ArticlePDF AvailableLiterature Review

The Psychology of Change: Self-Affirmation and Social Psychological Intervention

Authors:

Abstract and Figures

People have a basic need to maintain the integrity of the self, a global sense of personal adequacy. Events that threaten self-integrity arouse stress and self-protective defenses that can hamper performance and growth. However, an intervention known as self-affirmation can curb these negative outcomes. Self-affirmation interventions typically have people write about core personal values. The interventions bring about a more expansive view of the self and its resources, weakening the implications of a threat for personal integrity. Timely affirmations have been shown to improve education, health, and relationship outcomes, with benefits that sometimes persist for months and years. Like other interventions and experiences, self-affirmations can have lasting benefits when they touch off a cycle of adaptive potential, a positive feedback loop between the self-system and the social system that propagates adaptive outcomes over time. The present review highlights both connections with other disciplines and lessons for a social psychological understanding of intervention and change.
Content may be subject to copyright.
PS65CH13-Cohen ARI 31 October 2013 13:5
The Psychology of Change:
Self-Affirmation and Social
Psychological Intervention
Geoffrey L. Cohen1and David K. Sherman2
1Graduate School of Education, Department of Psychology, and (by courtesy) Graduate School
of Business, Stanford University, Stanford, California 94305; email: glc@stanford.edu
2Department of Psychological & Brain Sciences, University of California, Santa Barbara,
California 93106; email: david.sherman@psych.ucsb.edu
Annu. Rev. Psychol. 2014. 65:333–71
The Annual Review of Psychology is online at
http://psych.annualreviews.org
This article’s doi:
10.1146/annurev-psych-010213-115137
Copyright c
2014 by Annual Reviews.
All rights reserved
Keywords
health, intervention, relationships, self-affirmation, stereotype threat
Abstract
People have a basic need to maintain the integrity of the self, a global sense
of personal adequacy. Events that threaten self-integrity arouse stress and
self-protective defenses that can hamper performance and growth. However,
an intervention known as self-affirmation can curb these negative outcomes.
Self-affirmation interventions typically have people write about core per-
sonal values. The interventions bring about a more expansive view of the
self and its resources, weakening the implications of a threat for personal in-
tegrity. Timely affirmations have been shown to improve education, health,
and relationship outcomes, with benefits that sometimes persist for months
and years. Like other interventions and experiences, self-affirmations can
have lasting benefits when they touch off a cycle of adaptive potential, a
positive feedback loop between the self-system and the social system that
propagates adaptive outcomes over time. The present review highlights both
connections with other disciplines and lessons for a social psychological un-
derstanding of intervention and change.
333
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
Click here for quick links to
Annual Reviews content online,
including:
• Other articles in this volume
• Top cited articles
• Top downloaded articles
• Our comprehensive search
Further
ANNUAL
REVIEWS
Watch a video lecture online
PS65CH13-Cohen ARI 31 October 2013 13:5
Self-integrity: the
perception of oneself
as morally and
adaptively adequate
Contents
INTRODUCTION............................................................... 334
The Pervasive Psychology of Self-Defense ....................................... 335
Self-Afrmation Theory ........................................................ 336
What AreSelf-Afrmations?.................................................... 337
Understandingthe Effects of Self-Afrmation.................................... 339
Cycles of Adaptive Potential: How Social Psychological Processes Such
as Self-Affirmation Propagate Through Time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
AFFIRMATIONINTERVENTIONS............................................. 342
Education...................................................................... 342
Health......................................................................... 347
Intergroup Conictand InterpersonalRelationships.............................. 352
IMPLICATIONS,QUALIFICATIONS, ANDQUESTIONS ..................... 354
Lessons........................................................................ 354
Moderators andBoundaryConditions........................................... 358
ConnectionsWithOther Research Areas........................................ 360
CONCLUSION.................................................................. 362
INTRODUCTION
In the 1940s, despite war shortages in finer meats and produce, many American homemakers
refused to purchase inferior but more abundant foods even when pressured with patriotic appeals.
But when Kurt Lewin (1997/1948) brought homemakers together in small groups to talk about
obstacles to serving the recommended foods—thus creating a new group norm around the desired
behavior—their purchase patterns changed. In the U.S. Civil Rights era, prejudice was widespread,
and opposition to equal rights proved tenacious in many quarters. But when Milton Rokeach (1973)
threatened Americans’ conception of themselves as compassionate—with a brief insinuation that
they valued their own freedom more than the freedom of others—their support for civil rights
strengthened in a lasting way.
Today many social problems afflict society—inequalities in education, health, and economic
outcomes; political polarization; and intergroup conflict. But these social problems share a psy-
chological commonality with the historical cases described above. The commonality is the notion
that barriers and catalysts to change can be identified and that social psychological interventions
can bring about long-term improvement.
This review has two purposes. First it looks at threats to, and affirmations of, the self as barriers
and catalysts to change. Threats and affirmations arise from the self’s fundamental motive: to be
morally and adaptively adequate, good and efficacious. How people maintain the integrity of the
self, especially when it comes under threat, forms the focus of self-affirmation theory (Steele 1988;
see also Aronson et al. 1999, Sherman & Cohen 2006). We provide an overview of self-affirmation
theory and review research in three areas where the theory has yielded impactful self-affirmation
interventions: education, health, and interpersonal and intergroup relationships.
A second purpose of this review is to address questions related to the psychology of change
raised by self-affirmation research. Increasingly, social psychological research demonstrates the
potential for brief interventions to have lasting benefits (Cohen & Garcia 2008, Garcia & Cohen,
2012, Walton & Cohen 2011, Wilson 2011, Yeager & Walton 2011). These interventions help
people to adapt to long-term challenges. For example, a series of 10-minute self-affirming
334 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
Cycle of adaptive
potential: a positive
feedback loop between
the self-system and the
social system that
propagates adaptive
outcomes over time
Psychological threat:
the perception of
environmental
challenge to one’s
self-integrity
exercises, which prompt people to write about core personal values, raised minority student
achievement in public schools, with effects that persisted for years (Cohen et al. 2006, 2009;
Sherman et al. 2013). How is this possible? How and when do social psychological interventions
such as self-affirmation spark lasting positive change? An impactful intervention acts like almost
any formative experience. It works not in isolation but rather like a turning point in a story, an
event that sets in motion accumulating consequences (Elder 1998). Timely interventions can
channel people into what we refer to as a cycle of adaptive potential. This is a series of
reciprocally reinforcing interactions between the self-system and a social system, such as a school,
that propagates adaptive outcomes over time (cf. Elder 1974, Wilson 2011). The self acts; the
social system reacts; and the cycle repeats in a feedback loop (Caspi & Moffitt 1995). We discuss
lessons for intervention and for a social psychological understanding of change.
The Pervasive Psychology of Self-Defense
Key to understanding the effects of affirmation is psychological threat,the perception of an en-
vironmental challenge to the adequacy of the self. Whether people see their environment as
threatening or safe marks a dichotomy that runs through research not only on self-affirmation but
also on attachment, stress, and coping (see Worthman et al. 2010). Psychological threat represents
an inner alarm that arouses vigilance and the motive to reaffirm the self (Steele 1988). Although
psychological threat can sometimes trigger positive change (Rokeach 1973, Stone et al. 1994), it
can also impede adaptive coping. People may focus on the short-term goal of self-defense, often
at the cost of long-term learning. Like a distracting alarm, psychological threat can also consume
mental resources that could otherwise be marshaled for better performance and problem solving.
Thus, psychological threat can raise a barrier to adaptive change.
Major life events, such as losing one’s job or receiving a medical diagnosis, can obviously give
rise to psychological threat. But the self-integrity motive is so strong that mundane events can
threaten the self as well and instigate defensive responses to protect it (Sherman & Cohen 2006).
When people make trivial choices, such as between two similarly appealing music albums, they tend
to defensively rationalize their selection (Steele et al. 1993). When partisans encounter evidence
that challenges their political views, they tend to reflexively refute it (Cohen et al. 2007). When
sports fans see their favorite team suffer a defeat, they experience it partly as their own and increase
their consumption of unhealthy comfort foods (Cornil & Chandon 2013; see also Sherman & Kim
2005). When people confront petty insults, they sometimes turn to violence and even homicide to
reassert an image of personal strength and honor in the minds of others (Cohen et al. 1996; see also
Baumeister et al. 1996). Although the objective stakes of many of these situations seem low, the
subjective stakes for the self can be high. That everyday events can bring about feelings of threat
and trigger extreme responses attests to the power and pervasiveness of the self-integrity motive.
