Priming a New
Identity
Priming a New Identity:
Self-Monitoring Moderates the Effects of Nonself Primes on
Self-Judgments and Behavior
Kenneth G. DeMarree
Richard E. Petty
S. Christian Wheeler
2005
|
Kenneth G. DeMarree
Ohio State University
Richard E. Petty
Ohio State University
S. Christian Wheeler
Stanford University
The first experiment reported here was part of a master’s thesis by Kenneth G.
DeMarree under the supervision of Richard E. Petty. This research received financial support from National Science
Foundation Grant BNS 95-20854 and National Institute of Mental Health Grant T32-MH19728. We thank Russ Fazio, Bob
Arkin, and Keith Payne, who served on the thesis committee; the 2001–2004 Group for Attitudes and Persua-sion; and
Monica Biernat for comments.
Correspondence concerning this article should be addressed to Kenneth G. DeMarree,
Department of Psychology, Ohio State University, 1885 Neil Avenue Mall, Columbus, OH 43210.
Abstract
When a construct is primed, people
often act in construct-consistent ways. Several accounts for this effect have been offered, including ideomotor
theory and a social functional perspective. The authors tested an additional perspective, the Active-Self account,
whereby primes can temporarily alter self-perceptions. In Study 1, non-African American participants reported
feeling more aggressive on an implicit measure following an African American prime. In Study 2, participants
reported feeling luckier on an implicit measure following a number 7 (vs. 13) prime. In both studies, these effects
were obtained only for low self-monitors, who are more likely to change self-conceptions in response to diagnostic
self-information and to use their internal states in guiding behavior. Study 3 showed that low self-monitors also
show larger behavioral effects of primes.
The activation of stereotypes and other social constructs has been shown to affect
the behavior of people who are not members of the targeted group (for reviews, see Dijksterhuis & Bargh, 2001;
Wheeler & Petty, 2001). In most cases, the behavior observed is consistent with the activated stereotype. One
of the better known examples of this is a study by Bargh, Chen, and Burrows (1996), in which young college students
were primed with the elderly stereotype by unscram-bling sentences containing words consistent with the stereotype
(e.g., wrinkle). Upon leaving the lab, participants were secretly timed as they walked down the hall. Individuals
primed with the stereotype took more time to walk down the hall (i.e., they walked slower) than did control
participants. College students also displayed impaired performance on a memory task when primed with the stereotype
(Dijksterhuis, Aarts, Bargh, & van Knippenberg, 2000). In other studies, White participants performed worse on
a math test (Wheeler, Jarvis, & Petty, 2001) or exhibited more aggressive
facial expressions (Bargh et al., 1996) following the activation of an African American prime, and college students
performed relatively well or poorly on Trivial Pursuit questions following the activation of the stereotypes of
professor or soccer hooligan, respectively (Dijksterhuis & van Knip-penberg, 1998).
Although stereotype priming has been the most common method used in
prime-to-behavior studies, other means of activating traits have also been used. For example, Kay, Wheeler, Bargh,
and Ross (2004) used business objects as primes and found results comparable to those that would be expected from
trait or stereotype priming. For example, in one experiment, the questionnaires used to play an ulti-matum game
either were removed from an executive leather briefcase and completed with a business-style pen or were removed
from a backpack and completed with wooden pencils. Participants engaged in more competitive play in the business
condition than in the control condition. Comparable results were found with traditional priming paradigms (i.e.,
exposure to pictures of business objects prior to playing the game). Hence, similar effects may be obtained with a
variety of different means of activating trait constructs. For the sake of brevity, we will refer to primes capable
of activating trait constructs as “social construct primes.”
Although the finding that primes can influence overt behavior is now well
established, researchers have yet to agree on a single mechanism by which these effects occur. We review two
influen-tial perspectives, ideomotor activation and the social functional account, before providing an alternative
mechanism by which prime-to-behavior effects can occur—the Active-Self account.
Ideomotor Activation
The effect of social construct activation on behavior has gener-ally been
explained by using the theory of ideomotor action. Ideomotor theory proposes that in the absence of inhibitory
forces (Macrae & Johnston, 1998), merely thinking about a behavior is sufficient to produce that behavior
(Carpenter, 1874; James, 1890/ 1950). Support for this notion comes from research on the perception– behavior link,
which holds that the cognitive and neu-ral substrates for performing a behavior overlap with those used in
perceiving that behavior (Prinz, 1990). Numerous human and animal studies have shown that the same areas of the
brain—and in some cases, the same neurons—are active during both percep-tion (actual or imagined) and production of
a specific behavior (e.g., Decety, Jeannerod, Germain, & Pastene, 1991; Iacoboni et al., 1999; Kohler et al.,
2002; Rizzolatti & Arbib, 1998). In social psychology, researchers have argued that behavioral mimicry ef-fects
(in which participants mimic the nonverbal behavior of a confederate without doing so deliberately) are the
simplest instan-tiations of the ideomotor phenomenon (Chartrand & Bargh, 1999; Dijksterhuis & Bargh,
2001).
In the case of prime-to-behavior effects, however, there is no direct perception
or imagination of behavior. Instead, the activa-tion of a social construct has been hypothesized to activate
related traits and ultimately the behavioral representations that are implied by those traits. Once activated,
these behavioral representations then affect behavior by either producing the behavior in question or, more likely,
by modifying ongoing behavior in a manner consistent with the activated stereotype (Dijksterhuis & Bargh,
2001). For example, the activation of “elderly” may activate the trait “slow,” which then activates a behavioral
representation of “slow walking,” which can then affect walking speed (Bargh et al., 1996). It is relatively
straightforward how this mechanism can account for such simple motor behaviors as walking slowly, but it is less
clear how it can affect behaviors with no obvious motor component. For example, what motor representations cause a
person to answer fewer or more Trivial Pursuit questions correctly following a soccer hooligan or professor prime
(Dijksterhuis & van Knippenberg, 1998) or to exhibit impaired memory performance following an elderly prime
(Dijksterhuis, Aarts, et al., 2000)?
Social Functional Perspective
Several researchers have speculated about the evolutionary sig-nificance of the
perception– behavior link. In the domain of non-conscious mimicry, for example, Chartrand, Maddux, and Lakin (2004)
proposed that the tendency to mimic others has evolved to facilitate social interactions. These authors proposed
that mimicry may have played a survival role in the early evolution of our species and developed to serve a social
function because of its ability to facilitate and indicate liking (Chartrand et al., 2004; Lakin, Jefferis, Cheng,
& Chartrand, 2003). For example, mimicry of others has been shown to increase when a goal for affiliation is
active (Lakin & Chartrand, 2003), and being mimicked by another person can increase one’s liking of that person
(Chartrand & Bargh, 1999).
Recently, Kawakami, Dovidio, and Dijksterhuis (2003) ex-tended this social
functional perspective for mimicry behavior (Chartrand et al., 2004) to the
activation of stereotypes. They proposed that other instantiations of the perception– behavior link might also
serve a social-adjustive function, even when a social situation and direct perception of behaviors are not
involved. On the basis of a large body of research demonstrating a relationship between attraction and attitudinal
similarity (Byrne, 1961; New-comb, 1961) that indicated that behavior and self-disclosure indic-ative of
attitudinal similarity could serve to increase liking, Kawakami et al. (2003) argued that assimilative attitudinal
reports could be another socially adaptive outcome of subtle stereotype activation. That is, the activation of a
stereotype should produce stereotype-consistent attitudes as well as behavior because people “may . . . ‘socially
tune’ their attitudes to make them more con-sistent with their social environment (Lowery, Hardin, & Sinclair,
2001)” (Kawakami et al., 2003, p. 318). Results of four studies showed that stereotype priming can temporarily
produce attitudes consistent with those presumably held by the stereotype targets (Kawakami et al., 2003). For
example, following an elderly ste-reotype prime, participants endorsed more conservative attitudes, and following a
skinhead prime, participants endorsed more racist attitudes, compared with control participants.
This perspective offers a key advance over ideomotor theory. That is, the social
functional approach allows for changes in the expression of attitudes, judgments, and behaviors with no obvious
motor component. This is because the social functional account holds that activating social constructs causes
people to want to fit in with their social environment. Any expressed opinions, judg-ments, and behaviors that
would help the person fit in could therefore be modified.
The Active-Self Account
In this article, we examine the utility of a third possible mech-anism for
understanding the impact of nonself social construct primes on one’s own judgments and actions. This view, first
articulated by Wheeler and Petty (2001), suggests that the activa-tion of a nonself stereotype could potentially
bias a person’s self-representations in a prime-consistent manner, and behavior then follows from the activated
self. This could occur if the activated prime content were somehow linked to, or included in, a person’s
self-concept. A number of existing studies offer prelim-inary support for this notion.
In one pair of studies (Wheeler et al., 2001), White college students wrote an
essay about a day in the life of another student. This essay served as the priming manipulation, as one group of
participants was assigned to write their essays about Tyrone Walker, whom most assumed to be African American,
whereas others were assigned to write their essays about Erik Walker, whom most assumed to be White. Consistent
with the African American stereotype of academic underperformance, participants who wrote their essays about Tyrone
performed worse on a sub-sequent Graduate Record Examination math test compared with control participants. It is
interesting to note, though, that this effect was driven by people who spontaneously wrote their essays from the
first-person perspective. These individuals’ essays were written as if they were Tyrone Walker and they engaged in
a number of stereotypic acts. Thus, individuals who wrote their essays from the first-person perspective may have
spontaneously linked the activated trait concepts to their own
self-construals, thereby making stereotype-consistent behavior more likely.
Additional support can be found in a study by Hull, Slone, Meteyer, and Matthews
(2002). They replicated Bargh et al.’s (1996, Study 2) elderly-prime walking-speed study and also in-cluded a
measure of private self-consciousness (Fenigstein, Scheier, & Buss, 1975). Hull and colleagues’ earlier work
showed that people high in private self-consciousness are more likely to process information in a self-relevant
manner (Hull & Levy, 1979; Hull, Van Treuren, Ashford, Propsom, & Andrus, 1988). The results of this study
revealed greater stereotype activation effects among participants high in private self-consciousness. Thus, this
finding is consistent with the idea that priming effects are magni-fied when participants create self-stereotype
linkages during the priming phase of the experiment.
