Difficulty: Intermediate | Prerequisites: Lectures 2-3 on the self, basic knowledge of research methods and experimental design.
Big Picture
These lectures explore how people maintain a positive view of themselves and cope with existential threats. Terror management theory addresses how awareness of death shapes behaviour and ideology. The positive illusions literature (Taylor) challenges the traditional clinical view that mental health requires accurate self-perception. Social comparison and self-enhancement research examines how we evaluate ourselves relative to others. Together, these topics explain the motivated and cognitive machinery behind our self-protective biases.
TL;DR
When mortality is made salient, people cling more tightly to cultural worldviews and in-group belonging. Most people hold mildly inflated positive views of themselves, their control, and their futures, and these illusions are generally functional for mental health, motivation, and even physical health outcomes. The "better than average" effect arises from both motivated and cognitive (non-ego) processes.
Terror management theory (TMT)
The theory that awareness of one's own mortality creates existential anxiety, which people manage by investing in cultural worldviews that provide order, meaning, and a sense of permanence. In simple terms, we cope with knowing we will die by clinging to things that feel bigger and more lasting than ourselves.
Mortality salience
An experimental manipulation that makes thoughts of death more accessible, typically by asking people to describe their emotions about dying. Unconscious priming of death (e.g. being near a funeral home) tends to produce stronger effects than explicit, conscious reflection. Think of it as turning up the volume on death-related thoughts.
Cultural worldview defence
The tendency, under mortality salience, to more strongly endorse the beliefs, values, and symbols of one's own culture and to derogate outgroups whose existence challenges those beliefs.
Positive illusions (Taylor)
Three systematic biases most people hold: an unrealistically positive view of the self, an exaggerated sense of personal control, and unrealistic optimism about the future. Taylor argued these are functional and adaptive for mental health.
Better than average effect
The robust finding that most people rate themselves as above average compared to their peers on desirable traits and abilities. This is also called "illusory superiority" or the "Lake Wobegon effect." In simple terms, statistically, half of people must be below average, but almost nobody thinks they are.
Self-serving attributional bias
The tendency to attribute successes to internal, personal causes ("I studied hard") and failures to external, situational causes ("the test was unfair"). Think of it as taking credit for wins and deflecting blame for losses.
Comparative optimism
Judging negative events as less likely to happen to oneself than to one's peers, and positive events as more likely. In simple terms, "bad things happen to other people."
Social comparison (Festinger)
The process of evaluating one's own abilities and opinions by comparing them to those of other people. Originally conceived as an information-gathering process for accurate self-assessment.
Upward vs. downward comparison
Upward comparison is comparing oneself to someone doing better (can inspire or deflate). Downward comparison is comparing to someone doing worse (tends to enhance self-esteem).
Assimilation vs. contrast
Assimilation: incorporating information about another person into your self-view ("they're great, and I'm like them"). Contrast: distinguishing yourself from the other person ("they're great, but I'm not like them"). Think of it as absorbing versus bouncing off.
Confirmation bias
The tendency to search for, interpret, and recall information in ways that confirm pre-existing beliefs. Not ego-driven, just a cognitive shortcut that happens to serve self-enhancement.
The existential dilemma
Like other animals, humans have an instinct for self-preservation. Unlike other animals, we have the intelligence to know we will die. TMT proposes that this awareness creates existential anxiety, which we manage by investing in cultural worldviews that provide order, meaning, and permanence. Self-esteem matters in this framework because it signals "I am doing a good job living up to my cultural worldview."
We can transcend death through spiritual means (the soul, heaven) or secular means (offspring, achievements, identification with a collective that outlives us).
Mortality salience studies
German pedestrians stopped before vs. after passing a funeral home showed heavier prejudice after the funeral home.
Unconscious priming of mortality produces stronger effects than the original method (explicitly asking people to write about dying), possibly because explicit awareness triggers immediate conscious defences.
9/11 as a natural mortality salience induction
9/11 was psychologically devastating because it made mortality salient while simultaneously attacking the cultural worldview itself (the symbols of American power were destroyed). Greenberg argued this was a "bad day" because the usual anxiety buffer (fall back on cultural worldview) was under attack.
Predicted responses: increased patriotism, search for meaning, desire for justice and vengeance, increased in-group altruism, lashing out at those symbolically connected to the attackers, decreased prejudice toward African Americans ("we are all Americans now"), similar to Sherif's superordinate goal effect.
TMT and political leadership
Researchers presented students with three leader types: charismatic (embraces cultural worldview, gives a sense of transcendence), task-oriented (focuses on getting things done), and relationship-oriented (emphasises collective contribution). Under mortality salience, preference for the charismatic leader increased significantly, task-oriented stayed stable, and relationship-oriented decreased.
