Help With Journal Critique

The title for this Special Section is Origins of Children’s Self-Views, edited by Eddie Brummelman and Sander Thomaes

When Parents’ Praise Inflates, Children’s Self-Esteem Deflates

Eddie Brummelman University of Amsterdam and Stanford University

Stefanie A. Nelemans University of Leuven and Utrecht University

Sander Thomaes Utrecht University and University of Southampton

Bram Orobio de Castro Utrecht University

Western parents often give children overly positive, inflated praise. One perspective holds that inflated praise sets unattainable standards for children, eventually lowering children’s self-esteem (self-deflation hypothesis). Another perspective holds that children internalize inflated praise to form narcissistic self-views (self-inflation hypothesis). These perspectives were tested in an observational-longitudinal study (120 parent–child dyads from the Netherlands) in late childhood (ages 7–11), when narcissism and self-esteem first emerge. Supporting the self-deflation hypothesis, parents’ inflated praise predicted lower self-esteem in children. Partly supporting the self-inflation hypothesis, parents’ inflated praise predicted higher narcissism—but only in children with high self-esteem. Noninflated praise predicted neither self-esteem nor narcissism. Thus, inflated praise may foster the self-views it seeks to prevent.

“Terrific!” “Your drawing is amazing!” “You did incredibly well at this!” In current Western society, parents often give their children overly positive, inflated praise. About 25% of parents’ praise is inflated (Brummelman, Thomaes, Orobio de Castro, Overbeek, & Bushman, 2014). Parents may believe that by inflating their praise, they raise their chil- dren’s self-esteem (Damon, 1995). Does inflated praise actually succeed in raising children’s self- esteem? Or does it ironically backfire, and foster low self-esteem or even narcissism? To address these questions, we conducted an observational- longitudinal study in late childhood—the age at which individual differences in self-esteem and nar- cissism first emerge.

Praise to Raise Self-Esteem

Since the self-esteem movement emerged in the 1970s, Western society has become increasingly

concerned with building children’s self-esteem (Baumeister, Campbell, Krueger, & Vohs, 2003; Heine, Lehman, Markus, & Kitayama, 1999; Twenge & Campbell, 2009). Indeed, adults from Western countries view low self-esteem in children as a perva- sive and worrisome problem (Thomaes, Brummel- man, Bushman, Reijntjes, & Orobio de Castro, 2016) and are motivated to cure it through praise. Self- esteem interventions often rely on praise as one of their key components (O’Mara, Marsh, Craven, & Debus, 2006), and self-help books recommend parents to build children’s self-esteem by praising them (e.g., Talbot, 2009). Unsurprisingly, parents have internal- ized this belief. In one study, 87% of parents believed that children need praise in order to feel good about themselves (Brummelman & Thomaes, 2011).

Not all praise is alike, however. A large body of literature shows that praise comes in different shapes and sizes, with some more effective than others (for overviews, see Dweck, 2006; Hattie & Timperley, 2007; Henderlong & Lepper, 2002). For example, when children are praised for success at

This research was supported by The Netherlands Organization for Scientific Research (grant 431-09-022).

Correspondence concerning this article should be addressed to Eddie Brummelman, Research Institute of Child Development and Education, University of Amsterdam, P.O. Box 15780, 1001 NG Amsterdam, The Netherlands. Electronic mail may be sent to [email protected].

© 2017 The Authors Child Development © 2017 Society for Research in Child Development, Inc. All rights reserved. 0009-3920/2017/8806-0004 DOI: 10.1111/cdev.12936

Child Development, November/December 2017, Volume 88, Number 6, Pages 1799–1809

easy rather than difficult tasks, they may infer that they lack ability (e.g., “If you praise me for some- thing this easy, you must think I can’t do much bet- ter”; Barker & Graham, 1987). In addition, when children are praised for their ability (e.g., “You’re so smart!”) rather than their effort (e.g., “You worked so hard!”), they may infer that ability is a fixed, unchangeable trait (Mueller & Dweck, 1998). Conse- quently, they avoid challenging tasks; and when they fail, they infer lack of ability. Thus, subtle differences in praise can have considerable impact on children.

When parents want to raise children’s self-esteem, they might be especially likely to rely on overly posi- tive, inflated praise (Brummelman, Crocker, & Bush- man, 2016). Instead of telling children that they did well, they might tell them that they did incredibly well. Instead of telling children that their drawing is nice, they might tell them that their drawing is amaz- ing. People often use such extreme language in an attempt to persuade others (Hamilton & Hunter, 1998). Thus, parents may believe that by inflating their praise, they may persuade their children into feeling better about themselves. Accordingly, parents direct inflated praise particularly at those who seem to need it the most: children with low self-esteem (Brummelman et al., 2014).

Self-Inflation or Self-Deflation?

Does inflated praise succeed in raising children’s self-esteem? Perhaps not. Although direct empirical evidence is lacking, psychologists have theorized that inflated praise may backfire. The self-deflation hypothesis states that inflated praise may lead chil- dren to develop lower—not higher—levels of self- esteem (Brummelman, Crocker, et al., 2016). Accord- ing to this perspective, parents often give inflated praise to children with low self-esteem in an attempt to raise their self-esteem. Rather than raising self- esteem, however, inflated praise conveys to children that they should continue to meet very high stan- dards (Henderlong & Lepper, 2002). When children are told that they performed incredibly well, they may infer that they should perform incredibly well all the time (Baumeister, Hutton, & Cairns, 1990; Kanouse, Gumpert, & Canavan-Gumpert, 1981; Ryan, 1982). Struggles and setbacks are inevitable, so children may eventually fall short of the standards set for them and therefore feel down about them- selves. Thus, parents’ inflated praise may ironically lower children’s self-esteem over time.

