: Allyn & Bacon.

Collins, A. M., & Loftus, E. F. (1975). A spreading-activation theory ofsemantic processing. Psychological Review, 82, 407–428. doi:10.1037/0033-295X.82.6.407

Conte, H. R., Weiner, M. B., & Plutchik, R. (1982). Measuring deathanxiety: Conceptual, psychometric, and factor-analytic aspects. Journalof Personality and Social Psychology, 43, 775–785. doi:10.1037/0022-3514.43.4.775

Cooley, C. H. (1964). Human nature and the social order. New York, NY:Schocken Books. (Original work published 1964)

Cooper, D. P., Goldenberg, J. L., & Arndt, J. (2010). Empowering the self:Using the terror management health model to promote breast selfexamination. Manuscript submitted for publication, University of SouthFlorida.

Cox, C. R., Arndt, J., Pyszczynksi, T., Greenberg, J., Abdollahi, A., &Solomon, S. (2008). Terror management and adults’ attachment to theirparents: The safe haven remains. Journal of Personality and SocialPsychology, 94, 696–717. doi:10.1037/0022-3514.94.4.696

Cox, C. R., Goldenberg, J. L., Pyszczynski, T., & Weise, D. (2007).Disgust, creatureliness and the accessibility of death-related thoughts.European Journal of Social Psychology, 37, 494–507. doi:10.1002/ejsp.370

Cox, C. R., Reid-Arndt, S., Arndt, J., & Moser, R. (2010). Coping withcancer: The role of death thought accessibility in quality of life amongwomen diagnosed with breast cancer. Manuscript submitted for publi-cation, Texas Christian University.

Cozzolino, P. J. (2006). Death, contemplation, growth, and defense: Con-verging evidence of a dual-existential system? Psychological Inquiry,17, 278–287. doi:10.1080/10478400701366944

Crumbaugh, J. C., & Maholick, L. T. (1969). Manual of instruction for thePurpose in Life Test. Munster, IN: Psychometric Affiliates.

Dai, X., Wertenbroch, K., & Brendl, C. M. (2008). The value heuristic injudgments of relative frequency. Psychological Science, 19, 18–19.doi:10.1111/j.1467-9280.2008.02039.x

Das, E., Bushman, B. J., Bezemer, M. D., Kerkhof, P., & Vermeulen, I. E.(2009). How terrorism news reports increase prejudice against out-groups: A terror management account. Journal of Experimental SocialPsychology, 45, 453–459. doi:10.1016/j.jesp.2008.12.001

717DEATH-THOUGHT ACCESSIBILITY

Page 20

De Houwer, J. (2003). A structural analysis of indirect measures ofattitudes. In J. Musch & K. C. Klauer (Eds.), The psychology of evalu-ation: Affective processes in cognition and emotion (pp. 219–244).Mahwah, NJ: Erlbaum.

Descutner, C. J., & Thelen, M. H. (1991). Development and validation ofa Fear-of-Intimacy Scale. Psychological Assessment: A Journal of Con-sulting and Clinical Psychology, 3, 218 –225. doi:10.1037/1040-3590.3.2.218

Dunkel, C. S. (2009). The association between thoughts of defecation andthoughts of death. Death Studies, 33, 356 –371. doi:10.1080/07481180802602808

Duval, T. S., & Wicklund, R. A. (1972). A theory of objective self-awareness. New York, NY: Academic Press.

Epstein, S. (1980). The self-concept: A review and the proposal of anintegrated theory of personality. In E. Staub (Ed.), Personality: Basicand current research (pp. 82–132). Englewood Cliffs, NJ: Prentice-Hall.

Epstein, S. (1994). Integration of the cognitive and the psychodynamicunconscious. American Psychologist, 49, 709–724. doi:10.1037/0003-066X.49.8.709

Fingarette, H. (2000). Self-deception. Los Angeles: University of Califor-nia Press. (Original work published in 1969)

Fiske, S. T., & Taylor, S. E. (1991). Social cognition. New York, NY:McGraw-Hill.

Florian, V., Mikulincer, M., & Hirschberger, G. (2001). An existentialview on mortality salience effects: Personal hardiness, death-thoughtaccessibility, and cultural worldview defence. British Journal of SocialPsychology, 40, 437–453. doi:10.1348/014466601164911

Florian, V., Mikulincer, M., & Hirschberger, G. (2002). The anxiety-buffering function of close relationships: Evidence that relationshipcommitment acts as a terror management mechanism. Journal of Per-sonality and Social Psychology, 82, 527–542. doi:10.1037//0022-3514.82.4.527

Fransen, M. L., Fennis, B. M., Pruyn, A. T. H., & Das, E. (2008). Rest inpeace? Brand-induced mortality salience and consumer behavior. Jour-nal of Business Research, 61, 1053–1061. doi:10.1016/j.jbusres.2007.09.020

Freud, A. (1966). The ego and the mechanisms of defense. New York, NY:International Universities Press. (Original work published 1936)

Friedman, M., & Rholes, W. S. (2007). Successfully challenging funda-mentalist beliefs results in increased death awareness. Journal of Exper-imental Social Psychology, 43, 794 – 801. doi:10.1016/j.jesp.2006.07.008

Friedman, M., & Rholes, W. S. (2009). Religious fundamentalism andterror management: Differences by interdependent and independent self-construal. Self and Identity, 8, 24–44. doi:10.1080/15298860801984788

Fritsche, I., Jonas, E., & Fankhanel, T. (2008). The role of control moti-vation in mortality salience effects on ingroup support and defense.Journal of Personality and Social Psychology, 95, 524 –541. doi:10.1037/a0012666

Fritsche, I., Jonas, E., Fischer, P., Koranyi, N., Berger, N., & Fleischmann,B. (2007). Mortality salience and the desire for offspring. Journal ofExperimental Social Psychology, 43, 753–762. doi:10.1016/j.jesp.2006.10.003

Gailliot, M. T., Schmeichel, B. J., & Baumeister, R. F. (2006). Self-regulatory processes defend against the threat of death: Effects ofself-control depletion and trait self-control on thoughts and fears ofdying. Journal of Personality and Social Psychology, 91, 49–62. doi:10.1037/0022-3514.91.1.49

Gailliot, M. T., Schmeichel, B. J., & Maner, J. K. (2007). Differentiatingthe effects of self-control and self-esteem on reactions to mortalitysalience. Journal of Experimental Social Psychology, 43, 894–901.doi:10.1016/j.jrp.2006.04.004

Gawronski, B., & Bodenhausen, G. V. (2005). Accessibility effect onimplicit social cognition: The role of knowledge activation and retrieval

experiences. Journal of Personality and Social Psychology, 89, 672–685. doi:10.1037/0022-3514.89.5.672

Goldenberg, J. L., & Arndt, J. (2008). The implications of death for health:A terror management health model for behavioral health promotion.Psychological Review, 115, 1032–1053. doi:10.1037/a0013326

Goldenberg, J. L., Arndt, J., Hart, J., & Routledge, C. (2008). Uncoveringan existential barrier to breast self-exam behavior. Journal of Experi-mental Social Psychology, 44, 260–274. doi:10.1016/j.jesp.2007.05.002

Goldenberg, J. L., Cox, C. R., Pyszczynski, T., Greenberg, J., & Solomon,S. (2002). Understanding human ambivalence about sex: The effects ofstripping sex of meaning. Journal of Sex Research, 39, 310–320. doi:10.1080/00224490209552155

Goldenberg, J. L., Pyszczynski, T., Greenberg, J., & Solomon, S. (2000).Fleeing the body: A terror management perspective on the problem ofhuman corporeality. Personality and Social Psychology Review, 4, 200–218. doi:10.1207/S15327957PSPR0403_1

Goldenberg, J. L., Pyszczynski, T., McCoy, S. K., Greenberg, J., &Solomon, S. (1999). Death, sex, love, and neuroticism: Why is sex sucha problem? Journal of Personality and Social Psychology, 77, 1173–1187. doi:10.1037/0022-3514.77.6.1173

Gray, J. (2004). Heresies: Against progress and other illusions. London,England: Granta Books.

Greenberg, J., Arndt, J., Schimel, J., Pyszczynski, T., & Solomon, S.(2001). Clarifying the function of mortality-salience induced worldviewdefense: Renewed suppression or reduced accessibility of death-relatedthoughts? Journal of Experimental Social Psychology, 37, 70–76. doi:10.1006/jesp.2000.1434

Greenberg, J., Pyszczynski, T., & Solomon, S. (1986). The causes andconsequences of a need for self-esteem: A terror management theory. InR. F. Baumeister (Ed.), Public self and private self (pp. 189–212). NewYork, NY: Springer-Verlag.

Greenberg, J., Pyszczynski, T., Solomon, S., Simon, L., & Breus, M.(1994). Role of consciousness and accessibility of death-related thoughtsin mortality salience effects. Journal of Personality and Social Psychol-ogy, 67, 627–637. doi:10.1037/0022-3514.67.4.627

Greenberg, J., Simon, L., Harmon-Jones, E., Solomon, S., Pyszczynski, T.,& Chatel, D. (1995). Testing alternative explanations for mortalityeffects: Terror management, value accessibility, or worrisome thoughts?European Journal of Social Psychology, 12, 417–433. doi:10.1002/ejsp.2420250406

Greenberg, J., Solomon, S., & Arndt, J. (2008). A uniquely human moti-vation: Terror management. In J. Shah & W. Gardner (Eds.), Handbookof motivation science (pp. 113–134). New York, NY: Guilford Press.

Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition:Attitudes, self-esteem, and stereotypes. Psychological Review, 102,4–27. doi:10.1037/0033-295X.102.1.4

Greenwald, A. G., McPhee, D. E., & Schwartz, J. L. K. (1998). Measuringindividual differences in implicit cognition: The Implicit AssociationTest. Journal of Personality and Social Psychology, 74, 1464–1480.doi:10.1037/0022-3514.74.6.1464

Gurari, I., Strube, M. J., & Hetts, J. J. (2009). Death? Be proud! The ironiceffects of terror salience on implicit self-esteem. Journal of AppliedSocial Psychology, 39, 494 –507. doi:10.1111/j.1559-1816.2008.00448.x

Haidt, J., McCauley, C. R., & Rozin, P. (1994). Individual differences insensitivity to disgust: A scale sampling seven domains of disgust elici-tors. Personality and Individual Differences, 16, 701–713. doi:10.1016/0191-8869(94)90212-7

Hansen, J., Winzeler, S., & Topolinski, S. (2010). When the death makesyou smoke: A terror management perspective on the effectiveness ofcigarette on-pack warnings. Journal of Experimental Social Psychology,46, 226–228. doi:10.1016/j.jesp.2009.09.007

Harackiewicz, J. M., Abrahams, S., & Wageman, R. (1987). Performanceevaluation and intrinsic motivation: The effects of evaluative focus,

718 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 21

rewards, and achievement orientation. Journal of Personality and SocialPsychology, 53, 1015–1023. doi:10.1037/0022-3514.53.6.1015

Harmon-Jones, E., Simon, L., Greenberg, J., Pyszczynski, T., Solomon, S.,& McGregor, H. (1997). Terror management theory and self-esteem:Evidence that increased self-esteem reduces mortality salience effects.Journal of Personality and Social Psychology, 72, 24–36. doi:10.1037/0022-3514.72.1.24

Hart, J., Shaver, P. R., & Goldenberg, J. L. (2005). Attachment, self-esteem, worldviews, and terror management: Evidence for a tripartitesecurity system. Journal of Personality and Social Psychology, 88,999–1013. doi:10.1037/0022-3514.88.6.999

Hayes, J., Schimel, J., Faucher, E. H., & Williams, T. J. (2008). Evidencefor the DTA hypothesis II: Threatening self-esteem increases death-thought accessibility. Journal of Experimental Social Psychology, 44,600–613. doi:10.1016/j.jesp.2008.01.004

Hayes, J., Schimel, J., Howard, A. L., Webber, D., & Faucher, E. H.(2010). When worldviews change and when they stay the same: Explor-ing accommodation vs. derogation in response to worldview threat.Unpublished manuscript, University of Alberta, Edmonton, Alberta,Canada.

Hayes, J., Schimel, J., & Williams, T. J. (2008). Fighting death with death:The buffering effects of learning that worldview violators have died.Psychological Science, 19, 501–507. doi:10.1111/j.1467-9280.2008.02115.x

Heidegger, M. (1962). Being and time. New York: Harper & Row. (Orig-inal work published 1927)

Heine, S. J., Proulx, T., & Vohs, K. (2006). The meaning maintenancemodel: On the coherence of social motivations. Personality and SocialPsychology Review, 10, 88–110. doi:10.1207/s15327957pspr1002_1

Hergenhahn, B. R. (2001). An introduction to the history of psychology(4th ed.). Belmont, CA: Wadsworth.

Higgins, E. T., & King, G. (1981). Accessibility of social constructs:Information processing consequences of individual and contextual vari-ability. In N. Cantor & J. Kihlstrom (Eds.), Personality, cognition, andsocial interaction (pp. 69–121). Hillsdale, NJ: Erlbaum.

Hirschberger, G. (2006). Terror management and attributions of blame toinnocent victims: Reconciling compassionate and defensive responses.Journal of Personality and Social Psychology, 91, 832– 844. doi:10.1037/0022-3514.91.5.832

Hirschberger, G., Florian, V., & Mikulincer, M. (2005). Fear and compas-sion: A terror management analysis of emotional reactions to personswith disabilities. Rehabilitation Psychology, 50, 246–257. doi:10.1037/0090-5550.50.3.246

Jessop, D. C., Albery, I. P., Rutter, J., & Garrod, H. (2008). Understandingthe impact of mortality-related health-risk information: A terror man-agement theory perspective. Personality and Social Psychology Bulletin,34, 951–964. doi:10.1177/0146167208316790

Jessop, D. C., & Wade, J. (2008). Fear appeals and binge drinking: A terrormanagement theory perspective. British Journal of Health Psychology,13, 773–788. doi:10.1348/135910707X272790

Jonas, E., & Fischer, P. (2006). Terror management and religion: Evidencethat intrinsic religiousness mitigates worldview defense following mor-tality salience. Journal of Personality and Social Psychology, 91, 553–567. doi:10.1037/0022-3514.91.3.553

Jonas, E., Martens, A., Kayser, D. N., Fritsche, I., Sullivan, D., & Green-berg, J. (2008). Focus theory of normative conduct and terror-management theory: The interactive impact of mortality salience andnorm salience on social judgment. Journal of Personality and SocialPsychology, 95, 1239–1251. doi:10.1037/a0013593

Judd, C. M., & Kenny, D. A. (1981). Estimating the effects of socialinterventions. Cambridge, England: Cambridge University Press.

Kasser, T., & Sheldon, K. M. (2000). Of wealth and death: Materialism,mortality salience, and consumption behavior. Psychological Science,11, 348–351. doi:10.1111/1467-9280.00269

King, L. A., Hicks, J. A., & Abdelkhalik, J. (2009). Death, life, scarcity,and value: An alternative perspective on the meaning of death. Psycho-logical Science, 20, 1459–1462. doi:10.1111/j.1467-9280.2009.02466.x

Koole, S. L., & Van den Berg, A. (2005). Lost in the wilderness: Terrormanagement, action orientation, and nature evaluation. Journal of Per-sonality and Social Psychology, 88, 1014–1028. doi:10.1037/0022-3514.88.6.1014

Kornblum, S., Hasbroucq, T., & Osman, A. (1990). Dimensional overlap:Cognitive basis for stimulus–response compatibility: A model and tax-onomy. Psychological Review, 97, 253–270. doi:10.1037/0033-295X.97.2.253

Kuhl, J. (1981). Motivational and functional helplessness: The moderatingeffect of state versus action orientation. Journal of Personality andSocial Psychology, 40, 155–170. doi:10.1037/0022-3514.40.1.155

Landau, M. J., Johns, M., Greenberg, J., Pyszczynski, T., Martens, A.,Goldenberg, J. L., & Solomon, S. (2004). A function of form: Terrormanagement and structuring the social world. Journal of Personality andSocial Psychology, 87, 190–210. doi:10.1037/0022-3514.87.2.190

Landau, M. J., Solomon, S., Greenberg, J., Cohen, F., Pyszczynski, T.,Arndt, J., . . . Cook, A. (2004). Deliver us from evil: The effects ofmortality salience and reminders of 9/11 on support for President GeorgeW. Bush. Personality and Social Psychology Bulletin, 30, 1136–1150.doi:10.1177/0146167204267988

Lerner, M. J. (1980). The belief in a just world: A fundamental delusion.New York, NY: Plenum.

Lifton, R. J. (1979). The broken connection: On death and the continuityof life. Washington, DC: American Psychiatric Press.

MacDorman, K. F. (2005, December). Mortality salience and the uncannyvalley. Proceedings of the Fifth IEEE-RAS International Conference onHumanoid Robots, 5, 399–405. doi:10.1109/ICHR.2005.1573600

Martens, A., Burke, B. L., Schimel, J., & Faucher, E. H. (in press). Samebut different: Meta-analytically examining the uniqueness of mortalitysalience effects. European Journal of Social Psychology.

Martens, A., Greenberg, J., Schimel, J., & Landau, M. J. (2004). Ageismand death: Effects of mortality salience and perceived similarity to elderson reactions to elderly people. Personality and Social Psychology Bul-letin, 30, 1524–1536. doi:10.1177/0146167204271185

Maxfield, M., Pyszczynksi, T., Kluck, B., Cox, C. R., Greenberg, J.,Solomon, S., & Weise, D. (2007). Age-related differences in response tothoughts of one’s own death: Mortality salience and judgments of moraltransgressions. Psychology and Aging, 22, 341–353. doi:10.1037/0882-7974.22.2.341

McGregor, I., Zanna, M. P., Holmes, J. G., & Spencer, S. J. (2001).Compensatory conviction in the face of personal uncertainty: Going toextremes and being oneself. Journal of Personality and Social Psychol-ogy, 80, 472–488. doi:10.1037//0022-3514.80.3.472

Mead, G. H. (1934). Mind, self, and society. Chicago, IL: University ofChicago Press.

Mikulincer, M., & Florian, V. (2000). Exploring individual differences inreactions to mortality salience: Does attachment style regulate terrormanagement mechanisms? Journal of Personality and Social Psychol-ogy, 79, 260–273. doi:10.1037//0022-3514.79.2.260

Mikulincer, M., & Florian, V. (2002). The effects of mortality salience onself-serving attributions: Evidence for the function of self-esteem as aterror management mechanism. Basic and Applied Social Psychology,24, 261–271. doi:10.1207/S15324834BASP2404_2

Mikulincer, M., Florian, V., Birnbaum, G., & Malishkevich, S. (2002). Thedeath-anxiety buffering function of close relationships: Exploring theeffects of separation reminders on death-thought accessibility. Person-ality and Social Psychology Bulletin, 28, 287–299. doi:10.1177/0146167202286001

Mikulincer, M., Florian, V., & Hirschberger, G. (2003). The existentialfunction of close relationships: Introducing death into the science of

719DEATH-THOUGHT ACCESSIBILITY

Page 22

love. Personality and Social Psychology Review, 7, 20–40. doi:10.1207/S15327957PSPR0701_2

Mikulincer, M., Florian, V., & Tolmacz, R. (1990). Attachment styles andfear of personal death: A case study of affect regulation. Journal ofPersonality and Social Psychology, 64, 817–826. doi:10.1037/0022-3514.58.2.273

Mori, M. (1970). Bukimi no tani [The uncanny valley]. Energy, 7, 33–35.Moser, R. P., Arndt, J., Han, P., Waters, E., Amsellem, M., & Hesse, B.

(2010). Avoidance of doctors, cancer-related mortality salience andassociated factors: United States prevalence estimates from a nationallyrepresentative survey. Manuscript submitted for publication, NationalCancer Institute, Bethesda, MD.

Murray, H. A. (2008). Explorations in personality (70th anniversary ed.).New York, NY: Oxford University Press. (Original work published1938)

Navarrete, D. C., & Fessler, D. M. T. (2005). Normative bias and adaptivechallenges: A relational approach to coalitional psychology and a cri-tique of terror management theory. Evolutionary Psychology, 3, 297–325.

