Intuition and the Correspondence Between Implicit and Explicit Self-Esteem

Abstract

Four studies tested whether the perceived validity of intuition increases the correspondence between implicit and explicit self-esteem. Studies 1 and 2 found, with 2 different measures of implicit self-esteem, that people who chronically view their intuition as valid have more consistent implicit and explicit self-esteem. In contrast, people with relatively low faith in their intuition had a negative relation between implicit and explicit self-esteem, suggesting that they may overcorrect their explicit self-views for the potential bias posed by implicit self-esteem. In Studies 3 and 4, participants who were induced to view their intuition as valid reported explicit self-views (self-evaluations made under time pressure, or state self-esteem) that were more consistent with their implicit self-esteem. These results suggest that people experience implicit self-esteem as intuitive evaluations. The correspondence between implicit and explicit self-esteem among individuals who view their intuition as valid may suggest that these individuals incorporate implicit self-esteem into their explicit self-views.

Full text

Intuition and the Correspondence Between Implicit
and Explicit Self-Esteem
Christian H. Jordan and Mervyn Whitfield
Wilfrid Laurier University
Virgil Zeigler-Hill
University of Southern Mississippi
Four studies tested whether the perceived validity of intuition increases the correspondence between
implicit and explicit self-esteem. Studies 1 and 2 found, with 2 different measures of implicit self-esteem,
that people who chronically view their intuition as valid have more consistent implicit and explicit
self-esteem. In contrast, people with relatively low faith in their intuition had a negative relation between
implicit and explicit self-esteem, suggesting that they may overcorrect their explicit self-views for the
potential bias posed by implicit self-esteem. In Studies 3 and 4, participants who were induced to view
their intuition as valid reported explicit self-views (self-evaluations made under time pressure, or state
self-esteem) that were more consistent with their implicit self-esteem. These results suggest that people
experience implicit self-esteem as intuitive evaluations. The correspondence between implicit and
explicit self-esteem among individuals who view their intuition as valid may suggest that these
individuals incorporate implicit self-esteem into their explicit self-views.
Keywords: intuition, implicit, explicit, self-esteem, attitudes
Imagine that you are deciding whether to buy a particular used
car. There are many factors you might consider when making your
decision: mileage, safety ratings, appearance, and so on. Imagine
further that you have considered these factors and have, overall, a
positive impression of the car. What if, at the same time, you have
a negative gut reaction? Something about the car makes you feel
uneasy. How might this affect your overall attitude toward the car?
In a sense, you have two distinct sources of evaluation that, in this
instance, push you in opposite directions: rational analysis and
intuition.
This distinction is relatively familiar in the context of problem
solving and decision making (Hammond, 1996; Hogarth, 2001).
But can these sources of evaluation inform models of self-esteem
(SE)? Does SE derive, on one hand, from deliberative (although
probably not unbiased) consideration of accomplishments and
failings and, on the other hand, from relatively intuitive reactions?
We argue that it does, and that this distinction maps closely onto
the distinction between explicit and implicit SE. Moreover, by
considering intuition, we may be better able to specify when, and
for what kinds of people, explicit and implicit SE correspond with
each other.
Implicit and Explicit Attitudes and SE
Attitudes can be activated automatically (Bargh, Chaiken, Ray-
mond, & Hymes, 1996; Fazio, Sanbonmatsu, Powell, & Kardes,
1986). Objects in the environment can trigger evaluations quickly,
efficiently, and with little conscious guidance. Some theorists have
thus proposed that people can hold two distinct attitudes toward
the same attitude object at explicit and implicit levels (e.g., Wil-
son, Lindsey, & Schooler, 2000). Implicit attitudes correspond,
roughly, to automatically activated attitudes and are considered to
be distinct from explicit attitudes, which are more deliberative and
conscious in nature. Implicit attitudes are measured indirectly,
often by reaction time measures that are difficult to control (Fazio
& Olson, 2003; Kihlstrom, 2004), whereas explicit attitudes are
measured directly by self-report scales. The same is true of self-
attitudes. People may hold distinct explicit and implicit SE (Farn-
ham, Greenwald, & Banaji, 1999; Greenwald & Banaji, 1995).
Although there is controversy about whether implicitly and explic-
itly measured attitudes represent distinct constructs (Fazio & Ol-
son, 2003; Petty, Wheeler, & Tormala, 2003), the two do demon-
strate some discriminant validity.
Implicit and explicit measures of SE, for example, are generally
uncorrelated (Bosson, Swann, & Pennebaker, 2000). They do,
however, predict relevant outcomes, albeit distinct ones. Explicit
SE has, for example, predicted reports of anxiety for participants
about to undergo a sensitive interview, whereas implicit SE has
predicted nonverbal indicators of anxiety (Spalding & Hardin,
1999). In addition, specific combinations of implicit and explicit
SE may predict unique outcomes. People with high explicit but
low implicit SE appear to be more defensive and self-enhancing
than people with high explicit and high implicit SE (Bosson,
Brown, Zeigler-Hill, & Swann, 2003; Jordan, Spencer, Zanna,
Hoshino-Browne, & Correll, 2003; Jordan, Spencer, & Zanna,
2005; McGregor & Marigold, 2003; McGregor, Nail, Marigold, &
Kang, 2005; Zeigler-Hill, 2006). People with low explicit but high
Christian H. Jordan and Mervyn Whitfield, Department of Psychology,
Wilfrid Laurier University, Waterloo, Ontario, Canada; Virgil Zeigler-Hill,
Department of Psychology, University of Southern Mississippi.
This research was supported by a grant from the Social Sciences and
Humanities Research Council of Canada to Christian H. Jordan. We thank
Ian McGregor, Steve Spencer, and Mark Zanna for insightful comments on
an earlier version of this article.
Correspondence concerning this article should be addressed to Christian
H. Jordan, Department of Psychology, Wilfrid Laurier University, 75
University Avenue East, Waterloo, Ontario, Canada N2L 3G5. E-mail:
cjordan@wlu.ca
Journal of Personality and Social Psychology Copyright 2007 by the American Psychological Association
2007, Vol. 93, No. 6, 1067–1079 0022-3514/07/$12.00 DOI: 10.1037/0022-3514.93.6.1067
1067

implicit SE, in contrast, may be particularly responsive to self-
relevant events. Among students with low explicit SE, those with
high implicit SE reported more positive self-views shortly after an
academic success and more negative self-views after a failure
(Logel, Spencer, Wood, Holmes, & Zanna, 2006).
Despite such findings, many aspects of implicit SE remain
poorly understood. How, if at all, do people experience implicit
SE? Are people ever aware of their implicit SE? It is widely
assumed that implicit attitudes and SE are nonconscious, but no
clear evidence supports this assumption (Fazio & Olson, 2003;
Gawronski, Hofmann, & Wilbur, 2006). Even if implicit attitudes
are automatic in the sense of being efficient and unintentionally
activated, this does not necessarily mean that they are noncon-
scious (Bargh, 1994; Moors & De Houwer, 2006). Indeed, some
recent models of implicit attitudes posit that people are generally
aware of them (Gawronski & Bodenhausen, 2006; Petty & Brin˜ol,
2006, in press). Understanding whether and how people experi-
ence implicit SE is thus a timely issue.
This issue, moreover, dovetails with another fundamental ques-
tion: “When, under what conditions, and for what kind of people,
are implicit and explicit measures related?” (Fazio & Olson, 2003,
p. 304; see also Nosek, 2005). If people are aware of their implicit
attitudes, they may incorporate them into their explicit, overtly
expressed attitudes, increasing the correspondence between the
two (Hofmann, Gschwendner, & Schmitt, 2005; Kihlstrom, 2004).
Whether awareness increases explicit–implicit attitude consistency
may depend, however, on whether people view their implicit
attitudes as valid (as discussed further below; see also Gawronski
& Bodenhausen, 2006; Hofmann, Gschwendner, & Schmitt,
2005). But in what sense might people be aware of their implicit
attitudes?
Intuition and Dual-Systems Theories
Numerous theorists have posited that people process informa-
tion through two independent, although interacting, cognitive sys-
tems (e.g., Barrett, Tugade, & Engle, 2004; Carver, 2005; Epstein,
1990; Hinton, 1990; Sloman, 1996; Smith & DeCoster, 2000;
Smolensky, 1988; Strack & Deutsch, 2004). One system is fast
acting and operates largely through automatic processes—
associative connections and spreading activation. This associative
system is highly efficient, but relatively inflexible. The other
system, in contrast, is slower and operates largely through con-
trolled processes—symbolic rules that rely on deliberative reason-
ing. This rule-based system is not efficient, but it is highly flexible.
Within this framework, the associative system may produce im-
plicit attitudes, whereas the rule-based system may produce ex-
plicit attitudes (Carver, 2005; Smith & DeCoster, 2000; Strack &
Deutsch, 2004).
People may thus experience explicit and implicit attitudes dif-
ferently. Reasoning within the rule-based system is believed to be
largely transparent and introspectively accessible. People are
thought to be generally aware of the processes that contribute to
their attitudes within this system and to be able to articulate, at
least roughly, reasons for their attitudes. One could, for example,
point to the importance of safety ratings and mileage in forming an
attitude toward a car. The associative system, in contrast, operates
preconsciously. People are sometimes aware of its products but not
of its processes (Sloman, 1996; Smith & DeCoster, 2000;
Smolensky, 1988). Responses generated by this system may sub-
jectively “pop” into one’s head and thus be experienced as intu-
ition (Hinton, 1990; Hogarth, 2001; Lieberman, 2000; Smith &
DeCoster, 2000; Smolensky, 1988). Implicit attitudes and SE may
thus be experienced as intuitive evaluations.
Some limited evidence supports the link between implicit atti-
tudes and intuition. Hofmann, Gawronski, Gschwendner, Le, and
Schmitt (2005) had judges rate attitude domains in terms of the
spontaneity with which people report attitudes. Specifically, they
rated, for each domain, how much respondents “rely on their gut
reactions when asked to report their attitudes” (p. 1373). In a
subsequent meta-analysis, they found higher correlations between
implicit and explicit attitudes in domains for which people rely
more heavily on gut feelings.
Thus, people may experience implicit attitudes as intuition, and
this awareness may allow greater consistency between implicit and
explicit attitudes. If people are conscious of implicit attitudes, they
may use them in forming their explicit attitudes (Hofmann,
Gschwendner, & Schmitt, 2005). A negative gut reaction to a car
may thus temper an otherwise positive attitude toward that car.
This would increase the correspondence between implicit and
explicit attitudes.
It is important to note, however, that implicit attitudes might
only influence explicit attitudes to the extent that they are per-
ceived to be valid (Gawronski & Bodenhausen, 2006; Petty &
Brin˜ol, 2006, in press). That is, if implicit attitudes, experienced as
intuitive evaluations, are perceived to be valid, they may be
incorporated into explicit attitudes. We thus examined whether the
perceived validity of intuition increases the correspondence be-
tween implicit and explicit SE We predicted that the more people
perceive their intuition to be valid, the greater would be the
correspondence between their implicit and explicit SE.
Study 1
Pelham et al. (2005) reported indirect evidence that intuition
increases the correspondence between implicit and explicit SE.
They found that women show greater implicit– explicit SE consis-
tency than do men and suggested that this finding reflects women’s
greater focus on affect and intuition, although they did not directly
measure this focus. It is thus desirable to more directly test the role
of intuition in the correspondence between implicit and explicit
self-esteem by measuring individual differences in how much
people trust their intuition. We did so in Study 1 by using Epstein
and Pacini’s (1999) Rational–Experiential Inventory (REI), which
includes a subscale measuring Faith in Intuition, or the extent to
which people rely on and are engaged by intuition. We expected
people higher in Faith in Intuition to show closer correspondence
between implicit and explicit SE.
We also examined the REI subscale measuring Need for Cog-
nition, although our predictions for it were less straightforward. On
the one hand, the processes we are specifying require cognitive
engagement and should be enhanced by Need for Cognition. The
more cognitively engaged people are, the more likely they should
be to actively rely on implicit SE when reporting explicit self-
views, if they trust their intuition. On the other hand, Need for
Cognition increases overall cognitive elaboration and may also
increase reliance on other sources of information when reporting
explicit SE. To the extent that these other considerations diverge
1068
JORDAN, WHITFIELD, AND ZEIGLER-HILL