Before the 2004 US presidential election at Rutgers (a Democratic-leaning state), baseline voting intention favoured Kerry. When mortality or terrorism was primed, support shifted completely to Bush. Both liberal and conservative students moved toward Bush under mortality salience. TMT researchers argued this was not simply rational ("Bush is better at security") but partly irrational, pointing to a parallel study in an Islamic republic where mortality salience increased support for suicide bombing.
Limiting factors of TMT
Secure attachment: people with secure attachment relationships show different (less defensive) responses to mortality salience. Attachment security acts as a reservoir against existential threat.
Older people are less susceptible, possibly because they have already come to terms with mortality or have accomplished more.
Intrinsic religiousness (valuing religion for its inherent meaning rather than social benefits) reduces susceptibility.
Post-traumatic growth: some people work through trauma and emerge with new understanding, rather than just defending against the threat.
Schimel's reversal study
All other TMT studies prime death and show cultural worldview defence. Schimel flipped it: threatening the cultural worldview (a website mocking Canada) increased death-thought accessibility. Canadians in the threat condition were faster to identify death-related words in a lexical decision task, confirming that cultural worldview threats make death more cognitively accessible.
Coping with 9/11 (Silver, Fredrickson)
Silver's national sample (2,729 people assessed before and after 9/11): active planning and acceptance predicted better coping; behavioural disengagement, denial, and self-blame predicted worse outcomes.
Fredrickson (broaden-and-build model): resilient people experienced more positive emotion post-crisis, which mediated the path from resilience to lower depression. The resilient did better after 9/11 because they maintained the capacity for positive emotion.
Peterson and Seligman (web survey): post-9/11 Americans endorsed gratitude, hope, kindness, leadership, love, spirituality, and teamwork more than before, values that promote belonging and connection.
Traditional clinical psychology held that mental health requires accurate perception of reality. Shelly Taylor challenged this by studying women with breast cancer and HIV-positive men.
Three positive illusions
Unrealistically positive self-views: we use more positive than negative traits to describe ourselves, process and recall successes more efficiently, and show self-serving attributional biases.
Illusion of control: gamblers believe they can influence dice, make bigger bets against poorly dressed opponents, and want four times as much for lottery tickets when they chose the numbers.
Unrealistic optimism: we see negative events as more likely for others than for ourselves.
Evidence for the better-than-average effect
When asked to rate themselves compared to peers, people consistently place themselves well above the 50th percentile. In dating relationships, 99 out of 100 people rated their partner as above the 50th percentile. When observers independently rated people on a task, the targets rated themselves higher than the objective observers did.
Are illusions functional?
Taylor argued yes, using Freud's criteria of "love and work":
Happiness: self-serving attributions are associated with happiness.
Caring for others: positive illusions foster better mood, and people in a good mood help others more and have better social relationships.
Productive work: people with positive views of themselves work harder and longer. Taylor noted that young social psychologists all think they will be the next great researcher, and even though most will not be, the illusion motivates greater effort.
Longitudinal evidence
Taylor studied students entering university, measuring self-esteem, perceived control, and optimism at baseline. Self-esteem and optimism predicted better psychological adjustment. Self-esteem predicted higher GPA in the second year, mediated by academic motivation. The effect on health symptoms was weak.
HIV/AIDS studies
Taylor's work with HIV-positive men (tested every six months for years) showed:
AIDS-specific optimism did not lead to riskier sexual behaviour, addressing the concern that illusions cause dangerous risk-taking.
More negative expectations about disease progression predicted earlier onset of symptoms, even controlling for biomedical indicators (T-cell counts, drug regimens).
Men diagnosed with AIDS who scored high on "realistic acceptance" died on average nine months sooner. Denial appeared to be functional.
Mechanism: stress response
Taylor's group proposed that people with positive illusions do not see as many things as stressful, manage stress better when they do encounter it, and recover faster. In a lab stress test (counting backwards), people high in self-enhancement showed less cardiovascular reactivity, faster recovery, and lower baseline cortisol. If this effect holds over weeks and years, the cumulative benefit for health is substantial.
Cultural variations
Optimism about the future is specifically about negative events. Even European Americans do not show an optimism bias for positive events.
Japanese participants thought both positive and negative events were less likely to happen to them.
Self-enhancement varies by trait domain: Americans self-enhance on individualistic traits; East Asians self-enhance on collectivistic traits (e.g. loyalty). The pattern follows what is valued in each culture.
The more important a trait is to the self within a given culture, the bigger the self-enhancement effect.
Epley and Dunning: "Feeling Holier Than Thou"
The question: is the better-than-average effect because we are overly generous toward ourselves, or unduly harsh about others?
Daffodil days study (Cornell): students predicted whether they would buy daffodils for charity. They overestimated themselves (predicted they would buy, but many did not). They also overestimated peers, though less so. The actual purchase rate was even lower. The evidence pointed to overestimating ourselves rather than underestimating others.
