Content uploaded by Sarah Tuck
Author content
All content in this area was uploaded by Sarah Tuck on Jul 20, 2014
Content may be subject to copyright.
Lockwoodet al.USING ROLE MODELS TO HARNESS MOTIVATION
TO DO OR NOT TO DO:
USING POSITIVE AND NEGATIVE
ROLE MODELS TO HARNESS MOTIVATION
Penelope Lockwood
University of Toronto
Pamela Sadler
Wilfrid Laurier University
Keren Fyman and Sarah Tuck
University of Toronto
We examined role model preferences among individuals contemplating ad-
ditive and subtractive behavior changes. Individuals considering engaging in
a potentially beneficial activity, an additive behavior, were more likely to
view positive models as effective motivators. Individuals considering abstain-
ing from a potentially deleterious activity, a subtractive behavior, were more
likely to view negative models as effective motivators. In addition, we exam-
ined the impact of participants’ regulatory focus on role model preferences. In
Studies 1 and 2, we measured promotion and prevention focus directly. In
Study 3, we used self–construals as an indirect measure of regulatory focus. In
all studies, promotion–focused individuals were especially likely to prefer
positive role models when contemplating additive behaviors; prevention–fo
-
cused individuals were especially likely to prefer negative role models when
contemplating subtractive behaviors.
The positive and negative examples set by high and low achievers can
have a significant impact on the people around them. Positive role
422
Social Cognition, Vol. 22, No. 4, 2004, pp. 422-450
This research was supported by a grant from the Social Sciences and Humanities Re
-
search Council of Canada (No. 410–99–1376) to Penelope Lockwood.
We are grateful to Ken Dion for his comments on an earlier version of this manuscript,
and to Susan Furs for her assistance with data collection.
Address correspondence to Penelope Lockwood, Department of Psychology, Univer
-
sity of Toronto, 100 St. George Street, Toronto, ON, Canada, M5S 3G3; E–mail may be sent
to lockwood@psych.utoronto.ca.
models, individuals who have achieved a high degree of success, can
encourage others to pursue similar accomplishments (Lockwood &
Kunda, 1997). For example, an outstanding athlete can inspire others to
exercise and thus improve their own physical condition. Negative role
models, individuals who have experienced some kind of failure or mis
-
fortune, can motivate others to avoid similar adversity (Lockwood,
2002). For example, an individual who has injured another person as a
result of drinking and driving can motivate others to avoid driving
while drunk.
The effectiveness of such positive and negative role models may be
influenced by the congruence between the desired behavior change
and the strategy emphasized by the model. Role models can boost
motivation through either additive or subtractive behavior changes.
Additive behaviors involve adopting a potentially beneficial activ
-
ity; for example, an individual can improve his or her health by exer-
cising more often. Subtractive behaviors involve abstaining from or
reducing a potentially deleterious activity; for example, a student
can increase his or her academic success by cutting back on the
amount of time he or she spends watching TV. Individuals’ inclina-
tions to make these behavior changes can vary depending on
whether the incentives for the changes are framed in terms of pro-
moting a positive outcome or preventing a negative outcome. Pre-
vention goals involve a concern with errors of commission, the
avoidance of actions that could lead to an undesirable endstate (Hig-
gins, 1997). Individuals seeking to avoid losses may therefore be es
-
pecially likely to engage in subtractive behaviors; by reducing a
behavior with potentially negative consequences, one can prevent
an unpleasant outcome. Promotion goals, in contrast, involve the
aim to avoid errors of omission, ensuring that one performs the ac
-
tions that will lead to a desirable endstate (Higgins, 1997). Individu
-
als seeking to achieve gains may therefore be especially likely to
perform additive behaviors; by engaging in a behavior with
potentially positive consequences, one can promote a successful
outcome.
Consistent with this possibility, research from the literature on
counterfactual thinking suggests that framing scenarios in preven
-
tion or promotion terms has implications for the kinds of behaviors
that individuals select in order to mentally undo a negative outcome.
In one study, for example, prevention counterfactuals, thoughts
USING ROLE MODELS TO HARNESS MOTIVATION 423
about how a negative outcome could have been avoided, were asso
-
ciated with subtractive behaviors, deleting an action that produced
the undesirable result. Promotion counterfactuals, thoughts about
how a more positive outcome could have been achieved, were asso
-
ciated with additive behaviors, the adoption of an action that could
have produced a more desirable result (Roese, Hur, & Pennington,
1999). Thus, the framing of the situation as a loss that could have
been avoided was associated with subtractive behaviors, whereas
the framing of the situation as a gain that could have been achieved
was associated with additive behaviors.
Similarly, additive and subtractive behavior changes may be associ
-
ated with different role model preferences. Sometimes, individuals
will seek encouragement from the example set by a highly successful
other, who illustrates the excellence that they may be able to achieve in
the future (cf. Aspinwall, 1997; Buunk, Collins, Taylor, Van Yperen, &
Dakof, 1990; Collins, 1996; Taylor & Lobel, 1989; Taylor, Wayment &
Carillo, 1996; Wood, 1989); at other times, they may try to motivate
themselves by focusing on the example set by a highly unsuccessful
other, who illustrates the problems that they must try to avoid in the
future (Buunk et al., 1990; Wood & VanderZee, 1997). We propose that
individuals will prefer role models who are congruent with their de-
sired behaviorchange. When considering additive behaviors, individ-
uals will show a preference for positive models, who represent a
potential future gain; if a student wishes to sustain his or her motiva
-
tion to go to the library more often, he or she will choose an exemplar
of success, such as another student with an outstanding academic re
-
cord. The successful other highlights a promotion strategy, the pursuit
of achievement, and will therefore be motivating for individuals seek
-
ing to engage in potentially beneficial activities. In contrast, when con
-
sidering subtractive behaviors, individuals will show a preference for
negative models, who represent a potential future loss; if a student
wishes to maintain his or her resolution to cut back on social activities
that interfere with schoolwork, he or she will choose an exemplar of
failure, such as another student who has dropped out of school. The
unsuccessful other highlights a prevention strategy, the avoidance of
failure, and will therefore be motivating for individuals seeking to ab
-
stain from potentially deleterious activities. Individuals are thus likely
to select models that fit the desired behavior change: The better the fit
between the behavior and the model, the more effective the model will
424 LOCKWOOD ET AL.
be perceived to be; by matching subtractive behaviors with negative
models, and additive behaviors with positive models, individuals can
attempt to maximize the motivation garnered from the models (cf.
Higgins, 2000).
In addition, recent research suggests that individuals’ chronic reg
-
ulatory goals will influence the relative effectiveness of positive and
negative role models (Lockwood, Jordan, & Kunda, 2002). Individu
-
als with strong promotion orientations tend to be sensitive to the
presence or absence of positive outcomes (Higgins, 1997); they are
especially likely to observe and recall information about success
achieved by other individuals (Higgins & Tykocinski, 1992), and are
also most likely to persist on tasks that are framed in terms of achiev
-
ing gains rather than avoiding losses (Shah, Higgins, & Friedman,
1998). In general, promotion-focused individuals favor a strategy of
pursuing success over a strategy of avoiding failure (Higgins, 1997;
1998). In past research, we found that these individuals are especially
motivated by positive role models, who provide a vivid example of
such success; the successful model highlights a strategy congruent
with their goals (Lockwood et al., 2002).