from implicit SE in their implications, they will dilute the corre-
spondence between implicit and explicit SE (Gawronski & Boden-
hausen, 2006). Indeed, Need for Cognition is associated with less
correspondence between implicit and explicit prejudices (Florack,
Scarabis, & Bless, 2001). We thus included Need for Cognition in
a relatively exploratory manner.
Method
Participants
Sixty-eight female undergraduates enrolled in psychology
courses participated in exchange for course credit. For this initial
study, we included only female participants to create a more
homogeneous sample, which should provide a more sensitive test
of our predictions (Aronson, Ellsworth, Carlsmith, & Gonzales,
1990). Three participants had error rates of more than 20% on the
Implicit Association Test (IAT; described below) and were ex-
cluded from analyses (Greenwald, McGhee, & Schwartz, 1998).
One participant scored more than 3 standard deviations from the
mean on a primary dependent measure and was also excluded.
1
Results are reported for the remaining 64 participants.
Materials and Procedure
Up to 8 participants at a time participated in sessions, although
they were seated at individual workstations and did not interact
with each other. Participants were met by a male experimenter who
described the study as an investigation of “personality and percep-
tual styles.”
Need for Cognition and Faith in Intuition. Participants first
completed the short form of the REI, including subscales measur-
ing Need for Cognition and Faith in Intuition (Pacini & Epstein,
1999).
2
For the Need for Cognition subscale, respondents rated
their agreement with 12 items such as “I have a logical mind,” “I
enjoy problems that require hard thinking,” and “I am not a very
analytical thinker” (reversed). For the Faith in Intuition subscale,
respondents rated their agreement with 12 items such as “I suspect
my hunches are often inaccurate” (reversed), “If I were to rely on
my ‘gut feelings,’ I would often make mistakes” (reversed), and “I
like to rely on my intuitive impressions.” Responses were made on
5-point scales ranging from 1 (definitely not true of myself)to5
(definitely true of myself). These subscales showed good reliability
(Cronbach’s ␣s ⫽ .85 for Need for Cognition and .84 for Faith in
Intuition). Scores on the items for each subscale were averaged
together.
Implicit SE. Participants next completed an IAT measure of
implicit SE (Greenwald et al., 1998; Greenwald & Farnham,
2000). The IAT is a computer-based, response-mapping procedure
that measures the extent to which respondents associate the self
with positive and negative affect. Participants categorized words,
as quickly and accurately as possible, along two dimensions: (a)
pleasant versus unpleasant words (i.e., holiday, warmth, friend,
smile, sunshine, gift, love, happy, party, and joy vs. agony, death,
disease, vomit, evil, cockroach, pain, stink, disaster, and garbage),
and (b) self versus object words (i.e., me, myself, it, and that).
Rather than contrasting self with other, as is commonly done in SE
IATs, we contrasted self with the more neutral category of object
(Jordan et al., 2003, 2005; McGregor & Jordan, 2007). IAT scores
should thus more clearly reflect evaluations of self rather than
evaluations of others (Karpinski, 2004). The target words appeared
in the center of the screen, and the category labels appeared in the
upper left and right corners of the screen, corresponding to the
response keys (e and i) used to indicate category membership.
There were seven blocks of trials altogether. Within each block,
words were presented in random order.
Blocks 1, 2, and 5 were practice blocks for which participants
made single categorizations (pleasant vs. unpleasant or self vs.
object; 20 trials each). In the remaining blocks, participants made
both sets of categorizations, on alternate trials, using a single set of
response keys. In Block 4, participants used one response key to
indicate whether a word belonged to the unpleasant or object
category and the other key to indicate whether the word belonged
to the pleasant or self category (40 trials). In Block 7, participants
used one response key to indicate whether a word belonged to the
unpleasant or self category and the other key to indicate whether
the word belonged to the pleasant or object category (40 trials).
Blocks 3 and 6 were practice blocks (20 trials each) for Blocks 4
and 7, although data from these blocks were used in calculating
IAT scores (Greenwald, Nosek, & Banaji, 2003).
The IAT is premised on the logic that individuals with high
implicit SE (i.e., strong associations between the self and positive
affect) should respond faster when self and pleasant share a re-
sponse than when self and unpleasant share a response because the
positive affect associated with the self interferes with responding
in the latter (but not in the former) condition. This facilitation
effect serves as an index of implicit SE. We calculated IAT scores
using the algorithm advocated by Greenwald et al. (2003), with
higher scores indicating higher implicit SE. We calculated split-
half reliability for the IAT using separate scores for the critical
blocks (4 and 7) and the practice blocks (3 and 6). This revealed
that the IAT had acceptable reliability (Spearman–Brown coeffi-
cient ⫽ .62).
Explicit SE. Last, participants completed two measures of
explicit SE: Heatherton and Polivy’s (1991) state SE scale and the
Rosenberg (1965) Self-Esteem Scale (RSES). The state SE scale
has 20 items (Cronbach’s ␣⫽.87) and asks participants to indicate
how they feel “right now” with respect to such items as “I feel
confident about my abilities” and “I feel others respect and admire
me.” Responses were made on 5-point scales ranging from 1 (not
at all)to5(extremely). The RSES is a 10-item measure of trait SE
(Cronbach’s ␣⫽.92). Items include “I feel that I am a person of
worth, at least on an equal basis with others,” and “All in all, I am
inclined to feel that I am a failure” (reversed). Responses were
made on 9-point scales ranging from 1 (very strongly disagree)to
9(very strongly agree). Scores on items for each scale were
averaged together to create measures of explicit SE.
1
We also conducted all analyses including this participant. The pattern
of results was no different with this outlier retained in the data.
2
For the revised version of the scale that we administered, Pacini and
Epstein (1999) preferred the terms rationality and experientiality for the
subscales measuring Need for Cognition and Faith in Intuition, respec-
tively. We retain the older terminology here, however, because it corre-
sponds more closely to our theoretical focus in this article.
1069
IMPLICIT AND EXPLICIT SELF-ESTEEM