By contrast, the self-inflation hypothesis states that inflated praise does not lower self-esteem but rather cultivates narcissism in children. According to this

perspective, children use feedback from others to form views of themselves; more specifically, they come to see themselves as they believe they are seen by others (Cooley, 1902; Harter, Waters, & Whitesell, 1998; Mead, 1934; Thomaes et al., 2010). When chil- dren are told that they performed incredibly well, they may infer that they are extraordinary—a core feature of the trait of narcissism (Millon, 1969; Twenge & Campbell, 2009; Young-Eisendrath, 2008). Once children believe they are extraordinary, they try to maintain this view of themselves by trying to garner external validation (Brummelman, Thomaes, & Sedikides, 2016; Morf & Rhodewalt, 2001). Indeed, narcissism is characterized by cravings for respect and admiration from others (Baumeister & Vohs, 2001). Thus, inflated praise from parents may encourage core narcissistic traits in children.

At first blush, the self-inflation and self-deflation hypotheses may seem contradictory. They are, how- ever, complementary, as narcissism and self-esteem are distinct dimensions of the self (Brummelman, Thomaes, et al., 2016). In fact, narcissism and self- esteem are virtually uncorrelated in childhood (Thomaes & Brummelman, 2016). Narcissistic chil- dren can be low in self-esteem, and children with high self-esteem can be low in narcissism. This con- tradicts the popular view that narcissism is an exag- gerated form of self-esteem.

How do narcissism and self-esteem differ? Narcis- sistic children feel superior to others, believe they are entitled to privileges, and want to be admired by others (Campbell & Miller, 2011; Thomaes & Brum- melman, 2016). When they get the admiration they want, they feel on top of the world; but when they do not, they want to sink into the ground. Narcissists often respond to such humiliation by lashing out angrily and aggressively (Thomaes, Bushman, Stegge, & Olthof, 2008). By contrast, children with high self-esteem are happy with themselves as a per- son but do not see themselves as superior to others (Harter, 2012). As Morris Rosenberg (1965) said, “When we deal with self-esteem, we are asking whether the individual considers himself adequate— a person of worth—not whether he considers himself superior to others” (p. 62). Rather than seeking admi- ration, children with high self-esteem want to estab- lish deep and intimate bonds with others (Thomaes, Stegge, Bushman, Olthof, & Denissen, 2008). They rarely explode in angry and aggressive outbursts (Donnellan, Trzesniewski, Robins, Moffitt, & Caspi, 2005) and are at reduced risk for anxiety and depres- sion (Sowislo & Orth, 2013). Thus, although narcis- sism entails unhealthy feelings of superiority, self- esteem entails healthy feelings of worth.

1800 Brummelman, Nelemans, Thomaes, and Orobio de Castro

Present Study

The present observational-longitudinal study investigated, for the first time, the self-inflation and self-deflation hypotheses. At baseline, we assessed children’s narcissism and self-esteem levels. Five weeks later, we conducted in-home observations of parent–child interactions to code parents’ inflated and noninflated praise. We then followed up on children’s narcissism and self-esteem levels for three additional, 6-monthly waves.

The study was timed in the critical phase of late childhood (ages 7–11), when individual differences in narcissism and self-esteem first emerge (Thomaes & Brummelman, 2016). Children this age have acquired the cognitive capacities to form global self-evaluations (Harter, 2012), which underlie self- esteem and narcissism. Also, they use social com- parisons to evaluate themselves (Ruble & Frey, 1991), enabling narcissistic self-views: “I’m better than others!” In addition, children this age readily use praise to evaluate themselves and to form stan- dards for their future performance (Dweck, 2002), which makes praise especially salient to them.

Method

Participants

Participants were 120 children (50% girls; ages 7– 11 at baseline; M = 8.86, SD = 0.85; 87% of Dutch origin) and their parent (88% mothers; ages 30–62 at baseline; M = 43.29, SD = 4.12; 97% of Dutch ori- gin) recruited through public elementary schools serving lower-to-upper middleclass communities in the Netherlands. Of all parents who were contacted, 56% provided active informed consent for them- selves and for their child and participated in the study. The parents self-identified as the child’s primary caregiver. All the children gave their own assent. In a previous article, we reported the corre- lation between children’s self-esteem at Wave 1 and parents’ inflated praise (Brummelman et al., 2014, study 2). For the present research, we followed up on these children for three additional waves, assess- ing both their self-esteem and their narcissism levels. Data were collected from autumn 2011 to spring 2013.

Procedure and Materials

In each of four 6-monthly waves, children completed questionnaires in their classrooms at school. During the assessments, nonparticipating

classmates worked on their regular schoolwork, independently and in silence. Narcissism was assessed using the 10-item Childhood Narcissism Scale (Thomaes, Stegge, et al., 2008). Sample items include: “I think it’s important to stand out” and “I am a very special person” (0 = not at all true to 3 = completely true; Cronbach’s a range for Waves 1–4 = .69–.84). We assessed narcissism as a normal or everyday personality trait on which youth and adults in the general population vary (Thomaes & Brummelman, 2016). Self-esteem was assessed using the six-item Global Self-Worth Subscale of the Self- Perception Profile for Children (Harter, 1985). Sam- ple items include: “Some kids are happy with them- selves as a person” and “Some kids like the kind of person they are” (0 = I am not like these kids at all to 3 = I am exactly like these kids; Cronbach’s a range for Waves 1–4 = .78–.84). For both scales, responses were averaged across items, with higher scores indicating higher narcissism and self-esteem levels, respectively.

An average of 5 weeks (36 days) after Wave 1, children and their parent participated in one in- home observation. Parents were asked to adminis- ter 12 mathematics exercises to their child (i.e., Exercises 5–16 from the Arithmetic subtest of the Wechsler Intelligence Scale for Children–III; Wech- sler, 1991). Using a stopwatch and score sheet, par- ents judged whether the child correctly completed each exercise within 30 s. Parents were not allowed to help the child complete the exercises. Research assistants left the room until the exercises were completed.