Nelson, L. J., Moore, D. L., Olivetti, J., & Scott, T. (1997). General andpersonal mortality salience and nationalistic bias. Personality and SocialPsychology Bulletin, 23, 884–892. doi:10.1177/0146167297238008

Neuberg, S. L., & Newsom, J. (1993). Personal need for structure: Indi-vidual differences in the desire for simple structure. Journal of Person-ality and Social Psychology, 65, 113–131. doi:10.1037/0022-3514.65.1.113

Nisbett, R., & Wilson, T. (1977). Telling more than we can know: Verbalreports on mental processes. Psychological Review, 84, 231–259. doi:10.1037/0033-295X.84.3.231

Norenzayan, A., Dar-Nimrod, I., Hansen, I. G., & Proulx, T. (2009).Mortality salience and religion: Divergent effects on the defense ofcultural worldviews for the religious and the non-religious. EuropeanJournal of Social Psychology, 39, 101–113. doi:10.1002/ejsp.482

Ogilvie, D. M. (1987). The undesired self: A neglected variable in person-ality research. Journal of Personality and Social Psychology, 52, 379–385. doi:10.1037/0022-3514.52.2.379

Ogilvie, D. M., Cohen, F., & Solomon, S. (2008). The undesired self:Deadly connotations. Journal of Research in Personality, 42, 564–576.doi:10.1016/j.jrp.2007.07.012

Proulx, T., & Heine, S. J. (2008). The case of the transmogrifying exper-imenter: Affirmation of a moral schema following implicit changedetection. Psychological Science, 19, 1294–1300. doi:10.1111/j.1467-9280.2008.02238.x

Proulx, T., & Heine, S. J. (2009). Connections from Kafka: Exposure tomeaning threats improves implicit learning of an artificial grammar.Psychological Science, 20, 1125–1131. doi:10.1111/j.1467-9280.2009.02414.x

Pyszczynski, T., Greenberg, J., & Solomon, S. (1999). A dual-processmodel of defense against conscious and unconscious death-relatedthoughts: An extension of terror management theory. PsychologicalReview, 106, 835–845. doi:10.1037/0033-295X.106.4.835

Pyszczynksi, T., Greenberg, J., Solomon, S., Arndt, J., & Schimel, J.(2004). Why do people need self-esteem? A theoretical and empiricalreview. Psychological Bulletin, 130, 435– 468. doi:10.1037/0033-2909.130.3.435

Pyszczynski, T., Greenberg, J., Solomon, S., & Maxfield, M. (2006). Onthe unique psychological import of the human awareness of mortality:Theme and variations. Psychological Inquiry, 17, 328 –356. doi:10.1080/10478400701369542

Pyszczynski, T., Solomon, S., & Greenberg, J. (2003). In the wake of 9/11:The psychology of terror. Washington, DC: American PsychologicalAssociation.

Pyszczynski, T., Wicklund, R. A., Floresku, S., Koch, H., Gauch, G.,Solomon, S., & Greenberg, J. (1996). Whistling in the dark: Exaggerated

consensus estimates in response to incidental reminders of mortality.Psychological Science, 7, 332–336. doi:10.1111/j.1467-9280.1996.tb00384.x

Rorschach, H. (1942). Psychodiagnostics. Bern, Switzerland: Bircher.(Original work published 1921)

Rosenblatt, A., Greenberg, J., Solomon, S., Pyszczynski, T., & Lyon, D.(1989). Evidence for terror management theory I: The effects of mor-tality salience on reactions to those who violate or uphold culturalvalues. Journal of Personality and Social Psychology, 57, 681–690.doi:10.1037/0022-3514.57.4.681

Routledge, C., Arndt, J., Sedikides, C., & Wildschut, T. (2008). A blastfrom the past: The terror management function of nostalgia. Journal ofExperimental Social Psychology, 44, 132–140. doi:10.1016/j.jesp.2006.11.001

Routledge, C., Ostafin, B., Juhl, J., & Sedikides, C. (2010). Death can bebad for your psychological health: Mortality salience undermines well-being and exacerbates maladaptive behavior for individuals with lowself-esteem. Manuscript submitted for publication, North Dakota StateUniversity.

Rutjens, B. T., van der Pligt, J., & van Harreveld, F. (2009). Things will getbetter: The anxiety-buffering qualities of progressive hope. Personalityand Social Psychology Bulletin, 35, 535–543. doi:10.1177/0146167208331252

Schimel, J., Hayes, J., Williams, T., & Jahrig, J. (2007). Is death really theworm at the core? Converging evidence that worldview threat increasesdeath-thought accessibility. Journal of Personality and Social Psychol-ogy, 92, 789–803. doi:10.1037/0022-3514.92.5.789

Schmeichel, B. J., Gailliot, M. T., Filardo, E., McGregor, I., Gitter, S., &Baumeister, R. F. (2009). Terror management theory and self-esteemrevisited: The roles of implicit and explicit self-esteem in mortalitysalience effects. Journal of Personality and Social Psychology, 96,1077–1087. doi:10.1037/a0015091

Schmeichel, B. J., & Martens, A. (2005). Self-affirmation and mortalitysalience: Affirming values reduces worldview defense and death-thought accessibility. Personality and Social Psychology Bulletin, 31,658–667. doi:10.1177/0146167204271567

Shehryar, O., & Hunt, D. M. (2005). A terror management perspective onthe persuasiveness of fear appeals. Journal of Consumer Psychology, 15,275–287. doi:10.1207/s15327663jcp1504_2

Silvia, P. J. (2001). Nothing or the opposite: Intersecting terror manage-ment and objective self-awareness. European Journal of Personality, 15,73–82. doi:10.1002/per.399

Simon, L., Greenberg, J., Harmon-Jones, E., Solomon, S., Pyszczynski, T.,Arndt, J., & Abend, T. (1997). Terror management and cognitive–experiential self-theory: Evidence that terror management occurs in theexperiential system. Journal of Personality and Social Psychology, 72,1132–1146. doi:10.1037/0022-3514.72.5.1132

Spencer, S. J., Zanna, M. P., & Fong, G. T. (2005). Establishing a causalchain: Why experiments are often more effective than meditationalanalyses in examining psychological processes. Journal of Personalityand Social Psychology, 89, 845–851. doi:10.1037/0022-3514.89.6.845

Srull, T. K., & Wyer, R. S., Jr. (1980). Category accessibility and socialperception: Some implications for the study of person memory andinterpersonal judgments. Journal of Personality and Social Psychology,38, 841–856. doi:10.1037/0022-3514.38.6.841

Tangney, J. P., Baumeister, R. F., & Boone, A. L. (2004). High self-controlpredicts good adjustment, less pathology, better grades, and interper-sonal success. Journal of Personality, 72, 271–322. doi:10.1111/j.0022-3506.2004.00263.x

Taubman–Ben-Ari, O. (2004). Intimacy and risky sexual behavior: Whatdoes it have to do with death? Death Studies, 28, 865–887. doi:10.1080/07481180490490988

Taubman–Ben-Ari, O. (in press). Is the meaning of life also the meaning

720 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 23

of death? A terror management perspective reply. Journal of HappinessStudies.

Taubman–Ben-Ari, O., Findler, L., & Mikulincer, M. (2002). The effectsof mortality salience on relationship strivings and beliefs: The moder-ating role of attachment style. British Journal of Social Psychology, 41,419–441. doi:10.1348/014466602760344296

Taubman–Ben-Ari, O., & Katz–Ben-Ami, L. (2008). Death awareness,maternal separation anxiety, and attachment style among first-timemothers: A terror management perspective. Death Studies, 32, 737–756.doi:10.1080/07481180802289622

Taubman–Ben-Ari, O., & Noy, A. (in press). Self-consciousness and deathcognitions from a terror management perspective. Death Studies.

Trapnell, P. D., & Campbell, J. D. (1999). Private self-consciousness andthe five-factor model of personality: Distinguishing rumination fromreflection. Journal of Personality and Social Psychology, 76, 284–304.doi:10.1037/0022-3514.76.2.284

Tulving, E., Schacter, D. L., & Stark, H. A. (1982). Priming effects inword-fragment completion are independent of recognition memory.Journal of Experimental Psychology: Learning, Memory, and Cogni-tion, 8, 336–342. doi:10.1037/0278-7393.8.4.336

Vail, K. E., Arndt, J., Pyszczynski, T., & Motyl, M. S. (2010). From dustto dust: The role of destruction and reconstruction in terror managementprocesses. Manuscript submitted for publication, University of Missouri.

van den Bos, K. (2001). Uncertainty management: The influence of un-certainty salience on reactions to perceived procedural fairness. Journalof Personality and Social Psychology, 80, 931–941. doi:10.1037/0022-3514.80.6.931

van den Bos, K., Poortvliet, P. M., Maas, M., Miedema, J., & van den Ham,

E. (2005). An enquiry concerning the principles of cultural norms andvalues: The impact of uncertainty and mortality salience on reactions toviolations and bolstering of cultural worldviews. Journal of Experimen-tal Social Psychology, 41, 91–113. doi:10.1016/j.jesp.2004.06.001

Vargas, P. T., Sekaquaptewa, D., & von Hippel, W. (2007). Armed onlywith paper and pencil: “Low-tech” measures of implicit attitudes. In B.Wittenbrink & N. Schwarz (Eds.), Implicit measures of attitudes (pp.103–124). New York, NY: Guilford Press.

Vess, M., Routledge, C., Landau, M. J., & Arndt, J. (2009). The dynamicsof death and meaning: The effects of death-relevant cognitions andpersonal need for structure on perceptions of meaning in life. Journal ofPersonality and Social Psychology, 97, 728–744. doi:10.1037/a0016417

Webber, D. (2009). The motivating influence of death: Are individualdifferences in death-thought accessibility and defensive behavior re-lated? (Unpublished master’s thesis). University of Wyoming, Laramie.

Wegner, D. M. (1994). Ironic processes of mental control. PsychologicalReview, 101, 34–52. doi:10.1037/0033-295X.101.1.34

Wegner, D. M., & Smart, L. (1997). Deep cognitive activation: A newapproach to the unconscious. Journal of Consulting & Clinical Psychol-ogy, 65, 984–995. doi:10.1037/0022-006X.65.6.984

Zhou, X., Lei, Q., Marley, S. C., & Chen, J. (2009). Existential function ofbabies: Babies as a buffer of death-related anxiety. Asian Journal ofSocial Psychology, 12, 40–46. doi:10.1111/j.1467-839X.2008.01268.x

Zhou, X., Liu, J., Chen, C., & Yu, Z. (2008). Do children transcend death?An examination of the terror management function of offspring. Scan-dinavian Journal of Psychology, 49, 413–418. doi:10.1111/j.1467-9450.2008.00665.x

(Appendices follow)

721DEATH-THOUGHT ACCESSIBILITY

Page 24

App

endi

xA

Com

plet

eL

ist

and

Det

aile

dSu

mm

ary

ofth

eR

evie

wed

Mor

talit

ySa

lienc

eD

TA

Stud

ies

Rep

ort

Con

text

DT

Am

easu

reH

ypot

hesi

sSt

atis

tical

resu

lt

Gre

enbe

rget

al.