Greenwald (1980) likened the self to a totalitarian regime that suppresses and distorts infor-
mation to project an image of itself as good, powerful, and stable. However, unlike a totalitarian
regime, people can be self-critical. They sometimes denigrate themselves more than outside ob-
servers do and believe that others judge them more harshly than they actually do (e.g., Savitsky
et al. 2001). People can feel guilty for events they have little control over (Doosje et al. 2006). Al-
though they can spin idealized fantasies of their abilities, they can also give accurate self-appraisals
at moments of truth (Armor & Sackett 2006). Storyteller rather than totalitarian regime seems
an apt metaphor for the self. The self has a powerful need to see itself as having integrity, but
it must do so within the constraints of reality (Adler 2012, Kunda 1990, Pennebaker & Chung
2011, Wilson 2011). The goal is not to appraise every threat in a self-flattering way but rather to
maintain an overarching narrative of the self’s adequacy. A healthy narrative gives people enough
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 335
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
optimism to “stay in the game” in the face of the daily onslaught of threats, slights, challenges,
aggravations, and setbacks.
Successful social psychological interventions help individuals access this narrative process
through two avenues (see also Wilson 2011). One avenue is to encourage people to appraise
a difficult circumstance in a hopeful and nondefensive way that, in turn, sustains the perceived
adequacy of the self. Helping trauma victims make sense of their experiences promotes health
(Pennebaker & Chung 2011); helping students to interpret mistakes as an opportunity for
growth rather than evidence of incompetence improves their academic performance (Dweck 2008,
Walton & Cohen 2011, Wilson & Linville 1982, Yeager et al. 2014); and helping parents to see
their infants’ cries in a more sympathetic and less defensive light reduces abuse (Bugental et al.
2002). A second avenue for intervention focuses on changing not people’s appraisal of a specific
challenge but their appraisal of themselves. The present review addresses this second avenue and
the theory that it proceeds from, self-affirmation theory.
Self-Affirmation Theory
The postulate that people are motivated to maintain self-integrity rests at the center of self-
affirmation theory (Steele 1988; see also Sherman & Cohen 2006). Self-integrity is a sense of
global efficacy, an image of oneself as able to control important adaptive and moral outcomes in
one’s life. Threats to this image evoke psychological threat (see Steele 1988, Sherman & Cohen
2006). Three points about this motive merit emphasis.
First, the motive is to maintain a global narrative of oneself as a moral and adaptive actor (“I am
a good person”), not a specific self-concept (e.g., “I am a good student”) (cf. Aronson 1969). With
time, people may commit themselves to a particular self-definition (e.g., parent, teacher). However,
the self can draw on a variety of roles and identities to maintain its perceived integrity. Such
flexibility can be adaptive. People can flexibly define success in a way that puts their idiosyncratic
strengths in a positive light, establishing a reliable but realistic basis for self-integrity (Dunning
2005). The flexibility of the self-system can also promote adaptation, especially in dynamic social
systems. Lower animals have relatively simple goals that they try to meet. A mouse unable to
forage for food would be a failure. But humans have a unique ability to adapt to a vast range of
circumstances. For children and adults, the flexibility of the self-system may foster adaptation to
the wide array of challenges they face across cultures and over the lifespan (Worthman et al. 2010).
Second, the motive for self-integrity is not to be superior or excellent, but to be “good enough,”
as the term “adequate” implies—to be competent enough in a constellation of domains to feel that
one is a good person, moral and adaptive. An implication for intervention is that, to affirm the
self, an event need foster only a sense of adequacy in a personally valued domain, not a perception
of overall excellence.
Third, the motive for self-integrity is not to esteem or praise oneself but rather to act in
ways worthy of esteem or praise. Having people praise themselves (e.g., “I am lovable”) tends
to backfire among those who seem to need the praise most, low-self-esteem individuals, in part
because these “affirmations” lack credibility (Wood et al. 2009). People want not simply praise but
to be praiseworthy,not simply admiration but to be admirable,according to the values of their
group or culture (Smith 1759/2011; see also Leary 2005). An implication for intervention is that
rewards and praise are secondary to opportunities for people to manifest their integrity through
meaningful acts, thoughts, and feelings.
Although the flexibility of the self-system can be adaptive, it can also prove costly when people
cannot find constructive avenues to achieve self-integrity. The self may then seek out alternative
domains in which to invest itself. A disadvantaged student may want to succeed in school but,
336 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
Self-affirmation: an
act that manifests one’s
adequacy and thus
affirms one’s sense of
global self-integrity
Values affirmation
intervention: an
activity that provides
the opportunity to
assert the importance
of core values, often
through writing
exercises
distrustful that society will reward his or her efforts, find other niches to exert control and gain
respect; this is one explanation for the draw of gang membership and violent behavior (Matsuda
et al. 2013). However, the flexibility of the self-system can also be harnessed for positive ends.
People can import into a threatened domain the sense of personal integrity that they feel in another.
Thus they can sustain a global sense of adequacy while adaptively confronting a specific threat.
For a wide range of challenges, this is what self-affirmation interventions enable people to do.
What Are Self-Affirmations?
A self-affirmation is an act that demonstrates one’s adequacy (Steele 1988; see also G.L. Cohen &
J. Garcia, manuscript in preparation). Although big accomplishments such as winning a sports con-
test can obviously affirm one’s sense of adequacy, small acts can do so as well. Examples of events
that although small from the perspective of an outsider can be subjectively “big” (Yeager & Walton
2011) include a stressed employee who cares for his children or merely reflects on the personal im-
portance of his family; an ill resident of a nursing home who enacts a small measure of control over
daily visitations (Schulz 1976); and a lonely patient who, receiving a personal note from her doctor,
realizes that others care for her (Carter et al. 2013). Even small inputs into the self-system can have
large effects, because a healthy self-system is motivated to maintain integrity and generate affirm-
ing meanings (Steele 1988; see also Sherman & Cohen 2006). Many events in a given day are seen
as relevant to the self in some way and this enables people to continually refresh their sense of ade-
quacy. But there are times when sources of self-affirmation may be few, or threats to the self may run
especially high. Times of high need can be identified, making possible well-timed self-affirmation
interventions. Stressful transitions and choice points, for example, mark such timely moments.
Self-affirmations given at these times can help people navigate difficulties and set them on a better
path. Their confidence in their ability to overcome future difficulties may grow and thus buttress
coping and resilience for the next adversity, in a self-reinforcing narrative (Cohen et al. 2009).
Self-affirmations bring about a more expansive view of the self and its resources. They can
encompass many everyday activities. Spending time with friends, participating in a volunteer
group, or attending religious services anchor a sense of adequacy in a higher purpose. Activities
that can seem like distractions can also function as self-affirmations. Shopping for status goods
(Sivanathan & Pettit 2010) or updating one’s Facebook page (Toma & Hancock 2013) afford
culturally prescribed ways to enact competence and adequacy. For people who value science,
simply donning a white lab coat can be self-affirming (see Steele 1988).
Although many inductions of self-affirmation exist, the most studied experimental manipulation
has people write about core personal values (McQueen & Klein 2006; cf. Napper et al. 2009).
Personal values are the internalized standards used to evaluate the self (Rokeach 1973). People
first review a list of values and then choose one or a few values most important to them. The list
typically excludes values relevant to a domain of threat in order to broaden people’s focus beyond it.
To buffer people against threatening health information, health and rationality might be excluded
from the list. Among patients with chronic illness, values related to family might be avoided
insofar as they remind patients of the burden they worry they place on relatives (see Ogedegbe
et al. 2012). People then write a brief essay about why the selected value or values are important
to them and a time when they were important. Thus, a key aspect of the affirmation intervention
is that its content is self-generated and tailored to tap into each person’s particular valued identity
(Sherman 2013). Often people write about their relationships with friends and family, but they
also frequently write about religion, humor, and kindness (Reed & Aspinwall 1998).
Table 1 provides excerpts from affirmation essays written by adolescents and adults in re-
search studies. As the examples illustrate, completing a values affirmation is not typically an act
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 337
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
Table 1 Excerpts from affirmation essays
Middle school participants
Dance is important to me, because it is my passion, my life. My second home is the dance studio, my second family is my dance
team. My family and friends are so important to me, even more than dance. My family, I can’t live without them. My friends, I am
my real self around them (and my sister). I can be silly, goofy, and weird and they don’t care, they accept me for who I am. ...And
for being creative, I LOVE being creative in dance. When I’m dancing or making a dance it takes me to another place.
Being creative is important to me all the time, because I can use things different kinds of ways and look at things differently. For
example, if I can combine the color of the clothes I’ll wear or make different kinds of use for the things I have. ...Being with my
family is what makes me happy because only your family understands you better than anyone and you can be yourself no matter
what and they would never criticize you. With my friends it’s not always the same we can fight sometimes or cry but it’s what
friendships are for so you should enjoy the moment and be happy with them.
If I didn’t have my family, I [wouldn’t] be raised right and if I didn’t have my friends I would be a boring person. If I didn’t have my
religion, I wouldn’t know what to do, I would be lost.
Music is important to me because it gives me a way to express myself when I’m mad, happy, or sad. I also think family and friends
are important because everything, like money, fame, happiness mean nothing if you don’t have loved ones to share it with. My
friends and family are important because I love them to death and they make me who I am. I also think religious values are very
important because if you don’t know what you believe in anybody can tell you anything and you’ll believe it.
