See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/234168335

Self-Doubt and Self-Esteem: A Threat from within

Article in Personality and Social Psychology Bulletin · March 2002

DOI: 10.1177/0146167202286010

CITATIONS READS
64 13,408

3 authors, including:

Anthony Hermann Geoffrey J Leonardelli

Bradley University University of Toronto
31 PUBLICATIONS 1,084 CITATIONS 54 PUBLICATIONS 2,629 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Anthony Hermann on 22 March 2014.

The user has requested enhancement of the downloaded file.

PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN
Hermann et al. / SELF-DOUBT AND SELF-ESTEEM

Self-Doubt and Self-Esteem: A Threat From Within

Anthony D. Hermann
Geoffrey J. Leonardelli
Robert M. Arkin
The Ohio State University

The impact on self-esteem of activating self-doubt was investi- both subscales (self-doubt, concern with performance),
gated in three studies. Individuals with enduring high self- self-doubt and fear of failure inspires the expenditure of
doubt were expected to be more threatened by an experimental effort to ensure successful outcomes. While exhibiting
induction of self-doubt (modeled on the ease of retrieval para- quite different behaviors, self-handicappers and over-
digm) than individuals low in enduring self-doubt, and their achievers share the experience of self-doubt, which is
self-esteem was predicted to decline. The predictions were sup- thought to inspire their distinct, but related, coping
ported when self-esteem was measured postexperimentally (Experi- styles. In sum, the evidence suggests that people such as
ment 1) and when it was measured both pre- and postexperimentally self-handicappers and overachievers engage in behav-
(Experiment 2). There was no comparable loss in self-esteem for iors designed to protect themselves from failure, or at
individuals low in self-doubt. A third experiment explored the minimum the self-attributional implications of failure,
phenomenology of low-self-doubt individuals and replicated the when motivated by feelings of self-doubt about ability.
finding that their level of self-esteem was unaffected by the induc-
The Link Between Self-Doubt and Self-Esteem
tion designed to produce doubt.
A typical assumption is that these protective behaviors
are linked to notions of competence and self-worth. As
Uncertainty about one’s ability in performance situa- Jones and Berglas (1978) put it, “Each [the handicapper
and the overachiever] is fearful that failure will implicate
tions suggests the prospect of failure and can prompt
defensive, protective behavior. Self-handicapping is a competence. Each has an abnormal investment in the
good example of this class of self-protective actions (e.g., question of self-worth” (p. 205). More recently, research
Arkin & Baumgardner, 1985; Arkin & Oleson, 1998; has shown that engaging in a protective behavior, such as
Berglas & Jones, 1978, Higgins, 1990). In the landmark self-handicapping, appears to be associated with self-
study, Berglas and Jones (1978) found that individuals esteem maintenance. For instance, after a failure, partic-
who faced doubts about their ability sought a handicap ipants who claimed self-handicaps had higher self-esteem
to their performance, protecting themselves from the than those who did not claim self-handicaps (Feick &
attributional implication that a failure, if it occurred, Rhodewalt, 1997). Thus, feelings of self-worth do appear
would reflect a clear lack of ability.
More recently, it has been found that some individuals Authors’ Note: The authors acknowledge with much gratitude the ef-
who harbor doubts about their abilities, but who also forts of several research assistants who helped with various portions of
have particularly strong concerns about performing suc- the data collection and analysis: Dan Miller, Bill Preston, Tiffany
Wheeler, Yolonda Haynes, Jarrod Williams, and Ann Marie Altman. Ap-
cessfully, adopt a related but opposite strategy of over- preciation is also extended to Michael Walker for statistical advice, to
achieving (Arkin & Oleson, 1998; Oleson, Poehlmann, Zakary Tormala for commenting on a previous draft of this article, and
Yost, Lynch, & Arkin, 2000). The Subjective Over- to the members of the Arkin Lab group for their thoughtful commen-
achievement Scale (SOS) was recently developed (Oleson tary. Communications should be addressed to Anthony Hermann, De-
et al., 2000) to assess both individual differences in self- partment of Psychology, Ohio State University, 1885 Neil Ave Mall,
Columbus, OH 43210-1222; e-mail: [email protected], or to Rob-
doubt and concern with performance outcomes. The ert M. Arkin, Department of Psychology, Ohio State University, 1885
Self-Doubt Subscale was designed to “capture a general Neil Ave Mall, Columbus, OH 43210-1222; e-mail: [email protected].
sense of feeling uncertain about one’s competence” PSPB, Vol. 28 No. 3, March 2002 395-408
(p. 500). For subjective overachievers, who score high on © 2002 by the Society for Personality and Social Psychology, Inc.

395
396 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

to be at stake when facing the prospect of failure (if it can A threat that is self-generated and strictly internal
be attributed to oneself). More generally, this research is presents a different set of alternatives for the self-doubter.
consistent with conclusions made elsewhere; namely, Introspection (Duval & Wicklund, 1972), recall of past
that as doubt associated with one’s important abilities experiences (e.g., Bem, 1967), prediction of future per-
increases, global self-esteem decreases (e.g., Pelham, formance (Bandura, 1997), meta-cognitions (e.g., Mischel,
1991; Pelham & Swann, 1989). 1998), and other self-generated cognitions that elicit
Self-doubt and self-esteem are moderately (and nega- self-doubt cannot be deflected behaviorally. Thus, the
tively) correlated (e.g., correlations range from –.44 to – usual protective steps taken by the individual high in self-
.68; Oleson et al., 2000). However, it is important to keep doubt are rendered ineffective and the threat cannot be
in mind that although low self-esteem tends to accom- dismissed and ultimately may affect self-esteem.
pany self-doubt, the two are conceptually distinct: Self- Meta-cognitions, those judgments we make about our
doubt refers to how certain a person feels about impor- judgments (Jost, Kruglanski, & Nelson, 1998), are capa-
tant abilities, whereas self-esteem refers to a global evalu- ble of having a potent impact on self-evaluation. To illus-
ation of oneself as a person. For example, an individual trate, consider a person’s certainty in his or her self-eval-
can have a negative global evaluation of his or herself uation as a musician. If heroic effort is spent to generate
(e.g., low self-esteem) but either be certain or uncertain support for that self-evaluation, and the effort is salient
about specific and global competencies. Threat to one“s and weighted heavily in one’s judgment, it is plausible
basic self-esteem should emerge when one has uncer- that features of the self-evaluation as a musician (e.g., tal-
tainty about abilities that are important and favorably ented, gifted, enjoying potential) may be undermined
evaluated (e.g., Pelham, 1991; Pelham & Swann, 1989). by the meta-cognitive cues. The present research is con-
Thus, the evidence suggests that feelings of self-doubt cerned with such meta-cognitive sources of information,
pose a threat to self-esteem. Indeed, it seems likely that particularly those associated with self-reflection while
those chronically high in self-doubt, as evidenced by performing a task. The prediction is that because it may
their greater propensity to engage in self-protective behav- be difficult or impossible to set aside or alleviate feelings
ior, are especially likely to interpret self-doubt as threat- of doubt stimulated by meta-cognitive cues, self-esteem
ening. It also follows that unless steps are taken to set is likely to be damaged or decline, at least temporarily,
aside or alleviate the feelings of doubt, self-esteem may when meta-cognitive cues to feelings of self-doubt are
be damaged and decline. The question posed here is elicited.
whether this threat to self-esteem, and any ensuing dam-
age and decline in self-esteem, is actually greater for indi- Retrieval Difficulty: An Internal Threat to Self-Esteem?
viduals who are enduringly high in chronic feelings of One particularly subtle and compelling meta-cogni-
self-doubt about their competence. tive cue, recall difficulty, has recently been investigated
Internal and External Threats to Self-Esteem by Schwarz and his colleagues (Schwarz, 1998; Schwarz,
Bless, Strack, Klumpp, Rittenauer-Schatka, & Simons
It is important here to distinguish globally between 1991). The essential finding is that the experience of
two sources of threat to self-esteem. Often, individuals recall difficulty exerts an effect on self-judgments even
see threats to their self-worth as originating from exter- when the content of the information recalled exerts its
nal sources. The prospect of a public failure is prototypical: own influence in an opposite, contrary direction. Spe-
it looms as a threat because its implications signify not cifically, Schwarz et al. (1991) used a clever methodology
only to one’s self but also to others that the individual is to investigate the psychological processes underlying the
incompetent (Jones, 1989). Both self-handicapping and use of the availability heuristic (Tversky & Kahneman,
overachievement, and other protective mechanisms (e.g., 1973). The availability heuristic is a mental shortcut peo-
withdrawal from the situation), can deflect the signifying ple often use in which a judgment is based on “the ease
implications of the outcome and protect and maintain with which instances or associations come to mind”
self-esteem. One long-term cost, however, is an enduring (Schwarz et al., 1991, p. 208). Schwarz et al. (1991) noted
feeling of self-doubt. Doubts sustained by such protec- that the literature has been unclear whether “ease” refers
tive steps set aside the threat but also undermine the to the number of instances available in memory or to the
diagnosticity of one’s performance. Similarly, the shy perceived difficulty of recalling them and argued that
individual can avoid social rejection by making no over- the latter matters most. They maintained that this is par-
tures but remains enduringly dubious about his or her ticularly so when the two types of information are in con-
social acceptability. However, this is the cost that self- flict. In the study, participants rated themselves less asser-
doubters seem willing to absorb to ensure that the imme- tively after they were instructed to recall 12 examples of
diate, short-term threat to self-esteem is neutralized. their own assertive behavior than after recalling 6 exam-
 Hermann et al. / SELF-DOUBT AND SELF-ESTEEM 397

