Child Development, May/June 2003, Volume 74, Number 3, Pages 769–782

Role of Affective Self-Regulatory Efficacy in Diverse Spheres

of Psychosocial Functioning
Albert Bandura, Gian Vittorio Caprara, Claudio Barbaranelli, Maria Gerbino, and Concetta Pastorelli

This prospective study with 464 older adolescents (14 to 19 years at Time 1; 16 to 21 years at Time 2) tested the
structural paths of influence through which perceived self-efficacy for affect regulation operates in concert with
perceived behavioral efficacy in governing diverse spheres of psychosocial functioning. Self-efficacy to regulate
positive and negative affect is accompanied by high efficacy to manage one’s academic development, to resist
social pressures for antisocial activities, and to engage oneself with empathy in others’ emotional experiences.
Perceived self-efficacy for affect regulation essentially operated mediationally through the latter behavioral
forms of self-efficacy rather than directly on prosocial behavior, delinquent conduct, and depression. Perceived
empathic self-efficacy functioned as a generalized contributor to psychosocial functioning. It was accompanied
by prosocial behavior and low involvement in delinquency but increased vulnerability to depression in
adolescent females.

The capacity for self-regulation is one of the core within the complex interplay of socioeconomic,
features of human agency in social cognitive theory familial, educational, and peer influences (Bandura,
(Bandura, 1999a, 2001). Perceived self-efficacy plays a Barbaranelli, Caprara, & Pastorelli, 1996b, 2001).
pivotal role in this process of self-management Self-efficacy beliefs are developed and strengthened
because it affects actions not only directly but also by mastery experiences, social modeling, and per-
through its impact on cognitive, motivational, deci- suasive forms of social influences. Diverse interven-
sional, and affective determinants. Beliefs of personal tion programs attest to the efficacy-enhancing
efficacy influence what self-regulative standards impact of these modes of influence (Bandura, 1997;
people adopt, whether they think in an enabling or Schunk, 1989). In cross-cultural studies, the func-
debilitating manner, how much effort they invest in tional role of efficacy beliefs and the processes
selected endeavors, how they persevere in the face of through which they operate are replicated in both
difficulties, how resilient they are to adversity, how individualist and collectivist cultural systems (Ban-
vulnerable they are to stress and depression, and dura, 2002). The present study extended this line of
what types of choices they make at important research to the role of affective self-regulatory
decisional points that set the course of life paths. efficacy in the management of the transitional
A growing body of research has documented the stressors of adolescence.
contributing role of self-efficacy beliefs in self- Adolescence is an especially taxing transitional
development, adaptation, and change at different phase that presents a host of new challenges
phases of the life course (Bandura, 1995, 1997). (Bandura, 1997; Eccles & Midgley, 1989; Fursten-
Children’s beliefs in their efficacy contribute beerg, Eccles, Elder, Cook, & Sameroff, 1999; Graber,
uniquely to variance in developmental outcomes Brooks-Gunn, & Peterson, 1996). Adolescents have
to manage major biological, educational, and social
role transitions concurrently. Learning how to deal
Albert Bandura, Department of Psychology, Stanford Univer- with puberty changes, differently structured school
sity; Gian Vittorio Caprara, Claudio Barbaranelli, Maria Garbino, environments and enlarged peer networks, and
and Concetta Pastorelli, Dipartimento di Psicologia, University of
emotionally invested partnerships and sexuality
Rome, ‘‘La Sapienza,’’ Rome, Italy.
The research reported in this article was supported by grants become important. With growing independence,
from the Grant Foundation, the Spencer Foundation, and the adolescents commonly experiment with risky activ-
Jacobs Foundation. We thank Laura Carstensen and James Gross ities, some of which may take antisocial forms.
for their helpful comments on an earlier version of this article. Gender differences in depression begin to emerge in
Correspondence concerning this article should be addressed to
adolescence, with girls exhibiting higher vulnerabil-
Albert Bandura, Department of Psychology, Stanford University,
Stanford, California 94305-2130, or Gian Vittorio Caprara, Dipar- ity to depression. This is also a time when older
timento di Psicologia, Universita Degli Studi di Roma, University
of Rome, ‘‘La Sapienza,’’ Via dei Marsi 78, 00185 Roma, Italy.
Electronic mail may be sent to [email protected] or r 2003 by the Society for Research in Child Development, Inc.
[email protected]. All rights reserved. 0009-3920/2003/7403-0008
770 Bandura et al.

