Content uploaded by Bertram Gawronski
Author content
All content in this area was uploaded by Bertram Gawronski on Dec 12, 2023
Content may be subject to copyright.
in press, Annual Review of Psychology 1
Unawareness of Attitudes, their Environmental Causes, and their Behavioral Effects
Bertram Gawronski1 & Olivier Corneille2
1 University of Texas at Austin, USA
2 UCLouvain, Belgium
Claims about unawareness are abundant in attitude research. The current article provides an analysis of evidence regarding
three aspects of an attitude for which people may lack awareness: (1) the attitude itself, (2) its environmental causes, and (3)
its behavioral effects. Our analysis reveals that, despite widespread claims of unawareness of the three aspects, strong
empirical evidence for these claims is surprisingly scarce. The article concludes with a discussion of the most likely aspects
of attitudes that people may be unaware of; their relation to contextual factors that might influence evaluative responses
outside of awareness; open questions about the (un)awareness of attitudes, their environmental causes, and their behavioral
effects; and methodological recommendations for future research that aims to provide more compelling evidence for aspects
of attitudes that may evade awareness.
Keywords: attitudes; awareness; attitude-behavior relations; consciousness; implicit measures; unconscious learning
A common definition specifies attitude as “a
psychological tendency that is expressed by evaluating
a particular entity with some degree of favor or
disfavor” (Eagly & Chaiken, 2007, p. 582). A major
theme in attitude research pertains to the unawareness
of different aspects of attitudes. Is it possible to hold an
attitude without being aware of that attitude? Can
environmental stimuli influence attitudes outside of
awareness? And can attitudes influence behavioral
responses in a manner that evades awareness? The
current article provides an analysis of evidence relevant
to these questions. To this end, we first describe the
conceptual framework to organize our analysis and then
review evidence pertaining to the three questions
above. Counter to the prevalence of claims about
unawareness in the attitude literature, our analysis
reveals that strong empirical evidence for these claims
is surprisingly scarce. We conclude our analysis with a
discussion of the most likely aspects of attitudes that
people may be unaware of; their relation to contextual
factors that might influence evaluative responses
outside of awareness; open questions about the
(un)awareness of attitudes, their environmental causes,
and their behavioral effects; and methodological
recommendations for future research that aims to
provide more compelling evidence for aspects of
attitudes that may evade awareness.
Conceptual Framework
An important aspect of the above-cited definition is
the distinction between attitude as a latent mental
construct and the behavioral expression of latent
attitudes in overt evaluative responses (Eagly &
Chaiken, 2007).
1
This distinction stipulates that
measures of evaluative responses should not be treated
1
Another important aspect is that attitudes can have affective,
cognitive, and motivational bases. However, whether a given attitude
arises from affective, cognitive, or motivational processes is an
as direct indicators of attitudes, because variance in
evaluative responses can be due to various other factors,
and behavioral expressions of attitudes can be disrupted
by non-attitudinal factors (see Calanchini, 2020; De
Houwer et al., 2013). Thus, when studying attitudes, it
is important to always consider the extent to which
observed differences in evaluative responses are driven
by genuine differences in underlying attitudes or by
other non-attitudinal factors.
Expanding on the definition of attitude, it is possible
to distinguish between three aspects of an attitude for
which people may lack awareness (Gawronski &
Bodenhausen, 2012): the attitude itself, its
environmental causes, and its behavioral effects (see
Figure 1). Although establishing unawareness can be a
difficult methodological endeavor, a common approach
to investigate unawareness of a psychological entity X
is to test whether participants can report X (Nisbett &
Wilson, 1977; Timmermans & Cleeremans, 2015). To
the extent that there is a discrepancy between X and
people’s self-report of X, unawareness would provide a
potential explanation for the observed discrepancy (see
Gawronski & Bodenhausen, 2015). However,
inferences of unawareness additionally require that
there is no alternative explanation that may account for
the observed discrepancy. For example, while some
discrepancies between X and people’s self-report of X
may reflect a genuine inability to report X, other
discrepancies may reflect a motivationally driven
unwillingness to report X, low correspondence between
measures, low measurement reliability, or low
sensitivity in capturing the to-be-measured constructs.
Although inferences of unawareness would seem
justified in cases involving a genuine inability to report
X, they would be premature and potentially
empirical question that goes beyond the definition of the attitude
construct.
in press, Annual Review of Psychology 2
unwarranted in the other cases. Thus, when interpreting
discrepancies between X and self-reports of X as
evidence for unawareness, it is critical to always rule
out alternative explanations for the observed
discrepancies.
Expanding on these considerations, unawareness
claims for the three aspects can be linked to specific
empirical questions. Regarding the presumed
unawareness of attitudes, the central question is
whether there is evidence for attitudes that people are
unable to report. Regarding the presumed unawareness
of the environmental causes of attitudes, the central
question is whether there is evidence for environmental
influences on attitudes when people are unable to report
the stimulus event that is responsible for their attitude
or the causal impact of a stimulus event on their
attitude. Finally, regarding the presumed unawareness
of the behavioral effects of attitudes, the central
question is whether there is evidence that attitudes can
influence behavior when people are unable to report this
behavior or the causal impact of their attitudes on their
behavior.
Unawareness of Attitudes
A widespread assumption in the attitude literature is
that people can have attitudes that they do not know
they have—or put differently: people often do not know
that they like or dislike something. In addition to being
of great theoretical and practical interest, the question
of whether people can be unaware of their attitudes
constitutes the most fundamental one for the current
analysis, because it is logically impossible to have
accurate knowledge about the environmental causes or
behavioral effects of an attitude if one is unaware of the
attitude itself (Gawronski & Bodenhausen, 2012).
To empirically establish unawareness of an attitude,
research requires an indicator of actual (dis)liking and
a measure of people’s beliefs about their (dis)liking.
The most common approach is to compare responses on
so-called “objective” and “subjective” measures of
(dis)liking, with discrepancies between the two being
interpreted as evidence for unawareness. A central
assumption underlying this approach is that the
objective indicator of (dis)liking can be used as a
normative criterion for judging the (in)accuracy of
participants’ subjective reports of their (dis)liking
(Kruglanski, 1989). But how do we know that objective
indicators capture a person’s (dis)liking of an object
better than the person’s subjective self-report? This
question highlights the delicate issue that inferences of
unawareness are often based on claims by researchers
that they know better what their participants like or
dislike than the participants themselves, which
fundamentally depends on the validity of the so-called
“objective” measure. To the extent that the validity of
the “objective” measure seems questionable, inferences
of unawareness would be based on a weak foundation,
which can lead to inaccurate attitude theories
(Gawronski & Bodenhausen, 2015) and potentially
harmful consequences at the individual and the societal
level (Cameron et al., 2010; Daumeyer et al., 2019).
Another important issue for the interpretation of
discrepancies between objective and subjective
measures is that the two measures have similarly high
reliability and sensitivity (Shanks & St. John, 1994),
use the same attitudinal stimuli (Gawronski, 2019), and
are not confounded with other factors that are different
from awareness, such as the timing of responding to the
focal stimuli (Moors, 2016). Otherwise, discrepancies
between objective and subjective measures may be
driven by any of these factors, which undermines
inferences of unawareness.
Physiological vs. Self-Report Measures
One approach based on the distinction between
objective and subjective measures is to treat
physiological responses to an object as objective
indicators of attitudes, and responses on self-report
measures as subjective indicators of participants’
beliefs about their attitudes (Cunningham et al., 2009;
Ito & Cacioppo, 2007). To the extent that responses on
the two measures diverge, participants are often
assumed to be unaware of the attitudes captured by the
physiological measure.
Although this may be the case, there are several
issues that undermine straightforward inferences of
unawareness from dissociations between physiological
and self-report measures. First, many physiological
measures capture responses to stimuli that are much
faster than responses on traditional self-report measures
(e.g., event-related potentials). In these cases,
discrepancies between physiological and self-report
measures may be driven by differences in the time-
window of measured responses rather than lack of
awareness (Cunningham et al., 2007; Moors, 2016).
Second, many physiological measures suffer from low
reliability, which is less common for traditional self-
report measures (Krosnick et al., 2005). This difference
can lead to systematic dissociations for simple
methodological reasons that have nothing to do with
lack of awareness (i.e., reliability problem; see Shanks
& St. John, 1994). Third, to serve as an “objective”
indicator of attitudes, a physiological measure must
capture responses along the valence dimension rather
than responses along other dimensions (e.g., arousal).
Thus, if responses on a physiological and a self-report
measure do not align, a potential explanation is that the
physiological measure captures responses along a
dimension that is different from valence (i.e., sensitivity
problem; see Shanks & St. John, 1994). Fourth,
although the latter problem can be addressed via
thorough validation of the physiological measure,
studies on the construct validity of physiological
in press, Annual Review of Psychology 3
measures require knowledge about the valence of the
utilized stimuli. Yet, this knowledge typically comes
from studies using self-report measures (e.g., Lang et
al., 2008), which leads to an inferential paradox for
research that relies on physiological measures to study
unawareness of attitudes. To ensure that a physiological
measure captures responses along the valence
dimension, responses on the measure must converge to
those on self-report measures. Yet, to demonstrate
unawareness of attitudes, responses on physiological
measures need to diverge from those on self-report
measures. Together, these issues create major problems
for research that aims to demonstrate unawareness of
attitudes via discrepancies between physiological and
self-report measures. These problems may at least
partly explain why recent research on unawareness of
attitudes has moved away from treating physiological
measures as objective indicators of attitudes.
Indirect vs. Direct Measures
A much more popular approach to study unawareness
of attitudes is to compare self-reported evaluations on
direct measures to responses on a particular type of
indirect measures, such as the Implicit Association Test
(IAT; Greenwald et al., 1998), the Evaluative Priming
Task (EPT; Fazio et al., 1995), and the Affect
Misattribution Procedure (AMP; Payne et al., 2005).
The background assumption underlying this approach
is that responses on indirect measures can be treated as
objective indicators of attitudes, whereas responses on
direct measures are subjective indicators of
participants’ beliefs about their attitudes. Thus, in line
with the argument outlined at the beginning of this
section, discrepancies between direct and indirect
measures have been claimed to reflect unawareness of
attitudes captured by indirect measures, which is
reflected in descriptions of the relevant instruments as
implicit measures and the attitudes captured by these
instruments as implicit attitudes (Gawronski &
Brannon, 2019; for a critique of the implicit
terminology, see Corneille & Hütter, 2020).
Inferences of unawareness from discrepancies
between direct and indirect measures suffer from the
same problems outlined for physiological measures.
First, responses on many indirect measures (e.g., IAT,
EPT) are much faster than responses on direct
measures. Hence, discrepancies between the two kinds
of measures may be driven by differences in the time-
window of measured responses rather than lack of
awareness (Cunningham et al., 2007; Ranganath et al.,
2008). Second, many indirect measures suffer from low
reliability, the only exception being the IAT and the
AMP (Gawronski & De Houwer, 2014; Greenwald &
Lai, 2020). Thus, for indirect measures with low
reliability, dissociations to direct measures may be
driven by differences in their reliability rather than lack
of awareness (Cunningham et al., 2001). Third, like
physiological measures, the validity of most indirect
measures has been established via stimuli of known
valence, and this knowledge came from studies using
self-report measures (e.g., Fazio et al., 1986;
Greenwald et al., 1998; Payne et al., 2005). This issue
creates the same inferential paradox described for
physiological measures. On the one hand, responses on
direct and indirect measures must converge to confirm
the construct validity of the indirect measure. On the
other hand, responses on the two kinds of measures
must diverge to demonstrate unawareness of attitudes.