(199

4,St

udy

4)Pa

rtic

ipan

tsw

ere

give

nM

Svs

.te

levi

sion

salie

nce.

Tho

sein

the

MS

cond

ition

then

read

ash

ort

pass

age

todi

stra

ctth

emfr

omth

ough

tsof

deat

h(v

s.di

dno

tre

ada

shor

tpa

ssag

e).

DT

Aw

asas

sess

edth

erea

fter

.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of20

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

term

sw

ere

coff

in,

grav

e,de

ad,

skul

l,co

rpse

,an

dst

iff.

DT

Ash

ould

behi

gher

amon

gpa

rtic

ipan

tsin

the

MS

cond

ition

,bu

ton

lyaf

ter

adi

stra

ctor

task

,co

mpa

red

toth

ose

inth

eM

Sno

dist

ract

orta

skco

nditi

onor

the

TV

cond

ition

.

Aon

e-w

ayA

NO

VA

reve

aled

asi

gnif

ican

tef

fect

for

cond

ition

onD

TA

,F

(1,

34)

�16

.83,

p�

.001

.K

eypa

irw

ise:

DT

Aw

ashi

gher

inth

eM

Sdi

stra

ctor

task

cond

ition

than

inth

eM

Sno

dist

ract

orta

skco

nditi

on,

t(34

)�

3.40

,p

�.0

1,or

the

TV

cond

ition

,t(

34)

�4.

27,

p�

.01.

Arn

dt,

Gre

enbe

rg,

Pysz

czyn

ski,

&So

lom

on(1

997,

Stud

y1)

Part

icip

ants

wer

egi

ven

MS

vs.

exam

salie

nce.

Of

thos

ege

tting

exam

salie

nce,

half

wer

esu

blim

inal

lypr

imed

with

the

wor

dde

ath

and

the

othe

rha

lfw

ithth

ew

ord

fiel

d.T

hen,

eith

erD

TA

was

mea

sure

dw

ithou

tde

lay

orw

orld

view

defe

nse

was

mea

sure

dw

ithde

lay.

Wor

d-fr

agm

ent

com

plet

ion

task

.Se

ven

wor

dsou

tof

26co

uld

beco

mpl

eted

eith

eras

deat

h-re

late

dw

ords

oras

neut

ral

wor

ds.

The

poss

ible

deat

h-re

late

dw

ords

wer

ebu

ried

,ki

lled

,sk

ull,

grav

e,de

ad,

coff

in,

and

stif

f.

Subl

imin

alde

ath

shou

ldpr

oduc

ean

imm

edia

tein

crea

sein

DT

Are

lativ

eto

subl

imin

alfi

eld,

and

both

MS

and

subl

imin

alde

ath

shou

ldin

crea

sew

orld

view

defe

nse

rela

tive

toco

ntro

laf

ter

ade

lay.

Aon

e-w

aybe

twee

n-su

bjec

tsA

NO

VA

reve

aled

high

erD

TA

inth

esu

blim

inal

deat

h(v

s.fi

eld)

cond

ition

,F

(1,

20)

�6.

78,

p�

.02,

and

high

erw

orld

view

defe

nse

inbo

thth

eM

San

dth

esu

blim

inal

deat

h(v

s.co

ntro

l)co

nditi

ons,

F(2

,33

)�

7.37

,p

�.0

1.

Arn

dt,

Gre

enbe

rg,

Pysz

czyn

ski,

&So

lom

on(1

997,

Stud

y3)

Part

icip

ants

wer

esu

blim

inal

lypr

imed

with

the

wor

dde

ador

pain

orw

ere

supr

alim

inal

lypr

imed

with

the

wor

dde

ad.

The

n,ei

ther

DT

Aw

asm

easu

red

with

out

dela

yor

wor

ldvi

ewde

fens

ew

asm

easu

red

with

dela

y.

Wor

d-fr

agm

ent

com

plet

ion

task

.Se

ven

wor

dsou

tof

26co

uld

beco

mpl

eted

eith

eras

deat

h-re

late

dw

ords

oras

neut

ral

wor

ds.

The

poss

ible

deat

h-re

late

dw

ords

wer

ebu

ried

,ki

lled

,sk

ull,

grav

e,de

ad,

coff

in,

and

stif

f.

Subl

imin

alde

adsh

ould

prod

uce

anim

med

iate

incr

ease

inD

TA

rela

tive

tosu

blim

inal

pain

,an

dbo

thsu

pral

imin

alan

dsu

blim

inal

dead

shou

ldin

crea

sew

orld

view

defe

nse

rela

tive

tosu

blim

inal

pain

afte

ra

dela

y.

Res

ults

indi

cate

sign

ific

antly

high

erD

TA

inth

esu

blim

inal

dead

cond

ition

(M�

2.46

)th

anin

the

subl

imin

alpa

inco

nditi

on(M

�1.

50),

p�

.05,

and

high

erw

orld

view

defe

nse

inth

esu

pral

imin

alan

dsu

blim

inal

dead

than

inth

esu

blim

inal

pain

cond

ition

,F

(2,

36)

�4.

24,

p�

.05.

Arn

dt,

Gre

enbe

rg,

Solo

mon

,et

al.

(199

7,St

udy

1)

Part

icip

ants

wer

egi

ven

MS

vs.

exam

salie

nce.

Tho

sein

the

MS

cond

ition

wer

eal

sopu

tun

der

high

cogn

itive

load

vs.

low

cogn

itive

load

.D

TA

was

mea

sure

dtw

ice

afte

rM

S:be

fore

the

dist

ract

ion

task

and

afte

rth

edi

stra

ctio

nta

sk.

Tw

oco

unte

rbal

ance

dw

ord-

frag

men

tco

mpl

etio

nta

sks.

For

each

task

,fo

urw

ords

out

of20

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.Fo

ron

em

easu

reth

ede

ath-

rela

ted

wor

dsw

ere

buri

ed,

kill

ed,

skul

l,an

dgr

ave.

For

the

othe

rm

easu

reth

ede

ath-

rela

ted

wor

dsw

ere

dead

,co

ffin

,m

urde

r,an

dst

iff.

Am

ong

part

icip

ants

inth

eM

Sco

nditi

onun

der

low

cogn

itive

load

,D

TA

shou

ldbe

high

afte

ra

dela

y,bu

tam

ong

thos

ein

the

MS

cond

ition

unde

rhi

ghco

gniti

velo

ad,

DT

Ash

ould

behi

ghim

med

iate

lyre

lativ

eto

ade

lay.

Con

ditio

n�

Tim

eof

Acc

essi

bilit

yM

easu

rein

tera

ctio

n,F

(4,

54)

�8.

20,

p�

.001

.K

eypa

irw

ise

com

pari

sons

:D

TA

was

high

eraf

ter

the

dela

yan

ddi

stra

ctio

nth

anit

was

imm

edia

tely

afte

rM

Sfo

rpa

rtic

ipan

tsw

ithlo

wco

gniti

velo

ad,

t(54

)�

3.16

,p

�.0

1.In

cont

rast

,D

TA

was

high

erbe

fore

ade

lay

and

dist

ract

ion

than

afte

ra

dela

yan

ddi

stra

ctio

nfo

rth

ose

with

high

cogn

itive

load

unde

rM

S,t(

54)

�3.

59,

p�

.001

.

(App

endi

ces

cont

inue

)

722 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 25

App

endi

xA

(con

tinu

ed)

Rep

ort

Con

text

DT

Am

easu

reH

ypot

hesi

sSt

atis

tical

resu

lt

Arn

dt,

Gre

enbe

rg,

Solo

mon

,et

al.

(199

7,St

udy

3)

Part

icip

ants

wer

egi

ven

MS

vs.

exam

salie

nce.

Tho

sein

the

MS

cond

ition

wer

egi

ven

ach

ance

tode

fend

thei

rw

orld

view

vs.

not

give

na

chan

ceto

defe

ndit.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of25

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

buri

ed,

dead

,gr

ave,

kill

ed,

skul

l,an

dco

ffin

.

DT

Ash

ould

behi

gher

amon

gpa

rtic

ipan

tsin

the

MS

cond

ition

who

wer

eno

tgi

ven

ach

ance

tode

fend

thei

rw

orld

view

than

thos

ew

how

ere

give

na

chan

ceto

defe

ndit.

Pla

nned

cont

rast

s:T

hose

inth

eM

S/no

defe

ndco

nditi

onex

hibi

ted

high

erD

TA

than

thos

ein

the

exam

-sal

ient

and

MS/

defe

ndco

nditi

ons,

t(32

)�

4.84

,p

�.0

1.T

here

was

nodi

ffer

ence

inD

TA

betw

een

the

MS/

defe

ndco

nditi

onan

dex

am-

salie

ntco

nditi

on,

t(22

)�

1.

Har

mon

-Jon

eset

al.

(199

7,St

udy

3)Pa

rtic

ipan

tsre

ceiv

edpo

sitiv

evs

.ne

utra

lpe

rson

ality

feed

back

and

wer

eth

engi

ven

MS

vs.

tele

visi

onsa

lienc

e.T

hey

com

plet

edtw

om

easu

res

ofD

TA

,on

eim

med

iate

lyfo

llow

ing

MS

and

the

othe

raf

ter

ade

lay.

DT

Aw

asm

easu

red

twic

e,be

fore

ade

lay

and

afte

ra

dela

y,w

itha

wor

d-fr

agm

ent

com

plet

ion

task

.Fo

urou

tof

26w

ords

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

onbo

thm

easu

res.

Ord

erof

pres

enta

tion

was

coun

terb

alan

ced.

DT

Ash

ould

behi

gher

for

thos

ein

the

MS

cond

ition

give

nne

utra

lfe

edba

ckre

lativ

eto

posi

tive

feed

back

,pa

rtic

ular

lyaf

ter

ade

lay.

Self

-Est

eem

�M

S�

Del

ayin

tera

ctio

n,F

�7.

55,

p�

.01.

Key

pair

wis

eco

mpa

riso

ns:

On

the

dela

yed

DT

Am

easu

re,

posi

tive

feed

back

redu

ced

DT

Afo

llow

ing

MS

rela

tive

tone

utra

lfe

edba

ck,

t�

5.31

,p

�.0

1.N

opa

irw

ise

com

pari

sons

onth

eim

med

iate

DT

Am

easu

rew

ere

sign

ific

ant

(ts

�1.