Politics is another really important thing to me because I love politics and I some day want to become a corporate lawyer and to
later become the first black president.
College participants
How can one get by without friendship or family? I know I couldn’t, I need that support, at times it can feel like the only thing I
have that’s real. At other times I don’t need it, but love and comfort from relationships is something that is always nice. ... Iwas
stuck in Keystone this winter and had no [way] of getting back home, I felt helpless ...I didn’t know what to do, so I called a
friend and they drove 2 hours out of their way to come help me out, without even thinking twice, without that friend I would of
had one bad night. Not the end of the world no, but when in need I fall back on my support, friends and family, without that
support I would never stop falling.
My relationship with my family is very important to me because it is my parents and brother who helped push me to be who I am
today. Without them, I probably wouldn’t have the patience and motivation to have applied for this university and be successful
here. Whenever I have a problem, it is my family I can go to to help me through it. My friends are also very important. If I didn’t
have the strong loving relationship with my friends from home, I wouldn’t be who I am today. My new friends that I have made
[here] are also a big part in my life because they make me smile every day.
My religious values are the foundation of my life; they guided me, helped me, and strengthened me in every aspect of my life. I have
always had a strong faith which has taught me to love others and led me to be a better person. I’ve found that I enjoy life to a
greater extent, worry less, and smile more than my friends who don’t have religious values. I believe this is because my faith has
taught me to be grateful for everything I have, to trust that everything will be fine, and to enjoy every day as if it were the last.
For me the sense of humor of someone is the most important thing. Every time someone makes me laugh it gives me comfort and
happiness. I think having a good sense of humor is the best quality that a person can have. It does not matter if a person is good
looking or not if they can make others laugh. Every time I meet someone I care if they have a good sense of humor or if they are
funny. That is why most of my friends are always laughing, because we all like to make jokes and laugh together. I even think that
laughing, making jokes and having a good sense of humor is what keeps us together as friends. Furthermore, our sense of humor is
what makes us unique as a group of friends.
of self-aggrandizement (consistent with Crocker et al. 2008, Shnabel et al. 2013). Rather, it is a
psychological time-out (Lyubomirsky & Della Porta 2010): a moment to pull back and regain
perspective on what really matters. As one college student wrote, “How can one get by without
friendship or family? I know I couldn’t, I need that support, at times it can feel like the only thing
I have that’s real.” Although the physical act of writing this essay is momentary, it can bring to
mind a lifelong source of strength. As Table 1 also illustrates, people often affirm themselves by
writing about their connections to other people and to purposes and projects outside themselves
338 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
(Shnabel et al. 2013; see also Crocker et al. 2008). Against this broadened conception of the self
in the world, a particular threat that confronts a person feels less dire.
Understanding the Effects of Self-Affirmation
Before focusing on self-affirmation interventions in education, health, and interpersonal and inter-
group relations, we summarize how affirmations affect psychology to create a moment of potential
change. Then we discuss how and when that change persists.
The psychology of self-affirmation. First, affirmations remind people of psychosocial resources
beyond a particular threat and thus broaden their perspective beyond it (Sherman & Hartson
2011). Under normal circumstances, people tend to narrow their attention on an immediate
threat (e.g., the possibility of failure), a response that promotes swift self-protection and, in the
face of acute dangers, survival (e.g., the fight-or-flight response) (see Pratto & John 1991, Tugade
& Fredrickson 2004). But when self-affirmed, people can see the many ordinary stressors of daily
life in the context of the big picture (Schmeichel & Vohs 2009, Wakslak & Trope 2009). A specific
threat and its implications for the self thus command less vigilance. Nonaffirmed participants saw
a psychologically threatening stimulus—a live but securely caged tarantula—as physically closer
to them than it actually was, but self-affirmed participants estimated its distance accurately, as
though the affirmation psychologically distanced the threat from the self (Harber et al. 2011).
Second, because a threat is seen in the context of an expansive view of the self, it has less impact
on psychological well-being (Cohen et al. 2009, Cook et al. 2012, Sherman et al. 2013). Among
self-affirmed minority students in a field experiment, a low classroom grade exerted less influence
on their long-term sense of belonging in school than it did for their nonaffirmed peers (Cook
et al. 2012). Likewise, when college students were self-affirmed, their attention was less absorbed
by ruminative thoughts about past failure (Koole et al. 1999).
Third, affirmations foster an approach orientation to threat rather than avoidance. If a threat
is seen as important and addressable (Vohs et al. 2013), affirmations make it less likely that people
will shrink away from the threat or deny its importance to themselves. Self-affirmed participants in
one study asserted that the threatening domain was more important to them than did nonaffirmed
participants (Cohen et al. 2007, study 1; see also Koole et al. 1999). People can thus better deal with
the threat in a constructive way, rather than spend mental energy on avoidance, suppression, and
rationalization (see Koole et al. 1999, Taylor & Walton 2011). For example, self-affirmed partic-
ipants were less likely to shun threatening health information that could benefit them (e.g., Klein
& Harris 2009, van Koningsbruggen et al. 2009; see also Taylor & Walton 2011). Self-affirmed
participants also showed greater attention to their errors on a cognitive task, as indexed by error-
related negativity, a neural signal of the brain’s error-detection system (Legault et al. 2012). This
pattern suggests greater engagement among affirmed individuals in learning from their mistakes.
Affirmations lift psychological barriers to change through two routes: the buffering or lessening
of psychological threat and the curtailing of defensive adaptations to it.
Buffering against threat. Self-affirmations can reassure people that they have integrity and that
life, on balance, is okay in spite of an adversity before them. Social relationships appear to have this
kind of power. When people were put in a stressful situation, such as receiving mild electric shocks,
those who felt they had social support in their lives, or those who simply had the chance to see a
picture of a loved one, experienced less fear, threat, and pain (e.g., Master et al. 2009; see also
Cacioppo & Patrick 2008). Likewise, when people were put under intense social evaluation—giving
an impromptu speech in front of a judgmental audience—those who had reflected on an important
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 339
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
Defensive
adaptations:
adaptations aimed at
protecting
self-integrity by
reducing psychological
threat or its effects,
including cognitive
adaptations (e.g., self-
serving biases) and
behavioral adaptations
(e.g.,
self-handicapping,
alcohol consumption)
personal value no longer displayed an elevation in the stress hormone cortisol (Creswell et al.
2005). Neither social support nor values affirmation eliminated the stressor. Rather they placed it
in a larger context of “things that truly matter for my adequacy.” Less encumbered by
psychological threat, self-affirmed people can better marshal their cognitive resources to meet the
demands of the task at hand, for example, solving more creative problems under pressure (Creswell
et al. 2013) or exerting self-control in a depleting situation (Schmeichel & Vohs 2009).
Reducing defensiveness. Self-affirmations also reduce defensive responses, adaptations to
protect the self from threat (for a review, see Sherman & Cohen 2006). These include the self’s
strategies of spin control, such as denying responsibility for failure and taking selective credit for
success. Defensive responses also include various other adaptations, such as denigrating others to
affirm the self, and engaging in denial, rumination, and even heavy drinking and other chemically
induced escapes (Steele et al. 1981) that help people to cope with threats to self. Such defensive
adaptations serve as a psychological immune system (Gilbert et al. 1998). Although these defenses
protect self-integrity in the short term, they can undermine growth and prove self-defeating in
the long term. One way that self-affirmations promote change is by curbing defensive reactions.
Studies show that defensive denial, bias, and distortion in one domain are lessened by affirmations
of self-integrity in another. For example, self-affirmed individuals were more open to a scientific
report linking their behavior to cancer risk (Sherman et al. 2000).
In summary, affirmations help people to maintain a narrative of personal adequacy in threaten-
ing circumstances. They thus buffer individuals against threat and reduce defensive responses to it.
The effect is catalytic. Forces that would otherwise be suppressed by psychological threat—such
as cognitive aptitude or persuasive evidence—are unleashed.
Cycles of Adaptive Potential: How Social Psychological Processes Such
as Self-Affirmation Propagate Through Time
The review so far has described the psychology of the affirming moment. But several of the
intervention studies to be discussed find that the effects of self-affirming writing activities can
persist, for instance improving the grades of at-risk minority students years later (Cohen et al.
2009, Sherman et al. 2013). In fact, social psychology has established that brief interventions can
have large and long-term effects when they address key psychological processes, as pioneered in the
studies of Lewin and Rokeach described in the introductory section of this review (for reviews, see
Ross & Nisbett 2011, Wilson 2011). In some cases, the effects of the intervention even grow. Time
does not necessarily weaken the influence of the past, but can, it seems, preserve and strengthen it.
A key feature of these interventions is that their effectiveness depends on the point in the process
in which they are introduced. For example, if teachers are led to expect certain incoming students
in their classroom to bloom intellectually, they elicit stronger performance from those students
(Rosenthal 1994). But the effects of this “high expectations” intervention disappear if it is delivered
only a short time after teachers have met their students and begun to form their own impressions
of them (for a meta-analysis, see Raudenbush 1984).