Results and Discussion
The two measures of explicit SE, the state SE scale and the
RSES, correlated highly (r ⫽ .78). We thus converted scores on
these scales into z scores and averaged them into a single index of
explicit SE. This index was not correlated with IAT scores. Table
1 presents descriptive statistics and correlations for measures in
this study.
To test whether trust in intuition moderates the correspondence
between implicit and explicit SE, we conducted a multiple regres-
sion analysis predicting explicit SE from IAT scores, Faith in
Intuition, and Need for Cognition. These predictors were centered
on their means, and interaction terms were generated by creating
cross-product vectors (Aiken & West, 1991). We conducted a
hierarchical regression, entering main effects first and higher order
interactions sequentially on subsequent steps. Consistent with past
research (Epstein, Pacini, Denes-Raj, & Heier, 1996), this analysis
revealed a main effect of Faith in Intuition: higher Faith in Intu-
ition was associated with higher explicit SE (␤⫽.30), t(59) ⫽
2.62, p ⫽ .01. The main effect of Need for Cognition, however,
was not significant (␤⫽.17), t(59) ⫽ 1.49, p ⫽ .14. None of the
interactions involving Need for Cognition were significant either
(ts ⬍ 1). We thus reduced our model by dropping these interaction
terms.
As predicted, however, the interaction between IAT scores and
Faith in Intuition predicted explicit SE (␤⫽.32), t(59) ⫽ 2.81,
p ⫽ .007. The sign of this interaction was positive, indicating that
as Faith in Intuition increased, the relation between implicit and
explicit SE became more positive (see Figure 1).
3
Testing at the
conventional point of 1 SD above the mean of Faith in Intuition
(Cohen & Cohen, 1983), there was a marginally significant posi-
tive relation between IAT scores and explicit SE (␤⫽.28),
t(59) ⫽ 1.77, p ⫽ .08. This effect became significant at 1.1 SD
above the mean (␤⫽.33), t(59) ⫽ 2.00, p ⫽ .05. Unexpectedly,
at 1 SD below the mean of Faith in Intuition, there was a signif-
icant negative relation between IAT scores and explicit SE (␤⫽
⫺.35), t(59) ⫽⫺2.18, p ⫽ .03. Thus, as Faith in Intuition
increased across women, the relation between implicit and explicit
SE became more positive. Women who trust their intuition dem-
onstrated a more positive correspondence between implicit and
explicit SE. This is consistent with the idea that such individuals,
through awareness of implicit SE as intuition, incorporate implicit
SE into their explicit self-views.
Notably, this effect was asymmetrical. People with high implicit
SE (1 SD above the mean) showed large differences in explicit SE,
as a function of Faith in Intuition (␤⫽.62), t(59) ⫽ 3.78, p ⬍
.001, but those with low implicit SE (1 SD below the mean)
showed no such differences (␤⫽⫺.01), t(59) ⬍ 1. This asym-
metry reflects, in part, the overall positive correlation between
Faith in Intuition and explicit SE. People with higher explicit SE
generally trust their intuition more than do people with low explicit
SE (Epstein et al., 1996). Thus, if implicit SE did not moderate the
relation between Faith in Intuition and explicit SE, all participants
with high Faith in Intuition would be expected to report relatively
high explicit SE. The present findings show, however, that
whereas this relation is amplified among people with high implicit
SE, it is attenuated for people with low implicit SE. Thus, the
relatively small differences in explicit SE observed for people with
low implicit SE may mask the extent to which they actually temper
their self-views as a function of Faith in Intuition.
Also notable is the unanticipated finding that women with
relatively low Faith in Intuition showed a negative relation be-
tween implicit and explicit SE. An admittedly speculative expla-
nation for this finding is that these women, because they have
relatively low faith in their intuition, are more likely to view their
intuition as a potential source of bias in evaluation. They may thus
adjust or “correct” their reports to control this bias (Hofmann,
Gschwendner, & Schmitt, 2005). In fact, they may overcompen-
sate, creating a negative relation between implicit and explicit SE
(Wegener & Petty, 1995). (For a description of an analogous
process affecting the relation between implicit and explicit preju-
dice, see Fazio & Olson, 2003; Nosek, 2005.)
Overall, these findings suggest that people experience implicit
SE as intuition and react differently to it depending on whether
they perceive it to be valid. People who trust their intuition may
incorporate implicit SE into explicit SE. People with less faith in
intuition, in contrast, may ignore their implicit SE, or even correct
for the presumed bias posed by implicit SE, when determining
their explicit self-views.
Study 2
In Study 2, we sought to replicate Study 1 and increase its
generalizability in two ways: (a) by using a different measure of
implicit SE and (b) by including male participants. We used the
Implicit Self-Evaluation Survey (ISES; Pelham & Hetts, 1999) as
a measure of implicit SE. Because it is a paper-and-pencil mea-
sure, we were able to test a much larger sample than in Study 1.
This afforded us greater statistical power, which was desirable
because a key simple effect—the relation between implicit and
explicit self-esteem for individuals high in Faith in Intuition—was
only marginally significant at the conventional testing point in
Study 1. By including male participants, we were also able to
examine whether gender uniquely predicts the correspondence
between implicit and explicit SE once trust in intuition is taken
into account (cf. Pelham et al., 2005).
3
This interaction is likewise significant when state and trait SE are
analyzed separately: For state SE, ␤⫽.34, t(59) ⫽ 2.99, p ⫽ .004; for trait
SE, ␤⫽.26, t(59) ⫽ 2.24, p ⫽ .03.
Table 1
Correlations and Descriptive Statistics for Measures of Explicit
Self-Esteem, Implicit Esteem, Faith in Intuition, and Need for
Cognition: Study 1
Measure/statistic 1 2 3 4
1. Explicit self-esteem —
2. Implicit self-esteem ⫺.05 —
3. Faith in Intuition subscale .38
**
⫺.10 —
4. Need for Cognition subscale .30
**
.06 .20 —
M 6.75/3.47 0.61 3.53 3.49
SD 1.35/0.57 0.35 0.56 0.62
Note. Descriptive statistics for the Rosenberg Self-Esteem Scale appear
on the left and those for the state self-esteem scale appear on the right in
the explicit self-esteem column (Column 1).
**
p ⬍ .001.
1070
JORDAN, WHITFIELD, AND ZEIGLER-HILL

Method
Participants
Four hundred eighty-seven undergraduates enrolled in psychol-
ogy courses participated in exchange for course credit. One par-
ticipant was excluded from analyses for failing to provide com-
plete information. Results are thus reported for the remaining 486
participants (100 men and 386 women).
Materials and Procedure
Up to 15 participants at a time participated in sessions,
although they completed the measures individually and did not
interact with each other. Participants were met by a female
experimenter who described the study as an investigation of
“personality processes.” They were informed that they would
complete a series of paper-and-pencil personality measures. In
addition to the measures described below, they completed a
number of other personality measures that are not relevant to
the present study.
Need for Cognition and Faith in Intuition. Participants com-
pleted the long form of the REI, including the Need for Cognition
and Faith in Intuition subscales. This measure was identical to that
used in Study 1, except that each subscale consisted of 20 items
rather than 12. Both subscales showed good reliability in the
present sample (Cronbach’s ␣⫽.88 for Need for Cognition and
.89 for Faith in Intuition).
Implicit SE. Participants also completed the ISES (Pelham &
Hetts, 1999). The ISES is a paper-and-pencil version of an eval-
uative priming measure. It is designed to measure how much one’s
self-concept associatively activates positive and negative affect.
There are four items on this measure, each consisting of two parts:
a self-priming statement followed by a series of word-fragment
completions. For each item, participants respond first to a self-
priming statement that requires them to think about themselves.
These self-priming statements are “It’s important to me to under-
stand myself as well as possible,” “Other people value my abilities
and opinions,” “The way other people treat me has a lot to do with
my own unique personality,” and “I am very sensitive to my inner
thoughts and feelings.” Participants indicate their agreement with
each statement on a 7-point scale with endpoints labeled 1 (not at
all true)to7(very true). These statements serve to prime the
self-concept.
Each priming statement is followed by a series of word-
fragment completions. For each item, participants complete one
word fragment three times in a row by forming three different
words in the order in which the words occur to them (e.g., “1.
__ICE 2. __ICE 3. __ICE”). Each fragment has a target com-
pletion that is either positive or negative. Two items have
positive target completions (i.e., “NICE” and “GOOD”) and
two have negative target completions (i.e., “MEAN” and
“HATE”). For participants who have high implicit SE, thinking
about the self should activate positive affect. This should, in
turn, render positive words more accessible to them and nega-
tive words less accessible. Hence, participants with high im-
plicit SE should complete positive target words early in the
sequence of word fragments and negative target words late, if at
all. Participants receive a score of 1 for each item if the target
completion appears in the first serial position, a score of 2 if it
appears in the second position, and a score of 3 if it appears in
the third position. If the target word is not listed, they receive
a score of 4. To calculate ISES scores, the average serial
position for the two positive target completions is subtracted
from the average serial position for the two negative target
completions. Higher scores thus reflect greater accessibility of
positive words than negative words when the self is primed, and
hence higher implicit SE. Consistent with previous research
(e.g., Bosson et al., 2000), the internal consistency of the ISES
was low (␣⫽.47).
Explicit SE. Participants next completed the RSES exactly as
in Study 1, except that they responded using 5-point scales ranging
from 1 (strongly disagree)to5(strongly agree). Cronbach’s alpha
for this measure was .83.
-1.5
-1
-0.5
0
0.5
1
1.5
Lowest Highest
Implicit self-esteem
Explicit self-esteem
Low FI (-1 SD)
Mean FI
High FI (1 SD)
Figure 1. Explicit self-esteem (z-score composite) as a function of Faith in Intuition (FI) and implicit
self-esteem (Implicit Association Test scores) in Study 1. Predicted values displayed are for the full range of
implicit self-esteem represented in the sample (– 0.21 to 1.45). Lines are depicted at the mean and at 1 standard
deviation above and below the mean of Faith in Intuition.
1071
IMPLICIT AND EXPLICIT SELF-ESTEEM