adolescents have to cope with the demands of of perceived self-efficacy to the regulation of one’s
emerging adulthood. How well adolescents develop affective life and its impact on psychosocial func-
and exercise their personal efficacy during this tioning. It is conducted within the social functional
formative period can play a key role in setting the perspective of social cognitive theory of emotion
course their life paths take (Bandura, 1997). (Bandura, 1986, 1992).
For years psychological theorizing and research In the interpersonal transactions of everyday life,
have centered on how the mind works in processing, socioculturally constructed expressive rules specify
representing, organizing, and retrieving information. the conditions under which certain types of emo-
To the extent that emotions were addressed, they tional displays are normative and others are deviant
were usually treated as consequences of actuating (Thoits, 1989). Expressions of positive and negative
events rather than as determinants of psychosocial affect generally have different social effects. Every-
functioning. A comprehensive theory must also day life is strewn with situational provocations and
address the role played by affect regulation in stressors that generate negative affect. Negative
human self-development and change. More recent affect is a natural part of everyday life requiring
lines of research have clarified the impact of affect effective self-management through self-regulatory
regulation on attentional, cognitive, and motiva- capabilities. Unrestrained venting of anger, dispara-
tional processes, and how failures in affect regula- ging others, and voicing jealousy would get one
tion give rise to emotional and psychosocial endlessly embroiled in social, if not legal, troubles. If
dysfunctions (Bower, 1992; Carstensen, 1992; Gross fear automatically triggered immobility or avoid-
& Munoz, 1995; Larsen, 2000; Nolen-Hoeksema, ance behavior, personal development and accom-
1991). Affect is often the basis of social ties and their plishments would be severely constrained because
durability that influences the course of lives (Ban- most significant pursuits involve some risks and
dura, 1986). Other studies have examined the evaluative consequences that are fear arousing.
development of emotional competence as reflected Unlike the often discordant and divisive effects of
in the ability to discern emotions, to understand the negative affect, positive affect promotes social con-
social consequences of one’s emotionally expressive nectedness and bonding. Expression of affection,
behavior, and to manage one’s emotional states liking, and joyfulness cultivates personal attractive-
(Mayer & Salovey, 1997; Saarni, 1999). ness. By fostering affiliative relationships, positive
Different conceptual models have been proposed affect can enhance cognitive functioning, help buffer
concerning the underlying structure of affective the perturbing effects of aversive experiences, and
experiences (Russel & Caroll, 1999; Watson & Tell- facilitate adaptive coping (Folkman & Moskowitz,
engen, 1985). Common among these models is an 2000; Fredrickson, 1998; Isen, 1987). Enabling suppor-
evaluative dimension representing positive and tive relationships enhances a sense of personal
negative affect. The regulation of affect has impor- efficacy that, in turn, influences the quality of
tant intrapersonal, communicative, and behavioral affective and behavioral functioning. Indeed, media-
functional value (Bandura, 1986; Caprara, 2002; tional analyses show that social support produces
Larsen, 2000). Therefore, the perceived self-efficacy beneficial outcomes only to the extent that it enhances
to manage these basic affective states was assigned a perceived coping self-efficacy (Bandura, 2002).
pivotal role in the posited causal structure tested in Affective states are often depicted as operating
the present study. directly on psychosocial functioning, with negative
It is one thing to possess self-regulatory skills but affect producing adverse effects and positive affect
another to be able to adhere to them in taxing and producing beneficial effects. Adaptive functioning
perturbing situations. A resilient sense of efficacy is requires discriminative regulation of affect. People
needed to overrule emotional and psychosocial differ widely in how well they manage the emotional
subverters of self-regulative efforts (Bandura, 1997; experiences of everyday life. A growing body of
Zimmerman, Bandura, & Martinez-Pons, 1992). The evidence indicates that perceived self-regulatory
present research, therefore, sought to clarify the efficacy is an important factor in the variable
structural paths of influence through which per- behavioral effects of negative affect. For example,
ceived self-efficacy for affect regulation operates in negative affect precipitates binge eating often in
concert with perceived self-efficacy for behavioral bulimics of low perceived self-regulatory efficacy,
regulation in governing diverse forms of adaptation but infrequently in those of high perceived self-
encompassing affective, prosocial, and transgressive regulatory efficacy (Love, Ollendick, Johnson, &
spheres of functioning. This program of research Schlezinger, 1985; Schneider, O’Leary, & Agras,
broadens and extends developmentally the analysis 1987). In coping with threats, individuals of high
 Affective Self-Efficacy 771

