- Is Self-Esteem Important?
- Is Self-Esteem a Unique Construct?
- What are the Developmental Origins of Self-Esteem?
- Applied Research: Changing Self-Views and the Impact on Important Life Outcomes
I am a social-developmental psychologist interested in understanding how personality and social factors influence a person’s developmental course from conception to death. I am interested in understanding how to raise children to grow up to be healthy, productive members of society; including, finding supportive relationships and having a family, supporting themselves and their family, and not bringing harm to others. As such, I am interested in the developmental origins, developmental course, and interrelations among self-esteem, achievement, and antisocial behavior, particularly how perceptions of the self affect a person’s thoughts, feelings, and behavior.
To address my research questions, I use a wide range of research designs, such as cross-sectional, experimental, longitudinal, behavioral genetic, and the analysis of archival datasets (including publishing an APA book on the analysis of secondary data analysis; Trzesniewski, Donnellan, & Lucas, 2010). I also integrate a wide range of statistical procedures into my research program, such as multilevel modeling (including growth modeling), meta-analysis, quantitative genetics, and structural equation modeling.
Below I describe my main lines of research.
Although self-esteem is generally assumed to be a beneficial attribute, an emerging and increasingly influential perspective suggests that self-esteem is, at best, unrelated to any important outcomes, and, at worst, might be associated with negative consequences such as narcissism and aggression (e.g., Baumeister, Campbell, Krueger, & Vohs, 2003; Crocker & Park, 2004; Seligman, 1994; Twenge, 2006). I have used several approaches to test these claims.
Claim #1: Self-esteem has a “dark side”. I tested this claim by examining the relation between self-esteem and antisocial behavior in three, diverse studies. Our findings showed a robust negative relation between self-esteem and antisocial behavior. This relation held for self and informant reports, participants from different nationalities (United States and New Zealand), and age groups (adolescents and college students; Donnellan, Trzesniewski, Robins, Caspi, & Moffitt, 2005, Psychological Science). Moreover, this relation held both cross-sectionally and longitudinally and controlling for potential confounding variables. Thus, self-esteem does not appear to have a dark side.
Claim #2: Self-esteem is unrelated to any important outcomes. Using longitudinal data spanning from early adolescence (age 11) to early adulthood (age 26), I found that adolescents with higher self-esteem had better mental and physical health, better economic prospects, and lower levels of criminal behavior during adulthood, compared to adolescents with low self-esteem (Trzesniewski, Donnellan, Moffitt, Caspi, & Robins, 2006, Developmental Psychology). These long-term consequences of self-esteem could not be explained by adolescent depression, gender, or socioeconomic status. Moreover, the findings held when the outcome variables were assessed using objective measures (e.g., criminal reports, graduation rates) and informant reports; therefore, the findings cannot be explained by shared method variance in self-report data. All told, this research suggests that high self-esteem during adolescence has positive real-world consequences during adulthood.
Claim #3: Too much focus on self-esteem has created a generation of overly confident, miserable narcissists. Recently, researchers have suggested that today’s youth are more narcissistic, have higher self-esteem, and are more miserable than youth of the past (e.g., Twenge, 2006). Furthermore, the argument made is that society’s over-emphasis on self-esteem in recent decades has created a generation who is qualitatively different from previous generations. Specifically, individuals born in the 1970s to 1990s are thought to have higher self-esteem, more inflated opinions of themselves, higher levels of narcissism, and perhaps paradoxically, more misery than previous generations. As a result, they are not prepared for the trials of the “real world” (e.g., Twenge, 2006). These claims have important implications for the self-esteem field. If intervention programs to raise self-esteem created a generation of helpless narcissists, then perhaps future research should not focus on self-esteem. To test the claims about today’s youth, I used the Monitoring the Future data sets, which were specifically designed to study generational changes. We looked for generational changes in narcissism and self-enhancement (Trzesniewski, Donnellan, & Robins, 2008, Psychological Science; Trzesniewski, Donnellan, & Robins, 2008, Journal of Research in Personality; Donnellan & Trzesniewski, 2009, Personality and Social Psychology Compass; Donnellan & Trzesniewski, 2009, Journal of Research in Personality) and generational changes in 31 other constructs (e.g., self-esteem, happiness, loneliness; Trzesniewski & Donnellan, 2010; Perspectives in Psychological Science; Trzesniewski & Donnellan, 2009, Psychological Science, see also Arnett, Trzesniewski, & Donnellan, 2013, Journal of Emerging Adulthood). We failed to find substantial changes in the thoughts, feelings, and behaviors of today’s youth compared to past generations. That is, we found little reason to conclude that the average member of Generation Me is dramatically different from members of previous generations. Today’s youth seem to be no more egotistical than previous generations and they appear to be just as happy and satisfied as previous generations. In fact, today’s youth seem to have psychological profiles that are remarkably similar to youth from the past 30 years. Thus, self-esteem levels today are the same as they were 30 years ago (4.04, SD = .71, in 1977 vs. 4.00, SD = .81 in 2006; r(self-esteem and year) = -.01; N = 174,481) and there appears to be no reason to condemn interventions aimed at promoting self-esteem. Indeed, well-designed self-esteem interventions have been shown to have positive effects on child outcomes (e.g., Haney & Durlak, 1998; O’Mara, Marsh, Craven, & Debus, 2006).