These two sets of studies suggest that linkages of the prime content to the self
increase the behavioral effects of a prime. One possible reason for this is that the self-prime linkages may be
associated with a change in self-perceptions (implicitly or explic-itly) following a prime. Preliminary evidence
for a change in self-perceptions comes from some research by Dijksterhuis et al. (1998, Study 1). In one
experiment, participants were primed with an intelligence-relevant stereotype (e.g., professor) or an
intelligence-relevant exemplar (e.g., Einstein). Following stereo-type activation, participants completed a Trivial
Pursuit task as a measure of intelligence-relevant behavior. Participants’ perfor-mance assimilated to that of a
category prime (i.e., they answered more questions correctly following a professor prime) but con-trasted away from
that of an exemplar prime (i.e., they answered fewer questions correctly following an Einstein prime). Of greater
relevance, in another study (Dijksterhuis et al., 1998, Study 3) participants completed a lexical-decision task
following the prim-ing manipulation in place of the test. The target words used in this task were neutral or
related either to intelligence or to stupidity. Immediately preceding each target to be judged, participants were
subliminally primed with one of two sets of words: self-related words (e.g., “I” or “me”) or neutral words (e.g.,
“it”). Results indicated that participants in the professor-prime condition re-sponded more quickly to
intelligence-related words, regardless of the self-relevance of the preceding word. A rather different pattern was
shown for the Einstein prime condition. Participants in the Einstein prime condition responded more quickly to
intelligence-related words but also responded more quickly to stupidity-related words when they were preceded by a
self-relevant prime. Although no behavioral measures were included in this study, a comparison with the Study 1
results indicates that the feature uniquely tied to the self (i.e., stupidity in the case of the Einstein prime)
corre-sponds to the pattern of behavior observed. This activation pattern is consistent with a prime-induced bias
in the self-concept that directs behavior.
On the basis of this literature, there is reason to postulate that either
conscious or nonconscious changes in currently accessible self-representations could occur following a prime. We
refer to this perspective as the active-self account (Wheeler, DeMarree, & Petty, 2005). There are several
means by which self-representations could change following a social construct prime, but two possibilities seem
particularly likely (Wheeler et al., 2005). The first possibility is that the activation of a social con-struct
could selectively activate prime-relevant content already contained within the
chronic self-concept, thereby making the Active Self-concept temporarily more construct consistent than it would
otherwise be. This selective increase in the accessibility of prime-relevant aspects of the self could then guide
behavior (see also Markman & McMullen, 2003; Mussweiler, 2003). Another possibility is that the activated
social construct-relevant content could be misattributed to the self, creating an expanded self-concept that would
include aspects of the activated social construct (or its opposite) that are not part of the chronic self-concept.
Stated simply, people could confuse accessible mental contents with their own characteristics and traits (for a
related argument, see Mark-man & McMullen, 2003), much as primed content is sometimes misattributed to the
characteristics of an impression-formation target (Higgins, Rholes, & Jones, 1977). Again, this expanded,
prime-consistent self-concept could then guide behavior. In most of the research conducted to date, in which
assimilation has been the dominant finding, this activated self-content has likely been consistent with the social
construct. There are also cases, such as when the prime invokes a comparison with an extreme exemplar, in which the
activated content could include prime-inconsistent information (e.g., Dijksterhuis et al., 1998; Schubert &
Ha¨fner, 2003).
This line of theorizing offers an alternative mechanism by which nonself
stereotypes can impact judgments and behavior. For ex-ample, consider the study described earlier in which
participants reported stereotype-consistent attitudes following a stereotype prime (Kawakami et al., 2003).
Participants in that study may have reported attitudes consistent with the primed stereotype, not be-cause of a
motivation to fit into their social environment as proposed by the social functional account, but rather because
their attitudes had actually changed in a stereotype-consistent manner, at least temporarily. These individuals
could have incorporated stereotype-consistent content (e.g., stereotype-relevant attitudes) into their active
self-concept. The social functional perspective does not prohibit a change in the self-concept, but it predicts
that this change, if it occurs, would occur in the service of fitting in with others. In contrast, the Active-Self
model does not have such a requirement. In fact, our account proposes that the change is not in the service of
fitting in with others, but rather in the service of being oneself. Put in functional terms, the Active-Self
account would be more consistent with a self-expressive function (e.g., Abelson & Prentice, 1989) than with a
social-adjustive function. The social functional perspective also presumably limits itself to assimilative
responses (i.e., acting consistent with activated con-tent in order to fit in), whereas the Active-Self perspective
allows assimilation or contrast to occur, depending on which prime-relevant elements (consistent or inconsistent)
are most active and incorporated into the self-concept.
The social functional and Active-Self perspectives map onto a distinction that has
been made in the social influence literature between normative and informational influence, respectively (Cialdini
& Trost, 1998; Deutsch & Gerard, 1955). In normative social influence (e.g., Asch, 1956), a person conforms
for the sake of gaining or maintaining positive regard from others. Informa-tional influence (e.g., Sherif, 1935),
on the other hand, is “influ-ence to accept information obtained from another as evidence about reality” (Deutsch
& Gerard, 1955, p. 629). Normative influ-ence occurs in situations in which fitting in is important, whereas
informational influence occurs in situations in which the correct response is
unknown or ambiguous. This final point is a critical one for our analysis, as we will use the individual difference
variable of self-monitoring (Snyder, 1974, 1979) to identify people for whom fitting in is important across
situations and who should thus be more likely to show behavioral effects of social construct primes according to
the social functional perspective. As we ex-plain further below, according to the social functional perspective,
individuals high in self-monitoring should be more susceptible to prime-to-behavior effects, but according to the
Active-Self view, it is low self-monitors who should be more susceptible.
Self-Monitoring
The self-monitoring construct (Snyder, 1974) was developed to describe differences
in the extent to which people monitor and control their behavior and the public image they present (Snyder, 1979).
High self-monitors are social chameleons who modify their behavior to best fit into their current social
environment. They are attentive to social cues and are capable of controlling their inter-personal behavior to
achieve their desired public image (see Gange-stad & Snyder, 2000). It has been hypothesized that “the
attitudes and behavior of high self-monitors may be especially likely to serve the social-adjustive function”
(Lavine & Snyder, 2000, p. 101), and considerable research supports this view (e.g., see De Bono & Harnish,
1988; Petty, Wheeler, & Bizer, 2000). High self-monitors are also more susceptible to normative social
influ-ence than are lows (Cialdini & Trost, 1998; Snyder & Monson, 1975). If assimilation to primed
constructs stems from a desire to fit into the social context as the social functional view posits, then greater
assimilation to a primed social construct—whether in atti-tudes, judgments, or behaviors—would be expected among
high than low self-monitors.
Low self-monitors, on the other hand, are not as affected by social demands and
instead look inward and rely more heavily on their own attitudes, traits, and beliefs to guide their actions. Their
behavior reflects who they think they really are, in that they demonstrate consistency between their attitudes and
behavior (Snyder & Swann, 1976; Snyder & Tanke, 1976; Zanna, Olson, & Fazio, 1980), between their inner
states and self-presentation (Ickes, Layden, & Barnes, 1978), and between their personality characteristics and
behavior (Lippa, 1978). Snyder and Campbell (1982) nicely summarized these tendencies when they wrote, “To live
one’s life according to the principled theory of self . . . would require low-self-monitoring individuals to pay
serious attention to their own internal states, dispositions, and personal characteristics in order to guide their
social behavior” (p. 191). Because low self-monitors are more likely to act like themselves, greater as-similation
to a primed social construct among low self-monitors would be expected to the extent that the primed material is
incor-porated into their currently active self-concept.
In addition, there is evidence that low self-monitors exhibit greater change in
their behavior and self-perceptions in response to perceived dispositionally diagnostic information than do high
self-monitors. For example, low self-monitors exhibit more attitude change after a freely chosen counterattitudinal
behavior than do high self-monitors (Snyder & Tanke, 1976), and they show greater changes in impression
formation behavior following false feed-back about their own dispositional characteristics (Fiske & von Hendy,
1992).
Paralleling these findings is the evidence that low self-monitors may be more
responsive to priming manipulations if subtle acti-vation of traits is perceived to convey information about their
self-characteristics or subjective feelings. Because individuals are unlikely to have perfectly accurate knowledge
about the self, the traits made accessible by a prime could be used as cues about one’s self-characteristics and
subjective states (Markman & McMullen, 2003; Mussweiler, 2003; Stapel & Koomen, 2001). Much as perceptions
of an impression formation target are biased by the accessibility of applicable traits (Higgins et al., 1977),
one’s active self-concept may also be biased by construct accessibility, because the construct accessibility is
misattributed to some aspect of the self. Primed constructs could be perceived to be self-relevant because the
subtle activation of social constructs may appear to originate from the self (e.g., Mussweiler & Neumann,
2000). If a subtly activated social construct were to be perceived as diagnostic self-information, it would exert
larger effects among low self-monitors just as does more explicit feedback that is perceived to be dispositionally
diagnostic (Fiske & von Hendy, 1992). Because the self-concepts of high self-monitors are affected by social
and not dispositional feedback (Fiske & von Hendy, 1992), high self-monitors would be expected to exhibit less
self-change in response to a priming manipulation, unless the prime was perceived to convey social information,
rather than information about the self.
Additionally, as noted above, if low self-monitors are more likely than high
self-monitors to use primed constructs to alter their self-concept in a prime-relevant manner, then it is likely
that low self-monitors would also show larger behavioral effects of a prime because their behavior is more tied to
their internal states (Ickes et al., 1978; Lippa, 1978; Snyder & Swann, 1976), but only if the primes biased
the active self-concept. Hence, each of these potential mechanisms (differential self-change and differential
self-behavior consistency) would be sufficient to lead to greater prime-induced behavioral change among low
self-monitors than among high self-monitors.
It may seem counterintuitive that low self-monitors, who habit-ually rely on
self-knowledge, would exhibit greater self-change in response to these types of manipulations than would high
self-monitors. However, as Fiske and von Hendy (1992) have pointed out, “[low self-monitors] may or may not be
accurate and stable in their self-perceptions, but they will rely on whatever self-perception they believe to be
accurate at the time” (p. 579, em-phasis added). In discussing their findings, Fiske and Hendy fur-ther noted
that
low self-monitors are not necessarily accurate about their own inner states and
abilities; they are merely more likely to base their behavior on their perceived inner characteristics. This is an
important distinc-tion. Just because one is attuned to one’s inner states (low self-monitors) or to the social
world (high self-monitors), this does not mean one is more accurate in that regard. Our data suggest that low
self-monitors may be susceptible to false feedback about their per-sonalities. (p. 589)
In the current research, we examined how high and low self-monitors react to
various primes in order to provide information about the likely mechanism behind at least some prime-to-behavior
effects. Notably, each of the three perspectives on social construct prime-to-behavior effects we have discussed
predicts a different pattern of results across levels of self-monitoring. The ideomotor action model, which does not posit any change in the self, predicts no moderation of
priming effects by self-monitoring. The social functional perspective predicts the greatest amount of assimilation
among high self-monitors, who are motivated to fit into their social environments, because this account holds that
primes implicitly communicate social demands. The Active-Self model predicts larger effects among low
self-monitors, who may be more likely to change their self-conceptions on the basis of primed material and who are
more concerned with acting on their internal states, because this account holds that primes can operate not by
communicating social demands, but instead by influencing the active self-concept.