Results and Discussion
Implicit and explicit SE were uncorrelated overall. Table 2
presents descriptive statistics and correlations for all measures in
Study 2. In addition, female participants (M ⫽ 3.60, SD ⫽ 0.60)
reported significantly more Faith in Intuition than did male par-
ticipants (M ⫽ 3.45, SD ⫽ 0.46), t⬘ (196.65) ⫽ 2.76, p ⬍ .01.
We conducted a multiple regression analysis predicting explicit
SE from ISES scores and Faith in Intuition to test whether a focus
on intuition moderates the correspondence between implicit and
explicit SE. As in Study 1, Need for Cognition was included in this
analysis. We also included gender (dummy coded: female ⫽ 0,
male ⫽ 1). Preliminary analyses indicated that the results were not
moderated by gender, and so it was retained only as a control
variable. We conducted a hierarchical regression analysis, entering
main effects on the first step and higher order interactions on
subsequent steps. Consistent with past research (Epstein et al.,
1996), this analysis revealed main effects of both Faith in Intuition
and Need for Cognition: Higher Faith in Intuition was associated
with higher explicit SE (␤⫽.13), t(470) ⫽ 2.59, p ⫽ .01; higher
Need for Cognition was also associated with higher explicit SE
(␤⫽.15), t(470) ⫽ 3.00, p ⫽ .003. The main effects of implicit
SE and gender were not significant (both ps ⬎ .50).
Of greater relevance to our predictions, and closely paralleling
the results of Study 1, the interaction between implicit SE and
Faith in Intuition significantly predicted explicit SE (␤⫽.14),
t(470) ⫽ 2.67, p ⫽ .008, with a positive sign indicating that as
Faith in Intuition increases, the relation between implicit and
explicit SE becomes more positive (see Figure 2). Indeed, testing
at 1 SD above the mean of Faith in Intuition, there was a significant
positive relation between ISES scores and explicit SE (␤⫽0.17),
t(470) ⫽ 2.58, p ⫽ .01. This is notable because the same effect was
marginally significant in Study 1 when tested at the same point.
Meta-analyzing across the two studies, moreover, revealed this
simple slope to be strongly significant overall (z ⫽ 2.64, p ⫽ .004).
In contrast, at 1 SD below the mean of Faith in Intuition, there was
a negative relation between ISES scores and explicit SE (␤⫽
⫺.17), t(470) ⫽⫺2.58, p ⫽ .01. As in Study 1, this effect was
asymmetrical, with greater observed differences in explicit SE
among participants with high implicit SE (⫹1 SD) as a function of
Faith in Intuition (␤⫽.33), t(470) ⫽ 4.87, p ⬍ .001, than among
those with low implicit SE (⫺1 SD; ␤⫽⫺.02), t(470) ⬍ 1. Again,
this asymmetry reflects, in part, the overall positive correlation
between Faith in Intuition and explicit SE, which was amplified for
people with high implicit SE but attenuated for those with low
implicit SE.
These results were further moderated by Need for Cognition:
There was a significant three-way interaction between implicit SE,
Faith in Intuition, and Need for Cognition in predicting explicit SE
(␤⫽.10), t(470) ⫽ 2.04, p ⫽ .04. This three-way interaction
indicates that the predicted two-way interaction between Faith in
Intuition and implicit SE was apparent for participants who were
high in Need for Cognition (⫹1 SD; ␤⫽.24), t(470) ⫽ 4.11, p ⫽
.001, and those who scored at the mean of Need for Cognition
(␤⫽.14), t(470) ⫽ 2.67, p ⫽ .008, but not for those who were low
in Need for Cognition (⫺1 SD; ␤⫽.03, t ⫽ .39, ns). Thus,
supporting the notion that the two information-processing systems
interact to produce responses (Epstein, 1994), some minimum
degree of rational engagement may be necessary for implicit SE to
influence explicit SE.
The results of Study 2 thus parallel those of Study 1 and further
illustrate that people who trust their intuition demonstrate closer
correspondence between implicit and explicit SE. Overall, indi-
viduals who trust their intuition showed a positive relation between
implicit and explicit SE, suggesting that they may incorporate their
implicit SE into their explicit self-views. Individuals with rela-
tively low trust in their intuition, in contrast, showed no relation or
even a negative relation between implicit and explicit SE. The
results of these two studies suggest that people experience implicit
SE as intuition. The results of Study 2 further suggest that cogni-
tive engagement may contribute to this effect. People who were
moderate to high in Need for Cognition showed more correspon-
dence between implicit and explicit SE to the extent that they
perceived their intuition to be valid. People who were low in Need
for Cognition, however, did not.
It is also interesting to consider these results in relation to
Pelham et al.’s (2005) finding that women demonstrate greater
implicit– explicit SE consistency than men. This gender difference
is variable across studies (Greenwald & Farnham, 2000; Riketta,
2005). Indeed, we found no such difference in Study 2. Gender did
not moderate implicit– explicit SE consistency overall ( p ⬎ .50).
We did, however, find that Faith in Intuition moderated the cor-
respondence between implicit and explicit SE for both men and
women. This pattern of results suggests that Pelham et al. were
correct to highlight the role of intuition in their findings. Because
women tend to show greater Faith in Intuition than men, they may
often show greater correspondence between implicit and explicit
SE. The operative variable, however, appears to be trust in intu-
ition, rather than gender per se.
Study 3
The results of Studies 1 and 2 suggest that dispositional trust in
intuition may lead to increased correspondence between implicit
and explicit SE. Over time and repeated experiences, individuals
who trust their intuition may incorporate their implicit SE into
their explicit SE. We also wondered, however, whether explicit
self-views are more dynamically affected by situational factors
that influence the degree to which people perceive their intuition as
valid. In addition, the correlational nature of Studies 1 and 2 makes
it unclear whether trust in intuition causally affects the correspon-
dence between implicit and explicit SE. In Study 3, we thus
experimentally induced some participants to perceive their intu-
Table 2
Correlations and Descriptive Statistics for Measures of Explicit
Self-Esteem, Implicit Esteem, Faith in Intuition, and Need for
Cognition: Study 2
Measure 1234
1. Explicit self-esteem —
2. Implicit self-esteem .01 —
3. Faith in Intuition subscale .17
**
⫺.05 —
4. Need for Cognition subscale .19
**
.11
*
.21
**
—
M 4.12 0.76 3.57 3.56
SD 0.65 1.03 0.58 0.62
*
p ⬍ .05.
**
p ⬍ .001.
1072
JORDAN, WHITFIELD, AND ZEIGLER-HILL

ition as valid. We expected that these participants would report
explicit self-views that are more consistent with their implicit SE.
We reasoned, however, that standard measures of explicit SE
might not be sensitive enough to detect this effect because they are
highly stable and resistant to situational pressures (e.g., Rosenberg,
1965; Silber & Tippett, 1965). We thus sought a measure of
explicit SE that would be highly sensitive to the changes induced
by situational differences in the perceived validity of intuition. To
this end, we developed a measure of self-evaluations made under
time pressure (see also Koole, Dijksterhuis, & van Knippenberg,
2001). By having participants report their explicit self-views
quickly, we felt that intuition could more directly affect self-
reports. In addition, time pressure should prevent people from
defensively distorting self-views, which would make it more dif-
ficult for us to detect our predicted effects. We expected that
implicit SE would correspond more closely to self-evaluations
made under time pressure for individuals induced to trust their
intuition.
Method
Participants
Sixty-two undergraduates enrolled in psychology courses par-
ticipated in exchange for course credit. Three participants had error
rates of more than 20% on the IAT, and their data were excluded
from analyses. One participant’s self-evaluations were not re-
corded because of a technical failure. Results are reported for the
remaining 58 participants (18 men and 40 women). Gender did not
influence the results of this study and so is not discussed further.
Materials and Procedure
Up to 8 participants at a time participated in sessions, although
they were seated at individual workstations and did not interact
with each other. Participants were met by a male experimenter who
described the study as an investigation of decision making. They
were informed that they would complete a series of computer-
based reaction time tasks and some measures of personality and
affective experiences.
Implicit SE. Participants first completed the IAT to measure
implicit SE. As in Study 1, it showed acceptable reliability
(Spearman–Brown coefficient ⫽ .73).
Need for Cognition. Participants next completed Cacioppo
and Petty’s (1982) Need for Cognition scale.
Manipulation of the perceived validity of rational analysis or
intuition. After some practice making speeded self-evaluations
(described below), we induced participants to trust either rational
analysis or intuition. We did so by having participants make a
series of decisions either rationally or intuitively. Half of the
participants were told, “There is clear evidence that people who
adopt an intuitive approach to decision making are more successful
in many areas of their lives.” Such individuals were said to excel
in business, be popular, and have more successful romantic rela-
tionships. The remaining participants were told that “people who
adopt a rational approach to decision making” are more successful.
We told participants that we were interested in why intuitive (or
rational) decision makers are more successful and that we wanted
to see how well they made such decisions. To this end, they were
shown a series of fictitious personality profiles and were asked to
decide whether they would live with each person in a co-ed house.
They were asked to make decisions either rationally or intuitively,
depending on condition.
Participants in the intuitive condition were encouraged to use
gut feelings to decide, relying on how they felt overall about each
person. They were asked to work quickly and not think too much
about their decision. Participants in the rational condition were
instead encouraged to decide carefully, to write down each con-
sideration and why they felt it was important. They were encour-
aged to list pros and cons. All participants then saw seven person-
ality profiles presented by computer, four of which were for men.
Each profile included a short written description and a photograph
of the individual. Profiles included information that most people
would likely deem to be relevant to deciding (e.g., the person’s
agreeableness, responsibility, or finances), as well as information
3.5
3.7
3.9
4.1
4.3
4.5
4.7
4.9
Lowest Highest
Implic it self-esteem
Explicit self-esteem
Low FI (-1 SD)
Mean FI
High FI (+1 SD)
Figure 2. Explicit self-esteem as a function of Faith in Intuition (FI) and implicit self-esteem (Implicit
Self-Esteem Scale scores) in Study 2. Predicted values displayed are for the full range of implicit self-esteem
represented in the sample (–2.5 to 3.0). Lines are depicted at the mean and at 1 standard deviation above and
below the mean of Faith in Intuition.
1073
IMPLICIT AND EXPLICIT SELF-ESTEEM