Like any formative experience, a successful intervention is not an isolated event but rather a
turning point in a process (see Elder 1998). When well-timed and well-situated, it touches off
a series of reciprocally reinforcing interactions between the self-system and the social system
(see Figure 1). A positive feedback loop between these two powerful systems can drive adaptive
outcomes over time. We refer to this as a cycle of adaptive potential,because it increases the actor’s
potential to achieve adaptive outcomes. The cycle can take over and propagate adaptive outcomes
340 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
ab
cd
ef
Self
system
Adaptive
outcomes
Social
system
Figure 1
Cycle of adaptive potential, a positive feedback loop between the self-system and the social system that
promotes adaptive outcomes over time. Examples of paths include Path a: as a result of being self-affirmed,
the person (e.g., athlete, student) achieves more adaptive outcomes (e.g., better performance); Path b: as a
result of performing better, the person feels more self-affirmed; Path c: because the person performs better,
others (e.g., coaches, teachers) expect more of him/her; Path d: expecting more, others in the social system
draw out better performance from the person; Path e: others in the social system affirm the person’s self
through positive feedback, rewards, etc.; Path f: the person alters the social system through paths other than
adaptive outcomes (e.g., by seeking opportunities for practice or by selecting challenging courses).
Recursive process:
a process in which the
output feeds back as an
input
Interactive process:
a process in which the
output serves as an
input to an altogether
different process in a
system
Subjective construal:
the actor’s subjective
perception; even if the
objective environment
remains constant, the
subjective experience
of it can be changed
through intervention
long after the intervention has receded in time and memory (Caspi & Moffitt 1995, Walton &
Cohen 2011, Wilson 2011).
Three principles explain how and when processes propel themselves through time in cycles
of adaptive potential (Cohen et al. 2012, Garcia & Cohen 2012, Yeager & Walton 2011). First,
because of recursion, the output of a process such as self-affirmation can cycle back as its input, thus
perpetuating itself (Cohen et al. 2009, Wilson & Linville 1982). Better performance may affirm
the self, leading to still better performance, further affirming the self, and so on, as improvement
maintains or even builds on itself. Second, because of interaction,the output of a process can
interact with other processes in the environment. For example, if self-affirmed students perform
better, they may find themselves held to higher expectations by their teachers or placed in a
higher track, which could raise their performance further and open new opportunities. An early
advantage can thus channel people into subsequent experiences that perpetuate and broaden the
advantage. For long-term effects to occur, a process need not recur, as in recursion, but can instead
feed into altogether different processes. Indeed, many social environments abound with change
processes (G.L. Cohen & J. Garcia, manuscript in preparation). For instance, schools and cultures
produce massive change, as when they transform a kindergartner into an educated, civilized adult.
An intervention need not create new processes but may simply interface with these existing ones.
Effects then reverberate through the interconnected forces in a social system (Caspi et al. 1987;
Caspi & Moffitt 1995; Elder 1974, 1998; Lewin 1997/1948; Wilson 2011). Because of recursion
and interaction, the benefits of an intervention can maintain themselves through the progressive
accumulation of their own consequences (see Caspi et al. 1987).
Third, because of subjective construal, an intervention can trigger an enduring shift in
perception (Sherman et al. 2013; see also Ross & Nisbett 2011, Wilson 2011). Even if the
objective environment remains constant, the subjective experience of it may change. For example,
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 341
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
Latino American middle school children who affirmed core values saw instances of racial threat
as unrelated to their likelihood of success in school (Sherman et al. 2013). When affirmed, people
tend to narrate adversity as an isolated event rather than an indictment of their adequacy (Cohen
et al. 2009). Approach, rather than avoidance, becomes more likely, as does problem solving over
giving up. As these personal styles take hold, individuals may construe themselves as the kind of
person who can overcome difficulties, an identity that can then guide their behavior (cf. Freedman
& Fraser 1966). Their narrative of personal adequacy may strengthen, which may bolster coping
with the next adversity, further strengthening the narrative, in a repeating cycle. An intervention
may thus have lasting effects by changing the way people filter information about themselves and
their environment.
To illustrate these principles, consider the parable of the professional hockey player. Given
the many influences on a child’s likelihood of becoming a professional hockey player, it seems
surprising that birth date has a sizable impact (Gladwell 2008; for a recent comprehensive test,
see Addona & Yates 2010). Children with a birthday that falls soon after the cut-off date for
entry into this age-based sport have an advantage. As the oldest in their cohort, they tend to be
bigger and more adept than other children, and they may stand out as more talented as a result.
Their environment may be more affirming; they may score more, be given more opportunities
to practice, and be recruited to higher-caliber teams. These experiences affirm the children and
strengthen their self-confidence, love for the sport, and identity as a hockey player, which can fuel
their desire to practice and improve, evoking further affirmation and opportunity, in a repeating
interactive cycle (Figure 1). Analogous to affirmation interventions, a birthdate is not the sole
cause of children’s athletic fate. It is a trigger for a series of iterative interactions between the child
and a powerful system of athletic socialization that allocates more resources to higher performers.
The parable of the professional hockey player illustrates how recursion, interaction, and subjective
construal can turn a variable with no intrinsic causal power into a life-altering influence by putting
a person on a cycle of adaptive potential.
Like a fortuitous event, even a brief intervention can have a lasting impact if it is appropriately
situated and timed. It can then trigger a positive cycle or interrupt a negative one (Wilson &
Linville 1982). Indeed, the three domains where affirmation has had lasting benefit—education,
health, and relationships—are ones where problems emerge from a slow-moving accumulation of
costs. Each domain abounds with recursive, interactive processes that carry forward the influence
of timely experiences, both for ill and for good.
The influence of recursion, interaction, and subjective construal, and the affirmation inter-
ventions that tap into these processes, were tested using randomized experiments, many in field
settings such as schools and health care centers. The interventions did not eliminate the prob-
lems under study and, of course, were not expected to do so. But they did lead to positive and in
some cases lasting changes in academic performance, health, and the quality of interpersonal and
intergroup relations.
AFFIRMATION INTERVENTIONS
Education
Students want to think positively of themselves. But the daily stressors of school—tests, grades,
peer relations—can threaten their sense of personal adequacy. School can be especially threatening
for members of historically marginalized groups such as African Americans and Latino Americans
(Steele 2010). They may worry that they could be seen through the lens of a negative stereotype
rather than accorded respect and judged on their merits. Such vigilance is understandable and
342 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
even adaptive given the current and historical significance of race in America (Steele 2010,
Walton & Cohen 2011). Race, gender, immigration status, and other group memberships can
thus give rise to a repeated threat for entire groups in academic and work settings. As hundreds
of studies have shown,social identity threat—awareness that one could be devalued on the basis
of one’s group—can be stressful and undermine learning and performance (Inzlicht & Schmader
2011, Steele 2010, Steele et al. 2002). Even if an African American and a white student work in
the same classroom and receive similar instruction, their subjective experience may differ. For the
African American student, the prospect of being stereotyped as intellectually limited can render
the classroom more threatening.
Policy changes and education reform that eliminate this “threat in the air” are of paramount
importance (Steele 2010). Even partial closure of the achievement gap would make a large differ-
ence in the lives of many children and their families. Promisingly, lab studies demonstrate that
affirmations can improve the performance of students working under the specter of a negative
stereotype (Martens et al. 2006, Shapiro et al. 2012, Taylor & Walton 2011).
But in contrast to the lab, in real-world academic and work settings, social identity threat is not
acute but chronic (Cohen & Garcia 2008, Garcia & Cohen 2012, Yeager & Walton 2011). It recurs
in a multitude of daily experiences, such as learning new material, taking a test, getting help, and
making friends. Imagine an African American student who enters middle school with trepidation,
uncertain of whether he belongs and will be accepted by peers and teachers. He wants to achieve
academically. But, in the first week of school, he is called on by his teacher for his perspective
as a “black student.” Aware of being stereotyped, the student may feel that his fears have been
confirmed, and he may learn less and perform worse on the next exam. The student’s sense of
threat may then increase, harming performance further, in a recursive process that strengthens
with time. Increasingly subtle events may trigger perceived threat, with more mental energy spent
on vigilance rather than learning. If teachers fail to grasp the invisible forces at work, they may
see the student as limited, give him less support, and hold him to a lower standard. These could
exacerbate threat and undermine performance further.
However, a moment of validation at a threatening transition could improve a trajectory (Cohen
et al. 2009, Yeager & Walton 2011). If an affirmed student performed better early in an academic
transition, this could trigger a cascade of positive effects—greater self-confidence in the student,
higher expectations from the teacher—all of which could further affirm the student, relax vigilance,
and benefit performance, in a cycle of adaptive potential. Or, more modestly, a downward cycle
might be slowed or averted.