Study 1
In Study 1, we examined whether a primed stereotype would alter momentary
self-perceptions. Critically, we examined for the first time whether (and in which direction) self-monitoring would
moderate such changes in the active self-concept. We primed participants with the African American stereotype and
used an implicit measure of felt aggression as our dependent variable. Past research on stereotype-to-behavior
effects that used the African American stereotype found an increase in aggressive responses following the
activation of the stereotype (Bargh et al., 1996). For example, Bargh et al. (1996) found that participants who
were subliminally primed with pictures of African American faces dis-played more aggressive facial expressions
following a frustrating event than did control participants. For Experiment 1, we created an implicit measure of
aggression that could be administered easily in a computer lab. In designing this measure, we assumed that
participants who had been primed with the African American stereotype would feel more aggressive and thus would
choose more stimuli that are aggression relevant when they were told to use their own feelings in an ambiguous
perception task. The logic behind this measure is similar to that used to develop an implicit measure of mood
(Hass, Katz, Rizzo, Bailey, & Moore, 1992).
As noted above, if the prime is confused as being diagnostic self-content, the
reports of low self-monitors should be more affected by the prime than would the reports of high self-monitors. On
the other hand, if the prime is seen as conveying social information, the reports of high self-monitors should be
more affected by the prime than would the reports of low self-monitors, because of the motivations of high
self-monitors to fit in with social demands. Ideomotor theory would not predict any changes in the self-concept.
Because the underlying ideomotor processes are postulated to occur in an unmediated and automatic fashion, even in
animals without any self-awareness, there should be no biasing of the self-concept among either high or low
self-monitors, except potentially through self-perception processes that could follow behavior. Because there were
no behaviors in this study prior to the implicit self measure, ideomotor theory would predict no self-concept
change.
Method
Participants
One hundred thirty-two Ohio State University undergraduates who par-ticipated in
partial fulfillment of a course requirement were randomly assigned to one of the experimental conditions.1 Because
the effects of self-and nonself-stereotypes have sometimes been shown to differ (Shih, Ambady, Richeson, Fujita, & Gray, 2002) 14 African American participants and 6 participants who
did not indicate their race were deleted from the analyses. In addition, because of a computer error, the
aggression data from 1 participant were not available, leaving a total of 111 participants in the final sample (60
female, 51 male). Sessions were conducted in a computer lab with divided workstations, and each session included 2
to 8 participants.
Participants were informed that the study examined language and that they would
complete a short writing task with pen and paper, which would be followed by additional verbal tasks on the
computer. The instructions indicated that the computer would assign participants an essay topic.
Stereotype Prime
For the African American prime, participants were instructed to write “an essay
about a day in the life of an Ohio State University student, Tyrone Walker.” Control participants wrote their
essays about a presum-ably White student, Erik Walker. Participants were given 6 min to com-plete their essay. This
stereotype priming manipulation has been used successfully in prior research (Wheeler et al., 2001).
Self-Monitoring
All participants completed the 18-item Self-Monitoring Scale (Gange-stad &
Snyder, 1985; Snyder, 1974; Snyder & Gangestad, 1986), which assesses the degree to which people are motivated
and able to modify their self-presentations to suit a given situation. Participants were asked to indicate whether
each statement was or was not characteristic of them. Items were coded so that one point was allotted for each
answer consistent with high self-monitoring. The score was computed by adding the number of such
responses.2
Dependent Measure
Our dependent variable was an implicit aggression measure modified from an
existing implicit mood measure (Hass et al., 1992; Koole, Smeets, van Knippenberg, & Dijksterhuis, 1999).
Participants were told that the task was a measure of subconscious language perception and that a word would be
flashed on the screen so quickly that they could not consciously perceive it. They were then told,
Your subconscious will be able to perceive the presented word. Therefore, if you
select (by guessing, if necessary) a word that feels similar in meaning to the feeling you experience while the
word is being flashed, your subconscious will guide your decisions. Thus, go with your feelings.
1 As a pilot for future research, we included an additional manipulation of
anticipating an interaction for 37 of the first 50 participants. Because preliminary analyses revealed that this
manipulation was unsuccessful, it was dropped.
2 There is some controversy in the literature on the Self-Monitoring Scale.
Some of this debate concerns whether self-monitoring should be treated as a continuous or a discrete variable (see
Gangestad & Snyder, 1985; Miller & Thayer, 1989) or which version of the Self-Monitoring Scale should be
used. We chose to adopt Miller and Thayer’s (1989) recommendation to use the 18-item Self-Monitoring Scale and to
treat it as a continuous variable. Analyses using Gangestad and Snyder’s (1985) recommended split (10.5 on the
18-item scale) as well as those using the full 25-item scale produced the same pattern of results in all studies.
In addition, there is some debate as to whether the scale should use a true–false response format or a more
traditional Likert-type scale (see Gangestad & Snyder, 1991; Miller & Thayer, 1989). We used the original
true–false response format in these studies (Snyder, 1974).
The trials began with an asterisk serving as an orienting stimulus for 2,000 –5,000
ms, followed by the subliminal presentation (17 ms) of the target word. These words were actually nonword letter
strings (e.g., “ag-grimely”) that resembled the target words in appearance (e.g., aggressive, aggregate,
agriculture, and agreement). A mask of number signs (#s) then covered the word for 75 ms, and then the four
response options were presented and remained on the screen until the participants made their selection. Half of the
trials were target trials, in which one of the four response options was an aggression-relevant word (e.g.,
aggressive, beat, angry). The position of the aggression-relevant word was rotated, and the order of the trials was
randomized for each participant. We used aggression-related words because aggression and hostility are common
elements of the African American stereotype (Devine, 1989; Sagar & Schofield, 1980). By having participants
choose a word that “feels similar in meaning to the feeling [they] experience,” we hoped to detect any changes in
self-perception of their own aggressive feelings as a result of the activation of the stereotype and the extent to
which they used those self-perceptions in judgment. Following the aggression task, the Self-Monitoring Scale was
administered on the computer. Finally, participants completed an open-ended suspicion probe and were
debriefed.
Results
Analyses in all three studies followed the regression procedures outlined by Aiken
and West (1991). Accordingly, self-monitoring scores were mean centered by subtracting the mean of self-monitoring
from all observations to reduce multicollinearity con-cerns (Aiken & West, 1991). Initial analyses were
conducted with this mean-centered variable, and the cross-product of self-monitoring with condition provided the
interaction term for the model. If a significant interaction was found, it was decomposed by using a simple slopes
analysis. Self-monitoring was recentered at one standard deviation above and below the mean, and the interaction
term was recomputed with this recentered factor. The full model was then rerun, and the simple effect of prime for
high and low self-monitors was observed as the prime main effect in this recentered model (for complete details of
these procedures, see Aiken & West, 1991).
Scores on the implicit aggression measure were computed by summing the number of
aggression-relevant words selected. Ag-gression scores were then submitted to a Prime Self-Monitoring multiple
regression analysis. A significant effect of prime emerged (B 0.65), t(108) 2.10, p .04, such that participants who
wrote their essays about Tyrone selected more aggression-relevant words than did those who wrote about Erik. It is
important to note that this was qualified by the predicted Prime Self-Monitoring interaction (B 0.19), t(107) 2.15, p .03 (see Figure 1).
Decomposition of this interaction one standard deviation above and below the mean
(Aiken & West, 1991) indicated that there was a significant effect of prime among low self-monitors (B1.37),
t(107) 3.01, p .003, but not among high self-monitors (B 0.02, ns). That is, low self-monitors primed with the
African American stereotype displayed more felt aggression on the implicit measure than did those receiving the
control prime.
Discussion
This study demonstrated that assimilation to an activated ste-reotype occurs more
for low self-monitors than for high self-monitors. Low self-monitors primed with the African American stereotype
chose more aggression-relevant words to describe their current feelings than
did control prime participants. This effect was not observed among high self-monitors. This study shows for the
first time that self-monitoring can moderate the effects of a ste-reotype prime on self-perceptions. The effects
observed lend sup-port to the Active-Self notion, which holds that primes can alter self-perceptions and do so for
those who are most attuned to the self—those low in self-monitoring. These results are more difficult to explain
from the ideomotor perspective and from the perspec-tive that prime-to-behavior effects are due to a motivation to
fit into the social environment (Kawakami et al., 2003). The ideomo-tor perspective would not predict changes in
the self-concept for either high or low self-monitors. The social-adjustive perspective would predict that high
self-monitors, who are more likely to adjust their behavior to fit into their social environment, should show a
greater effect of the prime, whereas in this study, they showed no effect of the prime.
There are some potential alternative explanations for our find-ings on the
implicit aggression measure, however. First, it could be that low self-monitors might have a stronger or more
easily ac-cessible stereotype of African Americans stored in memory. Thus, one might argue that perhaps our
dependent measure did not assess self-perceptions but instead functioned as an implicit stereotyping measure.3
Although possible, we think that this alternative is unlikely because past research suggests that high
self-monitors, because of their greater reliance on behavioral scripts and proto-type information, have more
elaborate and accessible information regarding social groups (Schwalbe, 1991; Snyder, 1979; Snyder & Cantor,
1980). Thus, if this alternative account of our data was true, high self-monitors should exhibit larger stereotype
activation effects. Additionally, high self-monitors are more likely to apply stereotypes as well. For example,
high self-monitors experience more negative affect when an interviewer’s face does not match an occupational
stereotype (Larkin & Pines, 1994), and they are more likely to apply stereotypes (of affluence) when judging
other people (Christopher & Schlenker, 2000). In sum, the literature suggests that high self-monitors have more
elaborate stereotype representations and more frequent stereotype application than do those low in self-monitoring,
and this alternative therefore seems to be an unlikely account for our results. However, to further reduce the
likelihood of this concern, Study 2 uses a similar methodology, but instead of
a stereotype prime, it uses the com-mon association of the number 7 with good luck and the number 13 with bad luck
that should be comparable for nearly all people.
An additional alternative explanation concerns the possibility that high
self-monitors may be better than low self-monitors at bias correction processes. Although no participants reported
any sus-picion that the priming manipulation could have affected their responses on the implicit aggression task,
it is possible that high self-monitors are more skilled at identifying and correcting for potential biasing
influences than are low self-monitors. Theories of bias correction postulate that individuals must become aware of
the potential biasing agent as well as have the motivation and ability to correct their judgments for the perceived
bias (Petty & Wegener, 1993; Wilson & Brekke, 1994). Hence, Study 2 used a subliminal priming manipulation
and innocuous priming stimuli that should have reduced the likelihood of any possible correction.
Study 2
To address the alternative explanations offered for Study 1, we conducted another
study by using a similar dependent measure but a different priming procedure and a different primed construct.
Specifically, in this study, we used the numbers 7 and 13 as primes and an implicit measure of luck as the outcome
measure. We assumed that the associations between 7 and lucky and between 13 and unlucky were very common in
society, and that these associ-ations would be less likely to differ between high and low self-monitors and less
likely to activate social desirability concerns or motivations to correct. Also, by using a subliminal prime, we
further expected to reduce the likelihood that correction processes could operate.