perceived self-efficacy perform intimidating activ- demands, transgressive peer pressures, and em-
ities successfully despite anxiety arousal (Bandura, pathic feelings. Although perceived self-efficacy to
1997; Pajares & Valiante, 1997; Williams, 1995). regulate affective states can influence the latter
The mediating role of perceived self-efficacy is domains of functioning directly, for reasons already
also evident in the self-management of depression. given, most of its impact was hypothesized to be
Observational studies of interactions of clinically mediated through more behaviorally oriented effi-
depressed mothers with their infants revealed that cacy beliefs.
mothers’ beliefs in their parenting self-efficacy Affective states have a widely generalized impact
predict how competently they perform caretaking on judgments of personal efficacy. Experimentally
activities after controlling for social and marital induced negative affect diminishes perceived self-
support and severity of depression (Teti & Gelfand, efficacy across different spheres of functioning,
1991). The findings from these diverse lines of whereas positive affect enhances perceived self-
research underscore the influence of self-efficacy in efficacy (Kavanagh & Bower, 1985). The more
regulating the impact of affect. intense the induced affect, the greater is its impact
To test the generality of the self-efficacy theory of on self-efficacy beliefs (Forgas, Bower, & Moylan,
affect regulation, diverse psychosocial outcomes 1990; Salovey & Birnbaum, 1989). The higher the
were selected for study. They included prosocial affect-based sense of personal efficacy, the stronger is
behavior and antisocial conduct as behavioral the engagement in activities (Bandura, 1997; Kava-
effects, and depression as an emotional aspect of nagh, 1983). It was, therefore, predicted that per-
life. Figure 1 summarizes schematically the direct ceived self-efficacy to manage positive affect would
and mediated paths of influence in the posited foster perceived academic, social self-regulatory, and
structural model. The subsequent discussion pro- empathic efficacy. By contrast, a weak sense of
vides the conceptual rationale for each structural efficacy to manage negative affect would undermine
path. the behaviorally oriented efficacy beliefs.
The first segment of the posited model specifies The second segment of the posited structural
the impact of perceived efficacy to regulate affect on model specified the functional relations of behavior-
beliefs in ones’ capabilities to manage academic ally oriented efficacy beliefs to the different spheres

Efficacy Academic Depression 1 Depression 2

Negative
Affect Efficacy

Self-Regulatory
Delinquency 1 Delinquency 2
Efficacy

Efficacy Empathic
Positive Prosocial Prosocial
Affect Efficacy Behavior 1 Behavior 2

Figure 1. Posited causal structure through which perceived self-efficacy for affect regulation operating in concert with action-oriented
efficacy beliefs influence depression, delinquent conduct, and prosocial behavior.
772 Bandura et al.