Educational interventions. The first set of randomized field experiments tested the effectiveness
of values affirmation in lifting the achievement of African Americans (Cohen et al. 2006, 2009)
and Latino Americans (Sherman et al. 2013; for additional published replications, see Bowen et al.
2012, Harackiewicz et al. 2014, Miyake et al. 2010, Woolf et al. 2009). The research took place
at three middle schools with students in early adolescence, a key transition marked by feelings of
inadequacy and a quest for identity. Too many adolescents take a wrong turn and find themselves
ensnared in negative trajectories with lifelong consequences (Eccles et al. 1991, Moffitt et al.
2011). The schools were racially mixed such that roughly half the students were minority, that
is, African American or Latino American depending on the school, and roughly half were white.
Although the schools were located in middle-class neighborhoods, most minority students at
one site came from socioeconomically disadvantaged families (Sherman et al. 2013, study 1), an
important population given the widening socioeconomic gap in achievement (Reardon 2011).
Critically, in the experiments featured here, the academic environments provided material and
human resources to help students succeed, and indeed some had undertaken initiatives to advance
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 343
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
1.0 Pre-
intervention
Q1 Q2 Q3 Q4
1.5
2.0
2.5
3.0
3.5
4.0
Grade point average
Time point
White (armation)
White (control)
Latino American (armation)
Latino American (control)
Figure 2
Performance across the school year as a function of ethnicity and affirmation condition, with means and
error terms adjusted for baseline covariates and grade level. Middle school students completed a series of
writing exercises related to either their most important value in the affirmation condition or an unimportant
value/neutral topic in the control condition. Abbreviation: Q, quarter. Adapted from Sherman et al. (2013),
study 1.
the learning of underrepresented students. Thus, the social system was “ready” to respond and
reinforce better student performance once it occurred.
Each student was randomly assigned to complete either values affirmation exercises or control
exercises [for methodological details, see Cohen et al. 2006, 2009 (supplementary materials);
Sherman et al. 2013]. These were distributed by teachers as regular in-class assignments two to
five times over the year. Each took roughly 10 minutes to complete (see Table 1 for sample essays).
Although students likely expected teachers to read their essays, teachers knew neither students’
condition assignments nor the content of their essays. Because early outcomes matter more in a
recursive process, the interventions were given early in the year, typically the fourth week of school.
In most cases the exercises occurred right before an in-class exam so that their psychological effects
could be immediately channeled into better performance rather than decay before they could affect
a key outcome. Because novel rather than repetitive experiences have more emotional impact
(Lyubomirsky & Della Porta 2010), the nature of the exercises was varied throughout the year.
Students’ official grades were tracked for the next one to three years. The values affirmation
intervention significantly improved the grade point average (GPA) of the identity-threatened
groups, African American students in one school (Cohen et al. 2006, 2009) and Latino American
students in two others (Sherman et al. 2013), in their core courses (English, math, social studies,
and science). For instance, the affirmation halved the percentage of African American students who
received a D or F in the first term of the course in which the intervention was given (Cohen et al.
2006). Because the intervention benefited ethnic minority students but not white students, it closed
the achievement gap. The closure corresponded to roughly 30% for Latino and African American
students at two school sites (Cohen et al. 2009; Sherman et al. 2013, study 2) and 22% for econom-
ically disadvantaged Latino American students at another (see Figure 2) (Sherman et al. 2013,
study 1).
344 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
The intervention’s effects persisted and improved students’ trajectory through the rest of middle
school. Two years later, affirmed African Americans and Latino Americans continued to earn
higher GPAs than their nonaffirmed peers (Cohen et al. 2009, Sherman et al. 2013). At one site
where high school records were available, the intervention effect persisted into a third year, when
most students progressed into high school (Sherman et al. 2013, study 1). Students appeared to
carry the benefits with them into a new environment.
The affirmation did not boost GPA but rather slowed its decline. The downward trend common
among middle school students (Eccles et al. 1991) proved less steep among affirmed minorities. As
a result, the performance trajectories of affirmed and nonaffirmed students tended to diverge with
time (see Figure 2). This pattern is consistent with the notion that the intervention interrupted a
recursive cycle in which threat and poor performance fed on one another and worsened outcomes
over time. Also suggestive of a recursive process, the most threatened subgroup, those who would
be most undermined by threat and its accumulating consequences, derived the greatest lasting ben-
efit from affirmation. For example, among African American students, those who at baseline had a
history of poor performance and felt most unsure of their belonging at school showed almost a full
grade point benefit in GPA (Cook et al. 2012, study 1). These were the students whose academic po-
tential was most inhibited by psychological threat. Put differently, the intervention benefit was not
diffuse but concentrated among the most threatened, and sometimes hardest to reach, subgroup.
Would values affirmation help close the social-class achievement gap (Reardon 2011)? A ran-
domized field experiment that disentangled the effect of student social class from the effect of
student race found that it could (Harackiewicz et al. 2014). It focused on first-generation college
students, over 90% of whom were white. These students came from families where neither parent
had received a four-year college degree, a proxy for low socioeconomic status. Students with poor
socioeconomic backgrounds can face extra stress in college because of financial worries and a sense
of not fully belonging on campus (Stephens et al. 2012). For these students, two in-class values
affirmations improved their grades in an introductory biology course. The intervention cut the
achievement gap between them and their more financially advantaged peers in the course by
50%. Remarkably, first-generation students also proved more likely to enroll in the second course
in the biology sequence if they had been in the affirmation condition (86% did so) rather than in
the control condition (66% did so). In another study with low-income middle school students,
the standard affirmation did not affect students’ initial grades, but it did prevent a drop in grades
over the course of the year (Bowen et al. 2012).
Values affirmation has also helped another group of students who contend with social identity
threat: female college students enrolled in introductory physics, a gateway science course (Miyake
et al. 2010). Women who had been randomly assigned to complete values affirmation exercises in
their class performed better on their course exams and earned higher grades than did those who
completed control exercises. Once again, the greatest benefit accrued to the most threatened sub-
group: in this context, women who wondered relatively more if gender stereotypes about science
ability might be true1. Together, the results of these field experiments suggest that affirmation re-
moved a barrier that prevented the full expression of students’ potential (Walton & Spencer 2009).
1Another affirmation intervention with an ethnically heterogeneous sample of medical students in England found mixedeffects,
improving the performance of all students on their clinical evaluations and undermining the performance of white students
on a written examination (Woolf et al. 2009). However, at least two issues limit interpretability. First, it was unclear whether
identity threat contributed to the performance of the ethnic minority sample, a heterogeneous group consisting primarily
of Asian Indian, Pakistani, and Bangladeshi participants. Second, procedural changes made the affirmation evaluative in
nature. Students’ instructors evaluated their essays as “suitable” or “not suitable” for submission to each student’s portfolio
of learning. A few essays were singled out to be discussed in students’ tutorials. Beyond unblinding instructors to students’
condition assignment, the evaluative and public nature of the activity may have compromised its ability to be affirming.
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 345
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
Self-affirmation processes bear not only on underperformance but also on bullying and aggres-
sion. Young adolescents completed either affirmation or control tasks in school, with outcomes
measured several weeks or months after the intervention (Thomaes et al. 2009). Affirmation
lessened the extent to which students with grandiose self-views, a risk factor for aggression,
reacted to threat (i.e., a drop in their self-esteem) by hitting, name calling, rumor spreading,
or committing other antisocial acts as assessed by their classmates. In another field experiment,
self-affirmation increased prosocial behavior among students with a history of antisocial behavior,
as assessed by their teachers (Thomaes et al. 2012). In summary, psychological threat contributes
not only to underperformance but also to bullying and aggression, important problems that have
proved difficult to change. And timely self-affirmations can help remedy them.
Understanding the longevity of affirmation effects. How do affirmation intervention effects
persist over long periods of time, as they did in these studies? Recursion, interaction, and subjective
construal can channel people into a cycle of adaptive potential.
Feeling affirmed, a student may perform better on the next classroom test, and performing
better, the student may feel more affirmed, in a recursive process that lifts the student’s trajectory
and eventually becomes a continual source of self-affirmation. In one series of studies, much of
the effect of affirmation on second-year grades was mediated by its effect on first-year grades,
suggesting that early strong performance begot later strong performance (Cohen et al. 2009,
Harackiewicz et al. 2014). Also consistent with a recursive process was the finding that randomized
“affirmation boosters” in the second year did not increase the affirmation’s benefit. The processes
that the intervention set in motion in the first year sufficed to explain its continuing effect in the
second year (Cohen et al. 2009).
A cycle of adaptive potential is, by definition, interactive. Academic institutions teem with
processes that recognize, reinforce, and provide resources for student success. By interacting with
these processes, an intervention can trigger consequences that have causal power unto themselves.
For example, a student who excels early in the year may be seen by teachers as having greater po-
tential. The positive effects of high teacher expectancies on student performance (Rosenthal 1994)
may then take hold and propel the intervention effect through time (see Yeager & Walton 2011).