Method
Participants
Sixty-seven Stanford University students and staff, who were compen-sated with $10
for their participation in the experiment, were randomly assigned to a condition. Because of computer error,
complete data from 6 participants were not written to the hard drive, leaving 61 participants in the final sample
(37 female, 24 male). Sessions were conducted in a computer lab with divided workstations with 2 to 8 participants
per session. Participants were informed that the study examined how people make judgments about words and that
there would be two unrelated word experiments during the session.
Prime
Primes were subliminally presented to participants during a lexical-decision task.
The subliminal primes consisted only of numbers presented in the center of the screen: the number 7 in the lucky
prime condition and the number 13 in the unlucky prime condition. The number was presented for 17 ms before
presentation of a back mask (e.g., HENKOS) for 225 ms. Immediately following the back mask, a luck-unrelated word
(e.g., “some-thing”) or nonword (e.g., “botchef”) was presented until participants made a word or nonword judgment.
There was a total of 97 trials (49 word trials and 48 nonword trials), with a 150-ms intertrial
interval.
The score was computed by adding the number of responses indicating high levels of
self-monitoring.
Dependent Measure
Our dependent variable was an implicit luckiness measure similar to the aggression
measure used in Study 1. Instructions and procedures were identical to those in Experiment 1, with the exception
that the target words were related to luck and distracter words were similar in appearance to the luck words. For
example, in this experiment, one nonword presented subliminally was “bufky,” and the set of response options
included “lucky,” “lofty,” “ducky,” and “softy.” As in Study 1, half of the 28 trials were target trials, the
position of the luck-related word was rotated, and order of trials was randomized between participants. By having
participants choose a word that “feels similar in meaning to the feeling [they] experi-ence,” we hoped to detect
any changes in self-perception of their own feelings of luckiness as a result of the number prime and the extent to
which they used those self-perceptions in judgment. Following the lucki-ness task, the Self-Monitoring Scale was
administered on the computer. Finally, participants completed an open-ended, funneled suspicion probe and were
debriefed.
Results
Scores on the implicit luckiness measure were computed by summing the number of
luck-relevant words selected, and self-monitoring scores were centered at the mean (Aiken & West, 1991) prior
to analyses. Luckiness scores were then submitted to a Prime (7 vs. 13) Self-Monitoring multiple regression
analysis. This analysis revealed the presence of an outlier (univariate z 3.09, p .002) who selected luck-related
words on 12 of 14 possible trials. This outlier was poorly predicted by the regression analysis (standardized
residual 3.11, p .002; Studentized deleted residual 3.62,
p .0006). Regression analyses were therefore run both with and without
the outlier included in the analysis (Cohen, Cohen, West, & Aiken, 2003; McClelland, 2000). When the outlier
was included, neither the prime main effect (B
0.47), t(58) 0.94, p .35, nor the Prime Self-Monitoring
interaction (B 0.22), t(57) 1.50, p .14, attained
conventional levels of significance. When this one outlier was
discarded, the predicted Prime Self-Monitoring
interaction emerged (B 0.316), t(56) 2.31,
p .025.
Decomposition of this interaction at one standard deviation above and below the
mean of the self-monitoring distribution (Aiken & West, 1991) indicated that the main effect of priming was
significant.
Self-Monitoring
As in Study 1, all participants completed the 18-item Self-Monitoring Scale
(Gangestad & Snyder, 1985; Snyder, 1974; Snyder & Gangestad,
Figure 2. Luckiness scores as a function of prime and self-monitoring (plotted at
one standard deviation above and below the mean).
Significant for low self-monitors (B 1.36), t(56) 2.07, p .04,
but
not for high self-monitors (B 0.98), t(56) 1.39, p .17.
Discussion
This study again demonstrated that assimilation of self-perceptions to an
activated construct occurs more for low self-monitors than for high self-monitors. Low self-monitors primed with
the number 7 chose more luck-relevant words to describe their current feelings than did participants primed with
the number 13. This effect was not observed among high self-monitors. This study replicates and extends the basic
findings of Study 1.
This study rules out several alternative explanations that were possible
concerning the first study. Because this study used a subliminal prime, conscious correction among high
self-monitors seems unlikely, because one must be aware of a bias in order to consciously correct for it (Petty
& Wegener, 1993; Wilson & Brekke, 1994). Although automatic correction is possible (Wege-ner & Petty,
1997), this is likely to occur only with highly prac-ticed corrections, such as in situations in which bias
regularly occurs and is unwanted (e.g., see Maddux, Barden, Brewer, & Petty, 2005). Habitual correction for the
influence of lucky and unlucky numbers seems unlikely. In addition, the association be-tween numbers and luck seems
less likely to differ among high and low self-monitors because this association is unrelated to social contexts or
to the self.
Study 2 again provided results more consistent with the active-self predictions
than with those of ideomotor theory or social functional theory. Additionally, Study 2 extends the priming
lit-erature by showing that priming stimuli with social associations (i.e., symbols) can lead to self-concept
change much as do other types of social stimuli. Hence, this study shows that stimuli with social meaning cannot
only influence behavior and judgments (Kay et al., 2004) but also change momentary self-conceptions.
There is a potential alternative explanation for our findings on the implicit
aggression and luck measures, however. Research has shown that accessible constructs, such as stereotypes, are
easily processed and are likely to attract attention (Higgins, 1996; Mac-rae, Milne, & Bodenhausen, 1994;
Macrae, Stangor, & Milne, 1994). Low self-monitors, who are guided by their internal states, may have selected
more prime-consistent items on the basis of mere accessibility alone and not because they actually felt more
aggressive, despite our instructions to respond with their own feelings. High self-monitors, whose attention was
focused out-ward, may have shown a null effect because they did not notice any difference in accessibility. This
may be unlikely, as pointed out in the discussion of Study 1, because if these constructs were simply more
accessible and not linked to the self, then one would expect high self-monitors, not lows, to show larger priming
effects because of their increased reliance on stored social constructs. Nevertheless, Study 3 was conducted to
test and rule out the accessibility alternative.
Study 3
Both of our studies thus far have used a measure atypical of the prime-to-behavior
literature in that most research has examined some overt behavior or task performance that results from priming.
Indeed, only the study on attitudes mentioned earlier by Kawakami et al.
(2003) comes close to our focus on self-change that can result from a prime. Nevertheless, the results of our
studies were con-sistent with the idea that activated stereotypes can modify self-perceptions. Notably, this change
in self-perception was evident only among those individuals who were low in self-monitoring— individuals who are
vigilant in accessing and acting on self-knowledge. It is still possible, however, that self-monitoring mod-erates
the effects of stereotype activation on self-perceptions but not on overt action or task performance. Thus, we
conducted a third study to determine whether the self-monitoring effect would obtain on a more behavioral outcome
measure. In addition, Study 3 used two different stereotypes to ensure that our findings gen-eralized beyond the
constructs activated in Studies 1 and 2. In this study, we chose to use the professor and supermodel stereotypes,
which have opposing implications for information processing (Dijksterhuis et al., 1998; Dijksterhuis & van
Knippenberg, 1998). As noted earlier, past research has used tests such as those based on Trivial Pursuit questions
to assess information-processing be-havior. We instead chose to use an argument-quality manipulation within a
persuasion paradigm to assess information-processing activity (Petty & Cacioppo, 1986). This has the benefit of
gener-alizing the stereotype activation effect to another task.4
In contemporary persuasion research, the strength of arguments in a message is
often manipulated between participants in order to gauge the extent of message processing (e.g., Petty &
Cacioppo, 1979). Some participants receive strong arguments that have been pretested to elicit primarily favorable
thoughts about the focal product or issue when thinking is high, whereas other participants receive weak and
specious arguments that elicit primarily unfa-vorable thoughts. Because all arguments lead to the same conclu-sion
(e.g., “You should buy product X”), the attitudes of people who are not paying much attention to them (e.g.,
because they are distracted) may not differ between the strong and weak conditions. However, the attitudes of
people who are carefully scrutinizing the message (e.g., because personal relevance is high) will be more
influenced by the argument-quality manipulation (i.e., more favor-able attitudes following strong arguments than
following weak ones). In our priming paradigm, behaving like a professor would involve careful thought about the
arguments (just as it would involve performing well on a Trivial Pursuit game; see Dijkster-huis & van
Knippenberg, 1998), whereas acting like the stereotype of a supermodel would involve less thought. If low
self-monitors act like the primed concepts, we would expect their attitudes to be influenced by argument quality
more when primed with professor than with supermodel. The behavior of high self-monitors should not be as
influenced by the primes.
4 Although we acknowledge that this study does not involve overt motor
behavior, much of the research on behavioral priming effects, including a considerable amount of research on
ideomotor theory, has used dependent variables that are more mental than motor oriented. For example, depen-dent
measures in the automatic behavior literature have included test performance (Dijksterhuis et al., 1998;
Dijksterhuis & van Knippenberg, 1998; Haddock, Macrae, & Fleck, 2002; Schubert & Ha¨fner, 2003; Shih et
al., 2002; Shih, Pittinsky, & Ambady, 1999; Wheeler et al., 2001), deci-sions in ultimatum and social dilemma
games (Kay et al., 2004; Smeesters et al., 2003), forgetting (Dijksterhuis, Aarts, et al., 2000; Dijksterhuis,
Bargh, & Miedema, 2000), and attitude reports (Kawakami et al., 2003). Hence, although one could argue that
information processing is not behav-ioral, it is hardly atypical of the types of dependent measures used in this
literature, including those for which ideomotor and social functional ac-counts have been invoked. Indeed, the
large number of nonmotor behavior studies in this literature is noteworthy, given that ideomotor theory is, as the
name implies, a theory about motor behaviors. It is possible that simple motor behaviors and more complex behaviors
like information processing could each have a different mediating mechanism, and so it could be that the
Active-Self and ideomotor accounts are each applicable for a different range of dependent variables, but at this
point, such a claim is highly speculative. What is important from our perspective is that both the ideomotor and
social functional accounts have been applied to the kind of mental behavior we examined in Study 3 (see
Dijksterhuis & Bargh, 2001; Kawakami et al., 2003).
Method
Overview
In this investigation, we primed half of the participants with the profes-sor
stereotype and half with the supermodel stereotype by using an essay paradigm much like that used in Study 1.
Following the priming manipu-lation, participants viewed several advertisements, including the target ad for a
watch, and reported their attitudes. As noted earlier, we included an argument-quality manipulation in the target
ad to assess the extent of thinking in which participants engaged. The design of the study was a 2 (prime:
professor vs. supermodel) 2 (argument quality: strong vs. weak) Self-Monitoring between-participants
design.
Participants
Participants were 70 Stanford University students (29 male, 41 female) who
received $10 in compensation for their participation. Sessions were conducted in a classroom with approximately 10
–25 participants per session. All participants were told that the research involved two separate and unrelated
tasks.