This debate has proven quite controversial and comments by Dr. Twenge and several other prominent psychologists were solicited for our most comprehensive paper (Trzesniewski & Donnellan, 2010; Perspectives in Psychological Science). Our hope is to broaden and move forward this debate and help clarify issues about personality development and the robustness of cohort effects and help clarify the portrait of today’s young people.
Another claim made about self-esteem is that it is nothing more than depression, neuroticism, and/narcissism. This is a very important issue because if self-esteem is simply depression or narcissism in disguise, then future research should focus on the core trait and not on self-esteem (e.g., Baumeister et al., 2003; Watson et al. 2002). However, my previous research suggests self-esteem is a unique construct. For example, although self-esteem and narcissism are positively correlated, they show unique, sometimes opposite, patterns of relations with other constructs (Donnellan et al., 2005; Paulhus, Robins, Trzesniewski, & Tracy, 2005). Moreover, in current research, I am using behavioral genetic techniques to provide new insights into the relation between self-esteem and related constructs. First, I have found that depression, neuroticism, and negative affect combined accounted for only 36% of the variance in self-esteem. Second, I found unique genetic influences on self-esteem after accounting for depression, neuroticism, and negative affect, suggesting self-esteem is also distinguishable from these constructs at the genetic level. Third, I found that although self-esteem predicts changes in depression, depression does not predict changes in self-esteem (Orth, Robins, Trzesniewski, Maes, & Schmitt, 2009, Journal of Abnormal Psychology). Instead, stability of self-esteem is predicted by a combination of unique genetic and environmental factors. One environmental factor that predicts initial level of self-esteem in children and adults is the childhood relationship with parents, whereas an environmental factor that predicts change in self-esteem during adulthood is the relationship with the spouse or partner. Neither the parent nor spousal relationship had an environmental impact on the individual’s level or change in depression. Thus, self-esteem is influenced by unique genetic and environmental factors, providing further support for the distinction between self-esteem and depression.
Most of my previous work on the development of self-esteem has focused on mean-level and rank-order stability and change during childhood, adolescence, adulthood, and old age (Donnellan, Kenny, Trzesniewski, Lucas, & Conger, 2012, Journal of Research in Personality; Robins, Trzesniewski, Tracy, Potter, & Gosling, 2002, Psychology & Aging; Trzesniewski, Donnellan, & Robins, 2003, Journal of Personality and Social Psychology). I am following up on this work by testing hypotheses about predictors of change. For example, my colleagues and I have found that changes in health and SES account for the decline in self-esteem during old age (Orth, Trzesniewski, & Robins, 2010, Journal of Personality and Social Psychology). Currently, I am testing the role of parent-child closeness on change and stability of self-esteem during adolescence. Findings based on (a) parents and adolescents from two countries (US and Germany), (b) different measures of parenting (self-report, mother-report, father-report, and observations), and (c) a variety of longitudinal models (regressions, cross-lag models, parallel growth curves, enduring effects models, and latent difference models) show that parenting does not predict change in self-esteem (or vice-versa) even though the two are correlated within each time point. This raises questions about the developmental origins of self-esteem, given prominent theories of self-esteem all implicate the role of the relationship with important others. Therefore, I have turned to an earlier phase of the life span to find answers about how self-esteem is initially formed and the early predictors of self-esteem.