Moreover, when convergence is demonstrated for some
stimuli (e.g., flowers and insects) and divergence is
found for other stimuli (e.g., Black and White faces),
the observed differences across content domains open
the door for alternative explanations that do not involve
claims of unawareness. In line with this concern, some
have argued that discrepancies between direct and
indirect measures reflect unwillingness rather than
inability to report one’s personal attitudes (e.g., Dunton
& Fazio, 1997; Nier, 2005; M. A. Olson et al., 2007),
although claims that indirect measures are immune to
strategic control have been disputed (e.g., Calanchini,
2020; Corneille & Lush, 2023; Gawronski et al., 2007).
Fourth, direct and indirect measures tend to differ in
terms of various structural features, rendering
dissociations between the two kinds of measures
conceptually ambiguous (Payne et al., 2008). The
significance of this concern is supported by studies
showing that at least some dissociations between direct
and indirect measures disappear when confounds with
structural task characteristics are eliminated (e.g., Béna
et al., 2022).
Another concern is that deliberate evaluations on
direct measures are influenced by various response-
related factors that do not affect spontaneous
evaluations on indirect measures to the same extent
(Fazio, 2007; Gawronski & Bodenhausen, 2006). Thus,
dissociations between the two kinds of measures may
be driven by any of these response-related factors rather
than lack of awareness. Hahn at al. (2014) aimed to
address this ambiguity by asking participants to predict
their preferences for different social groups on several
IATs before they completed those IATs. Participants
showed high accuracy in predicting their IAT scores
regardless of their prior experience with the IAT,
regardless of how much information they received
about the IAT, and regardless of whether the IAT was
introduced as a measure of true beliefs or cultural
associations (see Gawronski et al., 2008; M. A. Olson
et al., 2009). Moreover, participants showed high
accuracy in predicting their IAT scores although self-
reported evaluations on direct measures showed the
same small correlations with IAT scores found in prior
research (for meta-analyses, see Cameron et al., 2012;
Hofmann et al., 2005). Together, these results pose a
in press, Annual Review of Psychology 4
challenge to the idea that indirect measures such as the
IAT capture attitudes that people do not know they
have.
A noteworthy aspect of Hahn et al.’s (2014) findings
is that participants showed high accuracy in predicting
their personal patterns of IAT scores at a within-
subjects level, but they were much less accurate in
predicting their scores on a particular IAT at a between-
subjects level. Put differently, although participants
were highly accurate in predicting their personal rank
order of preferences in the completed IATs (e.g., that
their preference for White over Black people was
stronger than their preference for White over Hispanic
people), they were less accurate in predicting how their
preference on a given IAT compares to that of the other
participants in the sample (e.g., that their preference for
White over Black people is stronger compared to the
majority of the other participants in the sample). This
difference is important, because it illustrates an inherent
problem of between-subjects approaches in research on
unawareness of attitudes. To obtain high convergence
between an objective and a subjective measure at a
between-subjects level, participants not only need to
know their attitude toward the focal object (e.g., how
much they like apples); they also need to know how
their attitude compares to the attitudes of the other
participants in the sample (e.g., how their liking of
apples compares to the liking of apples among the other
participants). Thus, low convergence in between-
subjects designs may not necessarily reflect
unawareness of the attitude; it may also reflect limited
knowledge about the attitudes of the other participants
(see Goffin & Olson, 2011; J. M. Olson et al., 2007).
This situation is different in within-subjects designs that
focus on attitudes toward multiple objects among
individual participants (e.g., Hahn et al., 2014; Hahn &
Gawronski, 2019). To obtain high convergence
between objective and subjective measures for multiple
objects at a within-subjects level, participants must
know how their attitude toward one object compares to
their attitude toward other objects (e.g., how much they
like apples compared to oranges, bananas, mangos,
etc.), but they do not have to know anything about the
other participants in the sample. These issues have to be
considered when interpreting findings of studies that
used between-subjects designs to establish unawareness
of attitudes (see Hahn & Goedderz, 2020).
The high level of accuracy in the prediction of IAT
scores appears to conflict with evidence that
participants tend to be rather surprised when they
receive feedback about their performance (e.g., when
they learn that they have a strong preference for White
over Black people; see Goedderz & Hahn, 2022).
Anecdotes of such surprise reactions have been
interpreted as evidence that people are unaware of the
attitudes captured by the IAT (e.g., Banaji, 2011;
Krickel, 2018), which seems difficult to reconcile with
the conclusion that high accuracy in the prediction of
IAT scores demonstrates awareness. To resolve this
apparent contradiction, it is worth noting that surprise
reactions in response to IAT feedback merely reflect a
discrepancy between participants’ verbal quantification
of their subjective preference (e.g., moderate
preference for White over Black people) and the
experimenters’ verbal quantification of the obtained
measurement score (e.g., strong preference for White
over Black people). Hence, participants may be
surprised about their IAT feedback, not because they
are unaware of their attitudes, but because the metric
underlying their verbal quantification does not match
the metric underlying the verbal quantification in the
feedback they receive (Gawronski, 2019). Consistent
with this argument, Hahn et al. (2014) found that,
although participants were highly accurate in predicting
their IAT scores, the metric underlying their verbal
quantifications “stretched” the metric commonly used
to convert numeric IAT scores into verbal feedback.
Because labeling conventions for what should be
considered a “weak,” “moderate,” or “strong” bias are
entirely arbitrary in the sense that there is no objective
basis to treat one metric as “correct” and another one as
“incorrect” (Kruglanski, 1989), interpretations of
surprise reactions as evidence for unawareness are
based on a questionable normative premise (Gawronski
et al., 2022a). These concerns are further supported by
evidence that the standard algorithm to calculate IAT
scores dramatically inflates their size (Wolsiefer et al.,
2017), suggesting that the mismatch in verbal
quantifications underlying surprise responses is rooted
in a systematic distortion of IAT feedback, not
unawareness of attitudes.
Revealed vs. Stated Preferences
An alternative to using indirect measures as
“objective” indicators of attitudes is to compare
evaluative responses on self-report measures to other
behavioral expressions of attitudes that do not involve
self-report (e.g., consumption behavior, purchasing
decisions). This approach is captured by the distinction
between stated and revealed preferences (De Corte et
al., 2021). For example, stated and revealed preferences
for popcorn would be discrepant if self-reported liking
of popcorn is unrelated to actual popcorn consumption
(e.g., Neal et al., 2011). Although the distinction is
more common in research on consumer behavior than
research on attitudes, discrepancies between stated and
revealed preferences have been interpreted as evidence
that people sometimes do not know what they like or
dislike (for examples in research on romantic attraction,
see Eastwick & Finkel, 2008). Although this may the
case, such inferences must be treated with caution in the
absence of further evidence. A widely accepted notion
in research on attitude-behavior relations is that
in press, Annual Review of Psychology 5
attitudes do not influence behavior in a direct,
unconditional manner and that behavior is influenced
by various other factors beyond attitudes (Ajzen &
Kruglanski, 2019; Fazio 1990). Although it is possible
that discrepancies between stated and revealed
preferences reflect people’s unawareness of their
attitudes, the known complexity of attitude-behavior
relations renders such inferences premature without
additional data that rule out other factors as the cause of
the observed discrepancies (e.g., social norms,
perceived behavioral control). Thus, inferences of
unawareness from discrepancies between stated and
revealed preferences have to be evaluated in the context
of extant theories of attitude-behavior relations and the
proposed factors that moderate attitude-behavior
relations. To our knowledge, there is no empirical work
that has systematically addressed these issues.
Discrepant Self-Reports
The preceding sections illustrate the problems of
treating physiological measures, indirect measures such
as the IAT, and measures of revealed preferences as
objective indicators of attitudes for inferences of
unawareness. An alternative to comparing responses
across measures that do versus do not involve self-
reports is to compare responses across two self-report
measures. One example involves potential
discrepancies between self-reported evaluations of
types and tokens (see Ledgerwood et al., 2018, 2020).
Whereas types are classes of objects, tokens are
individual instances of a class of objects. Operationally,
the distinction is captured by the difference between
measures involving evaluations of an abstract category
and measures involving evaluations of individual
exemplars of that category.
Discrepancies in self-reported evaluations of types
versus tokens may arise for several reasons. First, types
and tokens may be considered conceptually distinct
attitude objects, in that an abstract category is not the
same as the aggregate of multiple individual exemplars
of that category. From this perspective, discrepant
evaluations of a category and individual exemplars of
that category would reflect a simple lack of
measurement correspondence. Second, even when
types and tokens are deemed conceptually equivalent
(e.g., Ajzen & Fishbein, 1977), discrepancies may arise
when a given factor differentially affects responses
toward types versus tokens. For example, self-reported
evaluations of Black people as a social category may
differ from self-reported evaluations of individual
Black exemplars because people may be more
concerned about expressing negative evaluations of
Black people as a social category than about expressing
negative evaluations of individual Black exemplars
(Dovidio & Gaertner, 2004).
Finally, and most relevant for the current analysis,
discrepancies between self-reported evaluations of
types versus tokens may reflect lack of awareness. For
example, self-reported evaluations of an abstract
category (e.g., self-reported liking of Pinot Noir as a
type of red wine) may differ from self-reported
evaluations of exemplars of that category (e.g., self-
reported liking of specific tokens of Pinot Noirs in a
blind tasting) because people have genuinely inaccurate
beliefs about what they like and dislike (e.g., a person
may think they do not like Pinot Noirs, but they actually
do). In this case, self-reported evaluations of an abstract
category would reflect a person’s beliefs about their
(dis)liking of exemplars of that category, whereas self-
reported evaluations of individual exemplars of the
focal category would reflect the person’s actual
(dis)liking of exemplars of that category (Ledgerwood
et al., 2020). Unawareness of this kind may occur when
evaluations of abstract categories are based on
inductive inferences from concrete experiences with
individual exemplars (Alcser-Isais et al., 2022; Da Silva
Frost et al., 2023; Woiczyk & Le Mens, 2021) and these
inferences are distorted by sampling error or biases in
inductive reasoning (see Fiedler & Plessner, 2009). In
such cases, people may draw conclusions about their
liking of a category that does not accurately reflect their
liking of individual exemplars of that category.
Although empirical work along this line is still scarce,
the conceptual idea underlying these arguments raises
interesting questions about how people draw inferences
about their (dis)liking of an abstract category from their
(dis)liking of individual exemplars of that category, the
conditions under which these inferences can produce
beliefs about the (dis)liking of the category that do not
align with one’s actual (dis)liking of exemplars of that
category, and what such discrepancies can tell us about
people’s (un)awareness of their own attitudes (see
Ledgerwood et al., 2018).
Interim Conclusions
Several conceptual issues undermine inferences of
unawareness from discrepancies between responses on
physiological measures and responses on self-report
measures. Moreover, counter to a dominant narrative in
the literature, there is no evidence supporting the idea
that indirect measures such as the IAT, the EPT, and the
AMP would capture attitudes that people are unaware
of. If anything, the available evidence suggests the
opposite. Although it is possible that discrepancies
between stated and revealed preferences are driven by
unawareness of one’s attitudes, such interpretations
must be evaluated in the context of extant theories about
attitude-behavior relations. An alternative to comparing
responses across measures that do versus do not involve
self-reports is to infer unawareness from discrepancies
between two self-reports, one example being
discrepancies between self-reported evaluations of a
type and tokens of that type. However, such
discrepancies can also be driven by alternative factors
in press, Annual Review of Psychology 6
and compelling evidence for unawareness as a driving
force is still lacking.
Unawareness of Environmental Causes
Even when people are aware that they (dis)like an
object, they may not be aware why they (dis)like it.