In contrast, individuals with strong prevention orientationstend to be
sensitive tothe presence or absence of negative outcomes;they are espe-
cially likely to observe and recall information about failures experi-
enced by other individuals (Higgins & Tykocinski, 1992), and are also
more likely to persist on tasks that are framed in terms of avoiding
losses rather than achieving gains (Shah et al., 1998). Prevention-fo
-
cused individuals tend to prefer a strategy of avoiding failure over a
strategy of achieving success (Higgins, 1997, 1998). These individuals
are especially motivated by negative role models, who provide clear ex
-
amples of such failures; the unsuccessful models highlight strategies
that are congruent with their goals (Lockwood et al., 2002).
In the presentresearch, we propose that an individual’s regulatory fo
-
cus will be most influential in determining role model preferences
when there is congruence between regulatory focus and behavior type.
That is, promotion focus will be associated with a preference for posi
-
tive models, particularly when individuals consider additive behavior
changes: The promotion orientation and additive behavior type are
both associated with the achievement of successes or gains, and it is un
-
der these circumstances that a positive role model will be perceived to
be most useful. In contrast, prevention focus will be associated with a
USING ROLE MODELS TO HARNESS MOTIVATION 425
preference for negative models, particularly when individuals consider
subtractive behavior changes: The prevention orientation and
subtractive behavior are both associated with the avoidance of failure
or losses, and it is under these circumstances that a negative role model
will be perceived to be most effective.
In sum, role models are most likely to be perceived to be effective
motivators when both the type of behavior under consideration and
the individual’s regulatory concerns are congruent with the strategy
highlighted by the model. Individuals who are considering additive
behaviors should be especially likely to use positive models as a
means of harnessing their motivation, particularly to the extent that
they have a strong promotion focus; in contrast, individuals who are
considering subtractive behaviors should be especially likely to use
negative models as a means of harnessing their motivation, particu
-
larly to the extent that they have a strong prevention focus.
In this article, we focus on individuals’ preferences for role models,
rather than on the behavior changes that such role models may acti-
vate. Specifically, we are interested in whether individuals are more
likely to choose a positive or negative role model when they are try-
ing to engage their motivation for an additive or subtractive behav-
ior. It is important to note, however, that self–reported preferences
for role models may not always map onto actual behavior changes
following exposure to such role models. Individuals may show a
preference for positive models, for example, but might also be influ
-
enced by negative models. One’s beliefs that a specific role model
will or will not be motivating may not always be accurate; at times,
the impact of positive or negative role models on behavior may occur
outside individuals’ awareness (Nisbett & Wilson, 1977). Prefer
-
ences are nevertheless important, however, because they likely de
-
termine which kinds of role models individuals typically attend to
and seize upon in their day to day lives as a means of keeping them
-
selves engaged with their goals. To the extent that one expects to find
a negative model to be motivating when one is trying to quit smok
-
ing, for example, one may actively seek out the example of an indi
-
vidual who has experienced smoking–related health problems. By
understanding how individuals select role models, we gain insight
into how people seek to use successful and unsuccessful others
around them in an attempt to harness motivation over the longer
term.
426 LOCKWOOD ET AL.
In three studies, we examined whether the type of behavior (addi
-
tive or subtractive) and regulatory focus (promotion or prevention)
would determine participants’ preferences for positive relative to
negative role models. In Studies 1 and 2, we measured regulatory fo
-
cus directly, and examined preferences for academic (Study 1) and
health–related (Study 2) role models. In Study 3, we selected partici
-
pants from cultures known to differ in their emphasis on promotion
and prevention concerns (Lee, Aaker, & Gardner, 2000), and as
-
sessed their preferences for academic role models. In all studies, we
expected that participants considering additive behavior changes
would perceive positive models to be more motivating, and that par
-
ticipants considering subtractive behavior changes would perceive
negative models to be more motivating. We also attempted to repli
-
cate and extend previous research showing that participants’ own
regulatory orientations can influence role model preferences; specif-
ically, we examined whether participants’ regulatory concerns
would be most strongly associated with motivation by congruent
role models when the desired behavior change was also congruent
with their concerns. We expected that promotion focus would be
most strongly associated with a preference for positive role models
among individuals considering additive behavior changes, and
prevention focus would be most strongly associated with a
preference for negative role models among individuals considering
subtractive behavior changes.
STUDY 1: ADDITIVE AND SUBTRACTIVE BEHAVIORS ARE
ASSOCIATED WITH DIFFERENT ACADEMIC ROLE MODEL
PREFERENCES
In Study 1, we examined whether students would express a prefer
-
ence for positive or negative academic role models when considering
additive and subtractive behaviors. In order to achieve a desired aca
-
demic outcome, students may choose to focus on additive behaviors,
such as going to the library more frequently; alternatively, they may
choose to focus on subtractive behaviors, such as cutting back on so
-
cial activities that interfere with their schoolwork. Both kinds of be
-
haviors may have a positive impact on their grades. These different
kinds of behaviors, however, may prompt different choices of role
models. We expected that individuals would prefer role models that
were congruent with the desired behavior change: When consider
-
USING ROLE MODELS TO HARNESS MOTIVATION 427
ing adding potentially beneficial behaviors, participants would
show a preference for positive models; when considering subtract
-
ing potentially detrimental behaviors, participants would show a
preference for negative models.
In addition, we expected that regulatory focus would predict role
model preferences for behavior changes congruent with the focus:
Promotion focus would be especially strongly associated with a pref
-
erence for positive role models when individuals considered addi
-
tive behaviors; in contrast, prevention focus would be especially
strongly associated with a preference for negative role models when
individuals considered subtractive behaviors.
METHOD
PARTICIPANTS
Participants were 17 male and 51 female Introductory Psychology
students at the University of Toronto who received course credit for
their participation. Participants’ gender had no effects on any of the
variables and is therefore not discussed further.
PROCEDURE
Participants were told that the researchers were interested in finding
out how people motivate themselves: “Sometimes, we are motivated
to try to avoid becoming like a person who is experiencing problems,
and sometimes we are motivated to try to become like a person who
is experiencing success. Both positive and negative examples can
motivate us to change our behaviors. The following task involves an
-
swering questions about which kind of person is more likely to moti
-
vate YOU in different situations.” Participants were then asked to
indicate whether they would be more motivated by a positive or a
negative role model to perform a series of 12 behaviors. Six of the be
-
haviors were additive (e.g., develop better study habits, spend more
time studying, try to get more As) and six were subtractive (e.g.,
spend less time partying with friends, procrastinate less, stop falling
behind in readings). For each item, participants rated on a
seven–point scale whether they would be more motivated by a nega
-
tive or positive model to perform this behavior. The scale labels
ranged from –3 (negative role model is more motivating) to 0 (nega
-
428 LOCKWOOD ET AL.
tive and positive models are equally motivating) to + 3 (positive role
model is more motivating). A different example of a negative and a
corresponding positive role model was included for each item (e.g.,
“someone who is on academic probation for poor performance” vs.