The sessions were videotaped. Two independent trained coders, blind to variation in children’s level of narcissism and self-esteem, counted the number of times parents gave their child inflated and nonin- flated praise. We defined praise as spoken positive evaluations of the child’s traits, actions, or products (Henderlong & Lepper, 2002; Kanouse et al., 1981). Consistent with previous research (Brummelman et al., 2014), praise was considered inflated when it contained an adverb (e.g., extremely, incredibly) or adjective (e.g., amazing, fantastic) signaling a very positive evaluation. Both coders assessed all par- ent–child dyads. Intercoder agreement was high (Cohen’s j = .98). Discrepancies were resolved through discussion between coders.

Longitudinal Analyses

Due to camera failure, videos of four parent– child dyads were unavailable, causing these dyads to have missing values on praise. Due to attrition,

Inflated Praise, Self-Esteem, and Narcissism 1801

some children had missing values on narcissism and self-esteem (2, 5, 12, and 13 children on Waves 1–4, respectively). Little’s Missing Completely at Random test produced a normed v2 (v2/df) of 0.93, v2(62) = 57.787, p = .628, indicating that data were missing at random. Missing values were handled with full-information maximum likelihood.

We conducted path analyses in Mplus version 7 (Muth�en & Muth�en, 1998-2015) to examine longitu- dinal associations between praise (inflated, nonin- flated) and self-views (narcissism, self-esteem). We conducted separate analyses for inflated and non- inflated praise, as well as for narcissism and self- esteem. The models included 1-wave and 2-wave stability paths for narcissism and self-esteem; paths from narcissism and self-esteem at Wave 1 to inflated and noninflated praise; and paths from inflated or noninflated praise to narcissism and self-esteem at Waves 2, 3, and 4. To create parsi- monious models, all longitudinal parameters were constrained to be time invariant (Kline, 2005). Free- ing parameters of interest (i.e., longitudinal paths from praise to later self-views) did not significantly improve fit of any of the models.

We kept an interval between Wave 1 and the in-home observations to examine whether chil- dren’s self-views would predict subsequent paren- tal praise. We, therefore, modeled the paths from self-views (narcissism, self-esteem) at Wave 1 to praise (inflated, noninflated) as direct effects rather than as correlations. Modeling them as correlations did not change any of our findings. For one par- ent–child dyad, the in-home observation was inad- vertently scheduled before (rather than after) Wave 1. Because excluding this dyad did not change any of our findings, we retained it in our analyses.

Indirect Effects

We tested indirect effects using Mplus bias-cor- rected bootstrap confidence intervals (Hayes & Scharkow, 2013; MacKinnon, Lockwood, & Wil- liams, 2004) with 10,000 bootstrap draws.

Moderation

The consequences of praise may differ between children with high and low self-esteem (Brummel- man et al., 2014). We, therefore, explored whether children’s self-esteem at Wave 1 moderated the lon- gitudinal paths from praise (inflated, noninflated) to subsequent self-views (narcissism, self-esteem) by including an interaction between Wave 1 self-

esteem and praise (inflated, noninflated). All terms were centered before computing the interactions.

Model Fit

Model fit was assessed with the comparative fit index (CFI), the root mean squared error of approx- imation (RMSEA) and 90% confidence interval (CI), and the standardized root mean square residual (SRMR). CFI values ≥ .90, RMSEA values ≤ .08, and SRMR values < .10 indicate acceptable model fit, whereas CFI values ≥ .95, RMSEA values ≤ .05, and SRMR values < .08 indicate good model fit (Hu & Bentler, 1999; Kline, 2005; Muth�en, 2004). The comparative fit between nested models was tested with chi-square difference tests.

Results

Descriptive statistics and correlations are displayed in Table 1. Narcissism and self-esteem were virtu- ally uncorrelated, both within and across waves, attesting to their conceptual independence. Fre- quencies of inflated and noninflated praise were similarly uncorrelated.

We first tested the self-inflation and self-deflation hypotheses, respectively. After testing each hypoth- esis, we explored whether the effects of praise were moderated by children’s initial level of self-esteem. Goodness-of-fit statistics are displayed in Table 2.

Self-Deflation Hypothesis

We first tested the self-deflation hypothesis, which holds that parents’ inflated praise predicts lower self-esteem in children. The fully constrained baseline model demonstrated good fit to the data. Freeing longitudinal paths (i.e., paths from inflated praise to later self-esteem) did not significantly improve model fit.

Self-esteem at Wave 1 predicted inflated praise: Children with lower levels of self-esteem at Wave 1 received more inflated praise (b [95% CI] = �0.794 [�1.530, �0.099], b = �0.224). Inflated praise, in turn, predicted lower self-esteem at Waves 2, 3, and 4 (b [95% CI] = �0.029 [�0.052, �0.007], bs = �0.103 to �0.119). Importantly, the indirect path was signifi- cant: Self-esteem at Wave 1 predicted lower self- esteem at Waves 2, 3, and 4 partly through inflated praise (b [95% CI] = 0.023 [0.003, 0.063]; Figure 1).

We then tested whether children’s self-esteem at Wave 1 moderated the paths from inflated praise to later self-esteem (Waves 2–4). Constraining the

1802 Brummelman, Nelemans, Thomaes, and Orobio de Castro

moderation effects to be equal over time did not significantly worsen model fit. The fully constrained moderation model demonstrated good fit to the data. Children’s self-esteem at Wave 1 did not mod- erate the longitudinal paths from inflated praise to self-esteem at Wave 2, 3, and 4 (b [95% CI] = 0.012 [�0.023, 0.048]).

Thus, supporting the self-deflation hypothesis, children with lower self-esteem received more inflated praise from their parents. Parents’ inflated praise, in turn, predicted lower self-esteem in chil- dren over time, regardless of whether children started out with high or low self-esteem.

Self-Inflation Hypothesis

Next, we tested the self-inflation hypothesis, which holds that parents’ inflated praise predicts higher narcissism in children. The fully constrained baseline model demonstrated good fit to the data. Freeing longitudinal paths (i.e., paths from inflated praise to later narcissism) did not significantly improve model fit.

