Introduction
The social feedback children receive from significant others has a pervasive influence on their social development (e.g., Bowlby
1980; Cole et al.
2001; Harter
2006). Whereas children who receive frequent positive information about themselves are likely to develop positive self-views, those who receive recurring negative feedback tend to develop or maintain less favorable self-views (e.g., Bowlby
1980; Harter
2006). Importantly, however, children are not merely passive recipients of the feedback they obtain from others. Rather, they seek out and prefer some types of feedback over others.
According to self-verification theory (Swann
1983,
1990; Swann et al.
2002), self-views play a central role in the kinds of social feedback people seek out. The theory asserts that people work to confirm their firmly held self-views, even when they are negative. That is, people holding negative self-views are preferentially (but not exclusively) drawn to negative feedback because such self-confirming feedback bolsters confidence in the accuracy of negative self-views (i.e., “epistemic” reasons) and serves to avoid the negative consequences resulting from others forming overly positive appraisals (i.e., pragmatic reasons).
Among adults, the linkage between negative self-views and feedback seeking has been examined extensively. A series of studies (e.g., Swann et al.
1992,
1992) has provided robust empirical support for self-verification theory’s assertion that negative self-views consistently lead people to prefer negative over positive self-relevant feedback. For instance, Swann and colleagues observed that college students holding negative self-views, compared to those holding positive self-views preferred friends who evaluated them negatively, were more inclined to seek negative than positive feedback from their roommates, and preferentially solicited unfavorable feedback from their dating partners (
Swann et al. 1992). Moreover, participants scoring high in depressive symptoms as well as clinically depressed adults were more likely than controls to choose interacting with someone who appraised them negatively over the opportunity to avoid the encounter (Giesler et al.
1996; Swann et al.
1992).
Interestingly, Swann et al. (
1992) also showed that after being faced with self-discrepant positive feedback—feedback that threatened the subjective validity of their negative self-views—dysphoric participants subsequently preferred to receive feedback about their limitations. In contrast, non-dysphoric participants faced with self-discrepant negative evaluations preferred to receive feedback about their strengths. It thus appears that after being presented with self-discrepant positive feedback, dysphorics compensated for threats to their self-views by displaying a preference for negative feedback that likely served to reaffirm their low self-esteem. Consistent with this argument, in a study examining the spontaneous verbalizations of adults as they chose interaction partners (Swann et al.
1992), participants with negative self-views chose unfavorable partners predominantly for the epistemic reasons put forth by self-verification theory (e.g., “I feel more at ease with someone who can judge me for what I am”).
Relative to adults, few studies have examined the linkage between negative self-views and feedback seeking in children. Using a sample of children that were psychiatric inpatients at an academic medical center, Joiner et al. (
1997) showed that as children’s depression level increased, so did their self-reported preferences for receiving negative feedback. These findings were obtained using a questionnaire assessing feedback preferences in four distinct self-relevant domains (i.e., social, intellectual, athletic, and physical attractiveness). Two studies using a comparable methodology with community samples (Borelli and Prinstein
2006; Cassidy et al.
2003a,
b) indicated that the tendency for children higher in depression and lower in global self-worth to prefer more negative feedback extends to non-clinical samples.
A similar link between children’s self-views and their selection of peers was observed by Cassidy et al. (
2003a) in a study conducted among third and seventh graders. Specifically, findings revealed that children reporting negative self-views on specific competence domains (e.g., sports) were significantly more likely than their counterparts to display an interest in meeting a peer who evaluated them negatively on that domain. Interestingly, results provided preliminary evidence to suggest that self-verification strivings may be more pronounced among third graders, relative to seventh graders.
Unfortunately, the peer selection and feedback seeking preferences of children holding negative self-views are likely to maintain and/or exacerbate their negative self-views. That is, to the extent that children holding negative self-views preferentially seek out negative feedback and peers who appraise them negatively, their negative self-views are likely further strengthened and internalized.
However, a recent study by Reijntjes et al. (
2007) examining the linkage between depressive symptoms and feedback preferences among pre-adolescents provided preliminary evidence that children holding negative self-views may not always be more inclined than their peers to eschew positive feedback and undermine opportunities to improve their self-views. In this study, participants were first presented with a self-congruent or a self-discrepant global evaluation outcome and then self-reported feedback preferences (positive versus negative) were assessed. After being faced with an unfavorable self-congruent outcome, children displaying higher levels of depressive symptoms showed a stronger preference for more negative feedback, relative to their peers. In contrast, after receiving the self-discrepant success outcome children high in depressive symptoms were as likely as their peers low in depression to express a significant preference for
positive feedback.