“someone who has won an award for academic achievement";
“someone from your program who has been unemployed since
graduating last year” vs. “someone from your program who landed
a great job after graduating last year”). The additive behavior items
were combined into a single index (alpha = .70); the subtractive be
-
havior items were also combined into a single index (alpha = .71). We
had expected that the additive and subtractive items would load
onto two different factors; unexpectedly, however, a factor analysis
did not reveal a clear distinction between the additive and
subtractive item loadings. Nonetheless, the predicted mean
differences did emerge, as will be discussed below.
Participants also completed an 18–item measure of regulatory fo-
cus (Lockwood, Jordan, & Kunda, 2002). This measure includes a
promotion and a prevention subscale; both were reliable (promotion
alpha = .83, prevention alpha = .84). The subscales were not signifi-
cantly correlated, r = –.18, p = .13. In general, participants reported
higher promotion (M = 7.15) than prevention (M = 5.63) orientations,
t(67) = 6.23, p < .001.
RESULTS AND DISCUSSION
To test whether behavior framing would influence role model pref
-
erence, we conducted a repeated measures Analysis of Variance
(ANOVA), with behavior type (additive or subtractive) as the
within–participants variable. As expected, participants were more
likely to select a positive role model when the behavior change was
described in additive terms (M = 1.71) rather than subtractive terms
(M = .48), F (1, 67) = 68.94, p < .001.
Next, we assessed whether regulatory focus would be more pre
-
dictive of role model preferences when the focus was congruent with
the desired behavior change. That is, role model preference for addi
-
tive behaviors should be predicted chiefly by promotion focus,
rather than prevention focus. This relationship should be positive
because higher promotion focus should lead to a preference for a
positive role model for additive behaviors. In contrast, role model
preference for subtractive behaviors should be predicted chiefly by
USING ROLE MODELS TO HARNESS MOTIVATION 429
prevention focus, rather than by promotion focus. This relationship
should be negative because higher prevention focus should lead to a
preference for a negative role model for subtractive behaviors. To
evaluate these hypotheses, we estimated the structural equation
model shown in Figure 1. This just–identified model consists of two
simultaneous regressions, one for role model preference for additive
behaviors and the other for role model preference for subtractive be-
haviors. According to the congruency hypothesis for additive behav-
iors, path A should be significantly stronger than path C. Likewise,
for subtractive behaviors, path D should be significantly stronger
than path B. Structural equation modeling allows us to test whether
such pairs of paths are significantly different by setting those paths
equal to each other and evaluating model fit. As seen in the diagram,
promotion and prevention goals were allowed to covary, as were the
residuals for role model preference for additive and subtractive be
-
haviors, Z
1
and Z
2
.
The parameter estimates for this model are shown in Figure 2. As
predicted, role model preference for additive behaviors showed a
significantly positive association with promotion focus, but a
nonsignificant negative relationship with prevention focus. The for
-
mer path was significantly stronger than the latter path, χ
2
(1, N = 68)
= 13.48, p < .001. Role model preference for subtractive behaviors
showed a significantly negative association with prevention focus,
430 LOCKWOOD ET AL.
FIGURE 1. Structural equation model for role model preferences as a function of pro
-
motion and prevention goals.
as predicted. Contrary to our predictions, however, promotion focus
was positively associated with role model preference. These two
paths were significantly different, χ
2
(1, N = 68) = 16.47, p < .001.
In sum, both the framing of the desired behavior and individuals’
regulatory focus influenced participants’ preferences for positive
relative to negative role models as motivators. As expected, partici
-
pants indicated that the positive models would be more motivating
relative to negative role models when the behavior was additive
rather than subtractive. However, we found mixed support for our
hypothesis that regulatory focus would predict role model prefer
-
ences only for congruent behavior changes: As expected, prevention
predicted role model preferences only for the subtractive behaviors.
Promotion, in contrast, was associated with a preference for positive
models for both additive and subtractive behaviors. Promotion fo
-
cus may therefore be important in guiding role model preferences
for both additive and subtractive behaviors.
STUDY 2: ADDITIVE AND SUBTRACTIVE BEHAVIORS ARE
ASSOCIATED WITH DIFFERENT HEALTH–RELATED ROLE
MODEL PREFERENCES
In Study 1, we focused on academic role models, and students’ moti-
vation to engage in additive and subtractive behaviors relevant to aca-
demic success. We would expect to find similar effects, however, in
any domain in which a desired outcome can be achieved through ad
-
ditive and subtractive behaviors. For example, in order to achieve a
healthy, desirable body image, individuals frequently try to cut back
on high–fat foods, a subtractive behavior, or increase their physical ac
-
tivity, anadditive behavior; both forms ofbehavior are expected toim
-
prove one’s health and physical appearance. In line with our
reasoning that role models will be perceived as effective motivators
when they highlight a strategy congruent with the behavior, we
would expect that negative models will be viewed as more effective
when individuals consider reducing the fat in their diet; although one
could frame this activity as additive (i.e., going on a diet), the specific
behaviors involved are generally subtractive (i.e., decreasing amount
of food consumed, cutting back on fat, reducing calorie intake). In con
-
trast, we would expect that positive models will be viewed as more ef
-
fective when individuals contemplate increased exercise, an additive
USING ROLE MODELS TO HARNESS MOTIVATION 431
behavior
1
. We examined this possibility in Study 2.
We also used Study 2 to seek further support for our hypothesis
that regulatory focus will be most strongly related to role model pref-
erences when individuals are considering a behavior that is congru-
ent with the regulatory goal. In Study 1, contrary to our predictions,
promotion goals were associated with role model preferences for
both additive and subtractive behaviors. It may be the case that uni
-
versity students habitually consider academic motivation in promo
-
tion rather than prevention terms; these first–year students, who had
432 LOCKWOOD ET AL.
FIGURE 2. Role model preferences as a function of promotion and prevention goals
(Study 1).
Note. Because the parameters that are set equal are for the unstandardized variables,
the results are reported for variables in their raw score form. *p < .05, **p < .005, ***p <
.001.
1.Wecharacterizeddiet–relatedchangesas subtractive behaviors andactivity–relatedbe
-
haviors as additive behaviors because we believed that the diet/subtractive and exer
-
cise/additive distinction would reflect a natural separation in the way people think about
these health–related behaviors. That is, people who are trying to change their activity level
are especially likely to think in terms of starting an exercise program or joining a gym rather
than spending less time watching TV or spending less time sitting in front of a computer.
Onemay spend less time watching TV, but this will not necessarily be a healthy change if one
replaces this activity with other equally sedentary activities. Similarly, although improving
one’s eating habits may include additive behaviors, such as increasing fiber consumption,
the primary food–related change with which people wrestle is likely to be the reduction of
fatand/or calories in their diet. Indeed, although59% of Canadiansreport concern about the
level of fat in their diet, only 26% report concern about their carbohydrate and fiber intake
(Health Canada, 1999). Thus, individuals are more likely to try to engage in diet–related
health behaviors that are subtractive (reducing fat) rather than additive (increasing fiber).
extremely high averages when graduating from high school, may
not have been accustomed to think in terms of avoiding academic
failure because such failures have not been a relevant pastexperience
for them. In Study 2, we assessed role models in a domain in which
both promotion and prevention goals may be strong: Most individu-
als have at some point felt uncomfortable with their body shape; con
-
sequently, they may have goals to prevent an unhealthy body image
as well as to promote a more positive body image. Thus, health–re
-
lated prevention goals may be most strongly associated with restrict
-
ing food intake, whereas health–related promotion goals may be
most strongly associated with increasing physical activity.