Narcissism at Wave 1 did not predict inflated praise (b [95% CI] = 0.034 [�0.846, 1.009], b = 0.007) nor did inflated praise predict narcissism at Waves 2, 3, and 4 (b [95% CI] = 0.003 [�0.016, 0.025], b = 0.013; Figure 1). Consequently, there were no significant indirect effects from narcissism at Wave 1 to narcissism at later waves through inflated praise (b [95% CI] = 0.000 [�0.009, 0.011]).

We then tested whether children’s self-esteem at Wave 1 moderated the paths from inflated praise to later narcissism (Waves 2–4). Constraining the mod- eration effects to be equal over time did not signifi- cantly worsen model fit. The fully constrained

moderation model demonstrated good fit to the data.

The moderation was significant (b [95% CI] = 0.044 [0.019, 0.072]). Inflated praise predicted higher narcissism among children with high self- esteem (M + 1 SD) at Wave 1 (b [95% CI] = 0.041 [0.015, 0.073]) but not among children with low (M � 1 SD) or average self-esteem at Wave 1 (b [95% CI] = �0.012 [�0.037, 0.010] and b [95% CI] = 0.014 [�0.005, 0.036]), respectively.

Thus, partly supporting the self-inflation hypoth- esis, parents’ inflated praise predicted higher narcis- sism but only in children with high initial levels of self-esteem.

Auxiliary Analyses

Specificity

To examine the specificity of our findings, we repeated all analyses for noninflated praise (see Supporting Information). Neither self-esteem nor narcissism at Wave 1 predicted parents’ noninflated praise, nor did parents’ noninflated praise predict later self-esteem or narcissism. Also, none of the longitudinal paths (i.e., paths from noninflated praise to later self-esteem or narcissism) were mod- erated by children’s self-esteem at Wave 1. Thus, our findings are specific to inflated praise and do not generalize to noninflated praise.

Robustness

To examine the robustness of our findings, we repeated all analyses with children’s sex, age, and performance (i.e., the number of correct answers on

Table 1 Descriptive Statistics and Correlations

n M SD 2 3 4 5 6 7 8 9 10 11

1. Inflated praise 116 1.59 2.14 �.03 �.02 �.23* �.14 �.36** �.33** .01 .10 .12 �.00 2. Noninflated praise 116 4.72 3.36 — �.05 �.14 �.11 .03 �.07 .11 .15 .03 .11 3. Performance 116 11.07 1.05 — .23* .26** .05 .01 �.18 �.00 �.02 �.09 4. W1 self-esteem 118 2.12 0.61 — .46** .49** .45** �.02 .09 .08 .12 5. W2 self-esteem 115 2.21 0.59 — .64** .57** �.10 .05 �.13 �.01 6. W3 self-esteem 108 2.29 0.53 — .72** .09 .07 .05 .19 7. W4 self-esteem 107 2.32 0.53 — .06 .00 .03 .21* 8. W1 narcissism 118 1.08 0.47 — .44** .51** .49** 9. W2 narcissism 115 1.02 0.52 — .66** .60** 10. W3 narcissism 108 0.97 0.52 — .71** 11. W4 narcissism 107 1.07 0.52 —

Note. W1–W4 = Wave 1–Wave 4. Performance = the number of correct answers on the mathematics exercises. *p < .05. **p < .01.

Inflated Praise, Self-Esteem, and Narcissism 1803

the mathematics exercises) as time-invariant covari- ates (see Supporting Information). These variables did not predict inflated praise, noninflated praise, narcissism, or self-esteem in any of the models, nor did including them as covariates affect any of the longitudinal paths from praise to later self-views. This attests to the robustness of our findings.

Discussion

Does praising children in inflated ways help them develop higher self-esteem, as conventional wisdom

says it does? We addressed this question in an observational-longitudinal study timed in the criti- cal phase of late childhood. Consistent with the self-deflation hypothesis, parents’ inflated praise predicted lower—not higher—self-esteem in chil- dren over time. Partly consistent with the self-infla- tion hypothesis, parents’ inflated praise predicted higher narcissism over time but only in children with high initial levels of self-esteem. Attesting to the specificity of these findings, they did not emerge for parents’ moderately positive, nonin- flated praise. Together, these findings challenge conventional wisdom and show that inflated praise

Figure 1. Standardized associations (bs) between parents’ inflated praise and children’s self-esteem (a) and narcissism (b). Solid lines represent significant associations; dotted lines represent nonsignificant associations. Children’s self-esteem at Wave 1 predicted more inflated praise from parents, which in turn predicted lower self-esteem at Waves 2, 3, and 4. Children’s narcissism at Wave 1 did not predict inflated praise from parents, nor did inflated praise predict narcissism at Waves 2, 3, or 4.

Table 2 Goodness-of-Fit Statistics of Longitudinal Models Involving Self-Esteem, Narcissism, and Inflated Praise

Models v2(df) CFI RMSEA [90% CI] SRMR Dv2(df)

Inflated praise and self-esteem Baseline, constrained 5.187 (6) 1.000 .000 [.000, .110] .070 — Baseline, unconstrained 3.260 (4) 1.000 .000 [.000, .127] .066 1.93 (2) Moderation, constrained 4.975 (7) 1.000 .000 [.000, .090] .040 — Moderation, unconstrained 1.518 (5) 1.000 .000 [.000, .049] .020 3.46 (2)

Inflated praise and narcissism Baseline, constrained 7.057 (6) .994 .039 [.000, .130] .040 — Baseline, unconstrained 2.435 (4) 1.000 .000 [.000, .110] .022 4.62 (2) Moderation, constrained 16.678 (13) .980 .049 [.000, .109] .054 — Moderation, unconstrained 14.953 (11) .979 .055 [.000, .119] .053 1.73 (2)

Note. Baseline models include no interactions. Moderation models include an interaction between Wave 1 self-esteem and inflated praise. Constrained models include time-invariant longitudinal parameters. Unconstrained models include time-variant longitudinal parameters. CFI = comparative fit index; RMSEA = root mean squared error of approximation; 90% CI = 90% confidence interval; SRMR = standardized root mean square residual; Dv2 = Satorra-Bentler scaled chi-square difference test, comparing the fit of con- strained and unconstrained models; all Dv2 tests were nonsignificant, ps > .099.