There are several obvious ways in which people may be
unaware of the causes of their attitudes. First, there is
no uncaused cause, and distant ones often escape
understanding. Second, causal influences occur at
various levels (e.g., synaptic activity involved in
attitude learning; cultural determinants of systems of
preferences), and no individual can possibly hold
comprehensive knowledge about all of them. Third,
relatedly, people may be unable to introspect on the
various mechanisms (including the various
psychological mechanisms) driving their
phenomenological experiences and behavior. In this
very broad sense, people remain necessarily unaware of
the complex set of causes and mechanisms influencing
their attitudes.
In this section, we discuss how psychological
research has informed two more specific questions
about environmental causes of attitudes: (1) Can an
attitude be created when people are unable to report the
stimulus event that is responsible for it? (2) Provided
people are aware of the stimulus event, are there cases
where people are nevertheless unable to report that the
stimulus event influenced their attitude? Answering
these questions requires tight control over the stimulus
event that is responsible for the attitude, reliable
measures of evaluation, adequate measures of
awareness, and sensitive analytic procedures, all of
which raise significant methodological challenges (see
Newell & Shanks, 2014; Shanks et al., 2021).
Our analysis in this section focuses on evaluative
conditioning (EC) and mere exposure (ME), which are
frequently considered strong cases for attitude learning
without awareness. To address our first question
regarding unawareness of the stimulus event, we
organize our discussion around studies using
procedures that (1) weakened the strength of stimuli, (2)
weakened top-down attention to stimuli, and (3) linked
evaluative responses to measures of recollective
memory. Expanding on this analysis, we discuss our
second question regarding unawareness of the influence
of stimulus events. Although space constraints do not
permit elaborate discussions of measurement and
analytic issues, we regularly touch on them and provide
references to more comprehensive treatments of these
issues.
Unawareness of Stimulus Event
Evaluative Conditioning
EC procedures involve pairing a neutral stimulus (the
conditioned stimulus, or CS) with a stimulus of positive
or negative valence (the unconditioned stimulus, or
US). Following this CS-US pairing, the CS is typically
evaluated in line with the valence of the US, a
phenomenon known as EC effect (for a review, see
Moran et al., 2023). For example, pairing the logo of an
unfamiliar brand (i.e., neutral CS) with the picture of a
beautiful scenery (i.e., positive US) may result in more
positive evaluations of the CS. Conversely, pairing the
logo with a picture showing dental decay (i.e., negative
US) may elicit more negative evaluations of the CS. EC
effects have been found for various types of CSs (e.g.,
human faces, consumer products, kanji symbols,
abstract visual patterns, or meaningless letter strings).
They are frequently claimed to be driven by low-level
processes operating without awareness of the CS-US
pairings (e.g., Gawronski & Bodenhausen, 2006; Petty
et al., 2019). As we discuss here, however, evidence has
accumulated that is at odds with this conclusion.
Weak stimulus strength. Early studies claimed
successful EC effects when using briefly presented
stimuli, often called “subliminal” stimuli (e.g.,
Krosnick et al., 1992). These procedures are commonly
assumed to prevent participants from accurately
reporting the stimulus event. However, early
demonstrations with “subliminal” presentations have
been criticized for using flawed designs, inadequate
awareness checks, or no awareness checks at all (for a
review, see Sweldens et al., 2014). Recent studies
relying on stronger designs and adequate measures of
awareness have generally failed to support “subliminal”
EC. The most complete and controlled investigation of
short exposure effects was realized by Stahl et al.
(2016) in a series of 6 experiments involving 27
experimental conditions. These authors found EC
effects only for CS exposures associated with high CS
identification performance and high attention. Overall,
there is little to no evidence for EC effects when using
low-strength stimuli (for a detailed review, see
Corneille & Stahl, 2019).
Weak top-down attention. Although studies with
“subliminal” stimuli have long been considered the
strongest case for unconscious influences, such studies
have been criticized for their lack of ecological validity.
As Bargh (2022) pointed out, “subliminal stimuli are a
creation of 20th century technology, [and] the human
mind could not possibly have evolved to process them”
(p. 90). One solution to this problem is to use high
strength stimuli combined with low attention. Several
EC studies did so by asking participants to perform a
concurrent attention-demanding task while processing
CS-US pairs displayed on a computer screen (e.g.,
Mierop et al., 2017; Pleyers et al., 2009). EC effects
under attentional-load conditions are compared to EC
effects in a control condition in which participants do
not perform an attention-demanding task. Overall,
studies using this approach consistently found that EC
effects vanish to non-significance under attentional
in press, Annual Review of Psychology 7
load (for a detailed review, see Corneille & Stahl,
2019).
A limitation of studies using manipulations of
attentional load is that they confound awareness of CS-
US pairings with processing goals (e.g., processing
numeric values vs. listening to music). To address this
issue, Dedonder et al. (2014) compared effects of foveal
and parafoveal presentations. These authors presented
the USs in participants’ foveal eye region and paired
them with either foveal or parafoveal CSs. An EC effect
was found only for foveal but not parafoveal CSs, the
latter of which were less likely to enter awareness.
Expanding on concerns that brief-exposure studies
confound awareness with stimulus duration and that
attentional-load studies confound awareness with
processing goals, Hödgen et al. (2018) pointed out that
studies comparing effects of foveal and parafoveal
presentations confound awareness with spatial
proximity. In the latter type of studies, CS-US pairs are
presented closer to each other when both stimuli are
presented in the foveal region than when one of the two
stimuli is presented parafoveally. To address this issue,
Hödgen et al. relied on continuous flash suppression to
present the CSs outside of awareness. Here, participants
were presented with different visual information in their
left and right eye. One eye received “high-energy” US
information (i.e., the continuous flashing of a sequence
of US photos and colored pixel masks) while the other
eye received “low-energy” CS information (i.e., a
stationary grey shape of low visual contrast). In these
procedures, the high-energy information is dominant
and impairs awareness of the low-energy information.
A series of four experiments consistently failed to
obtain EC effects for suppressed CSs on both direct and
indirect measures.
Olson and Fazio (2001) relied on yet another
rationale: incidental learning. In a simulated
surveillance task, two Pokémon characters (CSs) were
incidentally paired with either positive or negative
stimuli (USs) on distractor trials that were irrelevant for
participants’ goal in the task (i.e., press the space bar
whenever they see a particular image). This way,
participants’ attention was pulled away from the
intentional processing of the CS-US pairs. Although the
procedure has been found to be effective in producing
significant EC effects, a high-powered replication of
Olson and Fazio’s (2001) original study found that EC
effects in the surveillance task are driven by a subset of
consciously encoded CS-US pairings (Kurdi et al.,
2022; Moran et al., 2021). Again, earlier conclusions of
unawareness could not be supported.
Lack of conscious recollection. Several studies
relied on high-strength stimuli and high attention
conditions and tested whether EC effects are observed
in the absence of conscious recollection of the CS-US
pairings. Although measures of recollective memory
are ambiguous about the role of (un)awareness during
exposure to CS-US pairings (Gawronski & Walther,
2012), research using such measures are still relevant
for the current question of whether people can (dis)like
something without being able to report the stimulus
event that is responsible for their (dis)liking. Depending
on the procedure and analytic approach, some studies
found evidence for memory-independent EC (e.g.,
Hütter & Sweldens, 2013; Jurchiș et al., 2020; Walther
& Nagengast, 2006; Waroquier et al., 2020) while other
studies did not (e.g., Kurdi et al., 2022; Mierop et al.,
2017; Pleyers et al., 2007; Stahl et al., 2009).
To reconcile the mixed findings and to address
methodological limitations of prior studies, Stahl and
colleagues (2023) developed a new procedure for
examining EC effects in the absence of feelings of
remembering the US valence. Following exposure to
CS-US pairs, participants were asked to use two buttons
sets, labeled SET 1 and SET 2. If they felt they could
remember the valence of the US paired with a given CS,
they were asked to use the buttons from SET 1 and to
press pleasant (vs. unpleasant) for reporting their
recollection of a positive (vs. negative) US. If, however,
they felt they could not remember the US valence, they
were asked to use the buttons from SET 2 and to press
pleasant (vs. unpleasant) to report liking (vs. disliking)
the CS. Compared to the approaches used in prior work,
two advantages of this procedure are that it (1) relates
evaluations to subjective memory states at the within-
person-within-item level (information criterion) and (2)
measures evaluative and memory judgments closely in
time (immediacy criterion). When validating and using
this new procedure, the authors found no evidence for
EC effects in the absence of feelings of remembering.
Mere Exposure
In ME studies, neutral stimuli typically void of
meaning (e.g., unfamiliar shapes of polygons) are
evaluated more positively when they have been
presented before than when they have not been
presented before, a phenomenon known as ME effect
(Zajonc, 1968). This effect is often considered another
compelling case for attitude learning without
awareness.
Weak stimulus strength. A recent meta-analysis
found a significant linear (and quadratic) ME effect at
durations of < 15 ms exposure (Montoya et al., 2017).
This result suggests that ME effects can be established
with low-strength stimuli. However, it is unclear
whether these effects were established without
awareness of the stimulus event. A minority of the
relevant effects came from unpublished raw data files
for which procedural information is lacking. All
remaining effects came from two articles that relied on
a unique “subliminal” procedure by Förster (2009).
This procedure reportedly presented sandwich-masked
stimuli for 10 ms or 14 ms in the center of computer
in press, Annual Review of Psychology 8
screens. Yet, based on the provided information, it
remains unclear whether the software and computer
monitors in these studies guaranteed such fast and
precise exposure durations. Most critically, as is often
the case in “subliminal” research, awareness measures
were either lacking in these studies or did not meet
reliability, immediacy, and sensitivity criteria (see
Newell & Shanks, 2014). Besides widespread
procedural concerns of this sort, ME effects at short
exposures also seem to depend on moderators that are
yet to be better understood. For example, Kawakami
and Yoshida (2019) did not find a significant ME effect
at 10 ms durations on a direct measure, but found one
on indirect measures (GNAT, IAT). Newell and Shanks
(2007) compared ME effects for short (40 ms) versus
long (400 ms) exposure durations and found a
significant effect on a direct measure only for the longer
exposure duration and when recognition performance
was best. Notably, when observed, dissociations
between evaluative judgments and memory judgments
in “subliminal” ME studies may also arise from
different decision strategies for the two kinds of
judgments. When these decision strategies are
swapped, the dissociation has been found to be
reversed, with above-chance recognition memory and
at-chance liking (Whittlesea & Price, 2001). Such
findings further complicate inferences of unawareness.
Weak top-down attention. In our discussion of EC
effects, we noted that the continuous-flash-suppression
procedure resolves several confounds in the study of
awareness. We also emphasized the importance of
including adequate awareness measures instead of
merely assuming that the procedure precludes
awareness. A ME study by de Zilva et al. (2013)
addressed both issues. Combining continuous flash
suppression with online identification measures, these
authors unexpectedly found that 36% (Experiment 1)
and 64% (Experiment 2) of their samples were aware of
the supposedly suppressed stimuli. This finding is
remarkable, because “none of these participants would
have been excluded on the basis of a traditional post-
exposure recognition test” (de Zilva et al., 2013, p. 6).
Moreover, a significant ME effect emerged only for
unsuppressed stimuli and for “suppressed” stimuli that
had entered awareness. However, if we relax the
stringency of the unawareness test, some studies lend
support for ME effects under conditions of weakened
attention, such as when attending to high-strength
stimuli presented as distractors (e.g., Hansen & Wänke,
2009) or when attending to visually suppressed high-
strength stimuli (Huang & Hsieh, 2013). In sum, there
is some evidence for a ME effect under conditions of
weakened attention, but it is not clear if those conditions
prevented awareness of the stimulus event at the time
of exposure.