ples. Conversely, participants rated themselves as more Method

assertive after they were instructed to recall 12 examples
PARTICIPANTS
of unassertive behavior than after recalling 6. Partici-
pants in these experiments tended to neglect a great The study included 123 students who participated in a
deal of relevant information in their memories in favor computer administered study titled “Confidence Training”
of paying attention to a meta-cognitive cue, the difficulty and received partial credit in their introductory psychol-
retrieving that information. ogy class. Data from three participants were discarded
Applied to self-doubt, the Schwarz et al. (1991) find- because they failed to follow instructions.
ing suggests that retrieval of instances of self-confidence PROCEDURE
should, ironically, lead individuals to feel less confident Participants were randomly assigned to recall either
(more doubtful) when they have difficulty in retrieving two or eight examples of self-confidence. Afterward,
the examples. This effect of the meta-cognitive cue might they completed self-report measures. All materials were
be equally potent for individuals predisposed to experi- presented via a software program designed to conduct
ence self-doubt and those not so predisposed. However, psychological experiments (Jarvis, 1998). In each of 15
consistent with our theorizing, an alternative hypothesis sessions, between 6 and 12 participants sat at individual
is that individuals with high levels of enduring self-doubt computer stations. Participants were informed that the
might be hypersensitive to retrieval difficulty compared purpose of the experiment was to develop materials for
to those low in self-doubt and thus find it more threaten- use in a counseling program designed to train clients to
ing. Therefore, we expected that after an experience of build self-confidence. Ostensibly to help develop realis-
retrieval difficulty, the self-esteem of individuals high in tic training scenarios, participants were asked to “list two
self-doubt would decrease, whereas the self-esteem of (eight) events in your life, which led you to feel confi-
individuals low in self-doubt would not. dent about your ability to perform in some important
area of your life.” A screen with either two or eight text
EXPERIMENT 1
boxes then appeared and participants were instructed to
type a brief description of no more than 250 characters
The following experiment was designed to test this for each event.1
hypothesis and consisted of a retrieval condition (two Next, students were asked to complete some general
examples, eight examples) × self-doubt (continuous) questions, purportedly to explore students’ interest in
between-participants design. An interaction was predicted: the training program. First, they were prompted to rate
The self-esteem of individuals high in self-doubt was themselves on confidence and uncertainty using 10-
expected to be lower after recalling eight examples of point scales (e.g., 1 = not at all uncertain, 10 = extremely
self-confidence than after two examples, whereas the uncertain). Following Schwarz et al.’s (1991) procedure,
self-esteem of individuals low in self-doubt was expected these ratings were included to verify that retrieval diffi-
to be unaffected by the number of examples recalled. culty was influencing ratings of confidence. Participants
Such an interaction would be consistent with the idea then completed the Rosenberg Self-Esteem Scale (RSE)
that individuals high in self-doubt are threatened by (Rosenberg, 1965), the Self-Doubt Subscale of the Sub-
retrieval difficulty but individuals low in self-doubt are jective Overachievement Scale (SOS-SD) (Oleson et al.,
not (even when the perceptions of recall difficulty are 2000), and a manipulation check that involved rating
the same for individuals high and low in self-doubt). how difficult it was to generate the requested number of
That is, all participants, regardless of their level of self- examples (1 = not difficult at all, 10 = extremely difficult).
doubt, should find the eight-example task to be more dif- Finally, participants were thoroughly debriefed.
ficult than the two-example task, even though only indi- RSE. This 10-item scale (Rosenberg, 1965) is designed
viduals high in self-doubt will be threatened by it. How- to measure global self-evaluation (e.g., “On the whole, I
ever, a second possible explanation for the interaction am satisfied with myself”). Participants responded to
on self-esteem exists: Perhaps retrieval condition could these items on a 6-point Likert scale (1 = disagree very
affect the perceptions of difficulty for individuals high in much, 6 = agree very much). After reversing the scores for
self-doubt but not for individuals low in self-doubt. Here, the negatively worded items, ratings were summed yield-
a Retrieval Condition × Self-Doubt interaction would be ing a potential range of scores from 10 (very low self-
evident not only on self-esteem but also on difficulty. The esteem) to 60 (very high self-esteem). Internal consistency
following experiment was designed to test both hypothe- was strong (α = .90).
ses by including measures of self-esteem and perceived Self-Doubt Subscale. This eight-item subscale of the
difficulty. SOS-SD (Oleson et al., 2000) is designed to measure
398 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