of psychosocial functioning. Perceived academic, when coactors’ joyful triumphs spell sorrowful loses
social self-regulatory, and empathic efficacy were for oneself and coactors’ sorrow spells joy for
hypothesized to affect depression, delinquent con- oneself, create counterempathic responsiveness
duct, and prosocialness both concurrently and (Englis, Vaughan, & Lanzetta, 1982; Lanzetta &
prospectively. Perceived academic self-efficacy plays Englis, 1989). Empathic responsiveness fosters pro-
a mitigating role in depression and transgression. A social behavior (Hoffman, 2001; Mussen & Eisen-
secure sense of academic self-efficacy reduces berg, 2001; Staub, 1971). Perceived self-efficacy for
vulnerability to depression by promoting academic empathic response can also deter delinquent con-
attainments and altering the construal and manage- duct by activating vicarious distress over the
ment of failure (Bandura, Pastorelli, Barbaranelli, & suffering of others and fostering social networks
Caprara, 1999). For individuals of high efficacy, conducive to harmonious relationships (Bandura,
failures, setbacks, and obstacles are viewed as 1992, 1999b; Feshbach & Feshbach, 1986; Miller &
surmountable and, therefore, spark redoubled effort Eisenberg, 1988).
rather than discouragement and despondency (Ban- To summarize the posited structural model,
dura, 1991). For individuals of low efficacy, failure perceived self-efficacy to regulate positive and
undermines motivation and breeds despondency. negative affect influences depression, delinquent
Difficulties in the academic sphere often result in conduct, and prosocial behavior both directly and
disengagement from academic activities and grav- mediationally by their impact on perceived aca-
itation to peers who favor transgressive pursuits demic self-efficacy, resistive self-regulatory efficacy,
(Dishion, 1990; Hinshaw, 1992; Jessor, Donovan, & and empathic self-efficacy. The inclusion of diverse
Costa, 1991; Patterson, Capaldi, & Bank, 1991). It is spheres of perceived self-efficacy and multiple
students’ beliefs in their academic capabilities rather developmental outcomes within the same design
than their actual academic performances that tend to permitted tests of cross-domain functional relations.
shape the course of their developmental trajectories
(Bandura, Barbaranelli, Caprara, & Pastorelli, 2001).
With regard to antisocial proclivities, a low sense of Method
academic efficacy increases risk by involvement in
Participants
transgressive activities and substance use (Bandura
et al., 1996a; Bandura, Barbaranelli, Caprara, Pastor- The participants were 464 older adolescents, 213
elli, & Regalia, 2001). males and 251 females, ranging in age from 14 to 19
As low sense of efficacy to ward off peer pressure years, with a mean age of 16 years at Time 1, and
to pursue detrimental activities creates vulnerability ranging in age from 16 to 21 with a mean age of 18
to troublesome social influences. Individuals who years at Time 2, 2 years later. To avoid potential
feel at a loss to manage the predicaments they get selection biases, all the students who were originally
caught in readily give in to inducements for in the 4th and 5th grades in one of two large
antisocial forms of conduct (Bandura et al., 1996b; elementary schools serving a community located
Bandura, Barbaranelli, Caprara, Pastorelli, et al., near Rome were the source of the participants when
2001). Perceived self-regulatory inefficacy predicts they were enrolled in the middle school at the Time 1
transgressive behavior after controlling for prior assessment. All of the families consented to have
level of transgressive behavior and quality of their children participate in the research. The project
familial relationships (Caprara, Regalia, & Bandura, included a staggered, multiple cohort design with
2002). three cohorts assessed at two time points. At the
People’s efficacy for being empathetic was also time of the second measurement, 95% of the
hypothesized to play an influential role in their participants were enrolled in several high schools
social and emotional lives. Empathic self-efficacy is serving the community. The remaining participants
not simply a reactive process of cognitive perspec- had left school and were employed in the commu-
tive taking but rather an active self-involvement in nity.
the emotional life of others (Bandura, 1986). Inter- The participants varied widely in socioeconomic
personal experiences during formative years, in background drawn from a community that repre-
which people experience joys and suffer pain in a sents a microcosm of the larger society. It contained
correlational way, create the foundation for empathic families of skilled workers, farmers, professionals,
responsiveness to the plight of others (Bandura, and local merchants and their service staffs. Eight-
1992, in press; McHugo, Smith, & Lanzetta, 1982). een percent were in professional or managerial
Conversely, discordant emotional experiences, as ranks, 40% were merchants or employees in various
 Affective Self-Efficacy 773