Lack of recollection. Hansen and Wänke (2009)
used a process-dissociation procedure to quantify the
respective contributions of familiarity and conscious
recollection to the ME effect. To do so, they compared
memory judgments for previously presented names of
unknown products in two experimental conditions
where familiarity and conscious recollection of these
names can be assumed to have converging versus
diverging effects on memory performance (see Jacoby,
1991). They found that repeated high-strength exposure
to the product names increased participants’ liking of
these names, and that this ME effect was associated
with feelings of familiarity but not with conscious
recollection. A noteworthy feature of Hansen and
Wänke’s study is that the authors experimentally
validated the functional independence of the
recollection and familiarity estimates provided by the
process-dissociation procedure. While the conscious
recollection estimate (but not the familiarity estimate)
was influenced by a manipulation of attention at
encoding, the familiarity estimate (but not the
conscious recollection estimate) was influenced by a
manipulation of figure-ground contrast. Together, these
findings suggest that repetition-induced feelings of
familiarity can influence one’s liking of a stimulus in
the absence of conscious recollection of being
previously exposed to this stimulus.
Unawareness of Causal Influence
The study by Hansen and Wänke (2009) represents
an interesting case for introducing our second question
regarding unawareness of the influence of stimulus
events. In that study, prior exposure to stimuli increased
liking of the stimuli without conscious recollection of
their prior presentation and, by implication, of the
influence of the stimulus event. Because people are
constantly exposed to a large number of events, it is
unlikely that they can consciously recollect them all,
and consciously weigh how much these events
collectively influenced their evaluation. At this point,
however, it is important to specify what the causally
effective event is. For example, participants may not
recollect their prior exposure to a stimulus, but they
may be perfectly aware that its processing is fluent. In
turn, they may rely on this meta-cognitive cue to draw
inferences about their liking of the stimulus
(Greifeneder & Schwarz, 2014). A good illustration is
provided by the ease-of-retrieval effect, whereby
mental contents and the subjective ease of their retrieval
can have opposite effects on evaluative judgments
(Schwarz et al., 1991). For example, asking participants
to recall five arguments (difficult experience) rather
than two arguments (easy experience) in favor of a
surgery fee can result in unfavorable evaluations of this
fee when people infer an unfavorable evaluation from
the difficulty of generating favorable arguments
(Greifeneder & Bless, 2007). For such effects to occur,
in press, Annual Review of Psychology 9
participants should not question the diagnostic value of
their feelings of difficulty. Hence, they should not
attribute it to the experimental manipulation. However,
they need to be aware of the feeling of difficulty itself.
As a case in point, ease-of-retrieval effects are typically
found when participants are asked to report their
subjective feelings before rather than after the judgment
at hand (Kühnen, 2010). Furthermore, participants also
need a naïve meta-cognitive theory for relating their
subjective experiences to this judgment (Schwarz,
2004). The question now becomes whether people are
aware of drawing causal inferences from their meta-
cognitive feelings when forming evaluations, which
may qualify as a meta-meta-cognitive question.
This begs the question of people’s knowledge about
stimulus-response relations. Over the past seven
decades, instruction-based replication studies have
shown that participants can accurately predict,
simulate, or produce many attitudinal phenomena based
on procedural information delivered in the original
studies (for a discussion, see Corneille & Béna, 2023).
A famous case is a study by Bem (1967), in which
observers were informed about the procedures used in
the classic cognitive dissonance study by Festinger and
Carlsmith (1959). Bem (1967) found that, based on this
procedural information alone, observers could
accurately estimate how the participants had completed
the evaluative ratings in the original study. Because the
observers in Bem’s study presumably did not
experience a state of cognitive dissonance, this
instruction-based replication study questioned the role
of arousal in cognitive dissonance effects. More
recently, instruction-based replications have been
reported on direct and indirect evaluative measures for
EC (e.g., De Houwer, 2006), ME (e.g., Van Dessel et
al., 2017), approach-avoidance (i.e., liking stimuli
better when they were approached rather than avoided;
e.g., Van Dessel et al., 2015), and vicarious evaluative
learning effects (i.e., liking stimuli better when they
were seen to elicit a positive rather than a negative
reaction in another person; e.g., Kasran et al., 2022).
These results suggest that participants hold causal
knowledge relating stimulus events (e.g., CS-US
pairings, stimulus repetitions) to evaluative responses.
A question worth examining in future research is
whether participants can verbally report this causal
knowledge (i.e., if they are aware of it) and, if so,
whether they are using it (and are aware of using it) to
inform their evaluations. Finally, it would be important
to know if participants use it (and are aware of using it)
because of compliance with experimental demands.
Indeed, whenever participants are aware of how
experimental procedures relate to responses,
experimental demand effects cannot be easily ruled out
(for discussions, see Corneille & Béna, 2023; Corneille
& Lush, 2022).
Given space limitations, we limited our discussion to
EC and ME. We chose to do so because these
procedures are typically considered low-level attitude
learning procedures (e.g., Petty et al., 2019). Other
paradigms may offer different conclusions, but they
face similar challenges. To illustrate, consider the
spreading-of-alternatives effect in studies using the
free-choice paradigm (Brehm, 1956). Here, participants
typically evaluate a chosen option more favorably than
a rejected option after making a choice, even when they
evaluated the two options similarly before making a
choice. It seems reasonable to assume that participants
in these studies are aware of the stimulus event (i.e., the
options they selected and rejected). Yet, when asked to
report the reason for their post-choice evaluations,
participants may fail to mention the influence of their
choice. However, in contrast to interpretations of the
latter finding as indicating lack of awareness, several
studies suggest that a spreading-of-alternatives effect
can be observed even when participants do not make a
choice (e.g., Chen & Risen, 2010; Gawronski et al.,
2007). These findings suggest that, when participants
“fail” to report the influence of their choice on their
evaluations, it may be because the choice itself had no
causal impact. This conclusion also explains why a
spreading-of-alternatives effect can be found among
participants with amnesia who cannot remember their
choice (Lieberman et al., 2011). Similar concerns apply
to inferences of unawareness in research on decision-
making more broadly, which have been discussed
extensively by Newell and Shanks (2014; see also
Shanks et al., 2021).
Interim Conclusions
Regarding our first question about unawareness of
stimulus events, the case for “unaware EC” is generally
unsupported for low-strength stimuli, and for high-
strength stimuli combined with weak attention. More
research is needed to reconcile the mixed evidence for
effects of high-strength stimuli combined with high-
attention but no conscious recollection. The case for
“unaware ME” is weaker than frequently stated for low-
strength stimuli, moderate for high-strength stimuli
combined with weak attention, and comparatively
strong for high-strength stimuli combined with high-
attention but no conscious recollection. Regarding our
second question about unawareness of the influence of
stimulus events, ME research suggests that people can
be unaware that a stimulus event influenced their
attitudes. This typically applies to situations where the
stimulus event(s) cannot be recollected at the evaluation
stage. However, when this is the case, a more proximal
“event” (e.g., a repetition-driven feeling of familiarity)
may be consciously used as a meta-cognitive cue
informing participants’ evaluations. Finally, studies on
instruction-based EC and instruction-based ME suggest
that people have causal knowledge relating stimulus
in press, Annual Review of Psychology 10
pairings and stimulus repetitions to evaluative
responses, but it remains unclear whether this
knowledge can be reported. More generally, although
people may sometimes report reasons for their attitudes
that are discrepant with those posited by the
experimenter, knowing who is mistaken in these cases
is often much less straightforward than experimenters
would like to believe (Corneille & Lush, 2022; Cotton,
1980; Kruglanski, 1989; Newell & Shanks, 2023).
Unawareness of Behavioral Effects
Even when people are aware of an attitude and the
environmental causes of that attitude, they may not be
aware of its behavioral effects. The general notion
underlying this idea is that attitudes may sometimes
influence behavior in a manner that evades awareness.
A frequently cited example of such effects are biases in
social behavior that arise from unrecognized influences
of intergroup attitudes (Fazio et al., in press). In general,
behavioral effects of attitudes can be deemed as being
outside of awareness either (1) when people are
unaware that they are engaging in the relevant behavior
or (2) when people are aware that they are engaging in
the relevant behavior, but they are unaware that the
behavior is influenced by their attitudes.
Unawareness of Behavioral Response
The first case involves instances where people are
unaware that they are engaging in the behavior that is
being influenced by their attitudes. Logically, a person
cannot be aware of the impact of their attitudes on a
given behavior if the person is unaware that they are
engaging in that behavior. Unawareness of this type is
likely limited to low-level, unintentional reactions and
less common for high-level, intentional actions. For
example, people may often be unaware of their
nonverbal expressions (e.g., Dovidio et al., 2002) and
visual attention (e.g., Roskos-Ewoldsen & Fazio, 1992)
in social interactions, but they are generally aware of
what they are doing when they hire a job candidate or
call the police on a suspicious person (Gawronski et al.,
2022b). Although there is evidence for attitudinal
influences on both low-level, unintentional reactions
and high-level, intentional actions (for a review, see
Fazio, 1990), a major problem for inferences of
unawareness in studies on low-level, unintentional
reactions is the lack of awareness checks in these
studies. We are not aware of any empirical work in this
area that included measures to probe participants’
awareness of the relevant behavior. Moreover, if one
were to include awareness checks in such studies,
asking participants about a specific behavioral reaction
can increase awareness of this reaction even when it
remains unrecognized in the absence of awareness
checks (Kouider & Dehaene, 2007; see also Fox et al.,
2011). These issues create a methodological problem
for studies that aim to provide empirical evidence for
the idea that attitudes can influence low-level,
unintentional behaviors that people do not know they
engage in. Thus, despite the intuitive plausibility of this
idea, there is currently no direct empirical evidence for
it. Another problem in this line of work pertains to the
underlying hypothesis that an observed behavioral
response is driven by attitudes rather than other non-
evaluative representations (e.g., semantic beliefs or
stereotypes). We will discuss this issue in more detail
after the following section on causal influences of
attitudes on behaviors that people know they are
engaging in.
Unawareness of Causal Influence
A second case involves instances where people are
aware that they are engaging in a specific behavior, but
they are unaware that the behavior is influenced by their
attitudes. The idea underlying the hypothesis of
unawareness in this case is that people are often fully
aware of what they are doing, but they may nevertheless
be unaware of the causal influence of their attitudes on
their actions. An example of such influences is the
impact of social category cues on action decisions
(Gawronski et al., 2022a). For example, in research on
gender bias in hiring decisions, participants are
generally aware that they are making a hiring decision,
but they may not be aware that their hiring decision is
influenced by their gender attitudes. Similarly, in the
real-world cases described under the hashtag
#LivingWhileBlack (see Griggs, 2018), people were
presumably aware of what they were doing when they
called the police on a Black person, but they may not
have been aware that their decision to call the police
was influenced by their racial attitudes.
Extant research suggests two potential mechanisms
by which attitudes may influence high-level, intentional
actions outside of awareness. First, attitudes may
influence high-level, intentional actions by influencing
the weighting of available information (Gawronski et
al., 2022a). For example, in a hiring scenario involving
a highly qualified man with superior credentials in
terms of a Criterion A and highly qualified woman with
superior credentials in terms of another Criterion B,
gender attitudes may lead a decision-maker to give
more weight to Criterion A than Criterion B, leading
them to hire the man and not the woman. Yet, in a
scenario where the credentials of the two candidates are
reversed, gender attitudes may lead the decision-maker
to give more weight to Criterion B than Criterion A,
leading to the same hiring decision regardless of who is
superior in terms of the two criteria (e.g., Norton et al.,
2004; Uhlmann & Cohen, 2005). Thus, to the extent
that attitudes can influence the weighting of
information outside of awareness, relevant evidence
would support the idea that people can be aware that
they are engaging in a high-level, intentional action
in press, Annual Review of Psychology 11
without being aware that the action is influenced by
their attitudes.