chronic individual differences in self-doubt about one’s no difference in retrieval condition for low-self-doubt
ability to perform important tasks (e.g., “As I begin an participants (β = .02, p = .79) but a reliable difference
important activity, I usually feel confident in the likely among high-self-doubt participants (β = –.28, p < .01).
outcome”). Participants responded to these items on the After recalling eight examples, the self-esteem of individ-
same 6-point scale used for the RSE. After reversing the uals high in self-doubt was lower than after recalling two
negatively worded items, ratings were summed yielding a examples.
potential range of scores from 8 (very low self-doubt) to 48 Difficulty ratings and self-esteem. To verify that perceived
(very high self-doubt). The scale exhibited adequate inter- difficulty was associated with the observed changes in
nal consistency (α = .84). high-self-doubt participants’ self-esteem, we performed
Results similar analyses on participants’ self-esteem scores but
substituted difficulty ratings for the retrieval condition
Hierarchical regression analyses were conducted to variable. Self-esteem scores were again submitted to a
determine whether scores on the Self-Doubt Subscale hierarchical regression analysis, but with the following
moderated the impact of retrieval condition (i.e., the predictors: self-doubt, difficulty, and the Difficulty × Self-
number of self-confidence examples recalled) on self- Doubt interaction term. Analysis yielded a significant
esteem. Following standard procedures (Cohen & Cohen, self-doubt main effect (β = –.72, p < .01), which was quali-
1983), self-doubt scores and retrieval condition were fied by a significant Difficulty × Self-Doubt interaction
entered first (to test for main effects) and the product of on self-esteem (β = –.26, p = .03). Simple slope analysis
the two was entered next (to test for interaction effects). (Aiken & West, 1991) indicated that at low self-doubt,
To simplify interpretability of the regression analysis perceived difficulty was uncorrelated with self-esteem
(Aiken & West, 1991), self-doubt scores were centered (β = .06, p = .50) but that at high self-doubt, perceived dif-
(i.e., the sample mean was set equal to zero). Retrieval ficulty was negatively correlated with self-esteem (β = –.19,
condition was dummy-coded: 0 for two examples and 1 p = .02). As shown in the right panel of Figure 1, much
for eight examples (Aiken & West, 1991). The interac- like retrieval condition, perceived difficulty only had an
tions were plotted using the predicted means for each impact on the self-esteem of those high in self-doubt;
retrieval condition at levels of self-doubt one standard those high-self-doubt participants who reported high
deviation above and below the mean of the Self-Doubt difficulty also reported lower self-esteem.
Scale for high and low self-doubt participants, respec-
tively. All analyses were conducted and all graphs were SELF-RATINGS
constructed in this manner, unless specified otherwise. Confidence. Analyses of participants’ self-ratings on the
MANIPULATION CHECK dimension of confidence yielded only a main effect of
self-doubt (β = –.57, p < .001) and no interaction. As par-
Analyses of participants’ ratings of retrieval difficulty
ticipants’ self-doubt scores increased, they rated them-
yielded a predicted main effect of retrieval condition (β =
selves as less confident.
.30, p < .001), as well as a main effect of self-doubt (β =
.37, p < .001), but no interaction (β = .03, p = .80). As Uncertainty. Analyses of participants’ self-ratings of
expected, recalling eight examples of past self-confi- uncertainty, however, yielded a main effect of self-doubt
dence was judged as more difficult (M = 4.87, SD = 2.59) (β = .34, p < .001) and a marginally significant interaction
than retrieving two examples (M = 3.37, SD = 2.15). In of self-doubt and retrieval condition (β = .24, p = .08).
addition, as level of chronic self-doubt increased, so did Simple effect analyses indicated that at low self-doubt,
perceived difficulty of retrieving examples. uncertainty decreased as number of examples increased
(β = –.12, p = .35), but at high self-doubt, uncertainty
SELF-ESTEEM
increased as number of examples increased (β = .20, p =
Analyses of participants’ scores on the RSE revealed .12), although neither simple effect was significant. Sim-
main effects of self-doubt (β = –.76, p < .001) and retrieval ple slope analyses revealed that self-doubt scores pre-
condition (β = –.13, p = .03), but these effects were quali- dicted uncertainty self-ratings in the eight-example con-
fied by the predicted interaction (β = –.23, p = .01). As dition (β = .46, p < .001) but not in the two-example
illustrated in the first panel of Figure 1, retrieval condi- condition (β = .15, p > .25). Uncertainty increased as self-
tion had little effect on the self-esteem scores of those doubt increased but only after participants recalled eight
participants with relatively low self-doubt but those high examples of past self-confidence (see Figure 2).
in self-doubt reported lower self-esteem after having to
CONTENT QUALITY: AN ALTERNATIVE EXPLANATION
recall eight examples. Simple effect analyses conducted
to assess retrieval condition differences for those high Although it is clear that difficulty was associated with
and low in self-doubt separately (Aiken & West, 1991) the effects found on self-esteem, an alternative explana-
confirmed this interpretation; these analyses revealed tion is possible. Perhaps the Retrieval Condition × Self-
 Hermann et al. / SELF-DOUBT AND SELF-ESTEEM 399

Figure 1 Predicted means of self-esteem as a function of retrieval condition and self-doubt (left panel) and difficulty and self-doubt (right panel):
Experiment 1.

result, each judge’s ratings were submitted separately to

regression analysis. However, because analysis revealed
no differences on judges’ ratings, only the analysis of
average judges’ ratings will be presented to simplify
presentation.
Content quality analyses. First, regression analyses were
conducted to determine if self-doubt and retrieval con-
dition interacted to predict the judges’ ratings. Analyses
yielded only a significant retrieval condition main effect
(β = –.33, p < .001). The judges’ rating of confidence in
the last two examples recalled in the eight-example con-
dition (M = 6.64, SD = 1.17) was lower than the content
Figure 2 Predicted means of the uncertainty ratings as a function of quality ratings of the two examples recalled in the two-
retrieval condition and self-doubt: Experiment 1.
example condition (M = 7.31, SD = .67).
Because retrieval condition had an impact on the
Doubt interaction could be explained by differences in judges’ perceptions of the examples’ quality, a new set of
the quality of the content recalled (Schwarz et al., 1991). analyses was conducted to investigate whether the qual-
Individuals high in self-doubt who recalled eight exam- ity of the examples also interacted with self-doubt to pre-
ples may have reported relatively lower self-esteem because dict participant’s self-esteem. Thus, self-esteem scores
were again submitted to a hierarchical regression analy-
their examples were less convincing than those they pro-
sis, but with the following predictors: self-doubt, judge’s
duced in the two-example condition than the examples
content quality ratings, and the Content Quality × Self-
produced by low-self-doubt participants.
Doubt interaction term. Analyses revealed only a main
Coding for content quality. Schwarz et al. (1991) ruled effect of self-doubt (β = –.75, p < .001). The interaction
out the possibility that decreasing quality of example term was not significant (β = .01, p = .82). Recalling more
content could account for the effect by having inde- examples led to poorer examples, but this decline in
pendent judges rate the content quality of the last two quality due to retrieval condition was not moderated by
examples (p. 198). Following Schwarz et al.’s procedure, self-doubt. Thus, although recalling more examples led
this alternative explanation was examined by first having to both higher difficulty ratings and poorer examples for
two independent judges rate the last two examples gen- all participants, the drop in self-esteem observed in high-
erated by all of our participants. Specifically, judges self-doubt participants was associated only with the per-
rated the events on “the level of confidence about abili- ceived difficulty of the task.
ties each event exhibits” (i.e., the two examples in the
Discussion
two-example condition and the last two examples in the
eight-example condition) using a 9-point scale (1 = not at Participants high in self-doubt appeared to have been
all confident, 9 = extremely confident). Interjudge reliability particularly sensitive to recalling eight examples of self-
was much lower than expected, r(119) = .54, p = .001. As a confidence. They appear to have taken the retrieval diffi-
400 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