00).

Sim

onet

al.

(199

7,St

udy

4)Pa

rtic

ipan

tsw

ere

gree

ted

bya

form

alvs

.in

form

alex

peri

men

ter

and

wer

egi

ven

MS

vs.

tele

visi

onsa

lienc

e.T

hey

com

plet

edtw

om

easu

res

ofD

TA

,on

eim

med

iate

lyfo

llow

ing

MS

and

the

othe

raf

ter

ade

lay.

DT

Aw

asm

easu

red

attw

ose

para

tetim

esw

ithw

ord-

frag

men

tco

mpl

etio

nta

sks.

DT

Am

easu

res

wer

eco

unte

rbal

ance

dan

dw

ere

give

nim

med

iate

lyfo

llow

ing

MS

and

afte

ra

dist

ract

ion

task

(i.e

.,de

lay)

.Fo

rbo

thm

easu

res,

six

out

of20

wor

dsco

uld

beco

mpl

eted

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.

DT

Ash

ould

behi

gher

amon

gth

ose

inth

eM

S/ex

peri

entia

lco

nditi

onre

lativ

eto

the

MS/

ratio

nal

cond

ition

,pa

rtic

ular

lyaf

ter

ade

lay.

Con

ditio

n�

Exp

erim

ente

r�

Del

ayin

tera

ctio

n,F

(1,

61)

�5.

56,

p�

.01.

Key

pair

wis

eco

mpa

riso

ns:

Tho

sein

the

info

rmal

expe

rim

ente

rco

nditi

onun

der

MS

exhi

bite

dhi

gher

DT

Aaf

ter

ade

lay,

t(29

)�

5.86

,p

�.0

1,an

dhi

gher

DT

Ath

andi

dth

ose

inth

eco

ntro

lco

nditi

on,

both

imm

edia

tely

and

afte

ra

dela

y,t(

29)

�7.

27,

p�

.001

,an

dt(

29)

�6.

76,

p�

.001

,re

spec

tivel

y.D

TA

did

not

diff

ersi

gnif

ican

tlyw

ithin

the

form

alex

peri

men

ter

cond

ition

(ts

�1.

00).

Arn

dtet

al.

(199

8,St

udy

2)Pa

rtic

ipan

tsco

mpl

eted

ade

ath

anxi

ety

scal

evs

.a

gene

ral

wor

ries

scal

ean

dth

enan

inte

rnal

vs.

exte

rnal

focu

sof

atte

ntio

ntr

eatm

ent.

Aft

era

fille

rm

easu

rew

asgi

ven,

DT

Aw

asas

sess

ed.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xou

tof

25w

ords

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath

wor

dsw

ere

buri

ed,

dead

,gr

ave,

kill

ed,

skul

l,an

dco

ffin

.

DT

Ash

ould

behi

gher

amon

gth

ose

who

com

plet

edth

ede

ath

anxi

ety

scal

ean

dar

eex

tern

ally

focu

sed

than

amon

gth

ose

who

com

plet

edth

ede

ath

anxi

ety

scal

ean

dar

ein

tern

ally

focu

sed.

Scal

e�

Focu

sin

tera

ctio

n,F

(1,

41)

�6.

12,

p�

.02.

Key

pair

wis

eco

mpa

riso

ns:

Part

icip

ants

who

wer

eM

Sex

tern

ally

focu

sed

show

edhi

gher

DT

Ath

anth

ose

who

wer

eM

Sin

tern

ally

focu

sed,

t(41

)�

3.87

,p

�.0

1;co

ntro

lin

tern

ally

focu

sed,

t(41

)�

4.06

,p

�.0

1;or

cont

rol

exte

rnal

lyfo

cuse

d,t(

41)

�3.

69,

p�

.05.

Mik

ulin

cer

&Fl

oria

n(2

000,

Stud

y2)

Part

icip

ants

com

plet

edat

tach

men

tsc

ales

and

wer

eth

engi

ven

MS

vs.

cont

rol.

For

thos

ein

the

MS

cond

ition

,D

TA

was

mea

sure

dim

med

iate

lyvs

.af

ter

ade

lay.

Heb

rew

wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of20

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath

wor

dsw

ere

the

Heb

rew

wor

dsfo

rm

ourn

ing,

deat

h,ca

dave

r,sk

elet

on,

grav

e,an

dm

urde

r.

DT

Ash

ould

behi

gher

for

anxi

ous-

ambi

vale

ntpa

rtic

ipan

tsim

med

iate

lyfo

llow

ing

MS

than

for

thos

ein

the

cont

rol

cond

ition

.

Atta

chm

ent

Styl

e�

Con

ditio

nin

tera

ctio

n,F

(4,

201)

�2.

73,

p�

.05.

Sim

ple

effe

cts

show

edan

xiou

s-am

biva

lent

part

icip

ants

had

high

erD

TA

inbo

thM

Sco

nditi

ons

than

inth

eco

ntro

lco

nditi

on,

F(2

,20

1)�

3.36

,p

�.0

5.

(App

endi

ces

cont

inue

)

723DEATH-THOUGHT ACCESSIBILITY

Page 26

App

endi

xA

(con

tinu

ed)

Rep

ort

Con

text

DT

Am

easu

reH

ypot

hesi

sSt

atis

tical

resu

lt

Mik

ulin

cer

&Fl

oria

n(2

000,

Stud

y3)

Part

icip

ants

com

plet

edat

tach

men

tsc

ales

and

wer

eas

sign

edto

MS/

defe

ndw

orld

view

,M

S/no

defe

ndw

orld

view

,or

leis

ure

activ

ities

/def

end

wor

ldvi

ew.

DT

Aw

asas

sess

edaf

ter

wor

ldvi

ewde

fens

eor

node

fens

e.

Heb

rew

wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of20

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath

wor

dsw

ere

the

Heb

rew

wor

dsfo

rm

ourn

ing,

deat

h,ca

dave

r,sk

elet

on,

grav

e,an

dm

urde

r.

DT

Ash

ould

behi

gher

amon

gav

oida

ntpa

rtic

ipan

tsin

the

MS/

node

fend

cond

ition

rela

tive

toth

ose

inth

eM

S/de

fend

cond

ition

.Fo

ran

xiou

s-am

biva

lent

part

icip

ants

,w

orld

view

defe

nse

shou

ldno

tre

duce

DT

Aw

hen

unde

rM

S.

Atta

chm

ent

Styl

e�

Con

ditio

nin

tera

ctio

n,F

(4,

201)

�3.

37,

p�

.05.

Sim

ple

mai

nef

fect

ssh

owed

that

DT

Aw

ashi

gher

for

avoi

dant

part

icip

ants

inth

eM

S/no

defe

ndco

nditi

onth

anth

eM

S/de

fend

and

cont

rol

cond

ition

s,F

(2,

201)

�16

.10,

p�

.01.

Anx

ious

-am

biva

lent

part

icip

ants

show

edno

sign

ific

ant

diff

eren

cein

DT

Abe

twee

nth

eM

S/de

fend

and

MS/

node

fend

cond

ition

s(p

�.1

0).

Flor

ian

etal

.(2

001,

Stud

y2)

Part

icip

ants

high

and

low

inpe

rson

alha

rdin

ess

wer

egi

ven

MS

vs.

tele

visi

onsa

lienc

e,an

dD

TA

was

asse

ssed

afte

ra

shor

tde

lay.

Heb

rew

wor

d-fr

agm

ent

com

plet

ion

task

.E

ight

of19

frag

men

tsco

uld

beco

mpl

eted

asa

deat

hw

ord.

The

poss

ible

deat

hw

ords

wer

eth

eH

ebre

ww

ords

for

deat

h,m

ourn

ing,

cada

ver,

grav

e,ki

llin

g,dy

ing,

grie

f,an

dsk

elet

on.

Doe

shi

ghpe

rson

alha

rdin

ess

redu

ceth

eef

fect

ofM

Son

DT

A?

Res

ults

indi

cate

don

lya

mai

nef

fect

ofM

S,F

(1,

116)

�8.

25,

p�

.01.

Pers

onal

hard

ines

sdi

dno

tm

oder

ate

this

effe

ct.

Gre

enbe

rget

al.

(200

1)Pa

rtic

ipan

tsw

ere

give

nM

Svs

.de

ntal

pain

;af

ter

ade

lay,

they

wer

egi

ven

ach

ance

tode

fend

vs.

not

defe

ndth

eir

wor

ldvi

ewaf

ter

thre

at.

Tho

sein

the

MS/

defe

ndco

nditi

onw

ere

put

unde

rhi

ghco

gniti

velo

advs

.lo

wco

gniti

velo

ad.

Fina

lly,

DT

Aw

asas

sess

ed.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of25

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

buri

ed,

dead

,ki

lled

,sk

ull,

grav

e,an

dco

ffin

.

DT

Ash

ould

belo

wer

for

thos

ein

the

MS/

defe

ndco

nditi

onvs

.M

S/no

defe

ndco

nditi

on,

rega

rdle

ssof

whe

ther

thos

ein

the

MS/

defe

ndco

nditi

onar

eun

der

high

cogn

itive

load

vs.

low

cogn

itive

load

.

Aon

e-w

ayA

NO

VA

reve

aled

anef

fect

for

cond

ition

,F

(3,

47)

�5.

28,

p�

.003

.K

eypa

irw

ise

com

pari

sons

:T

hose

inth

eM

S/de

fend

cond

ition

unde

rhi

ghco

gniti

velo

adha

dlo

wer

DT

Ath

anth

ose

inth

eM

S/no

defe

ndco

nditi

on,

t(44

)�

2.61

,p

�.0

1,an

ddi

dno

tdi

ffer

from

thos

ein

the

cont

rol

cond

ition

with

low

cogn

itive

load

,t(

44)

�1.

19,

p�

.24.

Mik

ulin

cer

&Fl

oria

n(2

002,

Stud

y3)

Part

icip

ants

wer

egi

ven

MS

vs.

cont

rol

and

rece

ived

failu

refe

edba

ck.

The

yth

enco

mpl

eted

aca

usal

attr

ibut

ion

scal

evs

.a

neut

ral

scal

e.D

TA

was

asse

ssed

ther

eaft

er.

Heb

rew

wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of20

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath

wor

dsw

ere

the

Heb

rew

wor

dsfo

rm

ourn

ing,

deat

h,ca

dave

r,sk

elet

on,

grav

e,an

dm

urde

r.