Second, attitudes may influence high-level,
intentional actions by influencing the interpretation of
ambiguous information (Gawronski et al., 2022a). For
example, if a White and a Black target person show the
same ambiguous behavior, racial attitudes may lead
perceivers to interpret the ambiguous behavior as
threatening when the target is Black, but as harmless
when the target is White (e.g., Duncan, 1976;
Hugenberg & Bodenhausen, 2003; Kunda & Sherman-
Williams, 1993; Sagar & Schofield, 1980). Moreover,
based on their differential interpretations of the
ambiguous behavior, perceivers may call the police on
the Black target, but not the White target. Thus, to the
extent that attitudes can influence the interpretation of
ambiguous information outside of awareness, relevant
evidence would support the idea that people can be
aware that they are engaging in a high-level, intentional
action without being aware that the action is influenced
by their attitudes.
Although there is considerable evidence for biased
weighting of mixed information and biased
interpretation of ambiguous information (Gawronski et
al., 2022a), there is very limited evidence that attitudes
can influence actions via the two mechanisms outside
of awareness. One potential reason for this lack of
evidence might be the difficulty of demonstrating
unawareness in these cases. Similar to the difficulties of
demonstrating unawareness of a low-level,
unintentional behavior, probing awareness of
attitudinal influences on high-level, intentional actions
can increase awareness of these influences even when
they remain unrecognized in the absence of awareness
checks (Kouider & Dehaene, 2007). Moreover, in areas
involving socially sensitive attitudes (e.g., gender
attitudes, racial attitudes), potential discrepancies
between actual and acknowledged influences may
reflect unwillingness rather than inability to report
attitudinal influences. For example, in cases where
gender attitudes influence hiring decisions via biased
weighting of mixed information, a person may be
unwilling to admit that they deliberately weighted the
candidates’ credentials in manner that rationalizes their
preference for a man over a woman. Similarly, in cases
where racial attitudes influence decisions to call the
police via biased interpretation of ambiguous
information, a person may be unwilling to admit that
they deliberately relied on the target’s race to
disambiguate the target’s behavior. In both cases, it
would be unwarranted to infer unawareness from
discrepancies between actual and acknowledged
influences of attitudes.
An alternative approach that avoids these issues is to
investigate people’s control over attitudinal influences
under conditions of high motivation and high ability to
control (Gawronski et al., 2022b). To the extent that an
attitudinal effect on behavior remains uncontrolled
under such conditions and statistical power for the
detection of a significant moderation is sufficiently
large, a plausible interpretation of the obtained null
effect is that the attitudinal effect remained
uncontrolled because participants were unaware of it
(see Strack & Hannover, 1996; Wegener & Petty, 1997;
Wilson & Brekke, 1994). To our knowledge, there is
only one study that has utilized this approach to probe
for unawareness of attitudinal effects. In a study by
Gawronski et al. (2003), German participants were
presented with ambiguous descriptions of either a
German-looking or a Turkish-looking young man and
asked to rate the target’s behavior along multiple
evaluative dimensions. In addition to the impression
formation task, the study included measures of ethnic
attitudes toward Germans and Turks and a measure of
motivation to control prejudiced reactions. The results
showed that participants rated the target’s behavior
more negatively for the Turkish-looking than the
German-looking target, and the size of this difference
increased as a function of participants’ attitudinal
preference for Germans over Turks. Interestingly, this
pattern was not moderated by motivation to control
prejudiced reactions, suggesting that ethnic attitudes
influenced the interpretation of ambiguous information
even for participants who were highly motivated to
control prejudiced reactions. Because the relevant
behavior (i.e., responses on a rating scale) was
relatively easy to control, these results are consistent
with the idea that attitudes biased participants’
interpretations of ambiguous behavior outside of
awareness. However, limitations of the study design
leave the obtained findings open to alternative
interpretations, one being that biased interpretations
influenced ethnic attitudes rather than vice versa
(because ethnic attitudes were measured after the task
to measure biased interpretations). Thus, although
Gawronski et al.’s (2003) findings are consistent with
the hypothesis that attitudes can influence behavior
outside of awareness, compelling evidence for this
hypothesis is still lacking.
Another obstacle in research on this question pertains
to the critical background assumption that an observed
behavioral response is driven by attitudes rather than
other non-evaluative representations (e.g., semantic
beliefs or stereotypes). Because this issue applies to
both unawareness of behavioral responses and
unawareness of causal influences, we discuss it in more
detail in our interim conclusions for this section.
Interim Conclusions
Although it seems intuitively plausible that attitudes
can influence behavior outside of awareness, stringent
tests of this idea require thorough awareness checks.
Yet, probing for awareness of attitudinal influences can
in press, Annual Review of Psychology 12
raise awareness of the focal influence, which can make
it difficult to provide evidence for unawareness
(Kouider & Dehaene, 2007). Moreover, even when
such evidence can be provided, it is critical to also
establish the hypothesized effect of attitudes on the
focal behavior and to rule out alternative interpretations
in terms of non-evaluative representations that tend to
be confounded with attitudes (e.g., semantic beliefs or
stereotypes). For example, when studying effects of
racial attitudes, it seems important to not only provide
positive evidence for the proposed role of racial
attitudes, but to also rule out alternative interpretations
in terms of racial stereotypes that tend to be confounded
with racial attitudes (see Amodio & Devine, 2006;
Phills et al., 2020). In correlational studies, such
confounds can lead to spurious associations between
racial attitudes and a focal behavior even when the focal
behavior is causally influenced by racial stereotypes
and not by racial attitudes.
The significance of this issue can be illustrated with
the presumed role of gender attitudes in hiring
decisions. Research on gender attitudes typically shows
a preference for women over men (Eagly & Mladinic,
1989), which conflicts with the commonly observed
preference for men over women in hiring decisions
(e.g., Moss-Racusin et al., 2012). Considering the
mismatching patterns of preferences in attitudes and
hiring decisions, a more plausible interpretation of the
latter is that the observed bias in favor of men over
women is driven by other non-attitudinal factors. In line
with this concern, some research suggests that gender
bias in hiring decisions arises from the (mis)fit between
stereotypical beliefs about men and women and
semantic beliefs about different occupations (Glick et
al., 1988). In contrast, the commonly observed
attitudinal preference for women over men seems to
have little impact on hiring decisions (Eagly &
Mladinic, 1994). Although attitudes can sometimes be
rooted in semantic beliefs (like prejudice can be rooted
in stereotypes), the two are conceptually and
empirically distinct (like prejudice and stereotypes are
conceptually and empirically distinct). Thus, when
studying whether attitudes can influence behavior
outside of awareness, it seems important to rule out
alternative interpretations in terms of non-attitudinal
representations that tend to be confounded with
attitudes. One possibility to do this is to test effects
across multiple target stimuli that are similar in valence
but different in terms of other non-evaluative properties
(e.g., Roskos-Ewoldsen & Fazio, 1992). Another
possibility is to investigate effects of newly created
2
A potential objection is that our conclusions are based on a treatment
of (un)awareness as a categorical property rather than a continuous
dimension. In response to this concern, it is worth noting that
categorical treatments tend to favor conclusions of unawareness (e.g.,
responses on trials may be categorized as “unaware” on a
attitudes that are not confounded with semantic beliefs
or non-evaluative stereotypes.
Conclusions
Our analysis suggests that, despite widespread claims
of unawareness in the attitude literature, strong
empirical evidence for these claims is surprisingly
scarce. One potential case of attitudes that evade
awareness involves discrepant self-reported evaluations
of a type and tokens of that type, but research on this
question is still very limited and alternative
interpretations of the observed discrepancies have not
yet been ruled out. A large body of findings speaks
against the idea that indirect measures such as the IAT
capture attitudes that people do not know they have.
Inferences of unawareness from discrepancies between
physiological and self-report measures and between
revealed and stated preferences are undermined by
conceptual, methodological, and theoretical issues.
Regarding the environmental causes of attitudes, the
available evidence suggests that ME effects can occur
without conscious recollection of the stimulus event,
although such effects may depend on decision strategies
applied at the judgment stage. Evidence is mixed for
ME effects arising from brief exposures or longer
exposures combined with low attention. Counter to the
common assumption that CS-US pairings can influence
attitudes without awareness of the pairings during
encoding, there is no compelling evidence for EC
effects under such conditions. The available evidence
for EC effects in the absence of subjective feelings of
remembering the CS-US pairings is mixed, at best.
Regarding the behavioral effects of attitudes,
inferences of unawareness are undermined by the lack
of awareness measures in prior research. Because
including such measures can raise awareness of
behavioral effects that may remain unrecognized in the
absence of awareness checks, alternative approaches
are needed to provide evidence for claims of
unawareness. To establish the postulated causal role of
attitudes, future research on this question should also
rule out effects of non-attitudinal factors such as
stereotypes and semantic beliefs.
2
It is worth noting that our review did not cover
research on contextual influences on evaluative
responses. Examples include an object’s position in a
series of options (Nisbett & Wilson, 1977), incidental
mood states (Schwarz & Clore, 2003), and primed
mental concepts (Loersch & Payne, 2011), all of which
have been claimed to influence evaluative responses
without awareness. However, similar to the conscious
dichotomous indicator, while the same responses would be classified
as “aware” on a continuous indicator; see Newell & Shanks, 2023).
Thus, if anything, a continuous treatment would suggest even greater
concerns about the scarcity of evidence for unawareness.
in press, Annual Review of Psychology 13
reliance on processing fluency in ME effects, the
relevant proximal factor in these studies may be a
conscious reliance on the response elicited by the
context (e.g., positive mood aroused by a sunny day)
when generating an evaluative response to a target
object (e.g., conscious reliance on mood when judging
one’s life satisfaction). More generally, unawareness
claims for various contextual influences have been
questioned on methodological and empirical grounds
(e.g., Adair & Spinner, 1981; Cotton, 1980; Hughes et
al., 2023; Newell & Shanks, 2023; White, 1980).
A notable limitation of our analysis is that it focused
on general effects without considering the possibility of
individual differences in (un)awareness. Although this
focus is consistent with the modal approach in this area,
it is possible that some people have better self-insight
than others (Schriber & Robins, 2012). For example,
although participants in Hahn et al.’s (2014) studies
were, on average, highly accurate in predicting their
IAT scores, there was considerable variability across
participants, in that some showed perfect accuracy
while others showed extremely poor performance in
predicting their IAT scores. Future research extending
the dominant focus on general effects to include
individual differences may help to gain deeper insights
into aspects of attitudes that may evade awareness.
Another question that we have not addressed yet is
how people gain awareness of an attitude, its
environmental causes, and its behavioral effects. The
reviewed evidence speaks only to the question of
whether people hold accurate beliefs about the three
aspects, but it does not address how people form such
beliefs (see Morris & Kurdi, 2023). Although
introspection is widely regarded as an important
mechanism to gain self-insight, it is not the only
mechanism and the extent to which people have
introspective access to psychological processes has
been disputed (Nisbett & Wilson, 1977; Wilson &
Brekke, 1994). An alternative way to gain knowledge
about one’s attitude toward an object, originally
proposed in Bem’s (1972) self-perception theory, is to
observe one’s evaluative responses to that object under
different situational conditions. Because awareness of
an attitude is a necessary precondition for awareness of
its environmental causes and its behavior effects, the
mechanisms by which people form beliefs about their
attitudes are central for all three aspects of attitudes.