that would likely be more peripheral (e.g., the person’s hobbies,
tastes in movies, or career aspirations). For example, one profile
indicated that “Trent” “treats everyone with the utmost respect and
decency” and that he “spends his weekends with his competitive
bowling team.” Participants indicated whether they would accept
each individual as a housemate. Although the instructions to be
intuitive or rational applied only to this task, we hoped our ma-
nipulation would induce a general trust in intuition or rational
analysis that would carry over into subsequent tasks (cf. Muss-
weiler, 2001).
Speeded self-evaluation task. At the beginning of the study,
participants were told that they would rate their own personalities
several times during the study. They were told that we were
interested in how their ratings related to their decision-making
styles and that for some of their responses they would be required
to respond quickly.
Before the manipulation, participants practiced the self-evaluation
task, although these trials were not identified as practice trials to
participants. Adjectives appeared, one at a time, at the center of their
computer screens. Participants indicated whether each adjective de-
scribed them, using a 4-point scale. The 1, 2, 8, and 9 keys at the top
of the keyboard corresponded to values of 1 (extremely uncharacter-
istic of me),2(somewhat uncharacteristic of me),3(somewhat
characteristic of me), and 4 (extremely characteristic of me), respec-
tively. Without time pressure, participants first responded to 20 ad-
jectives that were neutral in terms of their social desirability (based on
norms provided by Anderson, 1968), such as modern, musical, loyal,
and unique. This familiarized them with the response scale. They then
practiced evaluations under time pressure. They responded to another
29 neutral adjectives and were instructed to respond within1stothese
adjectives. If they did not, the message “Please try to respond faster!”
appeared for 2 s. After this practice, participants received the manip-
ulated focus on intuition or rational analysis described above. We
gave this practice before the manipulation so that the critical trials of
the self-evaluation task could be completed shortly after the manip-
ulation.
Indeed, immediately after the manipulation, participants com-
pleted another 10 neutral trials to refamiliarize themselves with the
task. They then completed the critical test trials, consisting of 10
positive adjectives (e.g., confident, likable, good, and secure) and
10 negative adjectives (e.g., incompetent, flawed, useless, and
insecure) derived from a measure of state SE (McFarland & Ross,
1982). Responses to negative adjectives were reversed, and re-
sponses to all adjectives were then averaged to create an index of
positive self-evaluation (Cronbach’s ␣⫽.82). All responses
longer than 1,300 ms (2.9% of responses) or shorter than 300 ms
(0.3% of responses) were first excluded.
Manipulation check. Last, participants completed a modified
REI. This scale was identical to that used in Study 1, except that
participants were asked to “please answer according to how you
feel right now at this moment. Do not worry about what you are
generally like or how you might have felt in the past.” We could
thus assess whether our manipulation affected Faith in Intuition
and Need for Cognition.
Results and Discussion
Following the manipulation, participants reported more Faith in
Intuition in the intuitive condition (M ⫽ 3.70, SD ⫽ 0.56) than in
the rational condition (M ⫽ 3.30, SD ⫽ 0.82), t(55) ⫽⫺2.16, p ⫽
.03. The manipulation appeared to primarily target intuition, how-
ever. Participants in the intuitive condition (M ⫽ 3.62, SD ⫽ 0.55)
and in the rational condition (M ⫽ 3.85, SD ⫽ 0.66) reported equal
Need for Cognition, t(55) ⫽ 1.45, ns.
Overall, participants showed positive IAT scores (M ⫽ .60,
SD ⫽ .50) and speeded self-evaluations (M ⫽ 3.06, SD ⫽ 0.31).
Speeded self-evaluations and IAT scores were not, overall, corre-
lated with each other (r ⫽ .15, ns). To test whether situational trust
in intuition moderates the correspondence between implicit and
explicit SE, we conducted a multiple regression analysis predicting
speeded self-evaluations from IAT scores and condition (effect
coded as intuitive ⫽ 1, rational ⫽⫺1). We also included prema-
nipulation Need for Cognition in our analyses. Because it did not
moderate any of the results, we dropped its interaction terms from
the model, retaining it only as a covariate.
We expected a stronger relation between IAT scores and self-
evaluations made under time pressure in the intuitive condition
relative to the rational condition. Consistent with this prediction,
the interaction between IAT and condition was significant (␤⫽
.28), t(53) ⫽ 2.35, p ⫽ .02. As can be seen in Figure 3, there was
a significant positive relation between IAT scores and self-
evaluations in the intuitive condition (␤⫽.55), t(53) ⫽ 3.04, p ⫽
.004, but not in the rational condition (␤⫽⫺.02), t(53) ⫽⫺0.10,
ns. Thus, consistent with the results of Studies 1 and 2, greater trust
in intuition increased the correspondence between implicit and
explicit SE. In this case, however, trust was experimentally in-
duced, demonstrating a causal role for trust in intuition. In addi-
tion, the self-evaluations made in this study were made under time
pressure, and so were highly efficient. They may thus be relatively
automatic, making the fact that they were influenced by our
manipulation particularly remarkable (cf. Koole et al., 2001).
Notably, participants in the rational condition showed no neg-
ative relation between implicit and explicit SE, as was found in
Studies 1 and 2 for participants with relatively low dispositional
faith in intuition. Earlier, we speculated that this negative relation
in Studies 1 and 2 might reflect a correction process in which some
individuals adjust their explicit SE for the potential bias posed by
implicit SE. If that is the case, there may be methodological
reasons why it did not occur in the rational condition of the present
study. First, the manipulation induced participants in this condition
to trust rational analysis, but did not actively discredit intuition.
Thus, participants in this condition were not likely to see intuition
as a source of bias, and so would not be motivated to correct their
self-evaluations for the influence of implicit SE. Second, correc-
tion processes are believed to be effortful, controlled processes that
should be disrupted by time pressure (Wegener & Petty, 1995).
Thus, even if some participants were motivated to correct their
self-evaluations, the time pressure in Study 3 likely prevented
them from doing so. In Study 4, we sought to remove these
methodological constraints on correction. We specifically discred-
ited intuition for some participants and encouraged them to over-
come the influence of intuition in their judgments. They also
reported their self-evaluations without time pressure.
It is also worth noting that Need for Cognition did not moderate the
results of Study 3. Study 2 suggested that some rational engagement
may be necessary for implicit SE to affect explicit self-views. In Study
3, however, even under conditions of time pressure, a focus on
intuition increased the correspondence between implicit and explicit
1074
JORDAN, WHITFIELD, AND ZEIGLER-HILL