types of businesses, 17% were skilled workers, 18% capability to express liking and affection toward
were unskilled workers, 5% were retired, and 2% others, to get oneself to express enthusiasm and
were unemployed. The community comprises a enjoyment, and to feel satisfaction with personal
homogenous Italian population. The families live accomplishments. The item ‘‘I can show liking for a
together in the community rather than being person toward whom I am attracted’’ assessed
segregated by residence and schools based on perceived efficacy to express fondness.
socioeconomic status. The socioeconomic diversity Perceived self-efficacy to regulate negative affect
of the sample and high residential integration adds was assessed by nine items in terms of perceived
to the generalizability of the findings. capability to manage negative affect in the face of
This community adheres to a stringent consent anxiety-arousing threats, anger provocation, rejec-
procedure for research conducted in the schools. A tion, and disrespect, and to control worrisome
research proposal must gain approval from a school ruminations when things go wrong. Other items of
council composed of parent and teacher representa- affective self-management measured perceived effi-
tives at the junior and high school levels. In addition, cacy to calm oneself in taxing situations and to
parents must give consent and children are free to recover quickly one’s emotional well-being after
decline to take part. Informed consent was obtained suffering perturbing experiences. ‘‘I can calm myself
from 100% of the families, with 88% of the sample in stressful situations’’ is a sample item.
reassessed at Time 2. The attrition was mainly due to
relocation from the area or absence from school at
Perceived Academic Self-Efficacy
the time of the assessments. In ANOVA, the latter
children did not differ significantly from their Participants’ beliefs in their efficacy to direct their
counterparts on any of the variables in the initial academic activities included high loading on 15
assessment, nor did the groups differ in the items measuring perceived efficacy to master differ-
covariance matrices as tested by the Box M test for ent academic areas of coursework; to fulfill personal,
homogeneity of covariance matrices. parents’, and teachers’ academic expectations; and to
The study was explained to parents and adoles- regulate their own learning activities. The items
cents as a project designed to gain better under- concerned with self-directed learning assessed chil-
standing of adolescent development. Participants dren’s efficacy to arrange environments conducive
were administered the sets of scales measuring the to learning, to plan and to organize their academic
variables of theoretical interest by three female activities, to use cognitive strategies to enhance
researchers during specially scheduled sessions in understanding and memory of the material being
a school. The set of five self-efficacy predictors was taught, to seek pertinent information and get
measured at Time 1, and the three psychosocial teachers and peers to help them with academic
domains of functioning were measured at both Time problems when needed, to motivate themselves to
1 and Time 2. do their schoolwork, to get themselves to complete
Participants’ beliefs in their personal efficacy were scholastic assignments within set deadlines, and to
measured for five domains of functioning. For the pursue academic activities when there are other
items in the different sets of efficacy scales, interesting things to do. The item ‘‘How well can
participants rated the strength of their belief in their you get teachers to help you when you get stuck on
capability to execute the designated activities, using schoolwork?’’ measured perceived self-efficacy to
a 5-point response format ranging from 1 (perceived enlist enabling social resources. The item ‘‘How well
incapability) to 5 (complete self-assurance in one’s can you study when there are other interesting
capability). things to do?’’ measured children’s perceived effi-
cacy to motivate themselves for academic pursuits in
the face of competing attractions.
Affective Self-Regulatory Efficacy
Affective self-regulatory efficacy was measured
Resistive Self-Regulatory Efficacy
by 14 items concerning perceived capability to
manage one’s emotional life. These included per- Perceived resistive self-regulatory efficacy cen-
ceived efficacy to discern one’s emotional states, tered on adolescents’ beliefs in their efficacy to ward
understand one’s feelings toward others, and man- off social inducements for transgressive conduct.
age the expression of positive and negative affect. This perceived capability was assessed by 10 items,
Perceived self-efficacy to manage positive affect which measured perceived efficacy to resist peer
was measured by five items in terms of perceived pressure to engage in high-risk activities involving
774 Bandura et al.