Indeed, fewer minority students were placed in their school’s remedial track or retained in grade
in the affirmation condition (3%) than in the control condition (9%); affirmed minorities also took
more advanced math courses (Cohen et al. 2009). When self-affirmed minority students performed
better, the system reacted by categorically reshaping their academic experience. Deflection from
failure channels such as remediation and entry into success channels such as advanced placement
courses can shape students’ academic trajectory (Grubb 2009, Steele 2010). The consequences of
the intervention can thus break free of their origin and propagate change through time.
Beyond recursion and interaction, subjective construal can fuel a cycle of adaptive potential.
Feeling affirmed and achieving more, students may become more hopeful. They may narrate ad-
versity in a more optimistic light. Consistent with this, several findings suggest that self-affirmed
minorities were more likely to see adversity as an isolated event rather than a threat to self (see
Cook et al. 2012; Sherman et al. 2013, study 2). Among nonaffirmed minority students, a poor
grade or a stressful day predicted a lower sense of belonging in school and a higher sense of racial
threat (see Cook et al. 2012, Sherman et al. 2013). Each adversity seemed to raise anew the ques-
tion of whether they belonged (Walton & Cohen 2011). Indeed, over two years, minority students
saw their sense of belonging in school fall, as if successive adversities reinforced the narrative that
they did not fit in (Cook et al. 2012). Likewise, among nonaffirmed first-generation college stu-
dents, uncertainty about whether they “had the right background” for their course grew over the
semester (Harackiewicz et al. 2014). But for self-affirmed students, these effects vanished. They did
346 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
not question their belonging. Self-affirmed minority students also displayed lower cognitive acces-
sibility of the racial stereotype, suggesting that they were less likely to filter classroom experience
through the lens of race (Cohen et al. 2006). To the extent that they were less vigilant to racial cues,
they may have had more mental resources for performance and learning (Taylor & Walton 2011).
Although the intervention was brief, its effects were repeatedly relived in subjective experience.
In summary, affirmation interventions can trigger a series of reciprocally reinforcing interac-
tions between the self-system and the social system that foster students’ adaptive potential over
time.
Health
Most people want to think of themselves as healthy, so it can be threatening to confront evidence
that challenges that assessment. Illness, frailty, and the prospect of death represent some of the
most powerful challenges to adaptive adequacy (see Greenberg et al. 1997). Health-related
threats can take many forms: evidence that one’s diet puts one at risk for heart disease, a screening
that reveals a vulnerability to cancer. Even a visit to the doctor’s office can be stressful (Havranek
et al. 2012). The effect can run in the reverse direction too. Prolonged stress can have serious
health costs that bring a person to the doctor’s office (Sapolsky 2004, Sterling 2004, Taylor 2010).
Self-affirmation interventions will not cure serious illness, but they can catalyze the impact of
other positive health-related forces, such as health information, postoperative care regimens, and
a person’s goal to live a healthful life.
Like education outcomes, health outcomes are also driven by recursive, interactive processes.
For example, stress and illness can exacerbate one another in a vicious cycle (Sapolsky 2004).
Failure to achieve a health goal may lead people to give up prematurely, worsening health (see
Logel & Cohen 2012, Vohs et al. 2013). Because these processes depend on continual feedback
loops, timely interventions could break a negative cycle and perhaps spark a positive one.
Research in this area uses the two routes of affirmation influence. One route uses affirmation
to reduce defensive rejection of threatening health information, the other to buffer people against
the effects of stress.
Reducing defensiveness. For many diseases, simple preventive steps, such as smoking cessation
or diet change, would save a profound number of lives if practiced on a large scale.
But imagine a middle-aged man whose doctor tells him that his diet puts him at risk for heart
disease and diabetes. Rather than admit to his dietary misdeeds, he may deploy defensive biases that
shield the self from blame but also block change (Sherman & Cohen 2006). He might denigrate
the evidence of a link between fat intake and heart disease or assure himself that drugs will undo
the damage. Rationalizing the threat away, he persists in his behavior. Persisting in his behavior,
he embraces his rationalization more, increasing resistance to change.
However, timely self-affirmation can open people up to threatening health information. With
their self-integrity less on trial, self-affirmed individuals can better confront threatening informa-
tion and evaluate it in a manner less tethered to their needs for self-integrity (Klein et al. 2011).
This effect has been confirmed in many studies (for reviews, see Harris & Epton 2009, 2010;
Sherman & Hartson 2011).
Affirmation has increased openness to information about the life-threatening habit of smoking.
Tobacco-related illness in the United States causes more deaths than HIV, car accidents, alcohol
and illegal drug use, suicides, and murders combined (CDC 2012). One study presented smokers
with graphic antismoking cigarette advertisements (Harris et al. 2007). Relative to participants
in a control condition, participants who first completed a self-affirmation—listing their desirable
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 347
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
qualities—saw the advertisements as more distressing, expressed greater confidence in their ability
to stop smoking, and had a stronger motivation to quit. This effect on motivation persisted one
week later.
When affirmed, people are also more likely to take the first steps toward positive behavioral
change. Another study focused on smokers low in socioeconomic status, the demographic group
with the highest rate of smoking (Armitage et al. 2008). Among affirmed participants, 59% took
leaflets about how to quit smoking, compared with only 37% of nonaffirmed participants. In yet
another study, people at risk for diabetes defensively rejected information about their health risk.
They declined the opportunity for a diabetes screening test, even though taking the test pre-
sumably would have helped them to acquire medical advice that could improve their health and
perhaps even prolong their lives (van Koningsbruggen & Das 2009). However, providing them
with the opportunity to reflect on important values before reviewing the information decreased
their denigration of the health message and increased their likelihood of agreeing to take a screen-
ing test (see also Howell & Shepperd 2012). Other small behavioral wins from self-affirmation
have been documented, such as increases in the number of college students purchasing condoms
after watching an AIDS education video (Sherman et al. 2000).
Two insights emerge from this research. First, affirmation enables more balanced information
processing. Without the self on trial, people are better able to evaluate evidence on its merits
(Correll et al. 2004, Klein et al. 2011). Thus, affirmation does not produce change by itself but
enables change to occur if evidence warrants it. Second, echoing the findings in education, af-
firmation promotes change among people under consistent psychological threat—people whose
behavior puts them at risk for a health condition and who thus have cause to feel that their self-
integrity is under threat (e.g., Griffin & Harris 2011, Harris & Napper 2005, van Koningsbruggen
& Das 2009; cf. Klein & Harris 2009).
Although three null findings have been published (Dillard et al. 2005, Fry & Prentice-Dunn
2005, Zhao et al. 2012)2, on the whole the research paints a picture of the affirmed mind as an open
mind (Correll et al. 2004). Whether the moment of openness then prompts enduring changes in
behavior, however, hinges on other factors.
Overcoming barriers to long-term behavior change. The effects of affirmation on attitude
and motivation tend to persist, even one month later (Harris et al. 2007, Harris & Napper 2005).
But long-term effects on behavior have been mixed (Harris & Epton 2010). In one positive study,
drinkers at a bar completed a brief affirmation—selecting and copying an affirming statement
that incorporated an implementation intention for how to deal with threat (e.g., by prompting
participants to write “If I feel threatened or anxious, then I will think about the things I value about
myself ”)—while others completed a standard affirmation, reflecting on acts of kindness (Armitage
et al. 2011). All participants then read a message that summarized the medical risks of alcohol
consumption. Relative to a control condition, both affirmations reduced consumption of alcohol,
as assessed by self-reports on a validated measure one month later. Between the two time points,
the percentage of participants drinking within healthful government-recommended levels jumped
2In two of these null reports, the authors raise questions about whether their manipulations successfully instantiated affirmation
because of either a novel affirmation method (Dillard et al. 2005) or failed manipulation checks for the majority of participants
(Fry & Prentice-Dunn 2005). In two of the reports (Dillard et al. 2005, Zhao et al. 2012), the null result may be due to
low statistical power, as the key comparisons, though not statistically significant, were in the predicted direction such that
affirmation led to more openness to threatening health messages among those for whom the message was most relevant.
By chance alone, periodic null effects would occur in a large set of studies. It is also possible that the effects of affirmation
depend on hidden moderators not assessed in these studies. For example, in Griffin & Harris (2011), affirmation effects were
moderated by trait defensiveness.
348 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
by an average of 32 percentage points in the affirmation conditions, more than double the change
observed in the control condition.
However, in some studies, affirmed participants fail to make the behavioral changes that they
predict they will (Harris et al. 2007, Harris & Napper 2005, Reed & Aspinwall 1998; for a review,
see Harris & Epton 2010). Indeed, changes in attitude and motivation seldom suffice for long-term
behavioral change (Lewin 1997/1948, Ross & Nisbett 2011). For many health-risk behaviors such
as smoking, many people want to stop but are unable to do so (CDC 2012).