Participants rated the extent to which they would be motivated to
engage in diet–related (cutting back on food) and exercise (increasing
physical activity) behaviors by role models in poor physical shape or
excellent physical shape. We also assessed participants’ health–re
-
lated promotion and prevention focus. We expected that participants
would view an out–of–shape model as most motivating when they
were contemplating eating less, but would view a very fit model as
most motivating when they were contemplating exercising more. We
also expected that participants’ regulatory focus would influence their
preferences for models: When contemplating additive behaviors, in
-
USING ROLE MODELS TO HARNESS MOTIVATION 433
FIGURE 3. Role model preferences as a function of promotion and prevention goals
(Study 2).
Note. Because the parameters that are set equal are for the unstandardized variables,
the results are reported for variables in their raw score form. *p < .05, **p < .005, ***p <
.001.
dividuals with strong, health–related promotion orientations should
view positive models as more effective relative to negative models;
when contemplating subtractive behaviors, individuals with strong
health–related prevention orientations should view negative models
as more effective relative to positive models.
METHOD
PARTICIPANTS
Participants were 30 male and 36 female Introductory Psychology
students at the University of Toronto who took part in the study for
course credit. Gender had no impact on any of the variables; there
-
fore gender is not discussed further.
PROCEDURE
As in Study 1, participants were asked to indicate the degree to
which they would be more motivated by a positive or a negative role
model to perform aseries of ten behaviors. Five of the behaviors were
additive (e.g., exercise more, be more active, go to a gym more often)
and five were subtractive (e.g., eat less, cut back on fatty foods, try to
keep my weight down). For each item, participants rated on a
seven–point scale whether they would be more motivated by a nega-
tive or a corresponding positive model (e.g., “a person in terrible
physical shape” vs. “a person who is in excellent physical shape”; “a
person who is weak and flabby” vs. “a person who is healthy and
toned”) to perform this behavior. Ratings were made on the same
seven–point scale used in Study 1, with higher scores indicating
greater motivation by positive relative to negative models. Both the
additive behavior subscale and the subtractive behavior subscale
were reliable (alphas = .82 and .86, respectively). A factor analysis us
-
ing an oblique rotation yielded two distinct factors: Additive items
formed one factor, with loadings ranging from .546 to .905;
subtractive items formed a second factor, with loadings ranging
from .501 to .903. The two factors were correlated, r = .36, and
together explained 63 percent of the variance.
Participants also completed a measure of health–related regula
-
tory focus. This measure was similar to the regulatory focus scale
used in Study 1, but was modified to target health–related concerns.
434 LOCKWOOD ET AL.
Six items were related to promotion focus (e.g., “My major health–re
-
lated goal right now is to increase my level of physical fitness,” “I am
more oriented toward improving my health than avoiding illness”)
and six items were related to prevention focus (e.g., “My major
health–related goal right now is to avoid experiencing health prob
-
lems,” “I am more oriented toward avoiding health problems than
achieving good health”). Both subscales were reliable (promotion al
-
pha = .81, prevention alpha = .80). The subscales were modestly cor
-
related, r = .30, p = .01. In general, participants reported higher
promotion (M = 5.95) than prevention (M = 3.92) health goals, t(65) =
8.00, p < .001.
RESULTS AND DISCUSSION
As in Study 1, we conducted a repeated measures ANOVA, with be-
havior type (additive or subtractive) as the within–participants vari-
able. As we had predicted, participants were more likely to select a
positive role model when the behavior change was described in ad-
ditive terms (M = 2.12) rather than subtractive terms (M = 1.09), F(1,
65) = 40.11, p < .001.
We next examined the possibility that regulatory focus would be
more predictive of role model preferences when this focus was con-
gruent with the desired behavior change. To test this possibility, we
estimated the structural equation model shown in Figure 1, utilizing
the same data analytic strategy we applied in Study 1. The parameter
estimates for this model are shown in Figure 3. As predicted, role
model preference for additive behaviors showed a significantly posi
-
tive association with promotion focus, but a nonsignificant negative
relationship with prevention focus. The former path was signifi
-
cantly stronger than the latter path, χ
2
(1, N = 66) = 9.87, p < .005. Also
consistent with our hypotheses, role model preference for
subtractive behaviors showed a significantly negative association
with prevention focus, but a non–significant positive relationship
with promotion focus. However, although coefficients for these
paths were in accordance with our predictions, the two paths were
not significantly different, χ
2
(1, N = 66) = 1.85, p = .17.
Overall, the results of Study 2 are consistent with those of Study 1.
In both studies, the perceived effectiveness of positive relative to
negative role models was determined by both the type of behavior,
USING ROLE MODELS TO HARNESS MOTIVATION 435
additive or subtractive, and participants’ regulatory orientation,
promotion or prevention. Participants were most likely to perceive a
positive model to be motivating when considering additive behav
-
iors, particularly to the extent that their promotion orientation was
strong. Participants were most likely to perceive a negative model to
be motivating when considering subtractive behaviors, and this
effect tended to be strongest when prevention orientation was
strong.
In both studies, participants showed a strong preference for posi
-
tive relative to negative models. Recent research suggests that the
strength of promotion and prevention orientations varies from cul
-
ture toculture; NorthAmericans tend to have an especially strong fo
-
cus on promotion (e.g., Lee et al., 2000). Consistent with this finding,
in the present studies, participants reported stronger promotion
than prevention orientations. In cultures in which prevention focus
is more strongly emphasized, however, preferences for negative role
models may be stronger. We examined this possibility in a third
study.
STUDY 3: SELF–CONSTRUALS PREDICT ROLE MODEL
PREFERENCES AMONG INDIVIDUALS CONSIDERING
ADDITIVE AND SUBTRACTIVE BEHAVIORS
In Study 3, we examined the importance of additive and subtractive
behaviors in determining role model preferences across different
cultural groups. We examined two groups that have been found to
differ in their emphasis on promotion and prevention goals: We
compared the role model preferences of North Americans, who tend
to have especially strong promotion goals, with those of East Asians,
who tend to have especially strong prevention goals (Elliot, Chirkov,
Kim, & Sheldon, 2001; Lee et al., 2000).
North American culture tends to emphasize the importance of in
-
dependent self–construals, focusing on the self as a separate, unique
individual (for reviews, see Markus & Kitayama, 1991; Triandis,
1989); individuals with strongly independent self–construals pursue
personal achievements that will help them stand out from their
group as especially successful individuals (Heine & Lehman, 1997).