SE, regardless of participants’ levels of Need for Cognition. This may
be because of the more statelike nature of the self-evaluations made in
this study. On any one occasion, if people are induced to trust their
intuition, they may incorporate implicit SE into explicit, state SE.
However, to see an effect on trait SE may require people to consider
their implicit SE and use it in forming explicit self-views over mul-
tiple occasions. This might only occur for people who have at least a
moderate Need for Cognition.
Study 4
Study 4 aimed to replicate Study 3 and extend it in several ways.
First, in Study 3 it is not clear which experimental condition
affected the correspondence between implicit and explicit SE. It is
possible that rational analysis disrupted the correspondence be-
tween implicit SE and speeded self-evaluations, rather than trust in
intuition increasing it. We thus added a condition in which partic-
ipants were told nothing about the validity of intuition. Relative to
this control condition, we expected participants induced to trust
intuition would report explicit SE that corresponds more closely to
implicit SE.
In addition, in Study 3 our manipulation artificially placed
intuition and rational analysis in opposition to each other. Rational
analysis and intuition may, however, be better conceptualized as
independent, but interacting, cognitive processes (Epstein, 1990).
Thus, in Study 4, we sought to manipulate only beliefs about the
validity of intuition, while encouraging all participants to view
rational analysis as valid. In this way, we could focus more
exclusively on the role of intuition in the correspondence between
implicit and explicit SE. We thus included three conditions in
Study 4: one in which participants were induced to perceive
intuition as valid, a control condition, and one in which partici-
pants were induced to perceive intuition as invalid.
Last, in Study 3, the measure of explicit self-views was some-
what unorthodox. We examined self-evaluations made under time
pressure because they were likely to be sensitive to situational
changes in self-views and unlikely to be distorted by defensive
responding. This measure was explicit in the sense that participants
knew they were reporting self-evaluations. Because the self-
evaluations were made under time pressure, however, they were
highly efficient, and in this sense may be relatively automatic. As
noted earlier, it is thus remarkable that we observed differences in
these evaluations as a function of our manipulation. In Study 4,
however, we wanted to test the robustness of our findings by
exploring whether the same effects would hold for self-evaluations
made without time pressure. We thus examined state SE as our
primary dependent measure.
Method
Participants
One hundred fifteen undergraduates enrolled in psychology courses
participated in exchange for course credit. Six participants had error
rates of more than 20% on the IAT, and their data were excluded from
analyses. Two participants did not complete the RSES (described
below), and 3 indicated that English was not their first language.
Another 2 participants were excluded because of their suspicions
about the experiment. Results are reported for the remaining 102
participants (35 men and 67 women). Gender did not influence the
results of this study and so is not discussed further.
Materials and Procedure
With a few exceptions (specified below), the procedure and
materials were identical to those of Study 3. Participants com-
pleted the RSES, as a measure of explicit SE, as part of a mass
pretesting session roughly 2 months before the experimental ses-
sion. This measure was identical to that used in Study 1. In the lab,
participants completed the IAT as a measure of implicit SE and
Cacioppo and Petty’s (1982) Need for Cognition scale, in that
order. These scales were identical to those used in Study 3. The
IAT again showed acceptable reliability (Spearman–Brown coef-
ficient ⫽ .61).
Manipulation of the perceived validity of intuition. We next
manipulated the perceived validity of intuition. This manipulation
1.5
1.7
1.9
2.1
2.3
2.5
2.7
2.9
3.1
Lowest Highest
Implicit self-esteem
Speeded self-evaluation
Trust intuition
Trust reas on
Figure 3. Speeded self-evaluation as a function of the manipulation of the validity of intuition and implicit
self-esteem (Implicit Association Test scores) in Study 3. Predicted values displayed are for the full range of
implicit self-esteem represented in the sample (– 0.68 to 2.16).
1075
IMPLICIT AND EXPLICIT SELF-ESTEEM

was identical to that used in Study 3, with the following excep-
tions. One third of participants were induced to perceive intuition
as valid, one third were induced to perceive intuition as invalid,
and one third (the control group) received no information about the
validity of intuition. All participants were told that people ap-
proach decisions through some combination of rational and intu-
itive strategies. Control participants were told nothing further.
Participants in the other two conditions were also told that, in
general, rational analysis improves the quality of decisions.
Participants induced to view intuition as valid were further told
that “paying attention to and trusting intuitions” improves deci-
sions and is related to success. In contrast, participants induced to
perceive intuition as invalid were told that “too much attention to
and relying too much on intuitions often undermines decisions.”
They were told that people who can overcome the influence of
intuition on their decisions are more successful. We told partici-
pants in these two conditions that we were interested in why
people who trust (or can overcome) intuition are more successful
and that we wanted to see how well they made decisions intuitively
(or by preventing intuition from affecting their decisions). Partic-
ipants in the control condition were told that we were interested
simply in observing how they normally make decisions. All par-
ticipants were then shown the personality profiles used in Study 3
and were asked to decide, for each person, whether they would live
with that person in a co-ed house. While deciding, they were asked
to “attend to and trust your intuitions” (trust intuition condition), to
“keep your intuitions from influencing your decisions” (overcome
intuition condition), or to “use whatever information seems most
relevant to you” (control condition).
Explicit state SE. After this manipulation, participants com-
pleted Heatherton and Polivy’s (1991) state SE scale, exactly as in
Study 1.
Manipulation check. Last, participants completed the modi-
fied REI that was used in Study 3 as a manipulation check.
Results and Discussion
The manipulation check suggested that the manipulation was
generally effective, F(2, 99) ⫽ 9.92, p ⬍ .001. Participants in the
trust intuition condition (M ⫽ 3.95, SD ⫽ 0.45) tended to report
more Faith in Intuition than those in the control condition (M ⫽
3.72, SD ⫽ 0.62), although not significantly so, t(99) ⫽ 1.49, p ⫽
.14. Likewise, participants in the overcome intuition condition
(M ⫽ 3.28, SD ⫽ 0.74) reported less Faith in Intuition than those
in the control condition, t(99) ⫽⫺2.98, p ⫽ .004. Thus, overall,
the effect of the manipulation on Faith in Intuition was in the
predicted direction for each condition but was somewhat weak. It
is worth noting that this relatively weak manipulation should only
make it more difficult for us to observe our predicted effects.
In contrast, the manipulation had little impact on Need for
Cognition, F(2, 99) ⫽ 1.62, p ⫽ .20. Participants in the trust
intuition condition (M ⫽ 3.31, SD ⫽ 0.59) did not report less Need
for Cognition than control participants (M ⫽ 3.37, SD ⫽ 0.51),
t(99) ⫽⫺0.41, p ⫽ .69. Likewise, those in the overcome intuition
condition (M ⫽ 3.55, SD ⫽ 0.58) did not report more Need for
Cognition than control participants, t(99) ⫽ 1.35, p ⫽ .18.
4
Overall, participants showed positive IAT scores (M ⫽ .70,
SD ⫽ .45) and state SE (M ⫽ 3.69, SD ⫽ 0.62). To test whether
our manipulation of the validity of intuition affected the corre-
spondence between implicit and explicit SE, we conducted a
multiple regression analysis predicting state SE from IAT scores
(centered) and condition (effect coded as trust intuition ⫽ 1, 0;
overcome intuition ⫽ 0, 1; control condition ⫽⫺1, ⫺1). We also
controlled premanipulation Need for Cognition and RSES scores.
The main effects were entered before the interaction terms. As
predicted, the interaction between IAT and condition was signif-
icant (R
2
change ⫽ .065), F(2, 94) ⫽ 3.60, p ⫽ .03.
Decomposing this interaction into its two effect-coded vectors
revealed that the interaction was significant when comparing the
trust intuition condition to the control condition (␤⫽.29), t(94) ⫽
2.36, p ⫽ .02, but not when comparing the overcome intuition
condition to the control condition (␤⫽⫺.05), t(94) ⫽⫺0.418,
p ⫽ .68. As can be seen in Figure 4, there was a significant positive
relation between IAT scores and state SE in the trust intuition
condition (␤⫽.52), t(94) ⫽ 2.10, p ⫽ .01, but not in the overcome
intuition condition (␤⫽⫺.14) or the control condition (␤⫽.10;
both ps ⬎ .30). Thus, consistent with the results of Study 3,
inducing participants to perceive their intuition as valid led to
greater correspondence between implicit and explicit SE. In this
case, this effect was observed on a more traditional measure of
state SE without time pressure.
In contrast, inducing participants to perceive their intuition as
invalid did not influence the relation between implicit and explicit
SE. This is consistent with the results of Study 3. In that study,
there were methodological factors that might explain why partic-
ipants in the rational condition did not correct their explicit self-
views for the potential bias of implicit SE. In Study 4, in contrast,
we specifically led participants to view their intuition as less valid.
We encouraged participants to try to overcome the influence of
intuition in their judgments, and yet we observed no negative
relation between implicit and explicit SE.
General Discussion
Across four studies, we found that greater perceived validity of
intuition was associated with greater consistency between implicit
and explicit SE. In Studies 1 and 2, people who chronically trust
their intuition demonstrated closer correspondence between im-
plicit and explicit SE. In Studies 3 and 4, people who were induced
to view their intuition as more valid reported explicit SE that
corresponded more closely with their implicit SE; this was true for
self-evaluations made under time pressure (Study 3) and for state
SE reported without time pressure (Study 4). These latter findings
highlight the causal role of trusting intuition in the correspondence
between implicit and explicit SE. When people view their intuition
as valid, they may incorporate their implicit SE more into their
explicit, overtly expressed self-views.
In addition, Studies 1 and 2 demonstrated that people with
relatively low dispositional trust in intuition have an inverse rela-
tion between implicit and explicit SE. These individuals may view
their implicit SE as a source of bias in their judgments and
consequently adjust their explicit SE away from their implicit SE.
4
This latter effect could be considered to be a trend by some standards, so
we also conducted all of the main analyses controlling postmanipulation Need
for Cognition. Because the results were nearly identical whether this variable
was controlled or not, we report the results without controlling for it.
1076
JORDAN, WHITFIELD, AND ZEIGLER-HILL