Yet, although self-perception theory has been
developed more than half century ago, research on this
question is still scarce. Thus, based on the current
conclusion that people seem to have much more insight
into the three aspects of attitudes than commonly
assumed in the literature, an important question for
future research is how people gain accurate knowledge
of an attitude, its environmental causes, and its
behavioral effects.
Going beyond studies that selectively focus on one of
the three aspects, another interesting direction for future
research involves investigations of their interplay. For
example, a recent line of work has started to examine
inductive generalizations from tokens to types
following exposure to CS-US pairings (e.g., Glaser &
Kuchenbrandt, 2017; Hütter et al., 2014; Reichmann et
al., 2023). To the extent that generalizations from
specific CSs to abstract types of CSs are prone to biases
in inductive reasoning (e.g., illusory correlations; see
Fiedler & Plessner, 2009), combining a focus on
environmental causes with a focus on discrepant
evaluations of a type and tokens of that type may
provide valuable insights into when and why people are
(un)aware of an attitude and its environmental causes.
Similarly valuable insights may be gained from other
potential combinations of research foci on the three
aspects of attitudes. Although the current review
suggests that empirical support for claims of
unawareness is much weaker than commonly assumed,
we do not rule out that novel integrative approaches
could provide more compelling evidence. Likewise, the
current analysis is agnostic about the possibility of
unrecognized influences in attitude paradigms that are
not typically seen as involving unawareness and were
therefore not discussed in this contribution. To support
future research along these lines, Table 1 provides a list
of methodological recommendations for sound
inferences of (un)awareness, summarizing key points
raised throughout this article. We hope that the current
analysis provides valuable directions for future studies
on the intriguing questions of whether it possible to
hold an attitude without being aware of that attitude;
whether environmental stimuli can influence attitudes
outside of awareness; and whether attitudes can
influence behavioral responses in a manner that evades
awareness.
References
Adair, J. G., & Spinner, B. (1981). Subjects’ access to
cognitive processes: Demand characteristics and
verbal report. Journal for the Theory of Social
Behaviour, 11, 31-52.
Ajzen, I., & Fishbein, M. (1977). Attitude-behavior
relations: A theoretical analysis and review of
empirical research. Psychological Bulletin, 84, 888-
918.
Ajzen, I., & Kruglanski, A. W. (2019). Reasoned action
in the service of goal pursuit. Psychological Review,
126, 774–786.
Alcser‐Isais, A. N., Smith, L. K., & Eastwick, P. W.
(2022). Inferring one's own attitude toward an
unknown attribute: The moderating role of
complexity in juice tasting. Journal of Consumer
Behavior, 21, 1378-1389.
in press, Annual Review of Psychology 14
Amodio, D. M., & Devine, P. G. (2006). Stereotyping
and evaluation in implicit race bias: Evidence for
independent constructs and unique effects on
behavior. Journal of Personality and Social
Psychology, 91, 652–661.
Banaji, M. R. (2011). A vehicle for large-scale
education about the human mind. In J. Brockman
(Ed.), How is the internet changing the way you
think? (pp. 392-395). New York: Harper Collins.
Bargh, J. A. (2022). The cognitive unconscious in
everyday life. In A. S. Reber & R. Allen (Eds). The
cognitive unconscious: The first half century (pp. 89-
113). New York: Oxford University Press.
Bem, D. J. (1967). Self-perception: An alternative
interpretation of cognitive dissonance phenomena.
Psychological Review, 74, 183‑200.
Bem, D. J. (1972). Self-perception theory. Advances in
Experimental Social Psychology, 6, 1-62.
Béna, J., Melnikoff, D. E., Mierop, A., & Corneille, O.
(2022). Revisiting dissociation hypotheses with a
structural fit approach: The case of the prepared
reflex framework. Journal of Experimental Social
Psychology, 100, Article 104297.
Brehm, J. W. (1956). Postdecision changes in the
desirability of alternatives. Journal of Abnormal and
Social Psychology, 52, 384-389.
Calanchini, J. (2020). How multinomial processing
trees have advanced, and can continue to advance,
research using implicit measures. Social Cognition,
38, s165-s186.
Cameron, C. D., Brown-Iannuzzi, J., & Payne, B. K.
(2012). Sequential priming measures of implicit
social cognition: A meta-analysis of associations
with behaviors and explicit attitudes. Personality and
Social Psychology Review, 16, 330-350.
Cameron, C. D., Payne, B. K., & Knobe, J. (2010). Do
theories of implicit race bias change
moral
judgments? Social Justice Research, 23, 272-289.
Chen, M. K., & Risen, J. L. (2010). How choice affects
and reflects preferences: Revisiting the free-choice
paradigm. Journal of Personality and Social
Psychology, 99, 573-594.
Corneille, O., & Béna, J. (2023). Instruction-based
replication studies raise challenging questions for
psychological science. Collabra, 9, Article 82234.
Corneille, O., & Hütter, M. (2020). Implicit? What do
you mean? A comprehensive review of the delusive
implicitness construct in attitude research.
Personality and Social Psychology Review, 24, 212-
232.
Corneille, O., & Lush, P. (2023). Sixty years after
Orne’s American Psychologist article: A conceptual
framework for subjective experiences elicited by
demand characteristics. Personality and Social
Psychology Review, 27, 83-101.
Corneille, O., & Stahl, C. (2019). Associative attitude
learning: A closer look at evidence and how it relates
to attitude models. Personality and Social
Psychology Review, 23, 161-189.
Cotton, J. L. (1980). Verbal reports on mental
processes: Ignoring data for the sake of the theory?
Personality and Social Psychology Bulletin, 6, 278-
281.
Cunningham, W. A., Packer, D. J., Kesek, A., & Van
Bavel, J. J. (2009). Implicit measurement of attitudes:
A physiological approach. In R. E. Petty, R. H. Fazio,
& P. Briñol (Eds.), Attitudes: Insights from the new
implicit measures (pp. 485-512). New York:
Psychology Press.
Cunningham, W. A., Preacher, K. J., & Banaji, M. R.
(2001). Implicit attitude measurement: Consistency,
stability, and convergent validity. Psychological
Science, 12, 163-170.
Cunningham, W. A., Zelazo, P. D., Packer, D. J., & Van
Bavel, J. J. (2007). The iterative reprocessing model:
A multilevel framework for attitudes and evaluation.
Social Cognition, 25, 736-760.
Da Silva Frost, A., Wang, Y. A., Eastwick, P. W., &
Ledgerwood, A. (2023). Summarized attribute
preferences have unique antecedents and
consequences. Journal of Experimental Psychology:
General. Advance Online Publication.
Daumeyer, N. M., Onyeador, I. N., Brown, X., &
Richeson, J. A. (2019). Consequences of attributing
discrimination to implicit vs. explicit bias. Journal of
Experimental Social Psychology, 84, Article 103812
De Corte, K., Cairns, J., & Grieve, R. (2021). Stated
versus revealed preferences: An approach to reduce
bias. Health Economics, 30, 1095-1123.
De Houwer, J. (2006). Using the Implicit Association
Test does not rule out an impact of conscious
propositional knowledge on evaluative conditioning.
Learning and Motivation, 37, 176-187.
De Houwer, J., Gawronski, B., & Barnes-Holmes, D.
(2013). A functional-cognitive framework for
attitude research. European Review of Social
Psychology, 24, 252-287.
De Zilva, D., Vu, L., Newell, B. R., & Pearson, J.
(2013). Exposure is not enough: Suppressing stimuli
from awareness can abolish the mere exposure effect.
PloS ONE, 8(10), e77726.
Dedonder, J., Corneille, O., Bertinchamps, D., &
Yzerbyt, V. (2014). Overcoming correlational
pitfalls: Experimental evidence suggests that
evaluative conditioning occurs for explicit but not
implicit encoding of CS–US pairings. Social
Psychological and Personality Science, 5, 250-257.
Dovidio, J. F., & Gaertner, S. L. (2004). Aversive
racism. Advances in Experimental Social
Psychology, 36, 1-52.
in press, Annual Review of Psychology 15
Dovidio, J. F., Kawakami, K., & Gaertner, S. L. (2002).
Implicit and explicit prejudice and interracial
interaction. Journal of Personality and Social
Psychology, 82, 62-68.
Duncan, B. L. (1976). Differential perception and
attribution of intergroup violence: Testing the lower
limits of stereotyping of Blacks. Journal of
Personality and Social Psychology, 34, 590-598.
Dunton, B. C., & Fazio, R. H. (1997). An individual
difference measure of motivation to control
prejudiced reactions. Personality and Social
Psychology Bulletin, 23, 316-326.
Eagly, A. H., & Chaiken, S. (2007). The advantages of
an inclusive definition of attitude. Social Cognition,
25, 582-602.
Eagly, A. H., & Mladinic, A. (1989). Gender
stereotypes and attitudes toward women and men.
Personality and Social Psychology Bulletin, 15, 543-
558.
Eagly, A. H., & Mladinic, A. (1994). Are people
prejudiced against women? Some answers from
research on attitudes, gender stereotypes, and
judgments of competence. European Review of
Social Psychology, 5, 1-35.
Eastwick, P. W., & Finkel, E. J. (2008). Speed-dating:
A powerful and flexible paradigm for studying
romantic relationship initiation. In S. Sprecher, A.
Wenzel, & J. Harvey (Eds.), Handbook of
relationship initiation (pp. 217-234). New York:
Psychology Press.
Fazio, R. H. (1990). Multiple processes by which
attitudes guide behavior: The MODE model as an
integrative framework. Advances in Experimental
Social Psychology, 23, 75-109.
Fazio, R. H. (2007). Attitudes as object–evaluation
associations of varying strength. Social Cognition,
25, 603-637.
Fazio, R. H., Jackson, J. R., Dunton, B. C., & Williams,
C. J. (1995). Variability in automatic activation as an
unobtrusive measure of racial attitudes: A bona fide
pipeline? Journal of Personality and Social
Psychology, 69, 1013-1027.
Fazio, R. H., Samayoa, J. A. G., Boggs, S. T., &
Ladanyi, J. (in press). What is implicit bias? In J. A.,
Krosnick, T. H., Stark, & A. L. Scott (Eds.), The
Cambridge handbook of implicit bias and racism.
Cambridge: Cambridge University Press.
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.
Festinger, L., & Carlsmith, J. M. (1959). Cognitive
consequences of forced compliance. Journal of
Abnormal and Social Psychology, 58, 203‑210.
Fiedler, K., & Plessner, H. (2009). Induction: From
simple categorization to higher-order inference
problems. In F. Strack & J. Förster (Eds.), Social
cognition: The basis of human interaction (pp. 93-
120). New York: Psychology Press.
Förster, J. (2009). Cognitive consequences of novelty
and familiarity: How mere exposure influences level
of construal. Journal of Experimental Social
Psychology, 45, 444-447.
Fox, M. C., Ericsson, K. A., & Best, R. (2011). Do
procedures for verbal reporting of thinking have to be
reactive? A meta-analysis and recommendations for
best reporting methods. Psychological Bulletin, 137,
316-344.
Gawronski, B. (2019). Six lessons for a cogent science
of implicit bias and its criticism. Perspectives on
Psychological Science, 14, 574-595.
Gawronski, B., & Bodenhausen, G. V. (2006).
Associative and propositional processes in
evaluation: An integrative review of implicit and
explicit attitude change. Psychological Bulletin, 132,
692-731.