culty they experienced to heart, and their self-esteem bility that the observed interaction is the result of an
dropped. This, of course, transpired despite the fact that increase in self-esteem for those recalling two examples
they were engaged in an activity that ostensibly could instead of a decrease in self-esteem for high-self-doubt
have boosted their self-regard. Under the same condi- participants recalling eight examples. Recalling two exam-
tions, however, participants low in self-doubt appeared ples may affirm self-worth for high-self-doubt individu-
not to generalize any difficulty they may have experi- als, but recalling eight examples may induce multiple
enced with the retrieval task to an evaluation of their processes, such as self-affirmation and difficulty in
core self, even though they reported that it was harder to retrieval, which cancel each other out, resulting in no
recall eight than two examples of confidence. Indeed, change in self-worth. In Experiment 2, we sought to rep-
there was no indication that low-self-doubt participants licate our basic findings and to address this alternative
were adversely affected by recalling eight examples at all. interpretation. The RSE was, therefore, administered in
Altogether, the evidence is consistent with the hypoth- Experiment 2 both before and after the manipulation to
esis that people high in self-doubt are more threatened assess change in self-esteem and shed additional light on
by retrieval difficulty than are individuals low in self- our findings.
doubt. The evidence on self-esteem and perceived diffi-
culty was inconsistent with the two alternative hypothe- EXPERIMENT 2
ses. First, the interaction on self-esteem was inconsistent
This experiment consisted of a retrieval condition (2
with the notion that the retrieval condition would influ-
examples or 12 examples) × self-doubt (continuous)
ence the self-esteem of all individuals, regardless of level
between-participants design. We predicted that individ-
of self-doubt. Second, all individuals reported relatively
uals high in self-doubt would show no change in their
greater difficulty after recalling eight than two examples,
self-esteem after recalling two instances of their confi-
and this effect was not moderated by self-doubt. Thus,
dence but would experience a decrease in their self-
this retrieval condition main effect is inconsistent with
esteem after recalling 12 instances of their confidence.
the notion that the retrieval condition produced the
However, because participants low in self-doubt are not
experience of difficulty only for participants high in self-
threatened by issues regarding their level of compe-
doubt. Together, then, the evidence on these measures tence, we predicted that their self-esteem would not
supports the idea that participants high in self-doubt drop even in the face of the subjective experience of
find self-doubt more threatening than participants low retrieval difficulty.
in self-doubt.
Method
ALTERNATIVE EXPLANATIONS
PARTICIPANTS
Analysis, however, also revealed a self-doubt main The study originally included 122 students but 1 indi-
effect on difficulty; the task was more difficult for individ- vidual was dropped for completing the materials incor-
uals high in self-doubt than for individuals low in self- rectly (resulting N = 121).
doubt. Plotting the predicted means on the difficulty rat-
ings revealed that low-self-doubt participants, on aver- PROCEDURE
age, rated recalling eight examples well below the mid- The procedure was identical to that of Experiment 1
point of the 10-point scale (M = 3.8). This main effect with the following five exceptions. First, the SOS-SD (β =
suggests an alternative explanation for the Retrieval .86) was administered (as part of a survey supposedly
Condition × Self-Doubt interaction on self-esteem. Per- given for another researcher) shortly before the experi-
haps participants low in self-doubt did not experience a mental variable was manipulated. This provided for a
sufficient level of difficulty to affect their judgments and, measure of chronic self-doubt that was truly independ-
in turn, their self-esteem was not affected. To address this ent of the experimental manipulation. Second, the RSE
concern in the second study, the difficulty of the “diffi- was administered both before (as part of the same sur-
cult condition” was raised by increasing the number of vey; β = .90) and after (β = .91) the induction, and third,
examples participants were asked to recall from 8 to 12. the self-ratings of uncertainty and confidence were elim-
In this way, we could examine whether a more difficult inated. This provided the opportunity to analyze self-
task would produce a similar drop in self-esteem for low- esteem change in the most direct and maximally sensi-
self-doubt participants. tive way and to determine under which conditions self-
In addition, attempts were taken in the second experi- doubt was associated with such change. Fourth, partici-
ment to clarify the nature of the interaction. We argue pants were asked to recall 12 examples of self-confidence
that recalling eight examples of confidence for high-self- rather than 8 to increase the subjective experience of dif-
doubt participants resulted in a drop in their self-esteem. ficulty in that condition, especially for the low-self-doubt
It is important to note, however, that it remains a possi- participants. Last, the quality of the examples was assessed
 Hermann et al. / SELF-DOUBT AND SELF-ESTEEM 401

not only by independent judges but also by the partici- condition on self-esteem stems from changes in self-
pants themselves. This enabled us to determine whether esteem subsequent to recalling 12 examples of past
the participants’ subjective ratings of quality, as well as confidence.
the objective quality, of the examples were associated
Effects unique to self-doubt? Although we had already
with self-doubt and changes in self-esteem.
controlled for preexperimental self-esteem scores when
Results we calculated our change scores, because measures of
self-doubt and self-esteem are correlated, it is possible
MANIPULATION CHECK that the effects on postexperimental self-esteem are
As in Experiment 1, hierarchical regression analysis explained equally well by participants’ preexperimental
was used to determine the impact of self-doubt and self-esteem scores as by their level of self-doubt. To verify
retrieval condition on the manipulation check and our that the Retrieval Condition × Self-Doubt interaction on
dependent measure. Analyses of participants’ rating of posttest self-esteem was associated uniquely with self-
retrieval difficulty yielded a main effect of retrieval con- doubt, we created a new individual difference predictor.
dition (β = .31, p < .001) and a main effect of self-doubt (β = We retained the residuals when predicting SOS-SD scores
.18, p = .04) but no interaction (β = .17, p = .18). On aver- with the RSE (which represents the unique variance of
age, retrieving 12 examples of past self-confidence was self-doubt: USD) and then submitted the variables to our
perceived as more difficult (M = 6.5, SD = 2.2) than standard set of regression analyses to determine whether
retrieving 2 examples (M = 5.1, SD = 2.5). As in Experi- USD interacted with retrieval condition to predict post-
ment 1, self-doubt had an independent impact on diffi- experimental self-esteem. Analyses for USD revealed
culty ratings; as level of chronic self-doubt increased, so only the predicted interaction (β = –.30, p = .01), indicat-
did perceived difficulty of retrieving examples. ing that the effect holds for self-doubt with the variance
associated with self-esteem partialed out.
SELF-ESTEEM CHANGE

RSE pretest scores were subtracted from RSE posttest Difficulty ratings and self-esteem. As with Experiment 1,
scores, and this difference score was used as a measure of we explored the role of perceived difficulty. In Experi-
self-esteem change. The difference score was submitted ment 1, an interaction between difficulty and self-doubt
to the same hierarchical regression analyses as were the indicated that the self-esteem of individuals high in self-
other dependent measures. Analysis revealed a margin- doubt decreased as difficulty increased but that difficulty
ally significant main effect of self-doubt (β = –.30, p = was unrelated to self-esteem of individuals low in self-
.06). As self-doubt increased, self-esteem decreased. How- doubt. To determine whether this interaction was repli-
ever, the predicted interaction of retrieval condition and cated, self-esteem difference scores were submitted to
self-doubt qualified the effect (β = –.32, p = .02). The pre- analysis, with difficulty and self-doubt as main effect pre-
dicted means are plotted in the left panel of Figure 3.2 dictors and Difficulty × Self-Doubt as the interaction pre-
For individuals high in self-doubt, retrieval condition dictor. The predicted means presented in the second
produced a marginally significant difference (β = –.24, panel of Figure 3 indicate a pattern similar to the Retrieval
p = .06). As is evident in the left panel of Figure 3, individ- Condition × Self-Doubt interaction on self-esteem change.
uals high in self-doubt reported lower self-esteem after At low self-doubt, self-esteem increased as difficulty
retrieving 12 examples of past confidence than after increased; at high self-doubt, self-esteem decreased as
retrieving 2. This difference was reversed for individuals difficulty increased. Analysis revealed, however, that the
low in self-doubt who reported marginally higher self- interaction was not significant (β = –.10, p = .28).
esteem after retrieving 12 examples than after 2 (β = .19, To gain more power in detecting the effects of self-
p = .12). doubt and perceived difficulty on participants’ self-esteem,
Of primary interest, however, was whether self-esteem the data from Experiments 1 and 2 were combined and
changed from baseline (i.e., if change was different from reanalyzed. A dichotomous study factor was included in
zero). After recalling two examples, no self-esteem change the analysis to examine whether some difference other
was evident for individuals either high or low in self- than statistical power (Cohen, 1988) could account for
doubt (ts < .73, ps > .46). After recalling 12 examples, the difference between Experiments 1 and 2. Posttest
however, individuals high in self-doubt experienced a self-esteem scores were submitted to analysis in a full-
decrease in self-esteem, and it was significantly different factorial three-way hierarchical regression analysis, with
from zero, t(116) = –1.95, p = .05. Furthermore, individu- perceived difficulty and self-doubt as continuous between-
als low in self-doubt experienced an increase in self- participant predictors and the study factor as a categori-
esteem after recalling many examples, and this increase cal predictor. Self-esteem scores, instead of difference
was also different from zero, t(116) = 2.67, p < .01. scores, were analyzed because only posttest self-esteem
Clearly, the interaction between self-doubt and retrieval scores were collected in the first experiment.
402 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

Figure 3 Predicted means of self-esteem change, adjusted for pretest self-esteem, as a function of retrieval condition and self-doubt (left panel)
and difficulty and self-doubt (right panel): Experiment 2.