DT

Ash

ould

behi

gher

amon

gth

ose

inth

eM

S/ne

utra

lsc

ale

cond

ition

rela

tive

toth

eM

S/ca

usal

attr

ibut

ion

cond

ition

.

Con

ditio

n�

Scal

ein

tera

ctio

n,F

(1,

96)

�5.

14,

p�

.05,

�2

�.1

2.Si

mpl

em

ain

effe

cts

show

edth

atfo

llow

ing

MS,

thos

egi

ven

anop

port

unity

tom

ake

caus

alat

trib

utio

nssh

owed

low

erD

TA

than

thos

eno

tgi

ven

anop

port

unity

,F

(1,

96)

�4.

43,

p�

.05,

�2

�.1

0.

Schm

eich

el&

Mar

tens

(200

5,St

udy

2)Pa

rtic

ipan

tsw

ere

give

na

self

-af

firm

atio

nvs

.no

affi

rmat

ion

and

then

MS.

Aft

era

5-m

inde

lay,

DT

Aw

asas

sess

ed.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of25

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

grav

e,bu

ried

,ki

lled

,sk

ull,

coff

in,

and

dead

.

DT

Ash

ould

behi

gher

amon

gth

ose

who

wer

eno

naff

irm

edre

lativ

eto

affi

rmed

afte

rM

S.

Sign

ific

ant

tte

st:

Aff

irm

edpa

rtic

ipan

tsha

dlo

wer

DT

Ath

andi

dno

naff

irm

edpa

rtic

ipan

ts,

t(52

)�

2.20

,p

�.0

3.

Jona

s&

Fisc

her

(200

6,St

udy

3)Pa

rtic

ipan

tsco

mpl

eted

am

easu

reof

intr

insi

c/ex

trin

sic

relig

ious

orie

ntat

ion

and

wer

egi

ven

MS

vs.

cont

rol.

Aft

era

dist

ract

orta

sk,

DT

Aw

asas

sess

ed.

Ger

man

wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of24

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.

DT

Ash

ould

behi

gher

for

extr

insi

cally

relig

ious

part

icip

ants

afte

rM

Sth

anfo

rin

trin

sica

llyre

ligio

uspa

rtic

ipan

ts.

MS

�In

trin

sic

Rel

igio

usO

rien

tatio

nin

tera

ctio

n,�

��

.45,

t(46

)�

2.10

,p

�.0

5.E

xtri

nsic

ally

relig

ious

part

icip

ants

show

edhi

gher

DT

Ath

andi

din

trin

sica

llyre

ligio

uspa

rtic

ipan

ts,

��

�.3

2,t(

46)

�2.

65,

p�

.05.

(App

endi

ces

cont

inue

)

724 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 27

App

endi

xA

(con

tinu

ed)

Rep

ort

Con

text

DT

Am

easu

reH

ypot

hesi

sSt

atis

tical

resu

lt

Frits

che

etal

.(2

007,

Stud

y2)

Part

icip

ants

wer

egi

ven

MS

vs.

dent

alpa

in.

Aft

era

dela

y,D

TA

and

offs

prin

gac

cess

ibili

tyw

ere

asse

ssed

.

Ger

man

wor

d-fr

agm

ent

com

plet

ion

task

.Si

xou

tof

24w

ords

coul

dbe

com

plet

edin

ade

ath-

rela

ted

man

ner,

and

one

wor

dco

uld

beco

mpl

eted

inan

offs

prin

g-re

late

dm

anne

r.T

hepo

ssib

lede

ath

com

plet

ions

wer

eth

eG

erm

anw

ords

for

grav

e,co

rpse

,gr

aves

tone

,co

ffin

,ob

itua

ry,

and

deat

h/de

ad.

The

poss

ible

offs

prin

gw

ord

was

chil

dren

.

MS

shou

ldin

crea

seD

TA

and

offs

prin

gac

cess

ibili

ty.

MS

led

tohi

gher

DT

A,

t(14

2)�

1.63

,p

�.0

5(o

ne-t

aile

d),

and

high

erof

fspr

ing

acce

ssib

ility

,F

(1,

139)

�4.

87,

p�

.05,

than

did

cont

rol.

Cox

etal

.(2

008,

Stud

y1)

Part

icip

ants

wer

egi

ven

MS

vs.

dent

alpa

in.

Aft

era

dela

y,th

eyth

ough

tab

out

apo

sitiv

eex

peri

ence

with

thei

rm

othe

rvs

.a

nega

tive

expe

rien

cew

ithth

eir

mot

her

vs.

apo

sitiv

eex

peri

ence

with

anac

quai

ntan

cevs

.a

neut

ral

expe

rien

cew

ithan

acqu

aint

ance

.D

TA

was

asse

ssed

ther

eaft

er.

Wor

d-fr

agm

ent

com

plet

ion

task

.Fi

vew

ords

out

of25

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

grav

e,ki

lled

,de

ad,

skul

l,an

dco

ffin

.

DT

Ash

ould

belo

wer

for

thos

ein

the

MS/

posi

tive

mot

her

prim

eco

nditi

onre

lativ

eto

othe

rM

S/pr

ime

cond

ition

s.

Salie

nce

�Pr

ime

inte

ract

ion,

F(3

,13

5)�

2.79

,p

�.0

4.T

here

was

anef

fect

ofpa

rent

prim

esth

atw

assi

gnif

ican

tw

ithin

the

mor

talit

yco

nditi

ons,

F(3

,13

5)�

5.78

,p

�.0

01,

but

not

the

dent

alpa

inco

nditi

ons

(F�

1).

Prim

ing

apo

sitiv

eex

peri

ence

with

one’

sm

othe

rle

dto

low

erD

TA

than

did

ane

gativ

eex

peri

ence

ora

posi

tive

orne

utra

lex

peri

ence

with

anac

quai

ntan

ce(F

s�

2.97

,ps

�.0

3).

Frits

che

etal

.(2

008,

Stud

y4)

Part

icip

ants

wer

egi

ven

MS

vs.

self

-det

erm

ined

deat

hsa

lienc

e(S

DD

S)vs

.de

ntal

pain

.A

fter

ade

lay,

DT

Aw

asas

sess

ed.

AG

erm

anle

xica

lde

cisi

onta

skin

clud

ing

deat

h-re

late

dan

dno

n-de

ath-

rela

ted

wor

dsw

asus

ed.

Fast

erre

actio

ntim

esin

dica

tehi

gher

DT

A.

The

deat

h-re

late

dw

ords

wer

eth

eG

erm

anw

ords

for

coff

in,

deat

h,fu

nera

l,co

rpse

,to

die,

and

grav

e.

DT

Ash

ould

behi

gher

inth

eM

San

dSD

DS

cond

ition

sth

anth

ede

ntal

pain

cond

ition

,and

ther

esh

ould

beno

diff

eren

cebe

twee

nth

eM

San

dSD

DS

cond

ition

s.

Plan

ned

com

pari

sons

reve

aled

sign

ific

antly

fast

erre

actio

ntim

esto

deat

hw

ords

inbo

thth

eM

San

dth

eSD

DS

cond

ition

sre

lativ

eto

the

dent

alpa

inco

nditi

on,

F(1

,83

)�

6.68

,p

�.0

1.N

ost

atis

tical

diff

eren

ceem

erge

dbe

twee

nth

eM

San

dSD

DS

cond

ition

s,F

(1,

83)

�0.

70,

ns.

Rou

tledg

eet

al.

(200

8,St

udy

2)Pa

rtic

ipan

tsfi

rst

com

plet

eda

mea

sure

ofno

stal

gia

and

then

wer

egi

ven

MS

vs.

dent

alpa

in.

Aft

era

dela

y,D

TA

was

asse

ssed

.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of28

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

grav

e,m

urde

r,bu

ried

,st

iff,

skul

l,an

dco

ffin

.

DT

Ash

ould

behi

gher

for

MS

part

icip

ants

not

pron

eto

nost

algi

ath

anfo

rM

Spa

rtic

ipan

tspr

one

tono

stal

gia.

Nos

talg

ia�

Con

ditio

nin

tera

ctio

n,b

�0.

64,

t�

�2.

37,

p�

.02.

With

inth

eM

Sco

nditi

on,

nost

algi

apr

onen

ess

was

asso

ciat

edw

ithlo

wer

DT

A,

b�

�0.

69,

t�

�3.

31,

p�

.002

.T

here

was

nodi

ffer

ence

with

inth

eco

ntro

lco

nditi

on,

b�

�0.

06,

t�

�0.

33,

p�

.74.

Rou

tledg

eet

al.

(200

8,St

udy

3)Pa

rtic

ipan

tsw

rote

abou

ta

nost

algi

cev

ent

vs.a

nor

dina

ryev

ent

and

wer

egi

ven

MS

vs.c

ontr

ol.A

fter

abr

ief

dela

y,th

eyco

mpl

eted

am

easu

reof

DT

A.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of28

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

grav

e,m

urde

r,bu

ried

,st

iff,

skul

l,an

dco

ffin

.

DT

Ash

ould

behi

gher

for

thos

ein

ordi

nary

even

t/MS

cond

ition

sth

anin

all

othe

rco

nditi

ons.

Pla

nned

cont

rast

s:T

hose

inM

S/no

n-no

stal

gia

cond

ition

sha

dhi

gher

DT

Ath

anth

ose

inal

lot

her

cond

ition

spo

oled

,F

(1,

73)

�7.

21,

p�

.01.

The

rew

asno

diff

eren

cebe

twee

nth

ede

ath

rem

inde

r/no

stal

gia

cond

ition

and

both

cont

rol

cond

ition

s,F

(1,

54)

�.1

8,p

�.6

5.Z

hou

etal

.(2

008)

Part

icip

ants

wer

egi

ven

afr

agm

ent

task

cont

aini

ngei

ther

deat

hw

ords

(MS)

orno

deat

hw

ords

(con

trol

)an

dw

ere

then

rate

don

thei

rat

titud

esto

war

dth

eC

hine

seon

e-ch

ildbi

rth

cont

rol

polic

y.

Chi

nese

wor

d-fr

agm

ent

com

plet

ion

task

.In

the

cont

rol

cond

ition

,al

lfr

agm

ents

wer

ene

utra

l.In

the

MS

cond

ition

,10

out

of30

frag

men

tsco

uld

beco

mpl

eted

only

asde

ath-

rela

ted

wor

ds.