Independent Variables
Priming manipulation. To conceal the purpose of the stereotype acti-vation task,
we told participants that the purpose of the first task was to examine the “role of hemispheric dominance on
creativity tasks.” The instructions indicated that participants would be assigned to write a cre-ative essay about
an assigned topic with either their dominant hand or nondominant hand in order to examine the relationship between
hemi-spheric brain activation and creativity. In fact, all participants were as-signed to write with their dominant
hands. Participants were instructed to write an essay about a day in the life of an individual, including his or
her behaviors, lifestyle, and appearance attributes. Half of the participants were instructed to write about a
professor, whereas the other half were instructed to write about a supermodel. All participants were told that they
would have 5 min to complete their essays. Aside from the topic manipulation, all participants received identical
instructions.
Message manipulation. Participants read two filler advertisements and then read
the target advertisement, which was for a wristwatch for men and women. Half of the participants received strong
arguments for the wrist-watch, whereas the other half of the participants received weak arguments for the
wristwatch. An example strong argument is, “The Chronotech is water resistant up to 350 feet and is suitable for
swimming and diving.” An example weak argument is, “The Chronotech is water resistant if not submerged and is
suitable for wear in all climate conditions.”
Self-Monitoring
As in Studies 1 and 2, self-monitoring was assessed with the 18-item scale
(Gangestad & Snyder, 1985; Snyder, 1974).
Dependent Variables
Following the priming task, the experimenter indicated that participants would be
moving on to Experiment 2. The experimenter told participants that they would be evaluating advertising copy for
different products and services. At that time, booklets containing the target and filler ads, attitude measures,
Self-Monitoring Scale, and suspicion checks were distributed.
Participants rated their attitudes toward the wristwatch on four 7-point semantic
differential scales anchored by bad and good, negative and positive, undesirable and desirable, and unfavorable and
favorable. Par-ticipants rated their attitudes toward the advertisement on these same scales as well as one item
that simply asked how much they liked the ad on a scale from 1 (not at all) to 7 (very much). Last, participants
rated the desirability of the watch’s features and the likelihood that the watch actually possessed these features.
Each item was anchored by a scale of 1 (not at all) to 7 (very much). These last two items were combined
multiplicatively to form a Likelihood Desirability index. This variable is based on research indi-cating that
attitudes are a multiplicative function of the perceived desir-ability and likelihood of attributes and serves as
an effective measure of belief-based attitudes (Fishbein & Ajzen, 1975). Scores on this index are higher when
participants perceive the products’ features to be very desir-able and very likely.
ResultsThe three attitude
measures were highly related ( .88), and each exhibited identical effects. These variables were therefore
standardized and averaged.5 Scores on the Self-Monitoring Scale were centered by subtracting the mean from each
person’s score (Aiken & West, 1991).
The attitude measure was subjected to a Prime X Argument X Quality Self-Monitoring multiple
regression analysis. There was a main effect of argument quality (B 1.25), t(65) 7.26, p
.0001, such that strong arguments produced more positive attitudes than did weak
arguments. The predicted three-way Prime Ar-gument Quality Self-Monitoring interaction was the only other effect to
emerge (B 0.30), t(61) 3.06, p .003. As depicted in Figure 3A, among low self-monitors, there was a
significantbPrime X Argument Quality interaction (B 1.51),
t(61) 3.19, p .002, such that low self-monitors primed with the
professor stereotype exhibited a larger argument-quality effect (consistent
with a thoughtful professor) (B 2.32), t(61) 6.10, p .0001, than did
low self-monitors primed with the supermodel stereotype (B 0.82), t(61) 2.91, p .005. Among high self-monitors, the
Prime Argument Quality interaction was not significant (B 0.49), t(61) 1.07, p .29 (see Figure 3B).
Discussion
This study extends the findings of the first two experiments by demonstrating
moderation by self-monitoring on participants’ information-processing behavior rather than a measure of
self-perception. Again, participants low in self-monitoring assimilated to the prime by thinking more about the
advertisement following the professor prime compared with the supermodel prime, whereas high self-monitors did not
demonstrate a significant effect of prime on behavior. The pattern among low self-monitors conceptually replicates
previous research that has examined the effects of activating these
stereotypes on such intellectual tasks as a Trivial Pursuit game (Dijksterhuis et al., 1998; e.g., Dijksterhuis
& van Knippenberg, 1998), but shows that such effects are moderated by self-monitoring.
In addition, this study helps to rule out the accessibility alter-native discussed
earlier, as simple accessibility of the primed constructs (professor or supermodel) cannot account for differ-ences
in information processing of an advertisement for a wrist-watch. Recall that in the first study, the prime and the
dependent variable contained related content, and as such, the accessibility of the prime alone may have caused the
preferential selection of prime-consistent content, despite our instructions to rely on their feelings. Study 3
used a dependent measure that was unrelated to the content of the specific constructs primed and, as such, is not
as amenable to a mere accessibility-based alternative.
General Discussion
The results from our three studies support the hypothesis that the activation of a
nonself social construct (e.g., African American, lucky, professor) can modify judgments and behavior by
influenc-ing the active self-concept. In Study 1, low self-monitors, who are more likely to change their
self-concept in response to diagnostic self-information and who are more likely to use their internal states in
guiding decisions and behavior, exhibited a pattern of responses consistent with the primed African American
stereotype, whereas high self-monitors did not. In an ambiguous perception task in which participants were asked to
rely on their feelings in making their judgments, low self-monitors selected more aggression-relevant stim-uli
following an African American prime than did control prime participants. High self-monitors showed no effect of the
prime.
In Study 2, we replicated the effect found in Study 1 by using a subliminal priming
manipulation and the construct of luckiness as activated by priming the numbers 7 or 13. This study also used the
ambiguous perception task wherein participants were asked to rely on their feelings, and as in Study 1, only low
self-monitors dem-onstrated an impact of the prime such that those primed with 7 felt luckier than those primed
with 13.
Study 3 allowed us to generalize these findings to additional stereotypes and to a
task-performance measure. Low self-monitors thought more about an advertisement following a professor prime
(consistent with the professor stereotype) than they did following a supermodel prime (consistent with the
supermodel stereotype). There were no significant effects of prime among high self-monitors.
Mechanisms of Self- and Behavioral Change
Self-Monitoring
Self-monitoring processes. The self-monitoring literature has largely focused on
high self-monitors and has emphasized low self-monitors considerably less (for exceptions, see, e.g., Arkin,
Gabrenya, Appelman, & Cochran, 1979; De Bono & Harnish, 1988; Fiske & von Hendy, 1992; Petty &
Wegener, 1998). Our data provide some additional insight into the cognitive processes of low self-monitors. In the
introduction, we posited two potential reasons for why one might expect low self-monitors to exhibit
prime-consistent self-concept and behavior. The first possibility was that because of their responsiveness to
dispositionally relevant information (Fiske & von Hendy, 1992), low self-monitors might be more likely to
modify their working self-concepts on the basis of the social construct prime. The second was that because of their
use of their internal states in guiding behavior, low self-monitors should be more likely to act in accordance with
any prime-induced change in the self.
Both processes are plausible contributors to behavioral priming effects, but our
data really address only the former mechanism with the implicit self-concept measures used in Studies 1 and 2. That
is, these measures indicated that there was a change in the accessible content of the self-concept among low but
not high self-monitors following a prime. A change in self-concept content is consistent with the differential
impact of potentially diagnostic dispositional information on the self. Furthermore, because the measures we used
were implicit, they may be less susceptible to any differences in the extent to which high and low self-monitors
rely on the self-concept in directing conscious judgments and behavior. These data thus provide additional support
for the idea that low self-monitors may sometimes be susceptible to situational influence, but only to the extent
that this influence is perceived as having implications for the self.
Parallels to other constructs. Our discussion of self-monitoring, and the
predictions we make, may sound somewhat reminiscent of the private self-consciousness findings discussed earlier
(Hull et al., 2002). Recall that Hull et al. found priming effects among participants high in private
self-consciousness, but not among those low in private self-consciousness. Our predictions would be redundant if
people high in private self-consciousness were also low in self-monitoring. Despite the surface similarity of these
measures and their predictions, several studies have found null or small
positive correlations between the constructs (e.g., rs .10 to .21; Lamphere & Leary, 1990; Penner & Wymer,
1983), whereas a negative correlation would be expected if low self-monitoring and high private self-consciousness
were empiri-cally equivalent. Additionally, factor analyses that have included measures of public- and private-self
phenomena have found that private self-consciousness and self-monitoring load on separate factors (Lamphere &
Leary, 1990; Penner & Wymer, 1983), fur-ther supporting their independence.
Not only are the constructs distinct ones empirically, but they also predict
different, albeit complementary, prime-to-behavior processes. Specifically, Hull et al. (2002) argued that the
self-consciousness effect is due to participants high in private self-consciousness engaging in self-relevant
processing of the prime content. By processing the prime content in self-relevant ways, they argued, primes that
are objectively inapplicable to the self can still exert an influence on behavior. This type of self-relevant
processing may be one means by which primes could induce self-change and, hence, is complementary with the
Active-Self framework (Wheeler et al., 2005).
Complementary to the process hypothesized by Hull et al. (2002), the private
self-consciousness findings have another po-tential explanation that is not applicable to the self-monitoring
predictions. Because self-relevant processing is deeper (e.g., Sy-mons & Johnson, 1997), high private
self-consciousness individ-uals could exhibit larger priming effects due to greater activation of the primed
construct, rather than due to greater change in the self-concept. Hence, each of these constructs predicts
independent processes, each of which could increase the magnitude of self- and behavioral change following a
prime.
The major difference between Hull et al.’s (2002) theoretical framework and ours
is that we predicted a change in the accessible content of the self-concept. The Active-Self framework can still
account for their findings if, for example, processing the prime content in self-referent ways entails a similarity
testing process, whereby commonalities between the self- and prime content are made more accessible (Mussweiler,
2003), or if perceivers imagine themselves in the primed role (Markman & McMullen, 2003), thereby making
self-concept assimilation more likely. According to the Active-Self account, if this type of self-relational
processing leads to greater activation of prime-consistent self-content, it could increase the assimilative effects
of the primes.
The reasons that self-monitoring may moderate the effects, however, are distinct
from those that Hull et al. (2002) suggested underlay private self-consciousness moderation. One reason we
discussed is that low self-monitors may be more likely to change their self-representations in response to a prime,
not because of deeper processing of the prime content, but rather because they may be more likely to misinterpret
the accessible prime content as diagnostic information about their actual characteristics and be-liefs. A second
reason is that low self-monitors should be more likely to act like a member of a primed stereotype group if that
stereotype is incorporated into their self-concept because low self-monitors are more likely to act on the basis of
their perceived inner states. Neither of these processes has been postulated or demonstrated for individuals high
in private self-consciousness.
In addition to the parallels with self-consciousness, self-monitoring may also
seem similar to individualism– collectivism, in that both high self-monitors and collectivists are concerned
with fitting in with their social environment, whereas individualists and low
self-monitors may both seem more likely to act on the basis of their internal states. However, if these dimensions
were the same, one might predict the opposite pattern of results, because a focus on the personal (vs. collective)
self-concept has been shown to lead to contrast in self-perceptions following exposure to exemplars (e.g., Stapel
& Koomen, 2001). Although there are numerous differences between the studies presented here and work on the
collective self-concept, we examined the extent to which self-monitoring overlapped empirically with individualism
and collectivism.