1804 Brummelman, Nelemans, Thomaes, and Orobio de Castro

can have unintended consequences for children’s self-development.

Theoretical Implications

Parents gave more inflated praise to children with lower self-esteem; parents’ inflated praise, in turn, predicted lower self-esteem over time. Strik- ingly, children’s initial low self-esteem predicted subsequent low self-esteem through eliciting inflated praise from parents. These findings support the transactional model of praise (Brummelman, Crocker, et al., 2016), which holds that parents’ well-meant attempts to counteract children’s low self-esteem by praising them in inflated ways can backfire. By giving inflated praise, parents may inadvertently pressure children to continue to dis- play exceptional performance, thereby eventually lowering children’s self-esteem in the face of inevi- table struggles and setbacks. Lowered self-esteem, in turn, may inspire parents to give even more inflated praise, thus potentially establishing a self- sustaining downward spiral.

More broadly, these findings emphasize the need for a transactional perspective on praise (Brummel- man, Crocker, et al., 2016). To date, praise has been studied primarily as a unidirectional process, with children as passive recipients of the praise. Consis- tent with transactional models of the self (Crocker & Brummelman, in press) and development (Samer- off & MacKenzie, 2003), whereas our findings show that children are not passive recipients of praise. They actively shape the praise they receive, which in turn shapes them. Such transactions cannot be revealed by studying merely the effects of praise on the child; they can be revealed only by studying transactions between the child and the environment (e.g., studying how children are praised by their parents, how the praise affects them, and how this feeds back into parents’ ways of praising them).

Are there alternative explanations for why par- ents’ inflated praise predicted lower self-esteem in children? Previous work shows that praising chil- dren for success at easy tasks can lower self-per- ceived competence (Graham, 1990; Meyer, 1992). Indeed, adults typically praise children who work hard (Weiner & Kukla, 1970). So when children are praised for success at an easy task, they may infer that the praiser thinks they did not have the ability to succeed without working so hard (Graham, 1990; Meyer, 1992). This may not fully explain our find- ings, however. First, in our study, children’s task was not explicitly labeled as “easy.” Second, par- ents’ inflated praise predicted lower self-esteem

regardless of children’s age and performance (see Supporting Information), even though older and better-performing children presumably found the task relatively easy. A more likely interpretation of our findings, therefore, is that inflated praise sets unattainable standards for children. This may not lower children’s self-esteem instantly (Brummelman et al., 2014) but only over time, when children encounter inevitable struggles and setbacks that make them realize they are unable to meet the stan- dards set for them (Brummelman, Crocker, et al., 2016).

Our findings raise the question: Why did inflated praise predict higher narcissism only in children with high self-esteem? Social-judgment theory (Sherif & Hovland, 1961; Wood, Perunovic, & Lee, 2009) holds that people compare persuasive messages (e.g., praise) with their current attitudes (e.g., self-esteem) and internalize only those messages that are close to their current attitudes: messages that fall within their latitude of acceptance. Thus, children with high and low self-esteem may feel similarly pressured by inflated praise to continue to display excellent perfor- mance. Yet only those with high self-esteem may internalize the praise to form narcissistic self-views (e.g., “Yes, indeed, I am incredible”). Those with low self-esteem may judge the praise as falling outside of their latitude of acceptance (e.g., “I’m not that incred- ible”). Thus, children internalize praise only when the praise does not clash with their current views of themselves.

Our study extends existing longitudinal research on the consequences of praise (Gunderson et al., 2013; Pomerantz & Kempner, 2013). It is the first longitudinal study on inflated praise, a type of praise that has rarely been studied empirically (Brummelman et al., 2014). Even though the differ- ence between inflated and noninflated praise can be subtle, it matters for children’s self-development. In our study, parents’ inflated praise predicted chil- dren’s lower self-esteem and higher narcissism, but their noninflated praise did not. Unlike inflated praise, noninflated praise does not entail an overly positive evaluation; it neither pressures children to perform extraordinarily well nor conveys to them that they are extraordinary individuals. Our find- ings, therefore, do not imply that parents should stop praising children altogether; they imply that parents should be careful not to praise children in inflated ways.

Additionally, our study is the first to examine whether praise may cultivate narcissism. Psycholo- gists often fear that praise inadvertently breeds nar- cissism in children (Baumeister et al., 2003; Millon,

Inflated Praise, Self-Esteem, and Narcissism 1805

1969; Twenge & Campbell, 2009). Direct empirical evidence was lacking, however. Our study demon- strates that praise does predict higher narcissism but only if phrased in an inflated way and directed at children with high self-esteem. More broadly, our findings resonate with the finding that narcis- sism is cultivated by parental overvaluation—parents seeing their child as an extraordinary and entitled individual, who have been shown to praise their child frequently (Brummelman, Thomaes, Nelemans, Orobio de Castro, & Bushman, 2015; Brummelman, Thomaes, Nelemans, Orobio de Castro, Overbeek, et al., 2015). However, overvaluing parents do not seem to praise with the purpose of raising their child’s self-esteem; rather, they do so with the purpose of making their child (and, by association, themselves) stand out. Indeed, overvaluing parents have narcissistic traits; they are ego-involved in their child’s performances and want to bask in the child’s reflected glory (Brummelman, Thomaes, Nelemans, Orobio de Castro, & Bushman, 2015; also see Grolnick, Gurland, DeCourcey, & Jacob, 2002). Thus, praising their child may repre- sent a concealed attempt to put themselves on a pedestal.