What may account for this interesting pattern of findings? We suggest that the motive to preferentially sample unfavorable feedback for the epistemic reasons specified by self-verification theory (i.e., corroborative feedback bolsters individuals’ confidence in the accuracy of their self-views and promotes intra-psychic coherence) may be weaker for children than for adults. It is only from late childhood that children become able to make global evaluations of themselves (e.g., “I am a worthless person”) that are at the core of negative self-views (e.g., Harter
2006). Hence, relative to adults, pre-adolescents’ negative self-views are less likely to be ingrained and crystallized into chronic states of self-doubt and self-derogation.
In line with this reasoning, a recent meta-analysis examining the stability of self-esteem across the life span showed that during late childhood and early adolescence self-esteem is more unstable than in any other developmental period (Trzesniewski et al.
2003). Because of this instability of children’s self-views, the self-discrepant success feedback may have temporarily suspended their negative self-views, which in turn may have attenuated efforts to verify their negative self-views with confirming negative feedback. Stated differently, we hypothesize that in a positive evaluative context children holding negative self-views are not strongly inclined to subsequently eschew positive feedback that is at odds with their negative self-views, because the favorable feedback is unlikely to engender the feeling that they may not know themselves well, thereby threatening their identity (feelings of “existential security”).
Although the findings of the Reijntjes et al. (
2007) study are at odds with predictions from self-verification theory, no firm conclusion can be drawn. First, feedback preferences were assessed using a forced-choice self-report item asking participants to indicate their preference for receiving favorable versus unfavorable feedback. However, self-reported feedback preferences may diverge from actually displayed feedback seeking preferences. For instance, according to the “immediate capitalization/later shifting” hypothesis advanced by Casbon et al. (
2005), immediately subsequent to receiving positive feedback individuals holding negative self-views may seek further positive feedback to foster the affective satisfaction evoked by the self-enhancing feedback and only later, after realizing that such feedback contradicts one’s self-views, shift to more negative feedback seeking. Several studies among adults have provided support for this argument (Swann et al.
1990; Hixon and Swann
1993), by showing that when people holding negative self-views are afforded more time to reflect on themselves, they are more likely to display a preference for negative self-verifying feedback. Hence, the dynamic social reality in which the feedback children receive may influence their subsequent feedback seeking is likely better captured by assessing actually obtained feedback over time.
Second, no effort was made to compare children holding strongly positive versus strongly negative self-views. Specifically, Swann has argued (e.g., Swann
1990; Swann et al.
1992) that the role of self-views should be examined by contrasting individuals whose self-views are strongly positive and strongly negative (e.g., those with scores in the top and bottom 20% of a large sample; see also Cassidy et al.
2003a,
b). However, by examining all participants, also those with intermediate levels of self-views, no extreme groups were contrasted. Moreover, although negative self-views are an important feature of depression (e.g., Beck
1967), a more direct index of children’s self-views is their sense of global self-worth (general self-esteem).
The present study was designed to test the hypothesis that the preference for negative feedback among children holding negative self-views will be significantly attenuated when their feedback seeking is assessed in the context of a social success experience, relative to a failure experience that primes their pre-existing negative self-views. Pre-adolescents were examined because self-esteem is least stable in this developmental period. To maximize the contrast between children holding positive versus negative self-views, following Cassidy et al. (
2003a,
b), children scoring in the upper and lower quintile on a widely used measure of general self-esteem comprised the positive versus negative self-view groups. Participants were first presented with an experimentally manipulated global peer evaluation outcome that was either positive or negative. Subsequently, an unobtrusive behavioral viewing time measure was employed to assess participants’ preference for positive versus negative self-relevant feedback during a fixed 5 min period. We expected that children holding negative self-views would display a significantly stronger interest in positive feedback when faced with the favorable evaluation outcome than they would after receiving the unfavorable evaluation outcome. In contrast, we expected that children holding positive self-views would display a robust preference for positive feedback across both outcomes.
Discussion
The major aim of the present study was to test the hypothesis that a social success experience leads to reduced self-verification strivings among pre-adolescents holding low levels of global self-esteem. In so doing, we went beyond previous work examining feedback seeking in children in two ways. First, unlike previous studies in which children indicated their preferences for receiving favorable versus unfavorable feedback, participants in the current study actually obtained feedback that was allegedly based on the personal information they had provided to their peer evaluators. Second, we employed an unobtrusive behavioral viewing time measure to address the limitation of previous studies’ sole reliance on self-reported feedback preferences.