Prisoner's dilemma study: people predicted they would cooperate more than they actually did. Their predictions about others were only insignificantly lower than what others actually did. People over-predict their own prosocial behaviour.
Base rate vs. case information: when given base rate data about what other Cornell students donated, people used it to predict a random peer's behaviour but never used it to adjust predictions about themselves. We think we have so much case-specific information about ourselves that population base rates do not apply.
Kruger: "Lake Wobegon Be Gone" (the below-average effect)
Kruger argued the better-than-average effect arises from egocentrism: we anchor on our own ability and fail to adjust for how good others are. If egocentrism is the mechanism, then for things people find difficult, you should see a below-average effect (people anchor on their own low ability and forget that everyone else is also poor).
He found this for chess, juggling, and computer programming. For easy tasks, people thought they were above average. For hard tasks, they rated themselves below average. In both cases, the correlation between self-rated skill and comparative judgement was high, while the correlation with perceived peers' ability was low. People are mindful of themselves but do not adjust for others.
There is still relative accuracy: someone who rates herself a 4 out of 10 gives a higher comparative estimate than someone who rates herself a 2, even though both incorrectly place themselves below average.
Motivated vs. non-motivated explanations
Motivated: most of us start with a positive self-view (average self-esteem is about 7/10). This positive view guides our search for evidence (confirmation bias, which is not ego-driven on its own), shapes how we evaluate that evidence, and may mean we do not even bother to search at all for many low-stakes judgements.
Non-motivated (cognitive):
Egocentrism: we focus on ourselves and insufficiently adjust for others.
Focalism: "how am I?" makes the self the focus. Reversing the question ("how is your classmate compared to you?") reduces but does not eliminate the bias.
Generalised group: comparing an individual to a vague, abstract group inflates the effect. The more you individuate the comparison target, the less bias you get.
Case vs. base rate: we use rich case information about ourselves but sparse or absent base rate information about others.
Both motivational and cognitive processes contribute. They are not the same thing, but they are difficult to separate empirically.
Social comparison (Festinger)
Festinger proposed that we use social comparison to evaluate our abilities and opinions. Later work added the relevant-attribute hypothesis: we do not compare to just anyone, but to people who are similar and therefore informative.
Superstars: relevance and attainability (Lockwood and Kunda)
When comparing to high-achieving people, two factors matter: is the superstar relevant to me, and is the superstar's achievement attainable?
Reading about a superstar in your own field makes you rate yourself more positively than reading about one in an irrelevant field.
First-year students found a high-achieving fourth-year relevant and attainable, and felt inspired. Fourth-year students found the same person relevant but unattainable, and showed no boost (slightly deflated, but not significantly).
People who believe intelligence is malleable (incremental theorists) are more inspired by superstars than those who believe intelligence is fixed (entity theorists).
Assimilation vs. contrast (Dijksterhuis)
Priming a stereotype leads to assimilation (acting like the stereotype). Priming a specific exemplar leads to contrast (acting opposite to the exemplar, who is used as a comparison standard).
Priming "university professor" improved trivial pursuit performance; priming "supermodel" worsened it.
Priming Einstein (exemplar) worsened performance; priming Claudia Schiffer (exemplar) improved it.
Evidence that Einstein triggers social comparison: after subliminal priming with "me," participants in the Einstein condition were faster to recognise "stupid" words, suggesting the self was activated in a negative comparison.
Personal vs. collective self and comparison
Which self is activated matters for comparison outcomes. When "I" is activated (personal self), contrast with others is likely. When "we" is activated (collective self), assimilation is likely.
Priming "we" and then encountering a successful comparison other makes you feel better (assimilation). Priming "I" and encountering an unsuccessful other also makes you feel better (contrast, downward comparison).
Methods for priming I vs. we: circling pronouns in paragraphs, writing about "I" or "we," or scrambled sentence tasks with individualistic or collectivistic words.
Temporal comparisons
We can compare ourselves to how we used to be rather than to other people. This is especially useful for young people, who are always progressing.
Young people primed with self-enhancement goals preferred temporal comparisons. When primed with evaluation goals, they preferred social comparisons.
We play with temporal distance for self-enhancement: we disparage a distant past self (contrast, "look how far I've come") and assimilate a recent past self ("I was good recently, and I'm still good"). People told the beginning of the term was "way back" rated that old self lower and their current self higher than people told it was "recent."
This works only if you believe change is possible. People who think nothing ever changes do not benefit from temporal distancing.
Review framework
We start by anchoring on the self (more accessible than others), so we tend to think we are good.
Information about others may arrive through deliberate social comparison or be thrust upon us.
Upward comparisons can inspire (if attainable, relevant, malleable beliefs) or deflate. Downward comparisons are self-enhancing.