We included measures of self-monitoring and individualism– collectivism as
ancillary measures in a data collection involving 423 total participants. The correlation between self-monitoring
and individualism in these samples was .17 ( p .001); between self-monitoring and collectivism, the correlation was .02 (ns).
These correlations are small in magnitude and suggest that, if anything, high
self-monitors are more likely to be high in individ-ualism, not in collectivism. These patterns of correlations are
inconsistent with the idea that self-monitoring is a proxy variable for collectivism.
Additionally, although we acknowledge the surface similarities between these
constructs, the predictions made by individualism– collectivism and self-monitoring differ. Whereas low
self-monitors have been shown to evince greater change in their self-beliefs in the face of information believed to
be dispositionally diagnostic (e.g., Snyder & Tanke, 1976), we are not aware of any literature relating this
tendency to individualism– collectivism. Whereas low self-monitors have been shown to evince greater self-behavior
consistency, this pattern has not been shown for individualism– collectivism. Rather, both individualists and
collectivists rely on their self-representations to guide behavior, but the types of self-representations
(individual vs. collective) that drive their respec-tive behavior differ. Our data do not address the question of
whether the primed stereotype is linked to the individual or col-lective self-representation, but the possibility
that it can be linked to either is one that future research should investigate.
Biased Activation Versus Expansion
What happens to the working self-concept of low self-monitors when a social
construct is activated? We have offered two possi-bilities and will discuss these briefly here (see also Wheeler et
al., 2005). One possibility is that social construct activation serves to make a biased portion of one’s chronic
self-concept accessible. Although the prime may not target one’s group memberships, features of the social
construct content could be contained in one’s chronic self-representation. For example, most non-African Amer-ican
individuals can recall instances in which they felt lazy, unin-telligent, or athletic, all of which are elements of
the African American stereotype. Hence, the mental representation of the self and the prime are almost certain to
share some features and therefore overlap to some extent. When the specific features of the self that are
consistent with the primed construct (but typically nondominant) are activated, the participants’ working
self-concept will contain prime-consistent features (Markman & McMullen, 2003; Mussweiler, 2003). We call this
potential mechanism the biased activation model (Wheeler et al., 2005).
Some support for the biased activation account can be found in research on the
relational self (Andersen & Chen, 2002; Baldwin, 1992) and transference (e.g., Hinkley & Andersen, 1996).
Accord-ing to this research, individuals have self-concept components representing the
self-with-relationship-partner for select important others (e.g., romantic partners, close friends). This
self-with-partner aspect of the self-concept is accessible when one interacts with or thinks about the significant
other. It is interesting to note that encounters with people who resemble a significant other can also activate the
relational schema associated with the significant other, a phenomenon known as transference (Andersen, Reznik,
& Manzella, 1996; Hinkley & Andersen, 1996). For example, when transference occurs, the working
self-concept resembles the self when with the relationship partner (Hinkley & Andersen, 1996). This
transference is very similar to the stereotype priming we have focused our discussion on thus far. A target similar
to a significant other can be seen as a prime for that significant other. This prime then activates a biased subset
of the perceiver’s self-schema—the self-when-with-other component—that is then active and capable of driving the
perceiver’s thoughts and behavior.
Another possible mechanism that may account for the change in self-concept, the
expansion model (Wheeler et al., 2005), involves the confusion of self and other (Aron, Aron, Tudor, & Nelson,
1991). By this account, social construct activation can sometimes create confusion regarding which accessible
material corresponds to the self and which material does not. Hence, although “lazy” may not be an element of a
person’s chronic self-concept, links between the self and the construct lazy could be created because of this
confusion of self and other. According to this model, the persistent content of a person’s actual self-concept
would not necessarily place a limitation on the effects of primes on behavior (see also Markman & McMullen,
2003). Virtually any accessible mental contents whose origins were unknown could be confused with the
self.
Some potential support for the expansion model comes from the literature on source
monitoring (Johnson, Hashtroudi, & Lindsay, 1993). Although this literature has generally focused on
memory-related phenomena (e.g., Dodhia & Metcalfe, 1999; Mather, John-son, & De Leonardis, 1999), recent
research has also examined the effects of source monitoring on judgment (Mussweiler & Neu-mann, 2000). For
example, Mussweiler and Neumann (2000) found that judgmental assimilation occurs when the source of accessibility
can be (mis)attributed to the self, and contrast occurs when accessibility is attributed to an outside source.
Although the effects of accessibility on judgments about others may not always be the same as the effects on one’s
behavior (Smeesters, Warlop, Van Avermaet, Corneille, & Yzerbyt, 2003), some research has found similar results
in these domains. For example, a distinct (Stapel & Koomen, 2001) exemplar prime is likely to produce contrast
in judgment about others (Stapel, Koomen, & van der Pligt, 1997), in perceptions of the self (Dijksterhuis et
al., 1998), and in behavior (Dijksterhuis et al., 1998). Thus, misattribution of the source of construct
accessibility to the self is one way in which primed content may cross the barrier into the working self-concept,
even in the absence of preexisting prime-consistent self-content.
High self-monitors, because of their constant monitoring of the social
environment, may be better source monitors, though this seems like an unlikely account for our data as Study 2 used
a subliminal prime that should have been difficult to detect, even for someone
skilled at source monitoring. Low self-monitors, in con-trast, are used to searching internally for self-relevant
aspects (e.g., attitudes, mood, motives) to guide their behavior. Thus, among low self-monitors, any internally
activated construct might be interpreted quickly as stemming from the self and treated accordingly.
An additional possibility is that these mechanisms may work in concert to produce
the self-change and behavioral effects ob-served. It may be the case, for example, that the extent or type of
chronic self- and social construct overlap (biased activation model) determines the likelihood of expansion to
include nonself prime content (expansion model) (for a more extensive discussion, see Wheeler et al.,
2005).
Although the models are conceptually distinct, distinguishing between these two
models empirically may prove to be difficult. In an idealized experimental world, it would be possible to find a
group of people who do not have any representation of stereotype-relevant traits in their self-concept. If a prime
still produced construct-consistent behavior, then the expansion model would be supported over the biased
activation model. Manipulations that vary the likelihood that the source of accessibility would be misattributed to
the self could also be used to explore the plausi-bility of the expansion model, and measuring the degree of
self-prime overlap prior to priming may offer insight into the biased activation model.
Other Proposed Mechanisms
Ideomotor Action
These studies call into question the degree to which the direct and simple
ideomotor activation model can account for all prime-to-behavior effects. The literature on this mechanism
(Dijksterhuis & Bargh, 2001; James, 1890/1950) provides no reason to expect that primes would affect
self-perceptions, unless such perceived change in the self were to result from self-perception-type infer-ences
following behavior. Because Studies 1 and 2 had no behav-iors prior to the self measure, this type of
self-perception mecha-nism would not be at work. Additionally, the ideomotor effect does not predict that changes
in behavior would be moderated by self-monitoring, unless the strength of activation differed across high and low
self-monitors. As discussed earlier, prior research suggests that, if anything, high self-monitors likely have
stronger activation, and so this also seems unlikely to account for our results.
Although ideomotor mechanisms seem unlikely to be the driver of our effects, it is
important to keep in mind that behavior is multiply determined, and the effects of stereotype activation on
behavior may be due to a number of mechanisms working in concert (Wheeler & Petty, 2001). Determining the
conditions under which each mechanism exerts its greatest influence on behavior may be a useful direction for
future research.
Social Function of the Perception–Behavior Link
What do these data mean for the social function that the perception– behavior link
has been hypothesized to play (Char-trand et al., 2004; Kawakami et al., 2003; Lakin et al., 2003)? Outside of
social psychology, the perception– behavior link has been hypothesized to be
the origin of a great deal of social behavior and, in particular, the acquisition of language (e.g., Kohler et al.,
2002; Rizzolatti & Arbib, 1998). Within social psychology, mimicry research has offered the most support for
this perspective (Chartrand & Bargh, 1999; Chartrand et al., 2004; Cheng & Chartrand, 2003; Dijksterhuis
& Bargh, 2001; Lakin & Chartrand, 2003; Lakin et al., 2003). Mimicry represents the purest form of the
perception– behavior link in that the perception is direct and is concurrent with the behavior.
In our paradigm, as in almost all studies we are aware of to date that have used
social construct primes, the perception was indirect and was separated from the behavior by time and by task (for
an exception, see Kay et al., 2004). This separation has two effects: First, it ensures that no actual behaviors
are observed. Second, it essentially removes any direct social implications, because the context in which the
construct is activated differs from that in which the behavior is measured. Hence, the primes do not com-municate
information about socially appropriate behavior and are, in fact, dissociated from the social context in which the
dependent variable is assessed. Perhaps this is partly why, in our studies, high self-monitors did not engage in
prime-consistent behavior. Indeed, consistent with the social-adjustive function of mimicry, high self-monitors are
more likely to engage in nonconscious mimicry (Cheng & Chartrand, 2003).
Although our results are inconsistent with the social functional account
predictions of behavioral priming effects, we do not dis-pute the possibility that mimicry may serve a
social-adjustive function. In fact, this idea has been well established (Chartrand et al., 2004; Cheng &
Chartrand, 2003; Lakin & Chartrand, 2003; Lakin et al., 2003). Instead, our data and the Active-Self
perspec-tive may offer limitations on the extent to which the social-adjustive function can be extrapolated to
automatic behaviors that result from primes, rather than from direct observation of another person’s actions.
Further research in which both social context and prime are manipulated could shed light on whether, and under what
conditions, social functional processes may operate in prim-ing contexts.
It may be useful to note that different functional accounts can be offered other
than the social-adjustive account discussed in this article. For example, a self-expressive functional account
(see, e.g., Abelson & Prentice, 1989) is entirely consistent with our Active-Self view, because the expressive
function account would also hold that prime-induced behavioral changes result from changes in the active
self-concept that are expressed via behavior. Hence, this account would also predict that low self-monitors, who
are likely to use the expressive function of attitudes, traits, and other internal states, should show the greatest
change in responses following a prime, so long as the primes affected the self-concept. Because we have little
evidence suggesting whether this process actually serves an expressive function or is a by-product of some other
process, we refrain from making strong claims about whether the process is truly functional for low
self-monitors.
Conclusion
Our three studies provide initial evidence that social construct priming effects
can involve a temporary change in self-perceptions for those individuals who are attuned to their inner states
(i.e., low self-monitors) and use them to guide their behavior. The consistent moderation by self-monitoring of the priming effects we obtained is not easily explained by
previous accounts of prime-to-behavior effects. Further research is needed to explore the conditions under which
the various mechanisms for behavioral priming effects are dominant. Future research should also explore the
underlying mechanism of the changed self-perception effect. Such research will be informative both to the
automaticity literature and the self and self-monitoring literatures.