As predicted, the feedback preferences of children holding negative self-views differed as a function of the valence of the preceding peer evaluation outcome. Specifically, whereas children holding positive self-views displayed a significant preference for positive over negative feedback across the two outcome conditions, children holding negative self-views evidenced a significantly stronger preference for positive feedback after receiving the positive self-discrepant evaluation outcome than after receiving the self-congruent negative evaluation outcome. In fact, after being faced with explicit favorable feedback, these children were as likely as their peers holding positive self-views to display a marked preference for positive feedback. This finding is at odds with a core postulate of self-verification theory stating that negative feedback seeking is driven by a need to verify negative self-views. Self-verification theory would therefore have predicted that after being faced with a self-discrepant favorable evaluation outcome children holding negative self-views display a preference for negative feedback to confirm their feelings of low self-worth.
Overall, the present results with regard to children holding negative self-views are at odds with self-verification theory, and fit better with the view that children holding negative self-views seek more negative feedback only in the context of a negative event that clearly matches their pre-existing unfavorable self-views. However, other studies with children (e.g., Cassidy et al.
2003a,
b) have shown that in the absence of a salient evaluative/performance event, children holding negative self-views report a general preference for negative self-verifying feedback. Taken together, contrary to research showing that among adults negative self-views consistently yield a preference for negative self-verifying feedback, it appears that in children the linkage between negative self-views and a preference for negative feedback is less robust.
One implication of the present findings is that providing low self-esteem children with a salient success experience may be an effective means to circumvent their typical inclination to pass up opportunities to be faced with disconfirming positive information about themselves that may serve to re-evaluate their self-view in a more positive light. Although research has shown that that at some level everyone desires favorable self-enhancing feedback (e.g., Taylor and Brown
1988), we acknowledge that some children holding negative self-views may engage in efforts to discount the positive feedback.
Several features of the present study deserve comment. First, only children’s feedback seeking behaviors were assessed, with no attention given to cognitive reactions such as perceptions of the accuracy of the feedback, or the reasons participants provide for their feedback preferences. Second, to maximize the credibility and ecological validity of the feedback seeking procedure, the provided feedback was a mixture of predominantly general feedback with more domain-specific social or physical feedback. Because most of the provided feedback was general in nature, perceived global self-worth served as the index of children’s self-views. However, previous work (e.g., Cassidy et al.
2003a,
b) has found differences with respect to specific versus global feedback seeking. Although the somewhat mixed feedback does not seem to compromise the validity of our main finding, namely that children low in global self-worth spend significantly less time viewing negative feedback after being faced with a positive versus a negative peer evaluation outcome, the mixing of general and specific feedback may have affected our results. For instance, it seems possible that a stronger match between the nature of the feedback and our measure of self-views would have yielded stronger differences between the two self-view groups. Third, it should be noted that all participants, including those holding negative self-views who were provided with a negative evaluation outcome, spent more time on positive than on negative feedback. These findings mirror those observed among adults (e.g., Swann et al.
1992). These authors note that “another consequence of the fact that people with negative self-views are motivated to attain favorable as well as unfavorable feedback is that they may sometimes display a
relative preference for unfavorable feedback” (p. 301). In the present study, children holding negative self-views faced with negative feedback also displayed only a relative preference for more negative feedback. Specifically, whereas their counterparts in the success condition spent significantly less time on negative than positive feedback, children in the failure condition spent as much time on negative feedback as they did on positive feedback. Finally, this investigation examined children’s feedback seeking after being provided a global peer evaluation outcome. Future research is needed to examine the extent to which the present findings extend to children’s feedback seeking subsequent to success and failure in other relevant domains (e.g., academics).
Notwithstanding these limitations, the findings from the present study contribute to the limited knowledge base on feedback seeking in children. Most importantly, our findings suggest that when faced with a salient success experience, children holding negative self-views may be apt to preferentially sample positive feedback. As discussed above, research has shown that people holding negative self-views are more likely to display a preference for negative self-verifying feedback when they have more time to reflect on the feedback received. Whereas children holding negative self-views were afforded ample time to realize that the positive feedback contradicted their negative self-views, they still displayed a significant preference for positive over negative feedback.
However, many questions remain. Below we offer several recommendations for advancing research. First, there is a need for more research with children examining self-verification’s postulate that those holding strongly positive or strongly negative self-views are preferentially drawn to feedback that confirms their self-conceptions because it engenders feelings of security and control. For instance, future studies could provide children with either self-discrepant or self-confirming feedback and assess their emotional, cognitive and physiological reactions. Second, to assess the effects of more versus less ingrained self-views, research is needed to examine if late adolescents are more likely than middle or early adolescents to display a motivated preference for self-verifying feedback. Finally, future studies should assess if feedback preferences are affected by the characteristics of the people who provide the feedback (“feedback source”). For instance, because children typically segregate into own-gender groups (Maccoby
1998), the gender of the feedback source may exert strong effects on feedback preferences.