This interpretation is speculative and has limitations. It is difficult
to square with the fact that people who were specifically induced
to distrust their intuition in Study 4 showed no inverse relation
between implicit and explicit SE. It may be that our manipulation
was not strong enough to induce the kind of correction processes
that are stimulated by dispositional reservations about intuition. In
addition, however, it is not clear how strong participants’ reserva-
tions about intuition really were in Studies 1 and 2. Our “low”
Faith in Intuition participants typically scored near the neutral
midpoint of the scale (but see Blanton & Jaccard, 2006). Clearly,
further research is needed to better understand this finding.
Overall, the results of the present studies provide a partial
answer to the question posed by Fazio and Olson (2003, p. 304):
“When, under what conditions, and for what kind of people, are
implicit and explicit measures related?” With respect to implicit
and explicit SE, one answer seems to be that people who trust their
intuition or those who are induced to trust their intuition report
explicit self-evaluations that correspond more closely to their
implicit SE. We strongly suspect that the same is true of other
attitudes. When people trust their intuition, we expect that their
implicit and explicit attitudes will, in general, be more closely
related to each other.
These results also have implications for understanding whether
and how people experience implicit SE. They suggest that people
experience implicit SE as intuition or gut feelings about the self.
Although we focused on the role of perceived validity of intuition
in the correspondence between implicit and explicit SE, our results
imply awareness of implicit SE. For people to judge the validity of
evaluations, they must logically be aware of them. Gawronski and
Bodenhausen (2006) argued persuasively that subjective assess-
ments of validity occur only within a propositional, rule-based
system. Associatively activated evaluations may be translated into
conscious propositions (e.g., “I feel badly about myself”; see also
Hofmann, Gschwendner, & Schmitt, 2005). Only then can people
judge whether these evaluations are valid—whether they person-
ally endorse them— or not. Our findings suggest that people who
perceive intuition as valid incorporate implicit SE into their ex-
plicit self-views. This suggests that people are aware of their
implicit SE and experience it as intuitive evaluations of the self.
Awareness of implicit SE may help to explain why individuals with
high explicit but low implicit SE are defensive (Bosson et al., 2003;
Jordan et al., 2003, 2005; McGregor & Marigold, 2003; McGregor et
al., 2005; Zeigler-Hill, 2006). Jordan and colleagues (2003, 2005)
suggested that implicit SE is preconscious, occasionally entering
awareness. When individuals with high explicit but low implicit SE
become aware of their relatively low implicit SE, this may create an
aversive inconsistency within the self (Jordan, Logel, Spencer, Zanna,
& Whitfield, in press). Thus, these individuals may be motivated to
affirm their positive explicit SE and deny their negative implicit SE,
causing them to react defensively. Although this reaction may foster
defensiveness, it may also be adaptive in some ways. It suggests that
these individuals are aware of having low implicit SE but perceive it
to be invalid. If they instead perceived it as valid, they might be at
greater risk for depression (Beevers, 2005; Jordan, Logel, Spencer, &
Zanna, 2006).
The Role of Need for Cognition and Cognitive
Elaboration
It is notable that Need for Cognition moderated the results of
Study 2. People who were moderate to high in Need for Cognition
showed closer correspondence between implicit and explicit SE to
the extent that they were higher in Faith in Intuition. People who
were low in Need for Cognition, however, did not. This finding
may reflect the interactive nature of associative and rule-based
processing commonly asserted in dual-system models of cognition.
It may also suggest that a minimal degree of cognitive engagement
is necessary for implicit SE to affect explicit SE, at least at a
dispositional level. This makes sense to the extent that such influ-
ence depends on people representing implicit SE propositionally
and then judging its validity (Gawronski & Bodenhausen, 2006;
Hofmann, Gschwendner, & Schmitt, 2005). It must be acknowl-
edged, however, that this moderation effect did not approach
significance in Study 1.
2
2.5
3
3.5
4
4.5
5
Lowest Highest
Implicit self-esteem
State self-esteem
Control condition
Overcome intuition
Trus t int uition
Figure 4. State self-esteem as a function of the manipulation of the validity of intuition and implicit self-esteem
(Implicit Association Test scores) in Study 4. Predicted values displayed are for the full range of implicit
self-esteem represented in the sample (–0.38 to 1.99).
1077
IMPLICIT AND EXPLICIT SELF-ESTEEM

In addition, the role of Need for Cognition in the correspon-
dence between implicit and explicit attitudes may be complicated.
It is likely not the case that greater cognitive elaboration will result
in uniformly closer implicit–explicit attitude consistency. Need for
Cognition has, for example, been associated with less correspon-
dence between implicit and explicit prejudice (Florack et al.,
2001). Gawronksi and Bodenhausen (2006) gave considerable
attention to the role of cognitive elaboration in the correspondence
between implicit and explicit attitudes. They noted that as cogni-
tive elaboration increases, people are likely to incorporate more
sources of information into explicit evaluations. If this additional
information is consistent with implicit attitudes, it should enhance
the relation between implicit and explicit attitudes (or leave it
unaltered). If additional information contradicts implicit attitudes,
however, it may reduce this correspondence.
Conclusions
People who view intuition as valid may incorporate implicit SE
into their explicit self-views. People with relatively low disposi-
tional faith in their intuition, in contrast, demonstrate a negative
relation between their implicit and explicit SE. This latter finding
may suggest that these individuals adjust or “correct” their explicit
self-views for the potential bias of implicit SE. Regardless of how
these findings are interpreted, however, they have important im-
plications for understanding the relation between implicit and
explicit SE. The two are normally uncorrelated. This lack of
correlation, however, does not mean that implicit SE and explicit
SE are independent. They are actually related, although in opposite
ways for different kinds of people. People who trust their intuition
have a positive relation between them, whereas people who do not
trust their intuition have either no relation or a negative relation.
The current findings may also have broad implications for
understanding implicit attitudes and SE. The adoption of the terms
implicit and explicit to describe relatively automatic and controlled
attitudes (by Greenwald & Banaji, 1995) may have inadvertently
shaped much of the theoretical debate about implicit attitudes.
Many researchers have thus asked whether people are aware of
implicit attitudes, perhaps because the term implicit implies lack of
awareness (Fazio & Olson, 2003). If different terms had been
adopted, however, such as intuitive and analytic (see Greenwald
and Banaji, 1995, footnote 1), this issue might have been ap-
proached differently. Definitions of intuition and associative pro-
cesses might suggest a refinement of the question of awareness.
We could ask, separately, whether people are aware of their
automatic evaluations and whether they are aware of the processes
that produce them. The present findings suggest that these ques-
tions may have different answers. People may be aware of their
implicit evaluations, but not where they come from. They may thus
experience implicit attitudes and SE as intuition.
References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting interactions. London: Sage Publications.
Anderson, N. H. (1968). Likeability ratings of 555 personality-trait words.
Journal of Personality and Social Psychology, 9, 272–279.
Aronson, E., Ellsworth, P. C., Carlsmith, J. M., & Gonzales, M. H. (1990).
Methods of research in social psychology (2nd ed.). Boston: McGraw-Hill.
Bargh, J. A. (1994). The four horsemen of automaticity: Awareness,
intention, efficiency, and control in social cognition. In R. S. Wyer Jr.,
& T. K. Srull (Eds.), Handbook of social cognition (2nd ed., pp. 1– 40).
Hillsdale, NJ: Erlbaum.
Bargh, J. A., Chaiken, S., Raymond, P., & Hymes, C. (1996). The auto-
matic evaluation effect: Unconditional automatic attitude activation with
a pronunciation task. Journal of Experimental Social Psychology, 32,
104 –128.
Barrett, L., Tugade, M. M., & Engle, R. W. (2004). Individual differences
in working memory capacity and dual-process theories of mind. Psy-
chological Review, 130, 553–573.
Beevers, C. G. (2005). Cognitive vulnerability to depression: A dual
process model. Clinical Psychology Review, 25, 975–1002.
Blanton, H., & Jaccard, J. (2006). Arbitrary metrics in psychology. Amer-
ican Psychologist, 61, 27– 41.
Bosson, J. K., Brown, R. P., Zeigler-Hill, V., & Swann, W. B. (2003).
Self-enhancement tendencies among people with high explicit self-
esteem: The moderating role of implicit self-esteem. Self and Identity, 2,
169 –187.
Bosson, J. K., Swann, W. B., & Pennebaker, J. W. (2000). Stalking the
perfect measure of implicit self-esteem: The blind men and the elephant
revisited? Journal of Personality and Social Psychology, 79, 631–643.
Cacioppo, J. T., & Petty, R. E. (1982). The need for cognition. Journal of
Personality and Social Psychology, 42, 116–131.
Carver, C. S. (2005). Impulse and constraint: Perspectives from personality
psychology, convergence with theory in other areas, and potential for
integration. Personality and Social Psychology Review, 9, 312–333.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/correlation
analyses for the behavioral sciences. Hillsdale, NJ: Erlbaum.
Epstein, S. (1990). Cognitive–experiential self-theory. In L. Pervin (Ed.),
Handbook of personality: Theory and research (pp. 165–192). New
York: Guilford Press.
Epstein, S. (1994). Integration of the cognitive and psychodynamic uncon-
scious. American Psychologist, 49, 709–724.
Epstein, S., & Pacini, R. (1999). Some basic issues regarding dual-process
theories from the perspective of cognitive-experiential self theory. In S.
Chaiken & Y. Trope (Eds.), Dual-process theories in social psychology
(pp. 462– 482). New York: Guilford Press.
Epstein, S., Pacini, R., Denes-Raj, V., & Heier, H. (1996). Individual
differences in intuitive– experiential and analytical–rational thinking
styles. Journal of Personality and Social Psychology, 71, 390 –405.
Farnham, S. D., Greenwald, A. G., & Banaji, M. R. (1999). Implicit
self-esteem. In D. Abrams & M. A. Hogg (Eds.), Social identity and
social cognition (pp. 230 –248). Oxford: Blackwell Publishers.
Fazio, R. H., & Olson, M. A. (2003). Implicit measures in social cognition:
Their meaning and use. Annual Review of Psychology, 54, 297–327.
Fazio, R. H., Sanbonmatsu, D. M., Powell, M. C., & Kardes, F. R. (1986).
On the automatic activation of attitudes. Journal of Personality and
Social Psychology, 50, 229 –238.
Florack, F., Scarabis, M., & Bless, H. (2001). When do associations matter?
The use of automatic associations toward ethnic groups in person judg-
ments. Journal of Experimental Social Psychology, 37, 518 –524.
Gawronski, B., & Bodenhausen, G. V. (2006). Associative and proposi-
tional processes in evaluation: An integrative review of implicit and
explicit attitude change. Psychological Bulletin, 132, 692–731.
Gawronski, B., Hofmann, W., & Wilbur, C. J. (2006). Are “implicit”
attitudes unconscious? Consciousness and Cognition, 15, 485– 499.
Greenwald, A. G., & Banaji, M. R. (1995). Implicit social cognition: Attitudes,
self-esteem, and stereotypes. Psychological Review, 102, 4 –27.
Greenwald, A. G., & Farnham, S. D. (2000). Using the Implicit Associa-
tion Test to measure self-esteem and self-concept. Journal of Personality
and Social Psychology, 79, 1022–1038.
Greenwald, A. G., McGhee, D. E., & Schwartz, J. L. K. (1998). Measuring
individual differences in implicit cognition: The Implicit Association
Test. Journal of Personality and Social Psychology, 74, 1464 –1480.
1078
JORDAN, WHITFIELD, AND ZEIGLER-HILL