Research on self-esteem in young children is difficult because it has long been believed that children younger than eight do not have global self-esteem (Harter, 1983; Harter & Pike, 1984), and the most popular measure of self-esteem for young children (the Self-Perception Profile for Young Children; SPPC; Harter, 1985) does not contain a subscale of global self-esteem. As a result, almost no research has been done examining self-esteem in children younger than eight. I have several lines of research aimed at addressing this gap in the literature. First, I am testing global self-esteem items with the goal of creating not only a scale that is reliable and valid for children as young as five, but that also has invariant items across the life span (currently there are no invariant life span self-esteem scales available). Preliminary findings show that the new invariant items are as reliable and have the same correlational pattern as existing self-esteem questions in adults and children eight and older, but reliability and correlations are lower for the younger children. To gain a better understanding of this age difference, I am testing qualitative differences in conceptualizations of self-esteem. Preliminary findings show that explanations for self-esteem are remarkably similar across age with the largest number of self-esteem explanations being in the categories of achievement, relationships, and psychological traits across ages. Thus, younger and older children appear to conceptualized self-esteem in a similar way, suggesting that the lower reliability and validity might have to do with understanding of the items rather than the construct not being present in young children.
Second, I have two longitudinal studies consisting of several assessments between eight months and 6 years that includes measures and observations of the parent-child relationship, temperament, and physiological reactions, and includes a measure of global self-esteem at age six. Hypotheses about the temperamental and parental influences on self-esteem are being tested, including a hypothesized interaction between parenting and stress reactivity. Given the robust genetic influence on self-esteem that is independent from depression and negative affect I have hypothesized that higher innate stress reactivity results in an environment full of negative experiences which, over time, get internalized into a negative self-view. However, I hypothesize that this effect can be ameliorated by a parent who teaches the child how to regulate their negative emotional reactions. A third longitudinal study is underway to gain a deeper understanding of the development of self-esteem. The first wave, currently being collected, is based on children aged 5 to 13. Measures of self-esteem and behavior are being collected from the children and one parent, narrative and interview data are also being collected from the parent and child, and parent-child interactions are being observed. This study will allow me to test hypotheses about how parental representations of self and of the child impact the way memories are discussed with the child and formed by the child, which in turn will affect the child’s representation of the self and evaluation of the self.
Applied Research: Changing Self-Views and the Impact on Important Life Outcomes
I have two lines of applied research that seeks to translate basic research findings into interventions. The first line of research focuses on changing the mindsets that individuals use to view themselves and those around them. Specifically, it has been suggested that people vary on a continuum from believing in fixed traits or behaviors to malleable traits and behaviors (Dweck, 1999). These fixed vs. growth mindsets impact how individuals approach and cope with challenges, and importantly, lab studies have found that these mindsets are changeable. My colleagues and I have shown that interventions focused on promoting a more malleable mindset about intelligence leads to higher grades during middle school (Blackwell, Trzesniewski, & Dweck, 2007, Child Development), promoting a more malleable mindset about personality leads to better coping and lower retaliation in response to being bullied during high school (Yeager, Trzesniewski, & Dweck, 2013, Child Development; Yeager, Trzesniewski, Tirri, Nokelainen, & Dweck, 2011, Developmental Psychology), and promoting a more malleable mindset about groups (e.g., Palestinians) in adults living in a retractable conflict environment promotes willingness to compromise for peace (Halperin, Crisp, Husnu, Trzesniewski, Dweck, & Gross, 2012, Emotion; Halperin, Russell, Trzesniewski, Gross, & Dweck, 2011, Science). New directions for this research include testing the bullying focused intervention in a group of adolescents with autism spectrum disorder and testing the cognitive mechanisms that account for the achievement effects (co-PI, currently funded by IES, $1,600,000).
Finally, I am working with the California State 4-H Youth Development Program to deliver and evaluate a positive youth development program that promotes growth mindsets, helps youth discover passions, and teaches goal management skills (Campbell, Trzesniewski, Nathaniel, Enfield, Erbstein, 2013, California Agricultural, Healthy Families and Communities special issue; PI, currently funded by UC-ANR, $600,000). This program is being delivered throughout the state through the 4-H club program and in partnership with afterschool programs. Preliminary results suggest the program is successful in promoting a range of positive outcomes. A more stringent test of the effects will be conducted through randomized-controlled trials over the next two years.