Gawronski, B., & Bodenhausen, G. V. (2012). Self-
insight from a dual-process perspective. In S. Vazire
& T. D. Wilson (Eds.), Handbook of self-knowledge
(pp. 22-38). New York: Guilford Press.
Gawronski, B., & Bodenhausen, G. V. (2015). Social-
cognitive theories. In B. Gawronski, & G. V.
Bodenhausen (Eds.), Theory and explanation in
social psychology (pp. 65-83). New York: Guilford
Press.
Gawronski, B., Bodenhausen, G. V., & Becker, A. P.
(2007). I like it, because I like myself: Associative
self-anchoring and post-decisional change of implicit
evaluations. Journal of Experimental Social
Psychology, 43, 221-232.
Gawronski, B., & Brannon, S. M. (2019). Attitudes and
the implicit-explicit dualism. In D. Albarracín & B.
T. Johnson (Eds.), The handbook of attitudes. Volume
1: Basic principles (2nd edition, pp. 158-196). New
York: Routledge.
Gawronski, B., & De Houwer, J. (2014). Implicit
measures in social and personality psychology. In H.
T. Reis, & C. M. Judd (Eds.), Handbook of research
methods in social and personality psychology (2nd
edition, pp. 283-310). New York: Cambridge
University Press.
Gawronski, B., Geschke, D., & Banse, R. (2003).
Implicit bias in impression formation: Associations
influence the construal of individuating information.
European Journal of Social Psychology, 33, 573-589.
Gawronski, B., LeBel, E. P., & Peters, K. R. (2007).
What do implicit measures tell us? Scrutinizing the
validity of three common assumptions. Perspectives
on Psychological Science, 2, 181-193.
Gawronski, B., Ledgerwood, A., & Eastwick, P. W.
(2022a). Implicit bias ≠ bias on implicit measures.
Psychological Inquiry, 33, 139-155.
in press, Annual Review of Psychology 16
Gawronski, B., Ledgerwood, A., & Eastwick, P. W.
(2022b). Reflections on the difference between
implicit bias and bias on implicit measures.
Psychological Inquiry, 33, 219-231.
Gawronski, B., Peters, K. R., & LeBel, E. P. (2008).
What makes mental associations personal or extra-
personal? Conceptual issues in the methodological
debate about implicit attitude measures. Social and
Personality Psychology Compass, 2, 1002-1023.
Gawronski, B., & Walther, E. (2012). What do memory
data tell us about the role of contingency awareness
in evaluative conditioning? Journal of Experimental
Social Psychology, 48, 617-623.
Glaser, T., & Kuchenbrandt, D. (2017). Generalization
effects in evaluative conditioning: Evidence for
attitude transfer effects from single exemplars to
social categories. Frontiers in Psychology, 8, Article
103.
Glick, P., Zion, C., & Nelson, C. (1988). What mediates
sex discrimination in hiring decisions? Journal of
Personality and Social Psychology, 55, 178-186.
Goedderz, A., & Hahn, A. (2022). Biases left
unattended: People are surprised at racial bias
feedback until they pay attention to their biased
reactions. Journal of Experimental Social
Psychology, 102, Article 104374.
Goffin, R. D., & Olson, J. M. (2011). Is it all relative?
Comparative judgments and the possible
improvement of self-ratings and ratings of others.
Perspectives on Psychological Science, 6, 48-60.
Greenwald, A. G., & Lai, C. K. (2020). Implicit social
cognition. Annual Review of Psychology, 71, 419–
445.
Greenwald, A. G., McGhee, D. E., & Schwartz, J. K. L.
(1998). Measuring individual differences in implicit
cognition: The Implicit Association Test. Journal of
Personality and Social Psychology, 74, 1464-1480.
Greifeneder, R., & Bless, H. (2007). Relying on
accessible content versus accessibility experiences:
The case of processing capacity. Social Cognition,
25, 853-881.
Greifeneder, R., & Schwarz, N. (2014). Metacognitive
processes and subjective experiences. In J. W.
Sherman, B. Gawronski, & Y. Trope (Eds.), Dual-
process theories of the social mind (pp. 314-327).
New York: Guilford Press.
Griggs, B. (2018, December 28). Living while black:
Here are all the routine activities for which police
were called on African-Americans this year. CNN.
Retrieved from
https://www.cnn.com/2018/12/20/us/living-while-
black-police-calls-trnd/index.html (July 26, 2023).
Hahn, A., & Gawronski, B. (2019). Facing one’s
implicit biases: From awareness to acknowledgment.
Journal of Personality and Social Psychology, 116,
769-794.
Hahn, A., & Goedderz, A. (2020). Trait-
unconsciousness, state-unconsciousness,
preconsciousness, and social miscalibration in the
context of implicit evaluations. Social Cognition, 38,
s115-s134.
Hahn, A., Judd, C. M., Hirsh, H. K., & Blair, I. V.
(2014). Awareness of implicit attitudes. Journal of
Experimental Psychology: General, 143, 1369-1392.
Hansen, J., & Wänke, M. (2009). Liking what’s
familiar: The importance of unconscious familiarity
in the mere-exposure effect. Social Cognition, 27,
161-182.
Hödgen, F., Hütter, M., & Unkelbach, C. (2018). Does
evaluative conditioning depend on awareness?
Evidence from a continuous flash suppression
paradigm. Journal of Experimental Psychology:
Learning, Memory, and Cognition, 44, 1641-1657.
Hofmann, W., Gawronski, B., Gschwendner, T., Le, H.,
& Schmitt, M. (2005). A meta-analysis on the
correlation between the Implicit Association Test and
explicit self-report measures. Personality and Social
Psychology Bulletin, 31, 1369-1385.
Huang, Y. F., & Hsieh, P. J. (2013). The mere exposure
effect is modulated by selective attention but not
visual awareness. Vision Research, 91, 56-61.
Hughes, S., Cummins, J., & Hussey, I. (2023). Effects
on the Affect Misattribution Procedure are strongly
moderated by influence awareness. Behavior
Research Methods, 55, 1558-1586.
Hütter, M., Kutzner, F., & Fiedler, K. (2014). What is
learned from repeated pairings? On the scope and
generalizability of evaluative conditioning. Journal
of Experimental Psychology: General, 143, 631-643.
Hütter, M., & Sweldens, S. (2013). Implicit
misattribution of evaluative responses: Contingency-
unaware evaluative conditioning requires
simultaneous stimulus presentations. Journal of
Experimental Psychology: General, 142, 638–643.
Hugenberg, K., & Bodenhausen, G. V. (2003). Facing
prejudice: Implicit prejudice and the perception of
facial threat. Psychological Science, 14, 640-643.
Ito, T. A., & Cacioppo, J. T. (2007). Attitudes as mental
and neural states of readiness: Using physiological
measures to study implicit attitudes. In B.
Wittenbrink & N. Schwarz (Eds.), Implicit measures
of attitudes (pp. 125-158). New York: Guilford Press.
Jacoby, L. L. (1991). A process dissociation
framework: Separating automatic from intentional
uses of memory. Journal of Memory and Language,
30, 513-541.
Jurchiș, R., Costea, A., Dienes, Z., Miclea, M., & Opre,
A. (2020). Evaluative conditioning of artificial
grammars: Evidence that subjectively-unconscious
structures bias affective evaluations of novel stimuli.
Journal of Experimental Psychology: General, 149,
1800-1809.
in press, Annual Review of Psychology 17
Kasran, S., Hughes, S., & De Houwer, J. (2022).
Learning via instructions about observations:
Exploring similarities and differences with learning
via actual observations. Royal Society Open Science,
9, Article 220059.
Kawakami, N., & Yoshida, F. (2019). Subliminal
versus supraliminal mere exposure effects:
Comparing explicit and implicit attitudes.
Psychology of Consciousness: Theory, Research, and
Practice, 6, 279-291.
Kouider, S., & Dehaene, S. (2007). Levels of
processing during non-conscious perception: a
critical review of visual masking. Philosophical
Transactions of the Royal Society B: Biological
Sciences, 362, 857-875.
Krickel, B. (2018). Are the states underlying implicit
biases unconscious? A neo-Freudian answer.
Philosophical Psychology, 31, 1007-1026.
Krosnick, J. A., Betz, A. L., Jussim, L. J., & Lynn, A.
R. (1992). Subliminal conditioning of attitudes.
Personality and Social Psychology Bulletin, 18, 152-
162.
Krosnick, J. A., Judd, C. M., & Wittenbrink, B. (2005).
The measurement of attitudes. In D. Albarracín, B. T.
Johnson, & M. P. Zanna (Eds.), The handbook of
attitudes (pp. 21-76). Mahwah, NJ: Erlbaum.
Kruglanski, A. W. (1989). The psychology of being
“right”: The problem of accuracy in social perception
and cognition. Psychological Bulletin, 106, 395–409
Kühnen, U. (2010). Manipulation checks as
manipulation: Another look at the ease-of-retrieval
heuristic. Personality and Social Psychology
Bulletin, 36, 47-58.
Kunda, Z., & Sherman-Williams, B. (1993).
Stereotypes and the construal of individuating
information. Personality and Social Psychology
Bulletin, 19, 90-99.
Kurdi, B., Hussey, I., Stahl, C., Hughes, S., Unkelbach,
C., Ferguson, M., & Corneille, O. (2022). Unaware
attitude formation in the surveillance task? Revisiting
the findings of Moran et al. (2021). International
Review of Social Psychology, 35, Article 6.
Lang, P. J., Bradley, M. M., & Cuthbert, B. N. (2008).
International affective picture system (IAPS):
Affective ratings of pictures and instruction manual.
Technical Report A-7. University of Florida,
Gainesville, FL.
Ledgerwood, A., Eastwick, P. W., & Gawronski, B.
(2020). Experiences of liking versus ideas about
liking. Behavioral and Brain Sciences, 43, Article
e136.
Ledgerwood, A., Eastwick, P. W., & Smith, L. K.
(2018). Toward an integrative framework for
studying human evaluation: Attitudes toward objects
and attributes. Personality and Social Psychology
Review, 22, 378-398.
Lieberman, M. D., Ochsner, K. N., Gilbert, D. T., &
Schacter, D. L. (2001). Do amnesics exhibit cognitive
dissonance reduction? The role of explicit memory
and attention in attitude change. Psychological
Science, 12, 135-140.
Loersch, C., & Payne, B. K. (2011). The situated
inference model: An integrative account of the effects
of primes on perception, behavior, and motivation.
Perspectives on Psychological Science, 6, 234-252.
Mierop, A., Hütter, M., & Corneille, O. (2017).
Resource availability and explicit memory largely
determine evaluative conditioning effects in a
paradigm claimed to be conducive to implicit attitude
acquisition. Social Psychological and Personality
Science, 8, 758-767.
Montoya, R. M., Horton, R. S., Vevea, J. L., Citkowicz,
M., & Lauber, E. A. (2017). A re-examination of the
mere exposure effect: The influence of repeated
exposure on recognition, familiarity, and liking.
Psychological Bulletin, 143, 459-498.
Moors, A. (2016). Automaticity: Componential, causal,
and mechanistic explanations. Annual Review of
Psychology, 67, 263-287.
Moran, T., Hughes, S., Hussey, I., Vadillo, M. A.,
Olson, M. A., Aust, F., … De Houwer, J. (2021).
Incidental attitude formation via the surveillance
task: A preregistered replication of the Olson and
Fazio (2001) study. Psychological Science, 32, 120-
131.
Moran, T., Nudler, Y., & Bar-Anan, Y. (2023).
Evaluative conditioning: Past, present, and future.