Temporal comparisons are good for the young; either way, distant past = contrast, recent past = assimilation.
Assimilation is beneficial when the target is positive; contrast is beneficial when the target is negative.
Students often think TMT claims people are always thinking about death. TMT is about what happens when mortality is made salient, not about chronic death anxiety.
"Positive illusions are always good" oversimplifies Taylor. Extreme illusions are likely harmful. Taylor acknowledges this and argues that mild, typical illusions are functional.
The better-than-average effect is not purely about ego. Cognitive (non-motivated) processes like egocentrism, focalism, and case vs. base rate contribute independently of self-esteem motives.
"Downward comparison always makes you feel better" ignores context. Downward comparison helps when it is spontaneous and the comparison target is not too close to you. If the target is someone you care about, it can be distressing.
⚠️ Know the three positive illusions (self-views, control, optimism about the future) and be able to provide evidence for each.
⚠️ Be able to distinguish motivated from cognitive explanations for the better-than-average effect and cite a study for each.
⚠️ Understand TMT's core logic: mortality salience, cultural worldview defence, and the role of self-esteem. Be able to describe the 2004 election study and the Schimel reversal.
⚠️ Know when assimilation vs. contrast occurs: stereotypes produce assimilation; exemplars produce contrast. This is a commonly tested distinction.
⚠️ Understand the role of attachment security, age, and intrinsic religiousness as limiting factors for TMT effects.
True or False: The better-than-average effect disappears for difficult tasks.
True (Kruger showed a below-average effect for hard tasks like chess and juggling).
Fill in the blank: Priming a ______ leads to assimilation, while priming a specific ______ leads to contrast.
Stereotype; exemplar.
True or False: Taylor found that AIDS-specific optimism led HIV-positive men to engage in riskier sexual behaviour.
False. There was no difference in risk behaviour.
Fill in the blank: TMT predicts that under mortality salience, people will prefer a ______ leader over a task-oriented or relationship-oriented leader.
Charismatic.
True or False: The Schimel study showed that threatening the cultural worldview decreased death-thought accessibility.
False. It increased death-thought accessibility.
Q: Explain how TMT accounts for the shift in voting intention toward Bush in the 2004 election study.
A: When mortality or terrorism was primed, students at a Democratic-leaning university shifted from favouring Kerry to favouring Bush. TMT explains this as an irrational defensive response: Bush's rhetoric about protecting America and its values resonated with the need to bolster cultural worldview under mortality threat. The charismatic, culturally affirming leader becomes more appealing when death is salient, regardless of rational policy analysis.
Q: A researcher argues that the better-than-average effect is entirely motivational. What cognitive (non-motivated) evidence would you cite to challenge this claim?
A: Kruger's below-average effect for difficult tasks shows that egocentrism (anchoring on the self without adjusting for others) drives the effect regardless of ego protection. People rate themselves below average on hard tasks, which is the opposite of what a purely motivational account would predict. Additionally, people show the better-than-average effect even when rating someone else compared to a group, not just themselves, suggesting the bias is partly about comparing an individuated target to an abstract group.
Q: Under what conditions do upward social comparisons inspire rather than deflate?
A: Lockwood and Kunda showed that upward comparisons inspire when the superstar is relevant to the comparer and the superstar's achievement seems attainable. First-year students were inspired by a successful fourth-year in their own field. Additionally, people who hold incremental (malleable) theories of intelligence are more inspired by superstars than entity (fixed) theorists.
Q: Explain the difference between assimilation and contrast in priming, and give one study demonstrating each.
A: Assimilation means incorporating the primed concept into your own behaviour (acting consistently with it). Contrast means acting opposite to the primed concept. Dijksterhuis showed that priming the stereotype "university professor" improved trivial pursuit scores (assimilation to the smart stereotype), while priming the exemplar Einstein worsened scores (contrast with a specific person used as a comparison standard).
TMT connects to the attachment literature: securely attached people show reduced TMT effects, linking back to Bowlby and adult attachment styles covered later. The positive illusions work connects to self-presentation and self-affirmation (Steele), covered in the next set of lectures, and to health psychology (SES, stress, allostatic load). Social comparison connects to self-monitoring (Snyder), where low self-monitors use internal attitudes to guide behaviour while high self-monitors attend to situational cues.
Tags: terror management theory, TMT, mortality salience, cultural worldview, Greenberg, positive illusions, Taylor, self-enhancement, better than average effect, self-serving bias, comparative optimism, illusion of control, social comparison, Festinger, Lockwood, Kunda, superstars, assimilation, contrast, Dijksterhuis, temporal comparison, egocentrism, Kruger, below average effect, Epley, Dunning, Fredrickson, broaden and build, 9/11, resilience, confirmation bias, PSY 320, personality and social psychology