Strengths, Limitations, and Future Directions

Strengths of this study include its combination of in-home observations and longitudinal follow- ups, its timing in late childhood, its parsing of inflated and noninflated praise, and its assessment of both self-esteem and narcissism. Our study also has limitations. First, we assessed praise at a single point in time. To examine whether parents’ inflated praise and children’s low self-esteem indeed establish a self-sustaining downward spiral (Brummelman, Crocker, et al., 2016), future research should assess both praise and self-esteem repeatedly over time. Second, we argued that inflated praise lowers children’s self-esteem by put- ting pressure on them to achieve very high stan- dards, but we did not assess those perceived standards. Future research should examine whether children’s perceived standards (e.g., “My father [mother] expects me to perform exception- ally well all the time”) mediate the adverse effects of inflated praise on self-esteem. Such perceived standards should be distinguished from perfor- mance goals, which are more narrowly defined as goals to demonstrate one’s ability (Ames & Archer, 1988). Third, our study was conducted in the Netherlands, a Western country. Parents from non- Western cultures, such as China, rarely praise their

children (Ng, Pomerantz, & Lam, 2007). Because non-Western parents place little value on self- esteem (Miller, Wang, Sandel, & Cho, 2002), they may be especially unlikely to praise their children in inflated ways. Future research should examine whether inflated praise is indeed more common in Western than non-Western countries.

Our study also generates novel research direc- tions. One direction is to examine how parents gauge children’s level of self-esteem to attune their praise accordingly. Do parents infer low self-esteem from children’s self-directed talk (e.g., “I’m no good . . .”), body language (e.g., collapsed posture), or confidence (e.g., their confidence in trying new tasks)? Do parents attune their praise deliberately, or does this happen automatically and outside of their conscious awareness? Addressing these ques- tions will inform interventions about how to effec- tively redirect parents’ praise. Another direction is to refine existing self-esteem interventions. Self- esteem interventions are often universal (i.e., target- ing children in the general population, rather than at-risk subgroups) and often rely on praise (O’Mara et al., 2006). Given our finding that inflated praise predicts lower self-esteem in the general popula- tion, self-esteem interventions may increase their effectiveness by relying on moderately positive, noninflated praise rather than overly positive, inflated praise.

More broadly, future research should try to iden- tify parenting practices that effectively raise chil- dren’s self-esteem. Parents often attempt to raise children’s self-esteem directly through praise (Brummelman, Crocker, et al., 2016). However, in our study, neither inflated nor inflated praise was effective in raising self-esteem, perhaps because praise is ultimately a judgment. As one expert put it, “the most notable aspect of a positive judgment is not that it is positive but that it is a judgment” (Kohn, 1993, p. 102). When children are praised fre- quently, they may start to judge themselves through the eyes of others—a process that does not lead to a stable and secure sense of worth (Deci & Ryan, 2000; Kernis, 2003). It may be more effective to raise children’s self-esteem indirectly, such as by building warm relationships with them (e.g., shar- ing joy with them, cuddling with them, and show- ing interest in their activities), which makes them feel accepted and valued (Brummelman, Thomaes, Nelemans, Orobio de Castro, Overbeek, et al., 2015). When children feel accepted and valued, they may gradually internalize the perception of them- selves as worthy individuals (Brummelman, Tho- maes, et al., 2016).

1806 Brummelman, Nelemans, Thomaes, and Orobio de Castro

Conclusion

Raising children’s self-esteem has become one of the main parenting goals in Western society (Baumeister et al., 2003; Heine et al., 1999). This is not surprising given that self-esteem contributes to children’s success and well-being (Orth & Robins, 2014). However, in their well-meant efforts to raise self-esteem, parents often give children overly posi- tive, inflated praise. Our study shows that, rather than raising self-esteem, inflated praise may culti- vate lower self-esteem and higher narcissism in children. Thus, what seems like common sense can lead parents to rely on types of praise that ironi- cally backfire.

References

Ames, C., & Archer, J. (1988). Achievement goals in the classroom: Students’ learning strategies and motivation processes. Journal of Educational Psychology, 80, 260–267. https://doi.org/10.1037/0022-0663.80.3.260

Barker, G. P., & Graham, S. (1987). Developmental study of praise and blame as attributional cues. Journal of Educational Psychology, 79, 62–66. https://doi.org/10. 1037/0022-0663.79.1.62

Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003). Does high self-esteem cause better perfor- mance, interpersonal success, happiness, or healthier lifestyles? Psychological Science in the Public Interest, 4, 1–44. https://doi.org/10.1111/1529-1006.01431

Baumeister, R. F., Hutton, D. G., & Cairns, K. J. (1990). Negative effects of praise on skilled performance. Basic and Applied Social Psychology, 11, 131–148. https://doi. org/10.1207/s15324834basp1102_2

Baumeister, R. F., & Vohs, K. D. (2001). Narcissism as addiction to esteem. Psychological Inquiry, 12, 206–210. https://doi.org/10.1207/S15327965PLI1204_2

Brummelman, E., Crocker, J., & Bushman, B. J. (2016). The praise paradox: When and why praise backfires in children with low self-esteem. Child Development Per- spectives, 10, 111–115. https://doi.org/10.1111/cdep. 12171

Brummelman, E., & Thomaes, S. (2011). [Parents’ beliefs about praise]. Unpublished raw data.

Brummelman, E., Thomaes, S., Nelemans, S. A., Orobio de Castro, B., & Bushman, B. J. (2015). My child is God’s gift to humanity: Development and validation of the Parental Overvaluation Scale (POS). Journal of Per- sonality and Social Psychology, 108, 665–679. https://doi. org/10.1037/pspp0000012

Brummelman, E., Thomaes, S., Nelemans, S. A., Orobio de Castro, B., Overbeek, G., & Bushman, B. J. (2015). Origins of narcissism in children. Proceedings of the National Acad- emy of Sciences of the United States of America, 112, 3659– 3662. https://doi.org/10.1073/pnas.1420870112

Brummelman, E., Thomaes, S., Orobio de Castro, B., Overbeek, G., & Bushman, B. J. (2014). “That’s not just beautiful—that’s incredibly beautiful!”: The adverse impact of inflated praise on children with low self- esteem. Psychological Science, 25, 728–735. https://doi. org/10.1177/0956797613514251

Brummelman, E., Thomaes, S., & Sedikides, C. (2016). Separating narcissism from self-esteem. Current Direc- tions in Psychological Science, 25, 8–13. https://doi.org/ 10.1177/0963721415619737

Campbell, W. K., & Miller, J. D. (Eds.). (2011). The hand- book of narcissism and narcissistic personality disorder: The- oretical approaches, empirical findings, and treatments. New York, NY: Wiley.