References
Abelson, R. P., & Prentice, D. A. (1989). Beliefs as possessions: A functional
perspective. In A. R. Pratkanis, S. J. Breckler, & A. G. Greenwald (Eds.), Attitude structure and function (pp.
361–381). Hillsdale, NJ: Erlbaum.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting interactions. Thousand Oaks, CA: Sage.
Andersen, S. M., & Chen, S. (2002). The relational self: An interpersonal
social– cognitive theory. Psychological Review, 109, 619 – 645.
Andersen, S. M., Reznik, I., & Manzella, L. M. (1996). Eliciting facial
affect, motivation, and expectancies in transference: Significant-other representations in social relations.
Journal of Personality and Social Psychology, 71, 1108 –1129.
Arkin, R. M., Gabrenya, W. K., Jr., Appelman, A. S., & Cochran, S. T. (1979).
Self-presentation, self-monitoring, and the self-serving bias in causal attribution. Personality and Social
Psychology Bulletin, 5, 73–76.
Aron, A., Aron, E. N., Tudor, M., & Nelson, G. (1991). Close relationships as
including other in the self. Journal of Personality and Social Psy-chology, 60, 241–253.
Asch, S. E. (1956). Studies of independence and conformity: I. A minority of one
against a unanimous majority. Psychological Monographs, 70, 1–70.
Baldwin, M. W. (1992). Relational schemas and the processing of social
information. Psychological Bulletin, 112, 461– 484.
Bargh, J. A., Chen, M., & Burrows, L. (1996). Automaticity of social behavior:
Direct effects of trait construct and stereotype activation on action. Journal of Personality and Social
Psychology, 71, 230 –240.
Byrne, D. (1961). Interpersonal attraction and attitude similarity. Journal of
Abnormal and Social Psychology, 62, 713–715.
Carpenter, W. B. (1874). Principles of mental psychology with their applications
to the training and discipline of the mind, and the study of its morbid conditions. New York: Appleton.
Chartrand, T. L., & Bargh, J. A. (1999). The chameleon effect: The perception–
behavior link and social interaction. Journal of Personality and Social Psychology, 76, 893–910.
Chartrand, T. L., Maddux, W. W., & Lakin, J. L. (2004). Beyond the perception–
behavior link: The ubiquitous utility and motivational mod-erators of nonconscious mimicry. In R. R. Hassin, J. S.
Uleman, & J. A. Bargh (Eds.), The new unconscious (pp. 334 –361). New York: Oxford University Press.
Cheng, C. M., & Chartrand, T. L. (2003). Self-monitoring without aware-ness:
Using mimicry as a nonconscious affiliation strategy. Journal of Personality and Social Psychology, 85, 1170
–1179.
Christopher, A. N., & Schlenker, B. R. (2000). The impact of perceived
material wealth and perceiver personality on first impressions. Journal of Economic Psychology, 21,
1–19.
Cialdini, R. B., & Trost, M. R. (1998). Social influence: Social norms,
conformity, and compliance. In D. T. Gilbert, S. T. Fiske, & G. Lindzey (Eds.), Handbook of social psychology
(4th ed., Vol. 2, pp. 151–192). New York: McGraw-Hill.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied
regres-sion/correlation analysis for the behavioral sciences. Mahwah, NJ: Erlbaum.
De Bono, K. G., & Harnish, R. J. (1988). Source expertise, source
attrac-
670 DEMARREE, WHEELER, AND PETTY
tiveness, and the processing of persuasive information: A functional approach.
Journal of Personality and Social Psychology, 55, 541–546.
Decety, J., Jeannerod, M., Germain, M., & Pastene, J. (1991). Vegetative
response during imagined movement is proportional to mental effort.
Behavioural Brain Research, 42, 1–5.
Deutsch, M., & Gerard, H. B. (1955). A study of normative and informa-tional
social influences upon individual judgment. Journal of Abnormal and Social Psychology, 51, 629 – 636.
Devine, P. G. (1989). Stereotypes and prejudice: Their automatic and controlled
components. Journal of Personality and Social Psychology, 56, 5–18.
Dijksterhuis, A., Aarts, H., Bargh, J. A., & van Knippenberg, A. (2000). On
the relation between associative strength and automatic behavior. Jour-nal of Experimental Social Psychology, 36,
531–544.
Dijksterhuis, A., & Bargh, J. A. (2001). The perception– behavior express-way:
Automatic effects of social perception on social behavior. In M. P. Zanna (Ed.), Advances in experimental social
psychology (Vol. 33, pp. 1– 40). San Diego, CA: Academic Press.
Dijksterhuis, A., Bargh, J. A., & Miedema, J. (2000). Of men and mack-erels:
Attention, subjective experience, and automatic social behavior. In H. Bless & J. P. Forgas (Eds.), The message
within: The role of subjective experience in social cognition and behavior (pp. 37–51). Philadelphia: Psychology
Press.
Dijksterhuis, A., Spears, R., Postmes, T., Stapel, D., Koomen, W., van
Knippenberg, A., & Scheepers, D. (1998). Seeing one thing and doing another: Contrast effects in automatic
behavior. Journal of Personality and Social Psychology, 75, 862– 871.
Dijksterhuis, A., & van Knippenberg, A. (1998). The relation between
perception and behavior, or how to win a game of Trivial Pursuit.
Journal of Personality and Social Psychology, 74, 865– 877.
Dodhia, R. M., & Metcalfe, J. (1999). False memories and source moni-toring.
Cognitive Neuropsychology, 16, 489 –508.
Fenigstein, A., Scheier, M. F., & Buss, A. H. (1975). Public and private
self-consciousness: Assessment and theory. Journal of Counseling and Clinical Psychology, 43, 522–527.
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior:
An introduction to theory and research. Reading, MA: Addison Wesley.
Fiske, S. T., & von Hendy, H. M. (1992). Personality feedback and situational
norms can control stereotyping processes. Journal of Person-ality and Social Psychology, 62, 577–596.
Gangestad, S. W., & Snyder, M. (1985). “To carve nature at its joints”: On the
existence of discrete classes in personality. Psychological Review, 92, 317–349.
Gangestad, S. W., & Snyder, M. (1991). Taxonomic analysis redux: Some
statistical considerations for testing a latent class model. Journal of Personality and Social Psychology, 61,
141–146.
Gangestad, S. W., & Snyder, M. (2000). Self-monitoring: Appraisal and
reappraisal. Psychological Bulletin, 126, 530 –555.
Haddock, G., Macrae, C. N., & Fleck, S. (2002). Syrian science and smart
supermodels: On the when and how of perception– behavior effects.
Social Cognition, 20, 461– 479.
Hass, R. G., Katz, I., Rizzo, N., Bailey, J., & Moore, L. (1992). When racial
ambivalence evokes negative affect, using a disguised measure of mood. Personality and Social Psychology Bulletin,
18, 786 –797.
Higgins, E. T. (1996). Knowledge activation: Accessibility, applicability, and
salience. In E. T. Higgins & A. W. Kruglanski (Eds.), Social psychology: Handbook of basic principles (pp.
133–168). New York: Guilford Press.
Higgins, E. T., Rholes, W. S., & Jones, C. R. (1977). Category accessibility
and impression formation. Journal of Experimental Social Psychology, 13, 141–154.
Hinkley, K., & Andersen, S. M. (1996). The working self-concept in
transference: Significant-other activation and self change. Journal of Personality and Social Psychology, 71, 1279
–1295.
Hull, J. G., & Levy, A. S. (1979). The organizational functions of the self: An
alternative to the Duval and Wicklund model of self-awareness.
Journal of Personality and Social Psychology, 37, 756 –768.
Hull, J. G., Slone, L. B., Meteyer, K. B., & Matthews, A. R. (2002). The
nonconsciousness of self-consciousness. Journal of Personality and Social Psychology, 83, 406 – 424.
Hull, J. G., Van Treuren, R. R., Ashford, S. J., Propsom, P., & Andrus, B. W.
(1988). Self-consciousness and the processing of self-relevant information. Journal of Personality and Social
Psychology, 54, 452– 465.
Iacoboni, M., Woods, R. P., Brass, M., Bekkering, H., Mazziotta, J. C., &
Rizzolatti, G. (1999, December 24). Cortical mechanisms of human imitation. Science, 286, 2526 –2528.
Ickes, W., Layden, M. A., & Barnes, R. D. (1978). Objective self-awareness and
individuation: An empirical link. Journal of Personality, 46, 146 –161.
James, W. (1950). The principles of psychology. New York: Dover Pub-lications.
(Original work published 1890)
Johnson, M. K., Hashtroudi, S., & Lindsay, D. S. (1993). Source monitor-ing.
Psychological Bulletin, 114, 3–28.
Kawakami, K., Dovidio, J. F., & Dijksterhuis, A. (2003). Effect of social
category priming on personal attitudes. Psychological Science, 14, 315– 319.
Kay, A. C., Wheeler, S. C., Bargh, J. A., & Ross, L. (2004). Material priming:
The influence of mundane physical objects on situational construal and competitive behavioral choice.
Organizational Behavior and Human Decision Processes, 95, 83–96.
Kohler, E., Keysers, C., Umilta`, M. A., Fogassi, L., Gallese, V., &
Rizzolatti, G. (2002, August 2). Hearing sounds, understanding actions: Action representation in mirror neurons.
Science, 297, 846 – 848.
Koole, S. L., Smeets, K., van Knippenberg, A., & Dijksterhuis, A. (1999). The
cessation of rumination through self-affirmation. Journal of Per-sonality and Social Psychology, 77,
111–125.
Lakin, J. L., & Chartrand, T. L. (2003). Using nonconscious behavioral mimicry
to create affiliation and rapport. Psychological Science, 14, 334 –339.
Lakin, J. L., Jefferis, V. E., Cheng, C. M., & Chartrand, T. L. (2003). The
chameleon effect as social glue: Evidence for the evolutionary signifi-cance of nonconscious mimicry. Journal of
Nonverbal Behavior, 27,
145–162.
Lamphere, R. A., & Leary, M. R. (1990). Private and public self-processes: A
return to James’s constituents of the self. Personality and Social Psychology Bulletin, 16, 717–725.
Larkin, J. E., & Pines, H. A. (1994). Affective consequences of
self-monitoring style in a job interview setting. Basic and Applied Social Psychology, 15, 297–310.
Lavine, H., & Snyder, M. (2000). Cognitive processes and the functional
matching effect in persuasion: Studies of personality and political be-havior. In G. R. Maio & J. M. Olson
(Eds.), Why we evaluate: Functions of attitudes (pp. 97–131). Mahwah, NJ: Erlbaum.
Lippa, R. (1978). Expressive control, expressive consistency, and the
correspondence between expressive behavior and personality. Journal of Personality, 46, 438 – 461.
Lowery, B. S., Hardin, C. D., & Sinclair, S. (2001). Social influence effects
on automatic racial prejudice. Journal of Personality and Social Psy-chology, 81, 842– 855.