Dea

thco

mpl

etio

nsw

ere

the

Chi

nese

wor

dsfo

rfu

nera

l,de

adbo

dy,

tom

b,co

ffin

,bu

ry,

fune

ral

proc

essi

ng,

skel

eton

,sk

ull,

sacr

ific

e,an

dca

ncer

.

Tho

sew

hoco

mpl

eted

the

deat

hfr

agm

ents

shou

ldre

port

low

erat

titud

esto

war

dth

eon

e-ch

ildpo

licy

than

shou

ldco

ntro

ls.

Sign

ific

ant

mai

nef

fect

ofco

nditi

on,

F(1

,54

)�

6.78

,p

�.0

5,su

chth

atth

ose

who

com

plet

edde

ath

frag

men

tsre

port

edlo

wer

attit

udes

tow

ard

the

one-

child

polic

y(M

�2.

7)th

andi

dth

ose

inth

eco

ntro

lco

nditi

on(M

�3.

6).

(App

endi

ces

cont

inue

)

725DEATH-THOUGHT ACCESSIBILITY

Page 28

App

endi

xA

(con

tinu

ed)

Rep

ort

Con

text

DT

Am

easu

reH

ypot

hesi

sSt

atis

tical

resu

lt

Nor

enza

yan

etal

.(2

009,

Stud

y3)

Rel

igio

usan

dno

nrel

igio

uspa

rtic

ipan

tsw

ere

give

nM

Svs

.te

levi

sion

salie

nce,

follo

wed

byth

ePA

NA

S,a

wor

ldvi

ewde

fens

em

easu

re,

ade

mog

raph

ics

ques

tionn

aire

,a

self

-est

eem

scal

e,an

da

mea

sure

ofD

TA

.

Wor

d-fr

agm

ent

com

plet

ion

task

.E

ight

of20

frag

men

tsco

uld

beco

mpl

eted

ina

deat

h-re

late

dm

anne

r.

MS

shou

ldpr

omot

em

ore

wor

ldvi

ewde

fens

eam

ong

nonr

elig

ious

part

icip

ants

,bu

tdo

relig

ious

vs.

nonr

elig

ious

part

icip

ants

diff

erin

leve

lof

DT

Afo

llow

ing

MS?

Alth

ough

MS

incr

ease

dw

orld

view

defe

nse

amon

gno

nrel

igio

uspa

rtic

ipan

ts,

with

rega

rdto

DT

A,

only

am

ain

effe

ctof

MS

was

obse

rved

,F

(1,

88)

�6.

81,

p�

.01.

The

rew

asno

mai

nef

fect

ofre

ligio

sity

onD

TA

,no

rdi

dre

ligio

sity

inte

ract

with

MS

topr

edic

tD

TA

.

Rut

jens

etal

.(2

009,

Stud

y1)

Part

icip

ants

wer

egi

ven

MS

vs.

dent

alpa

insa

lienc

ean

dw

ere

adm

inis

tere

da

mea

sure

ofD

TA

afte

ra

shor

tde

lay.

Fina

lly,

they

read

and

eval

uate

da

shor

tes

say

argu

ing

that

prog

ress

isa

myt

h.

Dut

chw

ord-

frag

men

tco

mpl

etio

nta

sk.

Ele

ven

of16

wor

dsco

uld

beco

mpl

eted

ina

deat

h-re

late

dm

anne

r.

DT

Aan

dw

orld

view

defe

nse

shou

ldbe

high

erin

the

MS

than

inth

ede

ntal

pain

cond

ition

.

MS

incr

ease

dbo

thD

TA

,F

(1,

31)

�4.

71,

p�

.05,

and

wor

ldvi

ewde

fens

e,F

(1,

32)

�9.

07,

p�

.01.

Rut

jens

etal

.(2

009,

Stud

y3)

Part

icip

ants

wer

eex

pose

dto

apr

o-pr

ogre

sses

say

vs.

cont

rol

and

wer

egi

ven

MS

vs.

tele

visi

onsa

lienc

e.A

fter

ade

lay,

DT

Aw

asas

sess

ed.

Dut

chw

ord-

frag

men

tco

mpl

etio

nta

sk.

Ele

ven

of16

wor

dsco

uld

beco

mpl

eted

ina

deat

h-re

late

dm

anne

r.

DT

Ash

ould

behi

gher

for

thos

ein

the

cont

rol

essa

y/M

Sco

nditi

onre

lativ

eal

lot

her

cond

ition

s.

Prog

ress

�Sa

lienc

ein

tera

ctio

n,F

(1,

85)

�4.

10,

p�

.05.

Pla

nned

cont

rast

s:T

hose

inth

eco

ntro

les

say/

MS

cond

ition

had

high

erD

TA

than

thos

ein

the

othe

rth

ree

cond

ition

s,t(

85)

�2.

62,

p�

.01.

Schm

eich

elet

al.

(200

9,St

udy

3)Pa

rtic

ipan

tsco

mpl

eted

mea

sure

sof

impl

icit

and

expl

icit

self

-est

eem

and

wer

egi

ven

MS

vs.

“mov

ing

away

”sa

lienc

e.A

fter

ash

ort

dela

y,D

TA

was

asse

ssed

and

am

easu

reof

self

-es

teem

defe

nse

was

adm

inis

tere

d.

Wor

d-fr

agm

ent

com

plet

ion

task

.Fi

veou

tof

20fr

agm

ents

coul

dbe

com

plet

edin

ade

ath-

rela

ted

man

ner.

MS

shou

ldpr

omot

ese

lf-

este

emde

fens

epr

imar

ilyam

ong

thos

ew

ithlo

wim

plic

it/hi

ghex

plic

itse

lf-

este

em.

Do

leve

lsof

impl

icit

vs.

expl

icit

self

-es

teem

pred

ict

DT

Afo

llow

ing

MS?

Alth

ough

thos

ew

ithlo

wim

plic

it/hi

ghex

plic

itse

lf-

este

emsh

owed

incr

ease

dse

lf-e

stee

mde

fens

efo

llow

ing

MS,

with

rega

rdto

DT

A,

only

am

ain

effe

ctof

MS

emer

ged,

t(64

)�

2.92

,p

�.0

1.T

here

wer

eno

mai

nef

fect

sof

self

-est

eem

,no

rw

ere

inte

ract

ions

obse

rved

.

Zho

uet

al.

(200

9,St

udy

2)Pa

rtic

ipan

tsvi

ewed

imag

esof

baby

anim

als

vs.

adul

tan

imal

san

dw

ere

give

nM

S.A

fter

ade

lay,

DT

Aw

asas

sess

ed.

Chi

nese

wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of28

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath

wor

dsw

ere

the

Chi

nese

wor

dsfo

rco

ffin

,de

adbo

dy,

fune

ral,

tom

b,sk

elet

on,

and

suic

ide.

DT

Ash

ould

behi

gher

for

thos

evi

ewin

gad

ult

anim

als

than

for

thos

evi

ewin

gba

byan

imal

s.

Am

ain

effe

ctfo

rco

nditi

onsh

owed

high

erD

TA

for

thos

ein

the

adul

tan

imal

cond

ition

than

the

baby

anim

alco

nditi

on,

F(1

,36

)�

5.10

,p

�.0

3.

Not

e.T

his

appe

ndix

empl

oys

the

follo

win

gab

brev

iatio

ns:M

S�

mor

talit

ysa

lienc

e;D

TA

�de

ath-

thou

ghta

cces

sibi

lity;

AN

OV

A�

anal

ysis

ofva

rian

ce;P

AN

AS

�Po

sitiv

eA

ffec

tNeg

ativ

eA

ffec

tSc

hedu

le.

Item

sin

this

appe

ndix

are

sort

edby

year

ofpu

blic

atio

nan

dth

enal

phab

etic

ally

with

inite

ms

publ

ishe

din

the

sam

eye

ar.

(App

endi

ces

cont

inue

)

726 HAYES, SCHIMEL, ARNDT, AND FAUCHER

Page 29

App

endi

xB

Com

plet

eL

ist

and

Det

aile

dSu

mm

ary

ofth

eR

evie

wed

Dea

th-A

ssoc

iati

onD

TA

Stud

ies

Rep

ort

Con

text

DT

Am

easu

reH

ypot

hesi

sSt

atis

tical

resu

lt

Gol

denb

erg

etal

.(1

999,

Stud

y2)

Part

icip

ants

eith

erhi

ghor

low

inne

urot

icis

mco

mpl

eted

aph

ysic

al(v

s.ro

man

tic)

aspe

ctof

sex

ques

tionn

aire

,an

dD

TA

was

asse

ssed

ther

eaft

er.

Wor

d-fr

agm

ent

com

plet

ion

task

.Fi

vew

ords

out

of25

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath

wor

dsw

ere

coff

in,

dead

,ki

lled

,sk

ull,

and

corp

se.

DT

Ash

ould

behi

gher

amon

gth

ose

high

inne

urot

icis

mw

hen

prim

edw

ithth

eph

ysic

alre

lativ

eto

the

rom

antic

aspe

cts

ofse

x.

Neu

rotic

ism

�Se

xPr

ime

inte

ract

ion,

F(1

,72

)�

4.17

,p

�.0

45.

Sim

ple

effe

cts

show

edth

atth

ose

high

inne

urot

icis

msh

owed

high

erD

TA

afte

rth

eph

ysic

alse

xpr

ime

than

the

rom

antic

sex

prim

e,F

(1,

72)

�5.

39,

p�

.024

.G

olde

nber

get

al.

(199

9,St

udy

3)

Part

icip

ants

eith

erhi

ghor

low

inne

urot

icis

mco

mpl

eted

aph

ysic

al(v

s.ro

man

tic)

aspe

cts

ofse

xqu

estio

nnai

rean

da

rom

antic

love

(vs.

food

)sa

lienc

em

anip

ulat

ion.

DT

Aw

asas

sess

edth

erea

fter

.

Wor

d-fr

agm

ent

com

plet

ion

task

.Fi

vew

ords

out

of25

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath

wor

dsw

ere

coff

in,

dead

,ki

lled

,sk

ull,

and

corp

se.

DT

Ash

ould

behi

gher

amon

gth

ose

high

inne

urot

icis

mw

hen

prim

edw

ithth

eph

ysic

al(v

s.ro

man

tic)

aspe

cts

ofse

x,bu

tth

ero

man

ticlo

vem

anip

ulat

ion

shou

ldbu

ffer

this

effe

ct.

Neu

rotic

ism

�Se

xPr

ime

�L

ove

Con

ditio

nin

tera

ctio

n,F

(1,

97)

�6.