Cooley, C. H. (1902). Human nature and the social order. New York, NY: Scribner’s.

Crocker, J., & Brummelman, E. (in press). The self: Dynamics of persons and their situations. In K. Deaux & M. Snyder (Eds.), Handbook of personality and social psychology (2nd ed.). New York, NY: Oxford University Press.

Damon, W. (1995). Greater expectations: Overcoming the cul- ture of indulgence in our homes and schools. New York, NY: Free Press.

Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determina- tion of behavior. Psychological Inquiry, 11, 227–268. https://doi.org/10.1207/S15327965PLI1104_01

Donnellan, M. B., Trzesniewski, K. H., Robins, R. W., Moffitt, T. E., & Caspi, A. (2005). Low self-esteem is related to aggression, antisocial behavior, and delin- quency. Psychological Science, 16, 328–335. https://doi. org/10.1111/j.0956-7976.2005.01535.x

Dweck, C. S. (2002). The development of ability concep- tions. In A. Wigfield & J. S. Eccles (Eds.), Development of achievement motivation (pp. 57–88). London, UK: Aca- demic Press. https://doi.org/10.1016/b978-012750053- 9/50005-x

Dweck, C. S. (2006). Mindset: The new psychology of success. New York, NY: Random House.

Graham, S. (1990). Communicating low ability in the classroom: Bad things good teachers sometimes do. In S. Graham & V. Folkes (Eds.), Attribution theory: Appli- cations to achievement, mental health, and interpersonal con- flict (pp. 17–36). Hillsdale, NJ: Erlbaum.

Grolnick, W. S., Gurland, S. T., DeCourcey, W., & Jacob, K. (2002). Antecedents and consequences of mothers’ autonomy support: An experimental investigation. Developmental Psychology, 38, 143–155. https://doi.org/ 10.1037/0012-1649.38.1.143

Gunderson, E. A., Gripshover, S. J., Romero, C., Dweck, C. S., Goldin-Meadow, S., & Levine, S. C. (2013). Par- ent praise to 1-to 3-year-olds predicts children’s moti- vational frameworks 5 years later. Child Development, 84, 1526–1541. https://doi.org/10.1111/cdev.12064

Hamilton, M. A., & Hunter, J. E. (1998). The effect of lan- guage intensity on receiver evaluations of message, source, and topic. In M. Allen & R. W. Preiss (Eds.),

Inflated Praise, Self-Esteem, and Narcissism 1807

Persuasion: Advances through meta-analysis (pp. 99–138). Cresskill, NJ: Hampton Press.

Harter, S. (1985). Manual for the self-perception profile for children: Revision of the perceived competence scale for chil- dren. Unpublished manuscript, Department of Psychol- ogy, University of Denver, Denver, CO.

Harter, S. (2012). Construction of the self: Developmen- tal and sociocultural foundations. New York, NY: Guilford.

Harter, S., Waters, P., & Whitesell, N. R. (1998). Relational self-worth: Differences in perceived worth as a person across interpersonal contexts among adolescents. Child Development, 69, 756–766. https://doi.org/10.1111/ j.1467-8624.1998.tb06241.x

Hattie, J., & Timperley, H. (2007). The power of feedback. Review of Educational Research, 77, 81–112. https://doi. org/10.3102/003465430298487

Hayes, A. F., & Scharkow, M. (2013). The relative trust- worthiness of inferential tests of the indirect effect in statistical mediation analysis: Does method really mat- ter? Psychological Science, 24, 1918–1927. https://doi. org/10.1177/0956797613480187

Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a universal need for positive self- regard? Psychological Review, 106, 766–794. https://doi. org/10.1037/0033-295X.106.4.766

Henderlong, J., & Lepper, M. (2002). The effects of praise on children’s intrinsic motivation: A review and syn- thesis. Psychological Bulletin, 128, 774–795. https://doi. org/10.1037/0033-2909.128.5.774

Hu, L.-T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional cri- teria versus new alternatives. Structural Equation Modeling, 6, 1–55. https://doi.org/10.1080/10705519909540118

Kanouse, D. E., Gumpert, P., & Canavan-Gumpert, D. (1981). The semantics of praise. In J. H. Harvey, W. Ickes, & R. F. Kidd (Eds.), New directions in attribution research (Vol. 3, pp. 97–115). Hillsdale, NJ: Erlbaum.

Kernis, M. H. (2003). Toward a conceptualization of opti- mal self-esteem. Psychological Inquiry, 14, 1–26. https://doi.org/10.1207/S15327965PLI1401_01

Kline, R. B. (2005). Principles and practice of structural equa- tion modeling (2nd ed.). New York, NY: Guilford.

Kohn, A. (1993). Punished by rewards: The Trouble with gold stars, incentive plans, A’s, praise, and other bribes. New York, NY: Houghton Mifflin.

MacKinnon, D. P., Lockwood, C. M., & Williams, J. (2004). Confidence limits for the indirect effect: Distri- bution of the product and resampling methods. Multi- variate Behavioral Research, 39, 99–128. https://doi.org/ 10.1207/s15327906mbr3901_4

Mead, G. H. (1934). Mind, self, and society. Chicago, IL: University of Chicago Press. https://doi.org/10.7208/ chicago/9780226516608.001.0001

Meyer, W. U. (1992). Paradoxical effects of praise and criticism on perceived ability. European Review of Social Psychology, 3, 259–283. https://doi.org/10.1080/ 14792779243000087

Miller, P. J., Wang, S. H., Sandel, T., & Cho, G. E. (2002). Self-esteem as folk theory: A comparison of European American and Taiwanese mothers’ beliefs. Parenting: Science and Practice, 2, 209–239. https://doi.org/10. 1207/s15327922par0203_02

Millon, T. (1969). Modern psychopathology: A biosocial approach to maladaptive learning and functioning. Philadel- phia, PA: Saunders.