The promotion–focused strategy of pursuing gains should be espe
-
cially appealing to highly independent individuals, who are bent on
achieving personal accomplishments. Thus, we would expect higher
436 LOCKWOOD ET AL.
levels of independence to be associated with a stronger promotion
orientation: In one study, for example, individuals who were primed
with independence, or who were members of cultural groups known
to emphasize independence, tended to view events framed in terms
of promoting gains as more important than events framed in terms of
preventing losses (Lee et al., 2000); another set of studies found that
individuals who scored higher on independence tended to report
more approach relative to avoidance goals (Elliot et al., 2001). We
would therefore expect that individuals with strongly independent
self–construals will choose positive role models as a mean of har
-
nessing their motivation; outstanding models provide an example of
the accomplishments for which they are striving, and highlight the
strategies needed to achieve such success. Indeed, previous research
found that European Canadian students, who have strongly inde
-
pendent self–construals, were more motivated by another student
who had received awards for outstanding achievements than by a
student who had been placed on probation due to poor academic
performance (Lockwood, Marshall, & Sadler, in press).
In contrast, East Asian cultures tend to emphasize the importance of
interdependent self–construals, focusing on the self as part of an inter-
connected set of social relationships; individuals with highly interde-
pendent self–construals are concerned with maintaining harmony
within their group, and avoid negative behaviors that could disrupt
their web of social relationships (Markus & Kitayama, 1991; Triandis,
1989). For such individuals, who are bent on avoiding behaviors that
might disturb social connections or disappoint significant others in
their lives, a prevention–focused strategy of avoiding problems
should beespecially appealing. We would therefore expect higher lev
-
els of interdependence to be associated with a greater emphasis on
prevention. Indeed, one set of studies found that individuals who
were primed with interdependence, or who were members of cultural
groups known to emphasize interdependence, tended to view events
framed in terms of preventing losses to be more important than events
framed in terms of promoting gains (Lee et al., 2000); another set of
studies found that higher levels of interdependence were associated
with more avoidance relative to approach goals (Elliot et al. 2001). We
would therefore expect that individuals with strongly interdependent
self–construals will be especially likely to choose negative role models
as a means of harnessing their motivation; the models provide an ex
-
USING ROLE MODELS TO HARNESS MOTIVATION 437
ample of the problems that they seek to avoid, and highlight the strate
-
gies needed to avert such difficulties in the future. Indeed, previous
research found that Asian Canadian students, who have strongly in
-
terdependent self–construals, were more motivated by an
unsuccessful than a successful student (Lockwood, Marshall, &
Sadler, in press).
Self–construals may be especially influential in determining role
model preferences to the extent that the self–construals are congruent
with the desired behavior change. Additive behaviors are associated
with promotion goals; independent self–construals, which foster such
promotion concerns, may therefore be especially predictive of a pref
-
erence for positive role models when individuals are considering
starting a new, potentially beneficial behavior. Subtractive behaviors,
in contrast, are associated with prevention goals; interdependent
self–construals, which foster such prevention concerns, may therefore
438 LOCKWOOD ET AL.
FIGURE 4. Role model preferences of European Canadians and Asian Canadians
considering additive and subtractive academic behaviors (Study 3).
be especially predictive of a preference for negative role models when
individuals are considering cutting back on a potentially deleterious
behavior. We examined this possibility in Study 3.
We selected participants from cultural groups previously found
to differ in self–construals: Asian Canadians and European Cana
-
dians (Heine, Lehman, Markus, & Kitayama et al., 1999). Student
participants from these groups indicated whether they would be
more motivated by negative or positive academic role models
when contemplating additive or subtractive behavior changes.
We expected that, as in Studies 1 and 2, participants would prefer
positive models when contemplating additive behaviors, and
negative models when contemplating subtractive behaviors. In
addition, we expected that Asian Canadian students would show
a general preference for negative role models, whereas European
Canadian students would show a preference for positive role
models. Finally, we predicted that self–construals would mirror
the pattern of findings obtained for regulatory focus in Studies 1
and 2: That is, independent self–construals would be most
strongly associated with a preference for positive models in addi-
tive situations, and interdependent self–construals wouldbe most
strongly associated with a preference for negative models in
subtractive situations.
METHOD
PARTICIPANTS
Participants were 45 Asian Canadian (20 males and 25 females) and
22 European Canadian (9 males and 13 females) Introductory Psy
-
chology students who received course credit for taking part in the
study. Asian Canadian participants were selected if they indicated
on a larger prescreening questionnaire that they were born in an East
Asian country (i.e., China, Hong Kong, Japan, Korea, Taiwan), and
identified their cultural background as East Asian; European Cana
-
dian participants were selected if they indicated that they were born
in Canada, and identified their cultural background as Western Eu
-
ropean. Asian Canadians had lived in Canada an average of 7.37
years, and European Canadians had lived in Canada since birth.
Asian Canadian (M = 20.36) and European Canadian (M = 20.59) par
-
ticipants did not differ in age, F <1. Participants’ gender had no effect
USING ROLE MODELS TO HARNESS MOTIVATION 439
on any of the dependent variables, and therefore is not discussed
further.
PROCEDURE
As in Study 1, participants were invited to take part in a study on adjust
-
ment to university life, and completed the same measure assessing
440 LOCKWOOD ET AL.
FIGURE 5. Role model preferences as a function of independent and interdependent
self construal (Study 3).
Note. Because the parameters that are set equal are for the unstandardized variables,
the results are reported for variables in their raw score form. *p < .05, **p < .005, ***p <
.001.
2. One additive item, “try harder to get more As” loaded relatively poorly onto both fac
-
tors; because these students had recently completed high school with very high averages,
and were expecting to excel academically, this item should perhaps have been worded in
-
stead as “try to get more A+s.” One subtractive item, “try to procrastinate less,” loaded
somewhat poorly onto the subtractive factor; possibly, students are accustomed to associ
-
ate procrastination–reduction with improving study skills because these two behaviors
are often grouped together in such activities as study skills courses.
3. In previous research, we found that these two self–construal subscales showed the
predicted pattern of associations with the regulatory focus scale used in Studies 1 and 2.
For example, in one study (N = 81), independence was positively correlated with the pro
-
motionsubscale (r = .22,p = .05)but not with the prevention subscale (r = –.12,p = .31);inter
-
dependence was positively correlated with the prevention subscale (r = .30, p = .007), but
notwith the promotion subscale (r = –.08, p = .46; Lockwood, Marshall,& Sadler, inpress).
whether they would be more motivated to perform additive and
subtractive academic behaviors by negative or positive role models. Be
-
cause the factor structure in Study 1 did not yield two distinct factors,
however, wordings for the items were altered slightly to emphasize the
additive or subtractive nature of the behaviors. For example, one addi
-
tive item worded “be focused on school” in Study 1 was changed to “fo
-
cus harder on school” in Study 2, making it clear that the behavior
involved starting something new. A factor analysis using an oblique ro
-
tation yielded two distinct factors: As predicted, additive items formed
one factor, with five of the six additive item loadings ranging from .691
to .769. Subtractive items formed a second factor, with five of the six
item loadings ranging from .551 to .838.
2
The two factors were modestly
correlated, r = .27, and together explained 49% of the variance. Overall,
this analysis provided evidence that the additive and subtractive
behaviors formed separate factors.
After completing the role model preference scale, participants
completed Singelis’s (1994) Independence/Interdependence Scale.