Analysis yielded a significant self-doubt main effect this study (i.e., using 12 examples instead of 8). Given
(β = –.74, p < .001), which was qualified by a Difficulty × this marginal interaction, it seems more likely that in this
Self-Doubt interaction (β = –.10, p = .02). Simple effects study, poorer examples in the difficult condition may
tests indicated that at low self-doubt, difficulty was have been responsible for the interaction between
uncorrelated with self-esteem (β = .06, p = .31); however, retrieval condition and self-doubt on self-esteem
at high self-doubt, difficulty was negatively correlated change. Support for this alternative explanation
with self-esteem (β = –.13, p = .04). No significant effect requires that the judges’ ratings of quality be positively
of the study factor was evident, whether alone or as a correlated with self-esteem change if content quality is
moderator of some other factor. truly accounting for the Retrieval Condition × Self-
CONTENT QUALITY Doubt interaction on self-esteem change. However, con-
tent quality was uncorrelated with self-esteem change,
Judges’ ratings. In Experiment 1, judges’ ratings of r(121) = .09, p = . 33. Thus, judges’ ratings of content
quality did not interact with self-doubt to predict partici- quality could not account for the Retrieval Condition ×
pants’ self-esteem. For this study, we again examined Self-Doubt interaction.
judges’ perceptions of the examples’ quality and fol-
lowed the same procedures to do so. Interjudge reliabil- Participants’ ratings. In this study, in addition to the
ity was higher than in the first study, r(121) = .77, p < .001; judges’ ratings after the fact, participants also rated their
the two judges’ ratings were thus averaged together to own examples. Specifically, participants rated the level
create one measure. Analyses revealed, as in Experiment of confidence they experience in the last two examples
1, a marginally significant main effect of retrieval condi- they generated using a 9-point scale (1 = not at all confi-
tion (β = –.17, p = .07). The judges rated the last two items dent, 9 = extremely confident) after completing the depend-
in the 12-example condition as exhibiting lower confi- ent measures. There was some evidence that judges’ and
dence (M = 6.56, SD = 1.29) than the 2 items in the two- participants’ ratings of quality differed; correlation of
example condition (M = 7.00, SD = 1.37). This main the two quality ratings indicated only a moderate rela-
effect was, however, qualified by a marginally significant tionship between the two ratings, r(121) = .31, p < .01.
interaction of self-doubt and retrieval condition (β = – Thus, it is possible that subjective ratings of quality would
.23, p = .09). Simple effects tests revealed that for individ- be associated with participants’ change in self-esteem,
uals high in self-doubt, judges’ ratings of confidence where judges’ ratings were not.
decreased as number of examples recalled increased (β Analysis of participants’ ratings of quality yielded only
= –.32, p = .01); however, for individuals low in self-doubt, a main effect of self-doubt (β = –.23, p = .01). Regardless
judges’ ratings of confidence were not associated with of the retrieval condition participants were in, the per-
number of examples recalled (β = –.01, p = .94). ceived quality of the examples recalled decreased as self-
This interaction on judges’ content quality ratings was doubt increased. No other effects emerged. Thus, partic-
not observed in Experiment 1 but may be the result of a ipants’ perceptions of quality could not account for the
harder task used in the “difficult” retrieval condition in interaction between retrieval condition and self-doubt.
 Hermann et al. / SELF-DOUBT AND SELF-ESTEEM 403

Discussion those low in self-doubt were not threatened by the

manipulation.
As observed in Experiment 1, participants high in self-
However, a nagging possibility remains that loss of
doubt reported lower self-esteem after recalling a rela-
self-esteem after difficulty recalling examples of self-con-
tively difficult number of examples of past self-confi-
fidence may be a universal sort of phenomenon, and a
dence. In addition, this effect was observed to be an
retrieval task of sufficient difficulty could induce individ-
actual decrease in self-esteem, clarifying the nature of
uals low in self-doubt to feel less confident and thus
differences observed among high-self-doubt participants
experience decreased self-worth. The predicted means
in Experiment 1. Rather than receiving a boost from
for the difficulty ratings made by participants high and
recalling 2 examples, those high in self-doubt suffered a
low in self-doubt suggest that participants high in self-
loss in self-regard after recalling 12. Furthermore, regres-
doubt found recalling 12 examples to be very difficult
sion analyses indicated that this drop in self-esteem was
(M = 7.3), whereas those low in self-doubt still rated the
not the result of a decline in the quality of the examples
task below the midpoint of the 10-point scale (M = 5.3). If
high-self-doubt participants recalled; rather, the interac-
this task was made more difficult (i.e., if the number of
tion of difficulty and self-doubt on self-esteem substanti-
examples of individuals had to recall was increased),
ates the notion that individuals high in self-doubt are
then individuals low in self-doubt may begin to experi-
threatened by difficulty produced by recalling examples
of one’s confidence. ence retrieval difficulty, feel their level of self-confidence
In sharp contrast to high-self-doubt participants, indi- drop, and experience a self-esteem drop as well.
viduals low in self-doubt reported higher self-esteem At the other extreme, individuals low in self-doubt
after recalling 12 examples of their confidence. More- might be immune to the effects of retrieval difficulty on
over, recalling a large number of examples served to con- their self-concept. When retrieving examples of self-con-
solidate and boost their already positive self-regard. This fidence, they may always give greater credence to the
boost for low-self-doubt participants was not observed in content of the examples retrieved regardless of the diffi-
Experiment 1 and may result from the increased sensitiv- culty involved. If the content of events recalled is indica-
ity in measurement afforded by a pretest/posttest design tive of what these individuals value and experience (i.e.,
or from the larger number of examples (up from 8 to 12) their self-confidence), then retrieving examples of past
that participants recalled in this study. Regardless, it is an confidence may affirm their self-concept and boost self-
indication that under some conditions, the self-esteem esteem. The number of examples recalled might then
of those low in self-doubt may be bolstered by recalling have a linear relationship with the self-regard for low-
memories of self-confidence. self-doubt participants despite the difficulty that may be
involved; as the number of examples increases, so does
POSSIBLE ALTERNATIVE EXPLANATIONS their self-esteem until some asymptotic level is reached.
One alternative explanation to our assertion that low- A third prediction—one that stems more directly
self-doubt participants did not experience threat when from our notion that retrieval difficulty is a threat only to
recalling a difficult number of examples stems from high-self-doubt individuals—would be that when retrieval
what is known about self-affirmation among those high difficulty is sufficiently strong, it affects the specific self-
in self-esteem. Those high in self-esteem (by definition) judgments of low-self-doubt individuals but does not
have more positive self-concepts and are also better able generalize to their global self-evaluation. It may be that
to fend off threats to the self by affirming their positive the self-concept of low-self-doubt participants is indeed
attributes and values (Steele, Spencer, & Lynch, 1993). It malleable, but because they lack the investment in pre-
may be that those low in self-doubt (who are also likely to serving perceptions of their abilities, they do not inter-
be high in self-esteem) experienced a threat when recall- pret undesirable self-concept shifts as implicating their
ing 12 examples but were better able to cope with this core self. In this case, low-self-doubt participants would
threat through affirming themselves by focusing on their be expected to rate themselves as less confident after
positive attributes. Although this notion of “self-affirma- recalling a difficult number of confidence examples but
tion after a threat” is consistent with the boost in self- their self-esteem would not change.
esteem observed in low-self-doubt participants, this expla- A third experiment was designed to focus exclusively
nation seems unlikely given that the interaction of retrieval on low-self-doubt participants and to explore the impact
condition and self-doubt remained when self-esteem was of retrieval cues on their self-esteem by using a paramet-
partialed out of the self-doubt predictor. Furthermore, ric extension of the experimental manipulation used in
we observed no indication of a threat in any post- Experiments 1 and 2. Participants in the following study
experimental dependent measures for those low in self- were asked to retrieve 8, 12, 16, or 20 examples of past
doubt, all of which suggests that the observed effects on self-confidence. As such, we could attempt to find a point
self-esteem were uniquely related to self-doubt and that at which the task became difficult for low-self-doubt par-
404 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