69,

p�

.01.

Key

pair

wis

eco

mpa

riso

ns:

Hig

her

DT

Aam

ong

thos

ew

ithhi

ghne

urot

icis

min

the

phys

ical

(vs.

the

rom

antic

)se

xpr

ime

cond

ition

,F

(1,

97)

�10

.35,

p�

.01.

With

inth

eph

ysic

alse

xpr

ime

cond

ition

,hi

gher

DT

Aam

ong

thos

ew

ithhi

ghne

urot

icis

min

the

food

(vs.

love

)co

nditi

on,

F(1

,97

)�

9.76

,p

�.0

1.Si

lvia

(200

1,St

udy

1)Pa

rtic

ipan

tsw

ere

assi

gned

toa

mir

ror-

indu

ced

self

-aw

aren

ess

cond

ition

vs.

ase

lf-n

ovel

tyse

lf-a

war

enes

sco

nditi

onvs

.a

cont

rol

cond

ition

.A

fter

som

efi

ller

scal

esw

ere

com

plet

ed,

DT

Aw

asas

sess

ed.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of15

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

coff

in,

grav

e,de

ad,

skul

l,co

rpse

,an

dst

iff.

Self

-foc

used

part

icip

ants

shou

ldha

vehi

gher

DT

Ath

anco

ntro

lpa

rtic

ipan

ts.

Pla

nned

cont

rast

s:T

hose

inth

em

irro

r-in

duce

dse

lf-a

war

enes

sco

nditi

on,

t(27

)�

2.9,

p�

.01,

and

self

-nov

elty

self

-aw

aren

ess

cond

ition

,t(

27)

�2.

0,p

�.0

5,ha

dhi

gher

DT

Ath

anth

ose

inth

eco

ntro

lco

nditi

on.

The

two

self

-aw

aren

ess

cond

ition

sdi

dno

tdi

ffer

from

each

othe

r(t

�1)

.Si

lvia

(200

1,St

udy

2)Pa

rtic

ipan

tsw

ere

assi

gned

toa

mir

ror-

indu

ced

self

-aw

aren

ess

cond

ition

vs.

aco

ntro

lco

nditi

on.

Aft

erso

me

fille

rsc

ales

wer

eco

mpl

eted

,de

ath-

and

life-

thou

ght

acce

ssib

ility

wer

eas

sess

ed.

Lif

e-an

dde

ath-

thou

ght

acce

ssib

ility

wer

em

easu

red

with

aw

ord-

frag

men

tco

mpl

etio

nta

sk.

Six

wor

dsou

tof

25co

uld

beco

mpl

eted

eith

eras

deat

h-re

late

dw

ords

oras

neut

ral

wor

ds,

and

anot

her

six

wor

dsco

uld

beco

mpl

eted

aslif

e-re

late

dw

ords

orne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

coff

in,

grav

e,de

ad,

skul

l,co

rpse

,an

dst

iff.

The

life-

rela

ted

wor

dsw

ere

live

,be

,li

fe,

exis

t,am

,an

dal

ive.

Self

-foc

used

part

icip

ants

shou

ldha

vehi

gher

DT

Aan

dlif

e-th

ough

tac

cess

ibili

tyth

ansh

ould

cont

rol

part

icip

ants

.

Self

-aw

are

part

icip

ants

had

high

erD

TA

than

did

non-

self

-aw

are

part

icip

ants

,F

(1,

38)

�5.

25,

p�

0.02

8.In

addi

tion,

self

-aw

are

part

icip

ants

had

high

erlif

e-th

ough

tac

cess

ibili

tyth

andi

dno

n-se

lf-

awar

epa

rtic

ipan

ts,

F(1

,38

)�

5.32

,p

�.0

27.

DT

Aan

dlif

e-th

ough

tac

cess

ibili

tyw

ere

sign

ific

antly

corr

elat

ed(r

�.3

5,p

�.0

3).

(App

endi

ces

cont

inue

)

727DEATH-THOUGHT ACCESSIBILITY

Page 30

App

endi

xB

(con

tinu

ed)

Rep

ort

Con

text

DT

Am

easu

reH

ypot

hesi

sSt

atis

tical

resu

lt

Gol

denb

erg

etal

.(2

002,

Stud

y1)

Part

icip

ants

wer

ere

min

ded

ofhu

man

sim

ilari

ties

toot

her

anim

als

vs.

hum

andi

stin

ctiv

enes

san

dw

ere

then

prim

edw

ithro

man

ticas

pect

svs

.ph

ysic

alas

pect

sof

sex.

Aft

era

brie

fde

lay,

DT

Aw

asas

sess

ed.

Wor

d-fr

agm

ent

com

plet

ion

task

.Fi

vew

ords

out

of25

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.

DT

Ash

ould

behi

gher

amon

gth

ose

inth

ehu

man

ssi

mila

rto

othe

ran

imal

s/ph

ysic

alse

xpr

ime

cond

ition

rela

tive

toth

ehu

man

ssi

mila

rto

othe

ran

imal

s/ro

man

ticse

xpr

ime

cond

ition

.

Cre

atur

elin

ess

�Se

xPr

ime

inte

ract

ion,

F(1

,11

0)�

5.07

,p

�.0

26.

Sim

ple

mai

nef

fect

ssh

owed

that

thos

ein

the

hum

ans

and

anim

als

sim

ilar/

phys

ical

sex

prim

eco

nditi

onha

dhi

gher

DT

Ath

anth

ose

inhu

man

san

dan

imal

ssi

mila

r/ro

man

ticse

xpr

ime

cond

ition

,F

(1,

110)

�4.

57,

p�

.035

.N

oot

her

com

pari

sons

wer

esi

gnif

ican

t.B

urri

s&

Rem

pel

(200

4,St

udy

6)

Part

icip

ants

read

abou

tsh

ield

bugs

,du

stm

ites,

orca

dave

rbe

etle

s.T

here

afte

r,D

TA

and

pref

eren

ces

for

ster

eoty

pic

vs.

coun

ters

tere

otyp

icta

rget

sw

ere

asse

ssed

.

Wor

d-fr

agm

ent

com

plet

ion

task

.T

hirt

een

wor

dsou

tof

39co

uld

beco

mpl

eted

with

ade

ath-

rela

ted

wor

d.T

hepo

ssib

lede

ath

com

plet

ions

wer

eco

ffin

,st

iff,

bone

s,bu

ry,

died

/dea

d,gr

ave,

rott

ing,

dece

ased

,de

caye

d,sk

ull,

corp

se,

tom

b,an

das

hes.

DT

Ash

ould

behi

gher

afte

rre

adin

gab

out

cada

ver

beet

les

rela

tive

tobo

thdu

stm

ites

and

shie

ldbu

gs.

Res

ults

indi

cate

da

mai

nef

fect

ofco

nditi

on,

F(2

,97

)�

4.37

,p

�.0

5,su

chth

atD

TA

was

high

erin

the

cada

ver

beet

leco

nditi

onth

anin

the

dust

mite

and

shie

ldbu

gco

nditi

ons.

Lan

dau,

Solo

mon

,et

al.

(200

4,St

udy

2)

Part

icip

ants

wer

esu

blim

inal

lypr

imed

with

911

vs.

WT

Cvs

.57

3.D

TA

was

asse

ssed

ther

eaft

er.

Wor

d-fr

agm

ent

com

plet

ion

task

.Si

xw

ords

out

of34

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.T

hepo

ssib

lede

ath-

rela

ted

wor

dsw

ere

grav

e,bu

ried

,m

urde

r,sk

ull,

coff

in,

and

stif

f.

DT

Ash

ould

behi

gher

for

thos

esu

blim

inal

lypr

imed

with

911

and

WT

Cco

mpa

red

to57

3.

Pla

nned

cont

rast

s:T

hose

give

nth

e91

1an

dW

TC

prim

eha

dhi

gher

DT

Ath

anth

ose

inth

e57

3co

nditi

on,

t(4

3)�

2.34

,p

�.0

3.

Mar

tens

etal

.(2

004,

Stud

y1)

Part

icip

ants

view

edpi

ctur

esof

old

adul

tsvs

.yo

ung

adul

ts.

DT

Aw

asas

sess

edth

erea

fter

.

Wor

d-fr

agm

ent

com

plet

ion

task

.E

ight

wor

dsou

tof

26co

uld

beco

mpl

eted

eith

eras

deat

h-re

late

dw

ords

oras

neut

ral

wor

ds.

The

poss

ible

deat

h-re

late

dw

ords

wer

egr

ave,

dead

,bu

ried

,ki

lled

,sk

ull,

coff

in,

corp

se,

and

stif

f.

DT

Ash

ould

behi

gher

for

thos

ein

the

old

adul

tpi

ctur

eco

nditi

onvs

.th

eyo

ung

adul

tpi

ctur

eco

nditi

on.

An

AN

OV

Ash

owed

that

thos

eex

pose

dto

pict

ures

ofol

dad

ults

had

high

erD

TA

rela

tive

toth

ose

expo

sed

topi

ctur

esof

youn

gad

ults

,F

(1,

91)

�3.

81,

p�

.05.

Tau

bman

–Ben

-A

ri(2

004,

Stud

y2)

Part

icip

ants

thou

ght

abou

tri

sky

sexu

albe

havi

orvs

.fo

odpr

efer

ence

s.D

TA

was

asse

ssed

ther

eaft

er.

Heb

rew

wor

d-fr

agm

ent

com

plet

ion

task

.Eig

htw

ords

out

of19

coul

dbe

com

plet

edei

ther

asde

ath-

rela

ted

wor

dsor

asne

utra

lw

ords

.The

poss

ible

deat

hw

ords

wer

eth

eH

ebre

ww

ords

for

mou

rnin

g,dy

ing,

deat

h,ca

dave

r,sk

elet

on,

bere

avem

ent,

grav

e,an

dki

lling

.

DT

Ash

ould

behi

gher

for

thos

ein

the

risk

yse

xual

beha

vior

cond

ition

vs.

the

food

pref

eren

ceco

nditi

on.

Am

ain

effe

ctfo

rco

nditi

onsh

owed

that

thos

ein

the

risk

yse

xual

beha

vior

cond

ition

had

high

erD

TA

(M�

1.06

)th

anth

ose

inth

eco

ntro

lco

nditi

on(M

�0.

43).

(App

endi

ces

cont

inue

)

728 HAYES, SCHIMEL, ARNDT, AND FAUCHER