3
RESULTS AND DISCUSSION
To test whether behavior framing and cultural background would in-
fluence role model preferences, we conducted a 2 × 2 mixed factorial
ANOVA with behavior type (additive or subtractive) as the
within–participants variable and cultural background (Western Euro-
pean or East Asian) as the between–participants variable. The main ef
-
fect of behavior type was significant, F (1, 65) = 58.39, p < .0001 (see
Figure 4). As in Study 1, participants reported a stronger preference
for positive relative to negative role models for additive behaviors (M
= 1.35) than for subtractive behaviors (M = .12). The main effect of cul
-
tural background was also significant, F (1, 65) = 4.74, p = .03. Euro
-
pean Canadian participants reported a stronger preference for
positive relative to negative role models (M = 1.14) than did Asian Ca
-
nadian participants (M = .54). The behavior type by cultural back
-
ground interaction was not significant, F < 1.
We next examined self–construals as predictors of role model pref
-
erences. Both the independent and interdependent self–construal
subscales were moderately reliable (alphas = .60 and .64, respec
-
tively). European Canadian participants reported marginally higher
independent self–construals (M = 4.87) than did Asian Canadian
participants (M = 4.54), F (1, 65) = 3.17, p = .08. In contrast, Asian Ca
-
USING ROLE MODELS TO HARNESS MOTIVATION 441
nadian participants reported marginally higher interdependent
self–construals (M = 4.85) than did European Canadian participants
(M = 4.53), F (1, 65) = 2.92, p = .09. The relatively weak differences in
self–construals between Asian Canadian and European Canadian
participants may be due to the fact that many of the Asian Canadian
participants had been living in Canada for several years; prolonged
exposure to North American cultural values may have resulted in in
-
creased independence and decreased interdependence among these
participants. We note, however, that recent research suggests that
East Asians living in a North American context may actually report
relatively high levels of interdependence when self–construals are
measured with subjective Likert scales, because they use North
Americans as their reference group (Heine, Lehman, Peng, &
Greenholtz, 2002). Thus, the weak differences in self–construals ob
-
tained in this sample may not be the result of acculturation, but
rather may simply reflect the fact that differences in self–construals
between East Asians and North Americans tend to be relatively
small (Oyserman, Coon, & Kemmelmeier, 2002). In any case, our hy-
potheses specifically concerned the relationship between
self–construals and preferences for role models, which can also be
examined without specific reference to cultural groups.
We expected that independent self–construals, which are closely
associated with promotion goals (Elliot et al., 2001; Lee et al., 2000;
Lockwood, Marshall, & Sadler, in press), would be especially impor
-
tant in additive situations; thus, role model preferences for additive
behaviors should be predicted chiefly by independent
self–construals, rather than interdependent self–construals. In con
-
trast, we expected that interdependent self–construals, which are
closely associated with prevention goals, would be especially impor
-
tant in subtractive situations; thus, role model preferences for
subtractive behaviors should be predicted chiefly by interdependent
self–construals, rather than by independent self–construals. To eval
-
uate these hypotheses, we estimated the structural equation model
shown in Figure 1, using independence in place of promotion, and
interdependence in place of prevention.
The parameter estimates for this model are shown in Figure 5. As
predicted, role model preference for additive behaviors showed a
significantly positive association with independent self–construals,
but a nonsignificant negative relationship with interdependent
442 LOCKWOOD ET AL.
self–construals. The difference between these paths approached sig
-
nificance, χ
2
(1, N = 67) = 3.19, p = .07. Also consistent with our hy
-
potheses, role model preference for subtractive behaviors showed a
significantly negative association with interdependent
self–construals, but a nonsignificant positive relationship with inde
-
pendent self–construals. The former path was significantly stronger
than the latter path, χ
2
(1, N = 67) = 5.06, p = .02.
Overall, Study 3 provides a conceptual replication of the findings
of Studies 1 and 2. As in Studies 1 and 2, participants reported a
greater preference for positive relative to negative role models when
considering additive rather than subtractive behaviors. In addition,
individuals from a culture that emphasizes promotion concerns, the
European Canadian participants, reported a greater preference for
positive relative to negative role models; individuals from a culture
that emphasizes prevention concerns, the Asian Canadian partici-
pants, reported a weaker preference for positive relative to negative
role models. Finally, interdependent self–construals, which are asso-
ciated with strong prevention concerns, predicted motivation by
negative models when participants contemplated subtractive be-
haviors; independent self–construals, which are associated with
strong promotion concerns, predicted motivation by positive mod-
els when participants contemplated additive behaviors. The congru-
ence between regulatory concerns and behavior type thus appears to
play an important role in determining the perceived motivating in-
fluence of negative and positive role models. Different cultures tend
to foster different self–construals, which in turn are associated with
different regulatory strategies. In more interdependence–oriented
cultures, which highlight prevention concerns, negative exemplars
may be especially effective when individuals are seeking subtractive
behavior changes; in more independence–oriented cultures, which
highlight promotion concerns, positive exemplars may be especially
effective when individuals are seeking additive behavior changes.
GENERAL DISCUSSION
When individuals consider engaging ina potentially beneficial activ
-
ity, an additive behavior, they are more likely to choose positive role
models as a means of harnessing their motivation; when individuals
consider reducing a potentially deleterious activity, a subtractive be
-
USING ROLE MODELS TO HARNESS MOTIVATION 443
havior, they are more likely to choose negative role models as a
means of sustaining their motivation. In addition, participants’ own
regulatory orientations influence their role model preferences: When
contemplating additive behaviors, such as increasing exercise or
studying harder, individuals with a stronger focus on promotion are
especially likely to select positive models; when considering
subtractive behaviors, such as cutting back on high–fat foods or re
-
ducing procrastination, individuals with a stronger focus on
prevention are especially likely to select negative role models.
Traditionally, social comparison research examining individuals’
preferences for comparison targets has suggested that, whereas indi
-
viduals will choose comparisons to better–off others when they are
concerned with self–improvement, they will choose comparisons to
worse–off others when they are seeking to self–enhance (e.g., Taylor
& Lobel, 1989; Wills, 1981). A more successful other demonstrates
achievements for which one can strive; in contrast, a worse–off other
provides reassurance by highlighting one’s own relative superiority.
The present studies, however, indicate that individuals may use ei-
ther better– or worse–off others to serve self–improvement goals,
and highlight the circumstances under which each form of compari-
son may be most effective. Better–off others may be particularly use-
ful guides for self–improvement for individuals who are seeking to
start new, potentially beneficial activities; worse–off others, in con-
trast, may be more useful for individuals who are seeking to avoid
potentially harmful activities. In addition, these studies highlight the
importance of regulatory focus in determining when individuals
will select upward and downward comparison targets as role mod
-
els. Past research indicates that promotion focus is associated with an
especially strong preference for positive relative to negative role
models (Lockwood et al., 2002); however, the present studies suggest
that this will be true only for additive behaviors. When individuals
are considering subtractive behaviors, prevention focus will be more
likely to determine role model preferences.