ticipants and to assess the impact of this difficulty on successful at increasing the perceived difficulty of the
their self-esteem. Extending the manipulation this way task for individuals low in self-doubt to a point at least
also afforded the opportunity to treat the independent exceeding the midpoint of the 9-point scale.
variable as a continuous variable, increasing the power SELF-ESTEEM CHANGE
of the analysis and creating the opportunity to test for
linear and curvilinear (i.e., quadratic and cubic) trends.3 To determine whether low-self-doubt participants’ self-
Experiment 3, then, was designed to investigate the phe- esteem changed as a function of the retrieval condition,
nomenology associated with low self-doubt. pretest self-esteem was subtracted from posttest self-
esteem, and this difference score was submitted to trend
EXPERIMENT 3
analysis, with retrieval condition as a between-partici-
pants factor.4 Neither linear nor curvilinear trends yielded
Method significant differences on self-esteem change, ps > .66.
PARTICIPANTS
However, difference scores for the whole sample were
significantly different from zero, F(1, 49) = 63.17, p <
The study included 57 students who participated for .001. Self-esteem increased (Mchange = 5.4, SD = 11.0), but
partial credit in an introductory psychology class. Partici- the number of examples did not moderate this increase
pants were selected to participate based on their score on (see Figure 4).
the SOS-SD administered as part of a mass prescreening 6
to 8 weeks prior to the experiment. Only those partici- SELF-RATINGS
pants in the lowest quartile of the Self-Doubt Scale distri- On confidence ratings, trend analysis indicated a sig-
bution (scores < 22) were recruited. Participants also nificant linear trend of number examples, t(50) = 2.29, p =
completed the Self-Doubt Scale again at the end of the .01. The ease of retrieval effect was evident; as number of
experiment. Two participants were removed from the examples increased, individuals reported that their self-
data set because their postexperiment self-doubt scores confidence decreased (see Figure 4). On uncertainty, no
were more than one standard deviation above the mean effects were significant, ps > .10.
of the entire population (M = 25.6, SD = 7.2); thus, they
CONTENT ANALYSIS
could no longer reasonably be considered in the low-self-
doubt category. Two independent judges rated the last two examples
PROCEDURE
every participant generated following the procedure
used in previous studies. Interjudge reliability was ade-
The procedure was essentially the same as that used in quate, r(54) = .64, p < .001; ratings were thus averaged
Experiment 2, the primary difference being the number together and then submitted to trend analysis. No trend
of examples participants were asked to recall and describe. was significant. Instead, the confidence exhibited by
Participants were randomly assigned to one of four con- examples across conditions (M = 5.8, SD = 1.3) was
ditions in which they had to recall 8, 12, 16, or 20 exam- slightly higher than the midpoint of the 9-point scale.
ples of confidence. The following dependent measures The quality of the last examples in each condition was
were included: (a) a self-rating of confidence, (b) a self- equivalent and the confidence expressed in the exam-
rating of uncertainty, and (c) the Rosenberg Self-Esteem ples remained relatively high overall.
Scale, which was measured both before and after the
retrieval manipulation (αpre = .64, αpost = .68). Discussion

Results The results of Experiment 3 yield insight into the phe-

nomenology of low-self-doubt individuals who confront
MANIPULATION CHECK both fairly easy and fairly difficult retrieval tasks. The evi-
As expected, perceived difficulty increased as the dence shows clearly that it was difficult for them to recall
number of examples recalled increased; mean difficulty 20 examples of self-confidence. Furthermore, this mag-
ratings for 8-example (M = 3.8, SD = 2.3), 12-example nitude of subjective difficulty was sufficient to produce
(M = 4.5, SD = 2.8), 16-example (M = 4.4, SD = 2.5), and the ease of retrieval effect (Schwarz et al., 1991). Spe-
the 20-example (M = 5.9, SD = 2.2) conditions were con- cifically, these low-self-doubt participants did experience
sistent with predictions. Trend analysis yielded a signifi- a decrease in confidence but did not when the task was
cant linear trend of retrieval condition on difficulty, less difficult (e.g., eight instances).
t(50) = 2.03, p = .02; quadratic and cubic trends were not It is interesting to note that self-ratings of confidence
significant. In addition, the 20-example condition was were influenced by retrieval condition in this study but
finally successful in leading individuals low in self-doubt uncertainty self-ratings were influenced by retrieval con-
to report average difficulty ratings above the midpoint of dition in Experiment 1. One possible reason for this dif-
the scale (i.e., 5.9 > 5.5). In sum, the manipulation was ference across studies is that individuals high and low in
 Hermann et al. / SELF-DOUBT AND SELF-ESTEEM 405