Macrae, C. N., & Johnston, L. (1998). Help, I need somebody: Automatic action
and inaction. Social Cognition, 16, 400 – 417.
Macrae, C. N., Milne, A. B., & Bodenhausen, G. V. (1994). Stereotypes as
energy-saving devices: A peek inside the cognitive toolbox. Journal of Personality and Social Psychology, 66, 37–
47.
Macrae, C. N., Stangor, C., & Milne, A. B. (1994). Activating social
stereotypes: A functional analysis. Journal of Experimental Social Psy-chology, 30, 370 –389.
STEREOTYPE PRIMES, SELF-MONITORING, AND THE SELF 671
Maddux, W. W., Barden, J., Brewer, M. B., & Petty, R. E. (2005). Saying no to
negativity: The effects of context and motivation to control prejudice on automatic evaluative responses. Journal
of Experimental Social Psychology, 41, 19 –35.
Markman, K. D., & McMullen, M. N. (2003). A reflection and evaluation model of
comparative thinking. Personality and Social Psychology Review, 7, 244 –267.
Mather, M., Johnson, M. K., & De Leonardis, D. M. (1999). Stereotype reliance
in source monitoring: Age differences and neuropsychological test correlates. Cognitive Neuropsychology, 16, 437–
458.
McClelland, G. H. (2000). Nasty data: Unruly, ill-mannered observations can ruin
your analysis. In H. T. Reis & C. M. Judd (Eds.), Handbook of research methods in social and personality
psychology (pp. 393– 411). Cambridge, England: Cambridge University Press.
Miller, M. L., & Thayer, J. F. (1989). On the existence of discrete classes in
personality: Is self-monitoring the correct joint to carve? Journal of Personality and Social Psychology, 57,
143–155.
Mussweiler, T. (2003). Comparison processes in social judgment: Mech-anisms and
consequences. Psychological Review, 110, 472– 489.
Mussweiler, T., & Neumann, R. (2000). Sources of mental contamination:
Comparing the effects of self-generated versus externally provided primes. Journal of Experimental Social
Psychology, 36, 194 –206.
Newcomb, T. M. (1961). The acquaintance process. New York: Holt, Rinehart &
Winston.
Penner, L. A., & Wymer, W. E. (1983). The moderator variable approach to
behavioral predictability: Some of the variables some of the time.
Journal of Research in Personality, 17, 339 –353.
Petty, R. E., & Cacioppo, J. T. (1979). Issue-involvement can increase or
decrease persuasion by enhancing message-relevant cognitive responses.
Journal of Personality and Social Psychology, 37, 1915–1926.
Petty, R. E., & Cacioppo, J. T. (1986). Communication and persuasion: Central
and peripheral routes to attitude change. New York: Springer-Verlag.
Petty, R. E., & Wegener, D. T. (1993). Flexible correction processes in social
judgment: Correcting for context-induced contrast. Journal of Experimental Social Psychology, 29,
137–165.
Petty, R. E., & Wegener, D. T. (1998). Matching versus mismatching attitude
functions: Implications for scrutiny of persuasive messages.
Personality and Social Psychology Bulletin, 24, 227–240.
Petty, R. E., Wheeler, S. C., & Bizer, G. Y. (2000). Attitude functions and
persuasion: An elaboration likelihood approach to matched versus mis-matched messages. In G. R. Maio & J. M.
Olson (Eds.), Why we evaluate: Functions of attitudes (pp. 133–162). Mahwah, NJ: Erlbaum.
Prinz, W. (1990). A common coding approach to perception and action. In O. Neumann
& W. Prinz (Eds.), Relationships between perception and action: Current approaches (pp. 167–201). Berlin,
Germany: Springer-Verlag.
Rizzolatti, G., & Arbib, M. A. (1998). Language within our grasp. Trends in
Neuroscience, 21, 188 –194.
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. Schubert, T. W., &
Ha¨fner, M. (2003). Contrast from social stereotypes in
automatic behavior. Journal of Experimental Social Psychology, 39,
577–584.
Schwalbe, M. L. (1991). Role taking, self-monitoring, and the alignment of conduct
with others. Personality and Social Psychology Bulletin, 17,
51–57.
Sherif, M. (1935). A study of some social factors in perception. Archives of
Psychology, 27, 1– 60.
Shih, M., Ambady, N., Richeson, J. A., Fujita, K., & Gray, H. M. (2002).
Stereotype performance boosts: The impact of self-relevance and the manner of stereotype activation. Journal of
Personality and Social Psychology, 83, 638 – 647.
Shih, M., Pittinsky, T. L., & Ambady, N. (1999). Stereotype susceptibility:
Identity salience and shifts in quantitative performance. Psychological Science, 10, 80 – 83.
Smeesters, D., Warlop, L., Van Avermaet, E., Corneille, O., & Yzerbyt, V.
(2003). Do not prime hawks with doves: The interplay of construct activation and consistency of social value
orientation on cooperative behavior. Journal of Personality and Social Psychology, 84, 972–987.
Snyder, M. (1974). Self-monitoring of expressive behavior. Journal of Personality
and Social Psychology, 30, 526 –537.
Snyder, M. (1979). Self-monitoring processes. In L. Berkowitz (Ed.),
Advances in experimental social psychology (Vol. 12, pp. 85–128). New York:
Academic Press.
Snyder, M., & Campbell, B. H. (1982). Self-monitoring: The self in action. In
J. Suls (Ed.), Psychological perspectives on the self (Vol. 1, pp. 185–207). Hillsdale, NJ: Erlbaum.
Snyder, M., & Cantor, N. (1980). Thinking about ourselves and others:
Self-monitoring and social knowledge. Journal of Personality and So-cial Psychology, 39, 222–234.
Snyder, M., & Gangestad, S. W. (1986). On the nature of self-monitoring:
Matters of assessment, matters of validity. Journal of Personality and Social Psychology, 51, 125–139.
Snyder, M., & Monson, T. C. (1975). Persons, situations, and the control of
social behavior. Journal of Personality and Social Psychology, 32,
637– 644.
Snyder, M., & Swann, W. B., Jr. (1976). When actions reflect attitudes: The
politics of impression management. Journal of Personality and Social Psychology, 34, 1034 –1042.
Snyder, M., & Tanke, E. D. (1976). Behavior and attitude: Some people are more
consistent than others. Journal of Personality, 44, 501–517.
Stapel, D. A., & Koomen, W. (2001). Let’s not forget the past when we go to
the future: On our knowledge and knowledge of accessibility effects. In G. Moskowitz (Ed.), Cognitive social
psychology: The Princeton Symposium on the Legacy and Future of Social Cognition (pp. 229 – 246). Mahwah, NJ:
Erlbaum.
Stapel, D. A., Koomen, W., & van der Pligt, J. (1997). Categories of category
accessibility: The impact of trait concept versus exemplar priming on person judgments. Journal of Experimental
Social Psychol-ogy, 33, 47–76.
Symons, C. S., & Johnson, B. T. (1997). The self-reference effect in memory: A
meta-analysis. Psychological Bulletin, 121, 371–394.
Wegener, D. T., & Petty, R. E. (1997). The flexible correction model: The role
of naive theories of bias in bias correction. In M. P. Zanna (Ed.),
Advances in experimental social psychology (Vol. 29, pp. 141–208). San Diego, CA:
Academic Press.
Wheeler, S. C., DeMarree, K. G., & Petty, R. E. (2005). The roles of the self
in priming-to-behavior effects. In A. Tesser, J. V. Wood, & D. A. Stapel (Eds.), On building, defending and
regulating the self: A psycho-logical perspective (pp. 245–271). New York: Psychology Press.
Wheeler, S. C., Jarvis, W. B. G., & Petty, R. E. (2001). Think unto others:
The self-destructive impact of negative racial stereotypes. Journal of Experimental Social Psychology, 37,
173–180.
Wheeler, S. C., & Petty, R. E. (2001). The effects of stereotype activation on
behavior: A review of possible mechanisms. Psychological Bulletin, 127, 797– 826.
Wilson, T. D., & Brekke, N. (1994). Mental contamination and mental
correction: Unwanted influences on judgments and evaluations. Psycho-logical Bulletin, 116, 117–142.
Zanna, M. P., Olson, J. M., & Fazio, R. H. (1980). Attitude– behavior
consistency: An individual difference perspective. Journal of Personal-ity and Social Psychology, 38, 432–
440.
Received February 19, 2004
Revision received May 13, 2005
Accepted May 19, 2005
Copyright and Other Protection. The Site and much of the
text, images, graphics, information and other content of the Site (collectively, the “Content”) are protected by
copyright, trademark and other laws. We or others own the copyright and other rights in the Site and the
Content. Much of the Research Content consists of articles and other
scholarly writings posted by authors and
researchers according to their understanding of their rights in that
material.
Some of the Posted Material consists of scholarly articles
deposited under a Harvard Open Access Policy (the “Open Access Policy Articles” or “OAP Articles”). Subject to the
other provisions of these Terms and Conditions, you may use,
reproduce, distribute, and display the OAP Articles for:
a. personal study;
b. teaching (including distribution of copies to students and
use in coursepacks and courseware programs);
c. research and scholarship (including computational research
uses such as data-mining and text-mining); and
d. provision of value-added services (including full-text
searching, cross-referencing, and citation extraction);
provided in each case that:
i. you will not sell or charge for any OAP Article (whether
or not any profit is involved) and will not sell advertising on the same page as any OAP Article;
ii. if you make an OAP Article available to others, you will
retain with the article its title, the name of the author(s), a reference to these Terms of Use, and any copyright
notice included on the original;
iii. if you make an OAP Article available on-line, you will
use reasonable efforts both to cite to the publisher’s definitive version of the article if it has been published,
and to provide a link to the publisher’s version if it is available on-line;
iv. you may not use a facsimile of the published version of
an article under these open access terms, unless the publisher so permits and that version has been made available
as an OAP Article on the Site;
v. you will not make any translation, adaptation or other
derivative work of an OAP Article, except that, as reasonably necessary in order to carry out a permitted use, you
may include the article in a collection or database, may change the technical format of the article, and may use
excerpts of the article for teaching or other permitted purposes, so long as, in each case, (A) your use of the
article or any portion thereof continues to comply with all provisions of these Terms of Use that are applicable to
OAP Articles, (B) you do not misrepresent the substance of the article, and (C) your derivative of the article
includes a citation, hyperlink, or similar reference to the original article and appropriately identifies the
nature of the derivative work (e.g., “abridged from_____”); and
vi. you may not sublicense or otherwise transfer your rights
in any OAP Article, and will only make OAP Articles available to others for use by them under these Terms of Use as
in effect from time to time.
If you have an article, report or research paper that you
understand or believe to be in violation of copyright please
contact us and we will remove it immediately while we investigate the circumstance.
This research is provided to interested parties who wish to broaden their understanding of subliminal priming
and all copyrights remain with the original authors. This page is provided in good faith with the
intention of promoting the studies and authors involved. If any author or copyright details are wrong or
need updating please contact us.
|