Behavior change not only has value in its own right but also helps to trigger changes in self-
construal that carry forward intervention effects. Once people act in a healthful way, they may
see themselves as a person who cares about health and then behave in ways congruent with that
identity (Freedman & Fraser 1966). To achieve long-term benefits with an affirmation, it may be
essential to ensure early behavioral wins. These wins also help to convince people that behavior
change is possible, a key condition for affirmation-induced change to occur (see Vohs et al. 2013).
Two strategies to promote long-term behavior change, and spur a cycle of adaptive potential,
have received suggestive support. One encourages people to form realistic, concrete plans for how
to implement their new health-related goals in specific situations (Gollwitzer & Sheeran 2006).
One study had college students review a report about the need to eat fruits and vegetables (Epton
& Harris 2008). The report provided easy-to-implement steps to incorporate these foods into
their diets (e.g., “Add extra vegetables to your pizza”). Although speculative, this may have been
a critical aspect of the intervention. When self-affirmed, participants ate more portions of fruits
and vegetables over the following week, as assessed by a validated daily diary instrument.
A second strategy to promote long-term behavior change reinstates affirmation at the
choice point, when the risk behavior occurs. Because affirmations help people postpone short-
term gratifications for the sake of long-term goals (Schmeichel & Vohs 2009), and reduce
stress that people might otherwise manage by drinking and smoking (Steele et al. 1981),
it may be important to time affirmations to moments of greatest vulnerability (O. Fotuhi,
S. Spencer, G.T. Fong, & M.P. Zanna, manuscript in preparation). For example, incorporat-
ing a specific implementation intention into the affirmation may be effective (“If I feel tempted to
drink alcohol, then I will think about the things I value about myself ”) (cf. Armitage et al. 2011).
Two recent unpublished studies used a recurring physical cue to keep the self-affirming value
salient after participants left the laboratory: a bracelet given to participants, inscribed with the
words, “Remember the values” (O. Fotuhi, S. Spencer, G.T. Fong, & M.P. Zanna, manuscript in
preparation), and a key chain with a secret compartment that contained a slip of paper on which
participants had written a phrase related to their most important value (G.M. Walton, C. Logel,
J.M. Peach, S.J. Spencer, M.P. Zanna, manuscript under review). Another technique has people
write about a value they share with a close other who supports their health goal (O. Fotuhi, S.
Spencer, G.T. Fong, & M.P. Zanna, manuscript in preparation). By mentally pairing a value with
their health goal in this way, smokers may be more likely to call to mind their value when they are
tempted to smoke. Interventions that used these tactics to increase the adhesiveness of a single
affirmation yielded long-term improvements in smoking cessation (O. Fotuhi, S. Spencer, G.T.
Fong, & M.P. Zanna, manuscript in preparation) and GPA (G.M. Walton, C. Logel, J.M. Peach,
S.J. Spencer, M.P. Zanna, manuscript under review).
Buffering against stress. Stress, especially when chronic, can threaten the integrity of the self
(Keough & Markus 1998). Indeed, stress often arises from events that call into question people’s
sense of adaptive adequacy, their perceived ability to meet the demands on them (see Sherman
& Cohen 2006). Given this, values affirmations may help buffer people against stress by allowing
them to anchor their sense of adequacy in another domain where it is less subject to question.
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 349
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
Physiological evidence of the stress-buffering effect of affirmation was found among college
undergraduates facing a real-world stressor—the midterm exam that they had singled out as most
stressful (Sherman et al. 2009; see also Creswell et al. 2005). Students in the treatment condition
completed two values affirmations before this midterm exam, whereas students in the control
condition completed neutral writing exercises. By the morning of their midterm, nonaffirmed
students showed a marked increase in epinephrine, as measured by urinary catecholamines, hor-
mones involved in the fight-or-flight response. Affirmed students did not show an increase in
epinephrine. Once again, the stress-reduction effect of the intervention was concentrated among
the most threatened: students worried about being disliked and seen as unintelligent.
Stress can harm health, especially when it is prolonged (Sapolsky 2004, Sterling 2004, Taylor
2010). Chronic stressors at sensitive periods can keep the stress system on alert for threat even
years later, a risk factor for cardiovascular disease (see Miller et al. 2009). This vigilance can create
a self-fulfilling prophecy wherein the expectation of threat readies people to perceive it, thereby
reinforcing the expectation of threat in a recursive loop that imposes an increasingly heavy toll on
health (Miller et al. 2009). Vigilance may help explain why various afflictions—poor sleep, obesity,
hypertension, heart disease—cluster among distressed populations (Sterling 2004; see also Walton
& Cohen 2011).
Interventions can promote health if they target the psychological “ringleader,” perceived
threat (see Cacioppo & Patrick 2008). The most effective way to do so is to change the social
environment—classrooms, work, home, society—so that it better provides everyday opportuni-
ties to manifest adequacy, such as through meaningful work and civic engagement (Sterling 2004).
But short of this, timely interventions that assuage threat at sensitive periods can help.
One study focused on women with early-stage (stage I and II) breast cancer (Creswell et al.
2007, a reanalysis of Stanton et al. 2002). The original study found that expressive writing, in
which patients reflected on their thoughts and feelings about their experience with breast cancer,
reduced self-reported illness symptoms at a three-month follow-up, relative to writing about facts
about cancer and treatment (for a review of expressive writing interventions, see Pennebaker &
Chung 2011). Content analyses suggested that the active ingredient in women’s essays was the act of
affirming the self. Patients who affirmed important values, such as relationships or religion, or who
reflected on valued personal qualities benefited most. Consistent with the stress-buffering effects
of self-affirmation, self-affirming writing was associated with less subjective distress immediately
after the writing exercises. By helping patients to see themselves as people with adaptive adequacy
in spite of their illness, or even because of their ability to cope with it, the writing exercise seemed
to lessen the stress of chronic illness and thus improved long-term health. People benefited from
the expressive writing not so much because it led them to reappraise their cancer but because it
helped them to reappraise themselves. As a cancer survivor said when reflecting on this research,
“This process of self-affirmation allowed me to find the strength and sense of self to keep getting
back up” (Norris 2008).
The role of self-affirmation as a buffer against stress and illness is highlighted in a study
that used a novel “meaning-making affirmation” (see Keough & Markus 1998). It encouraged
participants to extract a value-relevant meaning from their daily experience. College students
completed self-affirming writing exercises every other day on ten occasions during their winter
break. Participants wrote about their day’s experiences and put them into perspective with respect
to their most important value. They thus assigned a self-affirming meaning to what they might
otherwise construe as a hassle or stressor. For instance, they might see driving a sibling to work as
an expression of the value they place on family. Relative to a writing control group, self-affirmed
students reported less stress and better health, as assessed by a composite measure of illness, phys-
ical symptoms, and health center visits. As evidence that the exercises helped students to integrate
350 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
stressors into a self-affirming narrative, those who benefited most were students who reported
the greatest number of daily hassles that implicated their self-worth (Keough & Markus 1998).
The stress-buffering effects of self-affirmation can improve objective health outcomes by in-
terrupting negative cycles (Logel & Cohen 2012). Stress can undermine self-control, which can
increase the likelihood that one will fail to meet health-related goals, which can increase stress
further, in a repeating cycle. One study recruited women concerned with their weight. Consis-
tent with the effects of stress on self-control, preoccupation with one’s weight predicts distress,
poor eating, lack of exercise, and—ironically—weight gain (Haines et al. 2007, Logel & Cohen
2012). More generally, psychological threat can increase appetite for sugary and high-fat comfort
foods (see Cornil & Chandon 2013, Sterling 2004). Before being weighed and having their height
and waistline measured, participants completed a standard values affirmation or control exercise
(Logel & Cohen 2012). Approximately 2.5 months later, affirmed participants had lost more weight
than nonaffirmed participants. They also had lower body mass index (BMI) and a smaller waist
circumference, a health factor independent of BMI. Affirmation produced these effects, the au-
thors speculated, because it buffered women against weight-related stress and thus helped them to
tap into psychological resources to meet their weight-related goals (see also Schmeichel & Vohs
2009). Consistent with this speculation, affirmation both improved women’s working memory
performance—needed to maintain focus on long-term goals—and strengthened the association
between better memory and weight loss. Once people begin to achieve important health goals,
they may feel more affirmed and receive more affirming feedback from others, each of which could
reinforce the change and carry it forward through time.
Field experiments on patients with chronic health conditions. The February 2012 issue of
Archives of Internal Medicine featured a series of large randomized trials that found benefits of an
affirmation-informed intervention for at-risk individuals on a medical regimen. Medical treatment
poses many stressors, such as pain, aversive medical procedures, and separation from loved ones.
Severe or prolonged stress can interfere with treatment and recovery (Kiecolt-Glaser et al. 1998).
In the studies, all patients received a standard behavioral intervention to promote recovery
from their health problem. Patients in the treatment condition were also encouraged to reflect
on core values and proud moments whenever they encountered situations that made it difficult to
follow their medical regimen, and to remind themselves of small and often value-relevant events
that made them feel good, such as a beautiful sunset. In one study, affirmation increased prescribed
medication use among hypertensive African Americans, from 36% to 42% at a 12-month follow-
up (Ogedegbe et al. 2012). Another study focused on patients who had undergone an intervention
for a heart condition, for whom physical exercise marks a key step in recovery (Peterson et al.