ples of self-confidence they generated. Had they focused

on the content, instead of the properties of the retrieval
experience, their self-esteem might have been shored up
rather than assaulted. Their self-doubts might have been
assuaged rather than fueled. Instead, it appears that indi-
viduals high in self-doubt are quite sensitive to cues that
contribute to furthering feelings of doubt and which in
turn threaten self-esteem.
Individuals low in self-doubt exhibited the opposite
tendency; that is, they appeared to be affected by the
content of the examples they generated and their self-
esteem was unaffected by other properties of the retrieval
experience. Specifically, individuals low in self-doubt
Figure 4 Predicted means for standardized self-esteem change and reported an increase in self-esteem after recalling many
standardized confidence ratings, adjusted for pretest self-
esteem, as a function of retrieval condition for individuals examples of their confidence and they rated themselves
low in self-doubt: Experiment 3. as more certain, not less certain (Experiment 1), after
NOTE: Standardized self-esteem change scores have been adjusted so recalling many examples. Even when the ease of retrieval
that 0 equals no change.
effect was clearly produced successfully among individu-
als low in self-doubt (Experiment 3), they still experi-
self-doubt are schematic (Markus, 1977) on different enced an increase in self-esteem. In short, the self-regard
dimensions. Whereas individuals high in self-doubt may of low-self-doubt individuals showed no signs of being
be schematic about uncertainty, individuals low in self- threatened by the very same procedures that posed a
doubt may be schematic about confidence. As a result, clear threat and had a clear detrimental impact on the
ratings of confidence and uncertainty may have been dif- self-regard of participants high in self-doubt.
ferentially affected because these groups differ on dimen-
sions in which they typically evaluate themselves. The Relationship Between Self-Doubt and Self-Esteem
Nonetheless, the self-esteem of individuals low in self-
The present findings shed light on the dynamic rela-
doubt was unaffected by the retrieval condition. As
tionship between self-doubt and self-esteem. It has been
observed in Experiments 1 and 2, the self-esteem of low-
observed that people who experience self-doubt chroni-
self-doubt participants did not decrease under the con-
cally may have an “abnormal investment in the question
ditions that produced a decrease for those low in self-
of self-worth” (Jones & Berglas, 1978, p. 205). To use
doubt (e.g., recalling 8 and 12 examples). Indeed, there
Crocker’s terminology (Crocker & Wolfe, 2000), their
was an overall increase, regardless of experimental con-
feelings of self-worth are contingent. Those with self-
dition, in the self-esteem reported by these low-self-
doubt about competence (Jones, 1989) may find that
doubt participants. Together, the confidence ratings
their feelings of self-worth are often called into question
and the self-esteem findings suggest that individuals low
because of the centrality of that dimension to their every-
in self-doubt may be better at deflecting threats to their
day lives. And their feelings of overall self-worth appear
positive self-regard. Even when they experience a decrease
to be contingent on information that bears on their
in confidence, albeit a small one, this does not translate
judgments of competence (Crocker & Wolfe, 2000). As a
into any loss in feelings of self-worth.
result, those high in self-doubt are likely to find that their
self-worth hinges on cues that either (a) contribute to
GENERAL DISCUSSION
their feelings of doubt or (b) tend to set self-doubts
Collectively, the findings in these three experiments aside.
show that the self-worth of individuals high in self-doubt Consequently, in circumstances where cues to compe-
is more vulnerable to the threat posed by a situational tence are not present, or are irrelevant, doubt about
induction of doubt than is the case for individuals low in one’s overall worth should not be piqued and no threat
self-doubt. In the first two experiments, the task of retriev- to self-esteem is present. Self-esteem should remain fairly
ing a large number of examples of past self-confidence, stable. However, where cues to competence abound,
ironically, produced a decline in reports of global self- which is probably quite often, people who characteristi-
esteem. Yet, this loss in self-esteem was present only for cally experience doubt are likely to experience a threat
individuals high in self-doubt at the outset. The irony, of to self-esteem. Often, the threat can be managed. In the
course, is that those with high self-doubt could have case of self-handicapping and many other self-esteem
taken advantage of the content of the information embed- maintenance ploys, one’s self-esteem is protected when
ded in the 8 (Experiment 1) or 12 (Experiment 2) exam- the threat, inspired by the arousal of self-doubt, is set
406 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

aside. However, when this is not possible, a temporary in a wide array of self-esteem maintenance strategies,
decline in self-esteem should be observed. This is what such as self-handicapping (e.g., Harris & Snyder, 1986),
was found in the present experiments. where the goal of the act is to obscure the causes of
By contrast, individuals with little or no self-doubt behavior. Rather than risk the certainty that one’s lack of
appear to lack that abiding over-investment in the ques- ability is the cause of failure, it is preferable to obscure
tion of self-worth. Their judgments of self-worth are sta- the link between performance and behavior by introduc-
ble, less contingent on temporary events, and they should, ing causal ambiguity. The causal ambiguity should sus-
therefore, find it easier to shrug off information that is tain self-doubt. The present findings suggest that people
negative and that, otherwise, might be damaging to self- high in self-doubt also may contend with another, inter-
esteem. Individuals low in self-doubt simply do not seem nal source of ambiguity in their lives. Even when think-
to entertain the idea that their ability is in question. Con- ing about one’s strengths, rather than shortcomings,
sequently, it is exceedingly rare to draw them into ques- self-doubt may make cues about properties of the think-
tioning their feelings of self-worth. A subtle cue, such as ing process salient. This too would seem to contribute to
the properties of the retrieval experience, may simply go sustaining self-doubt.
unnoticed. Or, if noticed, the properties of the retrieval
experience (i.e., ease or difficulty) receive little or no The Breadth of the Impact of Meta-Cognitive Cues
weight. Persons low in self-doubt appear to focus their
The present findings raise questions about the speci-
attention squarely on the content of their thinking rather
than on these other cues. Unlike high-self-doubt individ- ficity versus generality of the impact of meta-cognitive
uals, who are quite sensitive to cues that contribute to cues. In the present studies, difficulty in retrieval had a
furthering feelings of doubt, low-self-doubt individuals more far-reaching and global impact than research has
are oblivious to them. been designed to reveal to date. Until now, research
using this paradigm has shown how retrieval difficulty
The Self-Perpetuating Nature of Self-Doubt affects an individual’s judgment in the same domain in
which the information is recalled (Schwarz, 1998). For
Although more speculative, the present findings sug-
instance, in the original study, recalling examples of
gest reasons why feelings of self-doubt might be self-per-
assertiveness or unassertiveness affected participants’
petuating. The meta-cognitive processes that serve as a
self-judgments on the same dimension, assertiveness-
cue to feelings of self-doubt are probably a common
unassertiveness (Schwarz et al., 1991). Recalling many
experience. Individuals high in self-doubt may spend a
health risk behaviors can affect perceptions of risk
good deal of time thinking about their level of confi-
(Rothman & Schwarz, 1998), generating many reasons
dence, recalling both past and present illustrations. One
to use public transportation affects attitudes about pub-
item on the SOS-SD scale is, “I often wish that I felt more
lic transportation (Waenke, Bless, & Biller, 1996), and so
certain about my strengths and weaknesses.” Endorsing
that statement may mean that much time in ones’ daily forth. In the present studies, recalling examples of self-
life is spent considering one’s level of ability, driven by confidence not only affected the self-perception of uncer-
feelings of uncertainty about it. The positive correlation tainty but also influenced global feelings of self-worth.
between a recent scale measuring ruminative tendencies That retrieval difficulty can influence not only specific
(Trapnell & Campbell, 1999) and the SOS-SD suggests self-evaluations but global ones as well opens the door to
this may be the case.5 considering how properties of one’s thinking can play a
Intuitively, it would be no surprise if individuals high role in forming and sustaining identity beyond specific
in self-doubt were found to spend inordinate amounts of judgments, on specific dimensions, where the self-evalu-
time dwelling on their competence shortcomings. More ation change is temporary. To illustrate, a belief about
interesting, though, effort spent trying to think of them- the effectiveness of one’s memory for certain events
selves possessing strengths in ability could have precisely could have either no impact, an impact on a restricted
the same effect. The intrusion of feelings of doubt, range of similar events that require recall (e.g., Strack &
thoughts about the properties of the retrieval process, Forster, 1998), or might generalize to judgments about
and so forth all could contribute to exacerbating self- one’s capacity for recall in general. Consider the absent-
doubt and damaging feelings of self-esteem. And, when minded professor trying to recall where his or her car is
self-esteem erodes, even temporarily, this too might con- parked. If youthful, the professor’s lapse might be taken
tribute directly to feelings of self-doubt. simply as an indication of a specific instance of being lost
In sum, a maladaptive cycle of self-doubt may be in thought. If much older, however, the professor’s lapse
inspired both by the behavioral and thinking styles of might lead him or her to question his or her capacity for
individuals with high self-doubt. Self-doubt is implicated recall and, perhaps, overall mental functioning.
 Hermann et al. / SELF-DOUBT AND SELF-ESTEEM 407