Interestingly, in all studies, participants reported a general prefer
-
ence for positive over negative role models. Even when contemplat
-
ing subtractive behaviors, participants expected to be about equally
motivated by positive and negative role models. The general lack of
preference for negative models may in part stem from social desir
-
ability concerns; participants may have felt uncomfortable admitting
444 LOCKWOOD ET AL.
that they would try to profit from the misfortunes of another person
by using the other’s failure to motivate themselves. In addition, par
-
ticipants in these studies were students at a North American univer
-
sity; in this context, promotion goals may have been especially
salient. Even the Asian Canadian participants in Study 3 had pre
-
sumably been exposed to a significant amount of North American
culture in the course of their studies, and may consequently have de
-
veloped stronger promotion orientations; thus, it is not altogether
surprising that these participants preferred positive to negative role
models.
Nevertheless, these studies do suggest that members of individu
-
alistic and collectivistic cultures may differ in how they use role
models to harness motivation. In cultures that emphasize independ
-
ence, individuals may work toward positive outcomes by adopting
activities that they perceive to be positively associated with success;
in order to sustain their motivation, they may focus on examples of
individuals who have attained outstanding achievements as a result
of adopting such behaviors. In cultures that emphasize interdepen-
dence, individuals may instead be more likely to seek positive out-
comes by cutting back on activities that will interfere with success; in
order to maintain their motivation, they may focus on examples of
individuals who have suffered hardship as a result of engaging in
such negative behaviors. Finally, individuals who have been ex-
posed to both individualistic and collectivistic cultures may use both
positive and negative examples to harness their motivation.
Rather than examining participants’ preferences for positive and
negative role models independently, the present studies forced par
-
ticipants to choose between the two: We asked them to indicate, for
each additive and subtractive behavior, the degree to which they
would find a positive model relative to a negative role model to be
motivating. We thus attempted to create a situation in which an indi
-
vidual had to settle on a specific exemplar as a motivational guide.
For example, someone who is trying to improve his or her physique
may tape to his or her mirror a photograph of an exemplar with a de
-
sirable figure, as an incentive to self–improve, or a photograph of an
exemplar with an undesirable figure, as an incentive to avoid failure.
We were specifically interested in which kind of role model individ
-
uals would find to be relatively more appealing in different circum
-
stances. Indeed, many campaigns that use role models to promote
USING ROLE MODELS TO HARNESS MOTIVATION 445
behavior changes highlight the example of either a positive or a neg
-
ative model, rather than both together; if one wishes to maximize the
effectiveness of such a program, it is important to know which model
is likely to be most motivating. Of course, it is also possible that indi
-
viduals might at times use both kinds of role models simultaneously
as a means of harnessing their motivation. In future research, it will
be useful to examine preferences for each kind of model
independently, and the extent to which individuals use multiple role
models to motivate themselves.
This research has practical implications for programs that seek to
change behavior in a target group. Some programs may be able to
achieve desired behavior changes by using models associated with ei
-
ther additive or subtractive behaviors. For example, a company that
seeksto boost worker productivitymay be able to enhance employees’
efforts by using a positive model, such as an employee who has re-
ceived accolades for his or her achievements; alternatively, the com-
panymaytrytoreducesocialloafingbyusinganegativemodel,such
as an employee who has been fired due to insufficient output. Both the
additive behavior change, increasing effort, and the subtractive be-
havior change, reducing social loafing, may lead to the desired result:
increased productivity.In the present research, we focusedon motiva-
tion in academic and health–related domains in which either additive
or subtractive behaviors can be beneficial. The present studies suggest
that, in such domains, individuals may be more likely to use positive
than negative models to harness their motivation, even when
considering subtractive behaviors.
Nevertheless, it is possible that in cases where programs are tar
-
geted at reducing a specific behavior, negative models may prove to
be especially effective. For example, campaigns designed to reduce
drinking and driving are primarily concerned with a subtractive be
-
havior: They are set up to motivate individuals to stop driving after
they have been drinking alcohol. To elicit this motivation, the pro
-
gram may benefit most from the use of a negative role model, such as
someone who has been involved in a serious accident as a result of
driving while drunk; this negative model may be effective in moti
-
vating other individuals to avoid a similar outcome by cutting back
on their alcohol consumption when they are planning to drive. It
may also be possible to reduce drinking and driving by emphasizing
additive behaviors, such as choosing a designated driver; to elicit
446 LOCKWOOD ET AL.
this kind of behavior, the program could be set up to illustrate the
benefits of this behavior by showing a positive model who has ar
-
rived home safely after using this strategy. In general, however,
when the primary purpose of the program is to motivate a
subtractive behavior, a negative model may be most effective.
This research also has specific implications for programs seeking
to alter health–related behaviors. The literature on health–related
communications has found that different kinds of activities can be
elicited through gain– and loss–framed messages (for reviews, see
Detweiler, Bedell, Salovey, Pronin, & Rothman, 1999; Rothman,
Martino, Bedell, Detweiler, & Salovey, 1999; Rothman & Salovey,
1997). Gain–framed messages, which emphasize the rewards associ
-
ated with a particular behavior, appear to be more effective in pro
-
moting behaviors that will ensure continued good health;
loss–framed messages, which emphasize the costs associated with a
particular behavior, appear to be more effective in motivating behav-
iors that will lead to the detection of health–related problems
(Rothman & Salovey, 1997). Positive role models serve as a form of
gain–framed message; they exemplify the benefits of good health. As
such, they may elicit additive, health–affirming behaviors, such as
increasing one’s use of sunscreen in order to maintain healthy skin
(e.g., Detweiler et al., 1999) or using a dental rinse to enhance one’s
dental health (e.g., Rothman et al., 1999). Negative role models, in
contrast, serve as a form of loss–framed message; they exemplify the
costs of not performing a particular behavior. As such, they may in
-
crease individuals’ motivation to engage in illness–screening behav
-
iors, such as obtaining a mammogram (Banks et al., 1995) or using a
dental rinse to detect plaque accumulation (Rothman et al, 1999). The
present research suggests that a negative role model may not only in
-
crease screening behaviors, but may also elicit subtractive behaviors,
such as decreasing one’s time spent in the full midday sun in order to
avoid skin cancer, or reducing one’s candy consumption to avoid
dental problems. Thus, both positive and negative role models may
elicit health–related behavior changes, but the form of the behavior
change may depend on whether the model represents a potential
health gain or loss.
In this research, we did not examine the actual impact of role mod
-
els on motivation; rather, we chose to focus on the kinds of models
that individuals would select as a means of harnessing their motiva
-
USING ROLE MODELS TO HARNESS MOTIVATION 447
tion to perform particular kinds of behaviors. Previous research sug
-
gests that individuals do indeed adopt positive and negative models
in their daily lives, as a means of boosting or maintaining their moti
-
vation in domains as diverse as academic success, athletic or artistic
pursuits, and interpersonal interactions (Lockwood et al., 2002).
These models may play an ongoing role in individuals’ lives as they
work to achieve their goals. The present studies suggest that both the
nature of the desired behavior change as well as individuals’ regula
-
tory focus will influence how such models are selected. We note,
however, that it is certainly possible that role models will not always
prompt the kinds of behavior changes that individuals intend when
they select these models. Indeed, although behavioral intentions do
predict behavior (e.g., Ajzen, 1991; 1996; Ajzen & Madden, 1986), this
relationship is imperfect at best. In future research, it will therefore
be important to examine how the adoption of role models influences
individuals’ additive and subtractive behavior changes over the long
term.