Morf, C. C., & Rhodewalt, F. (2001). Unraveling the para- doxes of narcissism: A dynamic self-regulatory process- ing model. Psychological Inquiry, 12, 177–196. https:// doi.org/10.1207/S15327965PLI1204_1

Mueller, C. M., & Dweck, C. S. (1998). Praise for intelli- gence can undermine children’s motivation and perfor- mance. Journal of Personality and Social Psychology, 75, 33–52. https://doi.org/10.1037/0022-3514.75.1.33

Muth�en, B. O. (2004). Mplus technical appendices. Los Angeles, CA: Muth�en & Muth�en.

Ng, F. F. Y., Pomerantz, E. M., & Lam, S. F. (2007). Euro- pean American and Chinese parents’ responses to chil- dren’s success and failure: Implications for children’s responses. Developmental Psychology, 43, 1239–1255. https://doi.org/10.1037/0012-1649.43.5.1239

O’Mara, A. J., Marsh, H. W., Craven, R. G., & Debus, R. L. (2006). Do self-concept interventions make a differ- ence? A synergistic blend of construct validation and meta-analysis. Educational Psychologist, 41, 181–206. https://doi.org/10.1207/s15326985ep4103_4

Orth, U., & Robins, R. W. (2014). The development of self-esteem. Current Directions in Psychological Science, 23, 381–387. https://doi.org/10.1177/0963721414547 414

Pomerantz, E. M., & Kempner, S. G. (2013). Mothers’ daily person and process praise: Implications for chil- dren’s theory of intelligence and motivation. Develop- mental Psychology, 49, 2040–2046. https://doi.org/10. 1037/a0031840

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton University Press.

Ruble, D. N., & Frey, K. S. (1991). Changing patterns of comparative behavior as skills are acquired: A func- tional model of self-evaluation. In J. Suls & T. A. Wills (Eds.), Social comparison: Contemporary theory and research (pp. 70–112). Hillsdale, NJ: Erlbaum.

Ryan, R. M. (1982). Control and information in the intrap- ersonal sphere: An extension of cognitive evaluation theory. Journal of Personality and Social Psychology, 43, 450–461. https://doi.org/10.1037/0022-3514.43.3.450

Sameroff, A. J., & MacKenzie, M. J. (2003). Research strategies for capturing transactional models of devel- opment: The limits of the possible. Development and Psy- chopathology, 15, 613–640. https://doi.org/10.1017/ S0954579403000312

Sherif, M., & Hovland, C. I. (1961). Social judgment: Assim- ilation and contrast effects in communication and attitude change. New Haven, CT: Yale University Press.

Sowislo, J. F., & Orth, U. (2013). Does low self-esteem predict depression and anxiety? A meta-analysis of

1808 Brummelman, Nelemans, Thomaes, and Orobio de Castro

longitudinal studies. Psychological Bulletin, 139, 213–240. https://doi.org/10.1037/a0028931

Talbot, J. C. (2009). The road to positive discipline: A parent’s guide. Los Angeles, CA: TNT.

Thomaes, S., & Brummelman, E. (2016). Narcissism. In D. Cicchetti (Ed.), Developmental psychopathology (Vol. 4, 3rd ed., pp. 679–725). Hoboken, NJ: Wiley.

Thomaes, S., Brummelman, E., Bushman, B. J., Reijntjes, A., & Orobio de Castro, B. (2016). Is it time to rethink the pervasiveness and nature of low self-esteem in children? Manuscript submitted for publication.

Thomaes, S., Bushman, B. J., Stegge, H., & Olthof, T. (2008). Trumping shame by blasts of noise: Narcissism, self-esteem, shame, and aggression in young adoles- cents. Child Development, 79, 1792–1801. https://doi. org/10.1111/j.1467-8624.2008.01226.x

Thomaes, S., Reijntjes, A., Orobio de Castro, B., Bushman, B. J., Poorthuis, A., & Telch, M. J. (2010). I like me if you like me: On the interpersonal modulation and regulation of preadolescents’ state self-esteem. Child Development, 81, 811–825. https://doi.org/10.1111/j. 1467-8624.2010.01435.x

Thomaes, S., Stegge, H., Bushman, B. J., Olthof, T., & Denissen, J. (2008). Development and validation of the Childhood Narcissism Scale. Journal of Personality Assessment, 90, 382–391. https://doi.org/10.1080/ 00223890802108162

Twenge, J. M., & Campbell, W. K. (2009). The narcissism epidemic: Living in the age of entitlement. New York, NY: Simon and Schuster.

Wechsler, D. (1991). Wechsler Intelligence Scale for Children (3rd ed.). San Antonio, TX: Psychological Corp.

Weiner, B., & Kukla, A. (1970). An attributional analysis of achievement motivation. Journal of Personality and Social Psychology, 15, 1–20. https://doi.org/10.1037/ h0029211

Wood, J. V., Perunovic, W. E., & Lee, J. W. (2009). Posi- tive self-statements: Power for some, peril for others. Psychological Science, 20, 860–866. https://doi.org/10. 1111/j.1467-9280.2009.02370.x

Young-Eisendrath, P. (2008). The self-esteem trap: Raising confident and compassionate kids in an age of self-impor- tance. New York, NY: Little, Brown.

Supporting Information

Additional supporting information may be found in the online version of this article at the publisher’s website:

Data S1. Specificity Analyses, Robustness Ana- lyses, and Moderation Analyses

Inflated Praise, Self-Esteem, and Narcissism 1809

This document is a scanned copy of a printed document. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material.