Annual Review of Psychology, 74, 245-269.
Morris, A., & Kurdi, B. (2023). Awareness of implicit
attitudes: Large-scale investigations of mechanism
and scope. Journal of Experimental Psychology:
General. Advance Online Publication.
Moss-Racusin, C. A., Dovidio, J. F., Brescoll, V. L.,
Graham, M. J., & Handelsman, J. (2012). Science
faculty’s subtle gender biases favor male
students. Proceedings of the National Academy of
Sciences, 109, 16474-16479.
Neal, D. T., Wood, W., Wu, M., & Kurlander, D.
(2011). The pull of the past: When do habits persist
despite conflict with motives? Personality and Social
Psychology Bulletin, 37, 1428-1437.
Newell, B. R., & Shanks, D. R. (2007). Recognising
what you like: Examining the relation between the
mere-exposure effect and recognition. European
Journal of Cognitive Psychology, 19, 103-118.
Newell, B., & Shanks, D. (2014). Unconscious
influences on decision making: A critical review.
Behavioral and Brain Sciences, 37, 1-19.
Newell, B. R., & Shanks, D. R. (2023). Open minded:
Searching for truth about the unconscious mind.
Cambridge, MA: MIT Press.
in press, Annual Review of Psychology 18
Nier, J. A. (2005). How dissociated are implicit and
explicit racial attitudes? A bogus pipeline approach.
Group Processes and Intergroup Relations, 8, 39-52.
Nisbett, R. E., & Wilson, T. D. (1977). Telling more
than we can know: Verbal reports on mental
processes. Psychological Review, 84, 231-259.
Norton, M. I., Vandello, J. A., & Darley, J. M. (2004).
Casuistry and social category bias. Journal of
Personality and Social Psychology, 87, 817-831.
Olson, J. M., Goffin, R. D., & Haynes, G. A. (2007).
Relative versus absolute measures of explicit
attitudes: Implications for predicting diverse attitude-
relevant criteria. Journal of Personality and Social
Psychology, 93, 907–926.
Olson, M. A., & Fazio, R. H. (2001). Implicit attitude
formation through classical conditioning.
Psychological Science, 12, 413-417.
Olson, M. A., Fazio, R. H., & Han, H. A. (2009).
Conceptualizing personal and extrapersonal
associations. Social and Personality Psychology
Compass, 3, 152–170.
Olson, M. A., Fazio, R. H., & Hermann, A. D., Sr.
(2007). Reporting tendencies underlie discrepancies
between implicit and explicit measures of self-
esteem. Psychological Science, 18, 287-291.
Payne, B. K., Burkley, M. A., & Stokes, M. B. (2008).
Why do implicit and explicit attitude tests diverge?
The role of structural fit. Journal of Personality and
Social Psychology, 94, 16-31.
Payne, B. K., Cheng, S. M., Govorun, O., & Stewart, B.
D. (2005). An inkblot for attitudes: Affect
misattribution as implicit measurement. Journal of
Personality and Social Psychology, 89, 277-293.
Petty, R. E., Briñol, P., Fabrigar, L. R., & Wegener, D.
T. (2019). Attitude structure and change. In R. F.
Baumeister & E. J. Finkel (Eds.), Advanced social
psychology (2nd Edition, pp. 117-156). Oxford:
Oxford University Press.
Phills, C. E., Hahn, A., & Gawronski, B. (2020). The
bidirectional causal relation between implicit
stereotypes and implicit prejudice. Personality and
Social Psychology Bulletin, 46, 1318-1330.
Pleyers, G., Corneille, O., Luminet, O., & Yzerbyt, V.
(2007). Aware and (dis)liking: Item-based analyses
reveal that valence acquisition via evaluative
conditioning emerges only when there is contingency
awareness. Journal of Experimental Psychology.
Learning, Memory, and Cognition, 33, 130-144.
Pleyers, G., Corneille, O., Yzerbyt, V., & Luminet, O.
(2009). Evaluative conditioning may incur
attentional costs. Journal of Experimental
Psychology: Animal Behavior Processes, 35, 279-
285.
Ranganath, K. A., Smith, C. T., & Nosek, B. A. (2008).
Distinguishing automatic and controlled components
of attitudes from direct and indirect measurement
methods. Journal of Experimental Social
Psychology, 44, 386-396
Reichmann, K., Hütter, M., Kaup, B., & Ramscar, M.
(2023). Variability and abstraction in evaluative
conditioning: Consequences for the generalization of
likes and dislikes. Journal of Experimental Social
Psychology, 108, Article 104478.
Roskos-Ewoldsen, D. R., & Fazio, R. H. (1992). On the
orienting value of attitudes: Attitude accessibility as
a determinant of an object’s attraction of visual
attention. Journal of Personality and Social
Psychology, 63, 198–211
Sagar, H. A., & Schofield, J. W. (1980). Racial and
behavioral cues in black and white children’s
perceptions of ambiguously aggressive acts. Journal
of Personality and Social Psychology, 39, 590-598.
Schriber, R. A., & Robins, R. W. (2012). Self-
knowledge: An individual-differences perspective. In
S. Vazire & T. D. Wilson (Eds.), Handbook of self-
knowledge (pp. 105-127). New York: Guilford Press.
Schwarz, N. (2004). Metacognitive experiences in
consumer judgment and decision making. Journal of
Consumer Psychology, 14, 332-348.
Schwarz, N., Bless, H., Strack, F., Klumpp, G.,
Rittenauer-Schatka, H., & Simons, A. (1991). Ease of
retrieval as information: Another look at the
availability heuristic. Journal of Personality and
Social Psychology, 61, 195-202.
Schwarz, N., & Clore, G. L. (2003). Mood as
information: 20 years later. Psychological Inquiry,
14, 296-303.
Shanks, D. R., Malejka, S., & Vadillo, M. A. (2021).
The challenge of inferring unconscious mental
processes. Experimental Psychology, 68, 113-129.
Shanks, D. R., & St. John, M. F. (1994). Characteristics
of dissociable human learning systems. Behavioral
and Brain Sciences, 17, 367-395.
Stahl, C., Béna, J., Aust, F., Mierop, A., & Corneille,
O. (2023). A conditional judgment procedure for
probing evaluative conditioning effects in the
absence of feelings of remembering. Behavior
Research Methods. Advance Online Publication.
Stahl, C., Haaf, J., & Corneille, O. (2016). Subliminal
evaluative conditioning? Above-chance CS
identification may be necessary and insufficient for
attitude learning. Journal of Experimental
Psychology: General, 145, 1107-1131.
Stahl, C., Unkelbach, C., & Corneille, O. (2009). On the
respective contributions of awareness of
unconditioned stimulus valence and unconditioned
stimulus identity in attitude formation through
evaluative conditioning. Journal of Personality and
Social Psychology, 97, 404-420.
Strack, F., & Hannover, B. (1996). Awareness of the
influence as a precondition for implementing
correctional goals. In P. M. Gollwitzer & J. A. Bargh
in press, Annual Review of Psychology 19
(Eds.), The psychology of action: Linking cognition
and motivation to behavior (pp. 579-596). New York:
Guilford Press.
Sweldens, S., Corneille, O., & Yzerbyt, V. (2014). The
role of awareness in attitude formation through
evaluative conditioning. Personality and Social
Psychology Review, 18, 187-209.
Timmermans, B., & Cleeremans, A. (2015). How can
we measure awareness? An overview of current
methods. In M. Overgaard (Ed.), Behavioral methods
in consciousness research (pp. 21-46). Oxford, UK:
Oxford University Press.
Uhlmann, E. L., & Cohen, G. L. (2005). Constructed
criteria: Redefining merit to justify discrimination.
Psychological Science, 16, 474-480.
Van Dessel, P., De Houwer, J., Gast, A., & Tucker
Smith, C. (2015). Instruction-based approach-
avoidance effects: Changing stimulus evaluation via
the mere instruction to approach or avoid stimuli.
Experimental Psychology, 62, 161‑169.
Van Dessel, P., Mertens, G., Smith, C. T., & De
Houwer, J. (2017). The mere exposure instruction
effect. Experimental Psychology, 64, 299-314.
Walther, E., & Nagengast, B. (2006). Evaluative
conditioning and the awareness issue: Assessing
contingency awareness with the Four-Picture
Recognition Test. Journal of Experimental
Psychology: Animal Behavior Processes, 32, 454-
459.
Waroquier, L., Abadie, M., & Dienes, Z. (2020).
Distinguishing the role of conscious and unconscious
knowledge in evaluative conditioning. Cognition,
205, Article 104460.
Wegener, D. T., & Petty, R. E. (1997). The flexible
correction model: The role of naive theories of bias
in bias correction. Advances in Experimental Social
Psychology, 29, 141-208.
White, P. (1980). Limitations on verbal reports of
internal events: A refutation of Nisbett and Wilson
and of Bem. Psychological Review, 87, 105-112.
Whittlesea, B. W. A., & Price, J. R. (2001).
Implicit/explicit memory versus analytic/nonanalytic
processing: Rethinking the mere exposure effect.
Memory & Cognition, 29, 234-246.
Wilson, T. D., & Brekke, N. (1994). Mental
contamination and mental correction: Unwanted
influences on judgments and evaluations.
Psychological Bulletin, 116, 117-142.
Woiczyk, T. K. A., & Le Mens, G. (2021). Evaluating
categories from experience: The simple averaging
heuristic. Journal of Personality and Social
Psychology, 121, 747-773.
Wolsiefer, K., Westfall, J., & Judd, C. M. (2017).
Modeling stimulus variation in three common
implicit attitude tasks. Behavior Research Methods,
49, 1193-1209.
Zajonc, R. B. (1968). Attitudinal effects of mere
exposure. Journal of Personality and Social
Psychology, 9, 1-27.
in press, Annual Review of Psychology 2
Table 1. Methodological recommendations for studies investigating the (un)awareness of attitudes, their
environmental causes, and their behavior effects.
1. Do not conflate unawareness with other features of automaticity (e.g., unintentionality, efficiency,
uncontrollability).
2. Be specific about the awareness question you are interested in (e.g., awareness of what? awareness
at which processing stage?).
3. Do not assume that the mere use of a particular procedure (e.g., short presentation times; indirect
measures) guarantees unawareness. Instead, check for unawareness in these procedures using
independent criteria.
4. Ensure that measures of evaluative responses and measures of awareness are held constant on
procedural factors (e.g., stimuli, timing of measured responses).
5. Ensure that measures of evaluative responses and measures of awareness have comparably high
reliability and sensitivity in capturing the to-be-measured constructs.
6. Rule out alternative explanations before inferring unawareness from unreported attitudes or
influences (e.g., unwillingness to report).
7. Consider that between-subjects approaches to studying (un)awareness confound self-insight with
knowledge about other participants in the sample.
8. Rule out spurious effects of attitudes driven by confounds with non-evaluative representations
(e.g., semantic beliefs, stereotypes).
9. Whenever possible, investigate unawareness questions using experimental approaches (e.g.,
manipulating attention level) rather than correlational designs (e.g., linking performance to
retrospective memory reports or responses in funnel debriefings).
in press, Annual Review of Psychology 3
Figure 1. Three aspects of attitudes for which people may lack awareness. First, people may be unaware
of the attitude itself, defined as psychological tendency that is expressed by evaluating a particular entity
with some degree of favor or disfavor. Second, people may be unaware of the environmental cause of the
attitude, including stimulus events that are responsible for an attitude and the causal influence of stimulus
events on an attitude. Third, people may be unaware of the behavioral effect of the attitude, including
behavioral responses that are influenced by the attitude and the causal influence of the attitude on
behavioral responses.