2012). The percentage of patients who achieved the recommended level of physical activity rose
from 37% in the control condition to 55% in the affirmation condition. In a third study on
young asthma patients, affirmation had a null effect on physical activity, although a positive effect
emerged among young patients who had had a serious medical episode that required hospitalization
or intensive care and who thus presumably faced the greatest psychological threat (Mancuso et al.
2012).
In these studies, the intervention also contained a positive mood component (e.g., patients
received happiness-inducing gifts in the mail), so it is unclear whether affirmation or positive
mood drove the effects. However, a randomized double-blind field experiment tested whether a
pure affirmation intervention could reduce patients’ stress in their interactions with their health
care provider and thus promote better patient-physician communication (Havranek et al. 2012).
Poor communication between patients and their physicians contributes to patient nonadherence
to prescribed care, a major health problem (Zolnierek & DiMatteo 2009). Approximately half of
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 351
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
patients fail to adhere to their doctor’s regimens for treatment or for prevention of illness (Sabat´
e
2003). Moreover, because accurate diagnosis hinges on patients’ ability to talk candidly about their
condition, it is essential to find ways to promote better communication between patients and their
physicians.
Communication tends to be less open and more strained when the patient and physician have
different racial backgrounds, for instance, when a minority patient meets with a white doctor. Such
race discordances predict poorer patient-physician communication, a cause of racial disparities
in compliance and health care (see Havranek et al. 2012). Although many factors contribute
to this problem, Havranek and colleagues (2012) speculated that social identity threat played
a role. Minority patients may worry that their doctor could stereotype them as uninformed or
unintelligent, a concern that may be fueled by the cold, awkward, or rushed demeanor of their
doctor. The patient’s unease and the doctor’s demeanor may reciprocally reinforce one another,
increasing tension and distrust. But perhaps a values affirmation at the beginning of this interaction
might put the patient at ease. The patient might then feel more empowered to take control of the
interaction, and the quality of the exchange might improve.
African American patients with a clinical diagnosis of hypertension participated in the study.
Most were socioeconomically disadvantaged. Patients were randomly assigned to complete either
a standard values affirmation task or a control task at their health care clinic immediately prior
to their appointment with a white health care provider. Patients remained unaware of the ratio-
nale for the activity, and health care providers remained unaware of the condition assignments
of their patients. Each meeting between patient and provider was audiotaped. As assessed by an
established coding protocol, the communications between the patient and the provider proved
superior among affirmed patients than among nonaffirmed patients. Affirmed patients gave and
requested more information about their medical condition. Their interactions evinced greater
attentiveness, warmth, and respect, and less depression and distress. Indeed, the aspects of the ex-
change that affirmation improved are predictive of patients’ adherence to their doctors’ prescribed
care (see Havranek et al. 2012).
As in other domains, affirmation acts as a catalyst in the health domain. It allows people to
take advantage of the opportunities for improvement available in a medical regimen or persuasive
information, and to deal more adaptively with stress. Many of the health-based affirmations also
seem to encourage a habit of self-affirming in challenging circumstances, which may help explain
the longevity of their effects. If success occurs at a timely moment, it can set in motion a cycle of
adaptive potential that benefits long-term health.
Intergroup Conflict and Interpersonal Relationships
People want to have rewarding relationships, and most want to resolve disagreements equi-
tably. But these motives are sometimes compromised by the competing motive to maintain
self-integrity. Because our social worth affects our sense of self-worth (Leary 2005), people may
be especially sensitive to perceived threats in their relationships. This vulnerability can cause
a harmful defensive style. Affirmation interventions can ease this defensiveness and put the
relationship on a better course. Indeed, affirmations have been shown to lessen defensive behavior
in interpersonal and intergroup encounters and to promote trust, compromise, and closeness (see
also Crocker et al. 2008).
Conflict. When people are affirmed in valued domains unrelated to a dispute, they are more open
to otherwise identity-threatening political information and less intransigent in negotiations (for
a review, see Cohen 2012). Among people in Israel and Bosnia, historically conflict-ridden areas,
352 Cohen ·Sherman
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
self-affirmed participants were more likely to acknowledge wrongdoings inflicted by their group
on the other side and to support reparations to those harmed ( ˇ
Cehaji´
c-Clancy et al. 2011). After
viewing presidential debates on the eve of the 2008 election, affirmed Democrats and Republicans
were less partisan in their evaluations of Barack Obama’s debate performance, and 10 days after
the election, previously affirmed Republicans thought that President Obama would govern in a
more balanced and objective fashion (Binning et al. 2010). Arab American men were able to induce
prejudiced people to consider their perspective of being unfairly treated in the wake of 9/11 when
they first asked them self-affirming questions, such as, “When were you really creative?” (Stone
et al. 2011). Affirmation—induced in this case by the target of prejudice and not a third-party
experimenter or interventionist—increased openness to a stigmatized other.
In negotiation, affirmation can lessen the escalation of commitment often observed as partisans’
identity becomes bound to the positions they advocate (Sivanathan et al. 2008). When affirmed,
people are less likely to derogate concessions offered by adversaries in a negotiation (Ward et al.
2011). In a negotiation over abortion rights, participants whose commitment to their political
views had been made salient proved responsive to affirmation. When affirmed, they made more
compromises in a negotiation over abortion legislation, and they left the negotiation with stronger
trust of the advocate on the other side of the issue (Cohen et al. 2007). Affirmation seems to provide
reassurance that a social or political identity, one that would otherwise fix people’s evaluations and
negotiation positions, is but one of many valued identities. In conflicts where downward spirals of
accusation and counteraccusation can descend into aggression (Kennedy & Pronin 2008), the
improved trust and open-mindedness brought about by affirmation holds promise for improved
relations.
Relationships. Relationships are so recursive and dynamic, and so intertwined with our sense of
self and well-being, that self-affirmation interventions could make a difference in relational matters.
One of the strongest predictors of marital distress is defensiveness (Gottman 2011). In arguments,
unhappy couples often sink into a mire of reciprocal retaliation. Each person’s defensiveness evokes
a response in kind from the other, a downward spiral of tit-for-tat that is not only mathematically
predictable, consistent with a recursive cycle, but also predictive of early divorce. By contrast,
happy couples seem to escape these recursive cycles by parrying with a surprising moment of
affirmation—a disarming compliment, an assurance of regard, a bit of shared humor. Indeed,
this pattern of disarming affirmation proved among the strongest predictors of marital stability
and happiness (Gottman 2011). Such affirming gestures seem especially beneficial for people who
feel insecure in their relationships (Overall et al. 2013). Observational studies thus suggest that
reducing threat can lessen defensiveness and help couples resolve disagreements constructively
(see also Murray et al. 2006).
But can one experimentally introduce such a turning point in a relationship? Although the
body of research is not yet as voluminous as in education and health, studies suggest that the
answer is yes. In one study, a standard affirmation reduced the defensive distancing strategies that
low-self-esteem people in long-term relationships use to regulate the risk of losing the affection
of their partner ( Jaremka et al. 2011). After a standard relationship threat induction—thinking
about personal traits they wanted to keep secret from their partner—affirmed low-self-esteem par-
ticipants proved less prone to derogate their partner and to contemplate relationship-sabotaging
behaviors such as starting arguments. Relational insecurity can also lead people to adopt a distant
but self-protective social demeanor, which can evoke rejection from others, deepening insecurity,
in a “self-fulfilling prophecy of social rejection” (Stinson et al. 2011, p. 1145). This negative spiral
can be interrupted, and a positive trajectory fostered, through values affirmation. When affirmed,
relationally insecure participants not only felt more confident in the regard of their family and
www.annualreviews.org Self-Affirmation and Social Psychological Intervention 353
Annu. Rev. Psychol. 2014.65:333-371. Downloaded from www.annualreviews.org
by Stanford University - Main Campus - Lane Medical Library on 01/03/14. For personal use only.
PS65CH13-Cohen ARI 31 October 2013 13:5
romantic partners, but also behaved with greater ease and positivity in an interaction with a stranger
several weeks later (Stinson et al. 2011).
Small positive shifts in how people construe their relationships can accumulate into enduring
changes in the narratives people craft about themselves and their relationships. These in turn can
promote warmer relationships, provide further affirmation, and trigger a cycle of adaptive potential
that could have far-reaching effects on trust, well-being, and health (Cacioppo & Patrick 2008).
More modestly, such shifts may enable couples to break free of downward cycles. Such recursive
possibilities received support in a series of studies that tested a novel intervention (Marigold et al.
2007, 2010; see also Finkel et al. 2013). It aimed to encourage low-self-esteem participants to
draw self-affirming meanings that, ordinarily, seem to elude them. College students in romantic
rela