Greenwald, A. G., Nosek, B., & Banaji, M. R. (2003). Understanding and
using the Implicit Association Test: I. An improved scoring algorithm.
Journal of Personality and Social Psychology, 85, 197–216.
Hammond, K. R. (1996). Human judgment and social policy: Irreducible
uncertainty, inevitable error, unavoidable injustice. New York: Oxford
University Press.
Heatherton, T. F., & Polivy, J. (1991). Development and validation of a
scale for measuring state self-esteem. Journal of Personality and Social
Psychology, 60, 895–910.
Hinton, G. E. (1990). Mapping part–whole hierarchies into connectionist
networks. Artificial Intelligence, 46, 47–76.
Hofmann, W., Gawronski, B., Gschwendner, T., Le, H., & Schmitt, M.
(2005). A meta-analysis on the correlation between the Implicit Asso-
ciation Test and explicit self-report measures. Personality and Social
Psychology Bulletin, 31, 1369 –1385.
Hofmann, W., Gschwendner, T., & Schmitt, M. (2005). On implicit–
explicit consistency: The moderating role of individual differences in
awareness and adjustment. European Journal of Personality, 19, 25– 49.
Hogarth, R. M. (2001). Educating intuition. Chicago: University of Chi-
cago Press.
Jordan, C. H., Logel, C., Spencer, S. J., & Zanna, M. P. (2006). Noncon-
scious self-esteem: Is there something you’re not telling yourself? In
M. H. Kernis (Ed.), Self-esteem issues and answers: A sourcebook of
current perspectives (pp. 60 –68). Philadelphia: Psychology Press.
Jordan, C. H., Logel, C., Spencer, S. J., Zanna, M. P., & Whitfield, M. L.
(in press). The heterogeneity of self-esteem: Exploring the interplay
between implicit and explicit self-esteem. In R. E. Petty, R. H. Fazio, &
P. Brin˜ol (Eds.), Attitudes: Insights from the new implicit measures. New
York: Erlbaum.
Jordan, C. H., Spencer, S. J., & Zanna, M. P. (2005). Types of high
self-esteem and prejudice: How implicit self-esteem relates to ethnic
discrimination among high explicit self-esteem individuals. Personality
and Social Psychology Bulletin, 31, 693–702.
Jordan, C. H., Spencer, S. J., Zanna, M. P., Hoshino-Browne, E., & Correll,
J. (2003). Secure and defensive high self-esteem. Journal of Personality
and Social Psychology, 85, 969–978.
Karpinski, A. (2004). Measuring self-esteem using the Implicit Association
Test: The role of the other. Personality and Social Psychology Bulletin,
30, 22–34.
Kihlstrom, J. F. (2004). Implicit methods in social psychology. In C.
Sansone, C. C. Morf, & A. T. Panter (Eds.), Handbook of methods in
social psychology (pp. 195–212). Thousand Oaks, CA: Sage.
Koole, S. L., Dijksterhuis, A., & van Knippenberg, A. (2001). What’s in a
name: Implicit self-esteem and the automatic self. Journal of Personality
and Social Psychology, 80, 669– 685.
Lieberman, M. D. (2000). Intuition: A social cognitive neuroscience ap-
proach. Psychological Bulletin, 126, 109–137.
Logel, C., Spencer, S. J., Wood, J. V., Holmes, J. G., & Zanna, M. P.
(2006). Guarded and responsive low self-esteem: How implicit self-
esteem functions among low explicit self-esteem individuals. Unpub-
lished manuscript, University of Waterloo, Waterloo, Ontario, Canada.
McFarland, C., & Ross, M. (1982). Impact of causal attribution on affec-
tive reactions to success and failure. Journal of Personality and Social
Psychology, 43, 937–946.
McGregor, I., & Jordan, C. H. (2007). The mask of zeal: Low implicit
self-esteem and defensive extremism after self-threat. Self and Identity 6,
223–237.
McGregor, I., & Marigold, D. C. (2003). Defensive zeal and the uncertain
self: What makes you so sure? Journal of Personality and Social
Psychology, 85, 838 –852.
McGregor, I., Nail, P. R., Marigold, D. C., & Kang, S.-J. (2005). Defensive
pride and consensus: Strength in imaginary numbers. Journal of Per-
sonality and Social Psychology, 89, 978–996.
Moors, A., & De Houwer, J. (2006). Automaticity: A theoretical and
conceptual analysis. Psychological Bulletin, 132, 297–326.
Mussweiler, T. (2001). Seek and ye shall find: Antecedents of assimilation
and contrast in social comparison. European Journal of Social Psychol-
ogy, 31, 499 –509.
Nosek, B. A. (2005). Moderators of the relationship between implicit and
explicit evaluation. Journal of Experimental Psychology: General, 134,
565–584.
Pacini, R., & Epstein, S. (1999). The relation of rational and experiential
information processing styles to personality, basic beliefs, and the ratio-
bias phenomenon. Journal of Personality and Social Psychology, 76,
972–987.
Pelham, B. W., & Hetts, J. J. (1999). Implicit and explicit personal and
social identity: Toward a more complete understanding of the social self.
In T. Tyler & R. Kramer (Eds.), The psychology of the social self (pp.
115–143). Mahwah, NJ: Erlbaum.
Pelham, B. W., Koole, S. L., Hardin, C. D., Hetts, J. J., Seah, E., & DeHart,
T. (2005). Gender moderates the relation between implicit and explicit
self-esteem. Journal of Experimental Social Psychology, 41, 84 –89.
Petty, R. E., & Brin˜ol, P. (2006). A metacognitive approach to “implicit”
and “explicit” evaluations: Comment on Gawronski and Bodenhausen
(2006). Psychological Bulletin, 132, 740–744.
Petty, R. E., & Brin˜ol, P. (in press). The meta-cognitive model (MCM) of
attitudes: Implications for attitude measurement, change, and strength.
Social Cognition.
Petty, R. E., Wheeler, C. S., & Tormala, Z. L. (2003). Persuasion and
attitude change. In T. Millon & M. J. Lerner (Eds.), Handbook of
psychology: Personality and social psychology (Vol. 5, pp. 353–382).
Hoboken, NJ: Wiley.
Riketta, M. (2005). Gender and socially desirable responding as modera-
tors of the correlation between implicit and explicit self-esteem. Current
Research in Social Psychology, 11, 14–28.
Rosenberg, M. (1965). Society and the adolescent self-image. Princeton,
NJ: Princeton University Press.
Silber, E., & Tippett, J. S. (1965). Self-esteem: Clinical assessment and
measurement validation. Psychological Reports, 16, 1017–1071.
Sloman, S. A. (1996). The empirical case for two systems of reasoning.
Psychological Bulletin, 119, 3–22.
Smith, E. R., & DeCoster, J. (2000). Dual-process models in social and
cognitive psychology: Conceptual integration and links to underlying mem-
ory systems. Personality and Social Psychology Review, 4, 108–131.
Smolensky, P. (1988). On the proper treatment of connectionism. Behav-
ioral and Brain Sciences, 11, 1–23.
Spalding, L. R., & Hardin, C. D. (1999). Unconscious unease and self-
handicapping: Behavioral consequences of individual differences in
implicit and explicit self-esteem. Psychological Science, 10, 535–539.
Strack, F., & Deutsch, R. (2004). Reflective and impulsive determinants of
social behavior. Personality and Social Psychology Review, 8, 220 –247.
Wegener, T. D., & Petty, R. E. (1995). Flexible correction processes in
social judgment: The role of naı¨ve theories in corrections for perceived
bias. Journal of Personality and Social Psychology, 68, 36 –51.
Wilson, T. D., Lindsey, S., & Schooler, T. Y. (2000). A model of dual
attitudes. Psychological Review, 107, 101–126.
Zeigler-Hill, V. (2006). Discrepancies between implicit and explicit self-
esteem: Implications for narcissism and self-esteem instability. Journal
of Personality, 74, 119 –143.
Received July 18, 2006
Revision received June 29, 2007
Accepted July 9, 2007 䡲
1079
IMPLICIT AND EXPLICIT SELF-ESTEEM