Conclusion Bem, D. J. (1967). Self-perception: An alternative interpretation of

cognitive dissonance phenomena. Psychological Review, 74, 183-
200.
The present studies shed some light on the experi- Berglas, S., & Jones, E. E. (1978). Drug choice as a self-handicapping
ence of self-doubt and suggest some interesting hypothe- strategy in response to noncontingent success. Journal of Personal-
ses about how it might be sustained in daily life. There is ity and Social Psychology, 36, 405-417.
already evidence that self-doubt is implicated in behav- Cohen, J. (1988). Statistical power analysis for the behavioral sciences.
Hillsdale, NJ: Lawrence Erlbaum.
ioral strategies (e.g., self-handicapping) that are designed Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation
to protect self-esteem but that do so at the cost of sustain- analysis for the behavioral sciences. Hillsdale, NJ: Lawrence Erlbaum.
ing self-doubt. It is particularly problematic to manage a Crocker, J., & Wolfe, C. T. (2000). Contingencies of self-worth. Unpub-
lished manuscript, University of Michigan, Ann Arbor.
threat to self-esteem when it is generated internally, how- Duval, S., & Wicklund, R. A. (1972). A theory of objective self awareness.
ever. Meta-cognitive cues that provoke feelings of self- New York: Academic Press.
doubt may be more readily noticed by high-self-doubt Feick, D. L., & Rhodewalt, F. (1997). The double-edged sword of self-
handicapping: Discounting, augmentation, and the protection
individuals, who are particularly sensitive to such cues, and enhancement of self-esteem. Motivation and Emotion, 21, 147-
than by low-self-doubt individuals, who are not. The 163.
result is that their feelings of doubt are underscored. Harris, R. N., & Snyder, C. R. (1986). The role of uncertain self-
Ultimately, their feelings of self-worth are shaken, at least esteem in self-handicapping. Journal of Personality & Social Psychol-
ogy, 51, 451-458.
temporarily. In turn, losses in self-esteem may contribute Higgins, R. L. (Ed). (1990). Self-handicapping: The paradox that isn’t.
to feelings of self-doubt about one’s competence. These New York: Plenum.
dynamics help explain the co-occurrence of self-doubt Jarvis, W.B.G. (1998). MediaLab Research Software (Verson 3.0)
[Computer software]. Philadelphia: Empirisoft Inc. (www.
and shaky levels of self-esteem already reported in the empirisoft.com)
literature. Jones, E. E. (1989). The framing of competence. Personality and Social
Psychology Bulletin, 15, 472-492.
Jones, E. E., & Berglas, S. (1978). Control of attributions about the
NOTES self through self-handicapping strategies: The appeal of alcohol
and the role of underachievement. Personality and Social Psychology
1. Pretesting revealed that students could spontaneously and with- Bulletin, 4, 200-206.
out extraordinary effort generate a median of five confidence exam- Jost, J. T., Kruglanski, A. W., & Nelson, T. O. (1998). Social meta-cog-
ples from their past. Consistent with Schwarz’s procedures (personal nition: An expansionist review. Personality and Social Psychology
communication, May 26, 1998), the easy and difficult conditions were Review, 2, 137-154.
set at the median minus 50% and the median plus 50%, respectively. Leonardelli, G. (1997). [Correlates with the Subjective Overachievement
2. Because posttest scores have a tendency to regress toward the Scale]. Unpublished raw data, The Ohio State University, Columbus.
mean of the distribution, difference scores are typically negatively cor- Markus, H. (1977). Self-schemata and processing information about
related with the pretest scores (Cohen & Cohen, 1983). To correct for the self. Journal of Personality and Social Psychology, 35, 63-78.
this artifact, pretest Rosenberg Self-Esteem Scale (RSE) scores were Mischel, W. (1998). Meta-cognition at the hype of social-cognitive psy-
entered as a covariate and were found to be negatively correlated with chology. Personality and Social Psychology Review, 2, 87-99.
the difference score (β = –.22, p < .05). The predicted means presented Oleson, K. C., Poehlmann, K. M., Yost, J. H., Lynch, M. E., & Arkin, R.
in Figure 2 are adjusted for the variance accounted for by pretest self- M. (2000). Subjective overachievement: Individual differences in
esteem. self-doubt and concern with performance. Journal of Personality,
3. Curvilinear versions of the first two hypotheses are also possible. 68, 84-86.
Perhaps self-esteem may increase but will asymptote or begin to decrease. Pelham, B. W. (1991). On confidence and consequence: The cer-
The following design provided a test for these curvilinear predictions. tainty and importance of self-knowledge. Journal of Personality and
4. As in Experiment 2, pretest self-esteem was entered as a covariate
Social Psychology, 60, 518-530.
to control for regression to the mean (see Note 2). As expected, it
Pelham, B. W., & Swann, W. B. (1989). From self-conceptions to self-
accounted for a significant portion of the variance, F(1, 50) = 184.86,
worth: On the sources and structure of global self-esteem. Journal
p < .001.
5. A recent correlational study of 646 student participants (Leonardelli, of Personality and Social Psychology, 57, 672-680.
1997) indicates that the Self-Doubt Subscale of the Subjective Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
Overachievement Scale (SOS-SD) and the rumination subscale of NJ: Princeton University Press.
Trapnell and Campbell’s (1999) Rumination-Reflection Questionnaire Rothman, A. J., & Schwarz, N. (1998). Constructing perceptions of
are positively correlated (r = .50, p < .001). vulnerability: Personal relevance and the use of experiential infor-
mation in health judgments. Personality and Social Psychology Bulle-
tin, 24, 1053-1064.
Schwarz, N. (1998). Accessible content and accessibility experiences:
REFERENCES
The interplay of declarative and experiential information in judge-
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and inter- ment. Personality and Social Psychology Review, 2, 87-99.
preting interactions. Newbury, CA: Sage. Schwarz, N., Bless, H., Strack, F., Klumpp, G., Rittenauer-Schatka, H.,
Arkin, R. M., & Baumgardner, A. H. (1985). Self-handicapping. In J. & Simons, A. (1991). Ease of retrieval as information: Another
H. Harvey & G. Weary (Eds.), Attribution: Basic issues and applica- look at the availability heuristic. Journal of Personality and Social Psy-
tions (pp. 169-202). San Diego, CA: Academic Press. chology, 61, 195-202.
Arkin, R. M., & Oleson, K. C. (1998). Self-handicapping. In J. M. Steele, C. M., Spencer, S. J., & Lynch, M. (1993). Self-image resilience
Darley & J. Cooper (Eds.), Attribution processes, person perception, and dissonance: The role of affirmational resources. Journal of Per-
and social interaction: The legacy of Ned Jones. Washington, DC: Amer- sonality and Social Psychology, 64, 885-896.
ican Psychological Association. Strack, F., & Forster, J. (1998). Self-reflection and recognition: The
Bandura, A. (1997). Self-efficacy: The exercise of control. New York: role of metacognitive knowledge in the attribution of recollective
Freeman. experience. Personality and Social Psychology Review, 2, 111-123.
 408 PERSONALITY AND SOCIAL PSYCHOLOGY BULLETIN

Trapnell, P. D., & Campbell, J. D. (1999). Private self-consciousness Waenke, M., Bless, H., & Biller, B. (1996). Subjective experience ver-
and the five-factor model of personality: Distinguishing rumina- sus content of information in the construction of attitude judg-
tion from reflection. Journal of Personality and Social Psychology, 76, ments. Personality & Social Psychology Bulletin, 22, 1105-1113.
284-304.
Tversky, A., & Kahneman, D. (1973). Availability: A heuristic for judg- Received October 30, 2000
ing frequency and probability. Cognitive Psychology, 5, 207-232. Revision accepted June 11, 2001

View publication stats