REFERENCES
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human
Decision Processes, 50, 179–211.
Ajzen, I. (1996). The directiveinfluence of attitudes on behavior. In P. M. Gollwitzer &
J. A. Bargh (Eds.), The psychology of action: Linking cognition and motivation to be
-
havior (pp. 385–403). New York: Guilford.
Ajzen, I., & Madden, T. J. (1986). Prediction of goal–directed behavior: Attitudes, in
-
tentions, and perceived behavioral control. Journal of Experimental Social Psy
-
chology, 22, 453–474.
Aspinwall, L. (1997). Future–oriented aspects of social comparison: A framework for
studying health–related comparison activity. In B. P. Buunk & F. X. Gibbons
(Eds.), Health, coping, and well–being: Perspectives from social comparison theory
(pp. 125–165). Hillsdale, NJ: Erlbaum.
Banks, S. M., Salovey, P., Greener, S., Rothman, A. J., Moyer, A., Beauvais, J., & Eppel,
E. (1995). The effects of message framing on mammography utilization. Health
Psychology, 14, 178–184.
Buunk, B. P., Collins, R. L., Taylor, S. E., Van Yperen, N. W., & Dakof, G. A. (1990). The
affective consequences of social comparison: Either direction has its ups and
downs. Journal of Personality and Social Psychology, 59, 1238–1249.
Collins, R. L. (1996). For better or worse: The impact of upward social comparisons on
self–evaluations. Psychological Bulletin, 119, 51–69.
Detweiler, J. B., Bedell, B. T., Salovey, P., Pronin, E., & Rothman, A. J. (1999). Message
framing and sunscreen use: Gain–framed messages motivate beach–goers.
Health Psychology, 18, 189–196.
448 LOCKWOOD ET AL.
Elliot, A. J., Chirkov, V. I., Kim, Y., & Sheldon, K. M. (2001). A cross-cultural analysis
of avoidance (relative to approach) personal goals. Psychological Science, 12,
505–510.
Health Canada (1999). Statistical report on the health of Canadians. Ottawa, ON: Health
Canada (Cat. No. 82–570–X1E).
Heine, S. J., & Lehman, D. R. (1997). The cultural construction of self–enhancement:
An examination of group–serving biases. Journal of Personality and Social Psy
-
chology, 72, 1268–1283.
Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a universal
need for positive self–regard? Psychological Review, 106, 766–794.
Heine, S. J., Lehman, D. R., Peng, K., & Greenholtz, J. (2002). What’s wrong with
cross–cultural comparisons of subjective Likert scales? The reference–group
effect. Journal of Personality and Social Psychology, 82, 903–918.
Higgins, E. T. (1997). Beyond pleasure and pain. AmericanPsychologist, 52, 1280–1300.
Higgins, E. T. (1998). Promotion and prevention: Regulatory focus as a motivational
principle. Advances in Experimental Social Psychology, 30, 1–46.
Higgins, E. T. (2000). Making a good decision: Value from fit. American Psychologist,
55, 1217–1230.
Higgins, E. T., & Tykocinski, O. (1992). Self–discrepancies and biographical memory:
Personality and cognition at the level of psychological situation. Personality
and Social Psychology Bulletin, 18, 527–535.
Lee, A.Y., Aaker, J.L., & Gardner, W.L. (2000). The pleasures and pains of distinct
self–construal: The role of interdependence in regulatory focus. Journal of Per-
sonality and Social Psychology, 78, 1122–1134.
Lockwood, P. (2002). Could it happen to you? Predicting the impact of downward
comparisonson the self. Journal of Personality and Social Psychology, 82, 343–358.
Lockwood, P., Jordan, C. H., & Kunda, Z. (2002). Motivation by positive or negative
role models: Regulatory focus determines who will best inspire us. Journal of
Personality and Social Psychology, 83, 854-864.
Lockwood, P., & Kunda, Z. (1997). Superstars and me: Predicting the impact of role
models on the self. Journal of Personality and Social Psychology, 73, 91–103.
Lockwood, P., Marshall, T., & Sadler, P. (in press). Promoting success or preventing
failure: Cultural differences in motivation by positive and negative role mod
-
els. Personality and Social Psychology Bulletin.
Markus, H. R., & Kitayama, S. (1991). Culture and the self: Implications for cognition,
emotion, and motivation. Psychological Review, 98, 224–253.
Nisbett, R. E., & Wilson, T. D., (1977). Telling more than we can know: Verbal reports
on mental processes. Psychological Review, 84, 231–259.
Oyserman, D., Coon, H. M., & Kemmelmeier, M. (2002). Rethinking individualism
and collectivism: Evaluation of theoretical assumptions and meta–analyses.
Psychological Bulletin, 128, 3–72.
Roese, N. J., Hur, T., & Pennington, G. L. (1999). Counterfactual thinking and regula
-
tory focus: Implications for action versus inaction and sufficiency versus ne
-
cessity. Journal of Personality and Social Psychology, 77, 1109–1120.
Rothman, A. J., Martino, S. C., Bedell, B. T., Detweiler, J. B. & Salovey, P. (1999). The
systematic influence of gain– and loss–framed messages on interest in and use
of different types of health behavior. Personality and Social Psychology Bulletin,
25, 1355–1369.
USING ROLE MODELS TO HARNESS MOTIVATION 449
Rothman, A. J., & Salovey, P. (1997). Shaping perceptions to motivate healthy behav
-
ior: The role of message framing. Psychological Bulletin, 121, 3–19.
Shah, J., Higgins, E. T., & Friedman, R. (1998). Performance incentives and means:
How regulatory focus influences goal attainment. Journal of Personality and So
-
cial Psychology, 74, 285–293.
Singelis, T. M. (1994). The measurement of independent and interdependent
self–construals. Personality and Social Psychology Bulletin, 20, 580–591.
Taylor, S. E., & Lobel, M. (1989). Social comparison activity under threat: Downward
evaluation and upward contacts. Psychological Review, 96, 569–575.
Taylor, S. E., Wayment, H. A., & Carillo, M. (1996). Social comparison, self–regula
-
tion, and motivation. In R. M. Sorrentino & E. T. Higgins (Eds.), Handbook of mo
-
tivation and cognition (pp. 3–27). New York: Guilford.
Triandis, H. C. (1989). The self and social behavior in differing cultural contexts. Psy
-
chological Review, 96, 506–520.
Wills,T. A. (1981). Downward comparison principles in social psychology. Psycholog
-
ical Bulletin, 90, 245-271.
Wood, J. V. (1989). Theory and research concerning social comparisons of personal
attributes. Psychological Bulletin, 106, 231–248.
Wood, J. V., & VanderZee, K. (1997). Social comparisons among cancer patients: Un-
der what conditions are comparisons upward and downward? In B. P. Buunk
& F. X. Gibbons (Eds.), Health, coping, and well–being: Perspectives from social
comparison theory (pp. 299–328). Hillsdale, NJ: Erlbaum.
450 LOCKWOOD ET AL.