the use of alcohol and drugs, sexual activity, theft of Depression

property, and various other types of transgressive
Participants rated their level of depression on the
activities that can get them into serious trouble.
20-item scale developed by Radloff (1977). The
For example, the following item assessed perceived
validity of this measure has been corroborated
self-regulatory efficacy to rebuff pressures exerted
(Radloff, 1977; Weissman, Sholomskas, Pottenger,
by peers to use drugs: ‘‘How well can you resist
Prushoff, & Locke, 1977). The items measure the
using drugs even if your friends push you to use
features that characterize depression, such as de-
them?’’
spondency, hopelessness, loss of appetite and inter-
The factor structures of the academic and resistive
est in pleasurable activities, sleep disturbance,
self-regulatory efficacy scales have been replicated
crying bouts, loss of initiative, and self-deprecation.
cross-nationally with Italian, Hungarian, and Polish
Participants rated how often over the past week they
children (Pastorelli et al., 2001). The predictive
experienced these aspects of depression using a 5-
validity of these forms of perceived self-efficacy
point response format. The alpha reliability coeffi-
has been verified in prior studies both cross-
cients were .88 for Time 1 assessment and .89 for
sectionally (Bandura et al., 1996a, 1996b; Caprara
Time 2 assessment.
et al., 1998) and prospectively (Bandura, Barbara-
nelli, Caprara, & Pastorelli 2001; Bandura, Barbar-
anelli, Caprara, Pastorelli et al., 2001; Bandura et al.,
1999; Caprara 2001; Caprara et al., 2002; Zimmerman Prosocial Behavior
et al., 1992). Participants rated on a 5-point response format
their prosocial behavior on a 24-item scale that
Empathic Self-Efficacy assessed their degree of helpfulness, sharing, con-
soling, supportiveness, and cooperativeness. ‘‘I try
Perceived empathic self-efficacy was measured by to help others’’ and ‘‘I try to console people who are
12 items in terms of perceived capability to sense sad’’ are sample items. This scale is an expanded
another person’s feelings and need for emotional version of the measure of prosocial behavior devel-
support, to discern coactors’ emotional expressions, oped by Caprara and Pastorelli (1993) for younger
to experience emotions from another person’s children. The concurrent validity of this measure has
perspective, to respond empathetically to others’ been corroborated in studies relating children’s self-
distress and misfortune, and to be sensitive to how ratings of prosocial behavior with their level of
one’s actions affect others’ feelings. The sample item prosocial behavior as rated by their parents’,
‘‘I can experience how a person in trouble feels’’ teachers’, and peers’ sociometric nominations
assessed perceived empathic capability for empathic (Caprara & Pastorelli, 1993). Factor analysis of the
distress. adolescent version revealed a single factor structure.
To investigate the dimensionality of the sets of The reliability coefficients were .94 for Time 1
self-efficacy items, a principal factor analysis with assessment and .95 for Time 2 assessment.
Oblimin rotation was performed. Only items loading
.40 or higher were considered for inclusion in a
factor. The actual item loadings in the factors ranged
Delinquent Behavior
from .45 to .78. The results revealed a five-factor
structure corresponding to the posited five domains Delinquent behavior was measured initially and 2
of self-efficacy functioning, each representing a years later by the Achenbach (1991) Delinquency
single factor. The percentage of the total variance scale. Both the reliability and predictive validity of
explained by these different self-efficacy scales was this measure of delinquent behavior are well
11% for empathic efficacy, 11% for academic efficacy, established (Achenbach, Howell, McConaughy, &
9% for resistive self-regulatory efficacy, 9% for Stanger, 1995a; Achenbach & McConaughy, 1996).
efficacy to regulate negative emotions, and 5% to The scale, comprising 11 items, assesses a wide
manage positive emotions. range of transgressive behaviors including aggres-
The alpha reliability coefficients for the self- sion, theft, cheating, lying, destructiveness, truancy,
efficacy factor scales were uniformly high. The and use of alcohol and drugs. Participants recorded
coefficients were .88 for regulating negative affect, whether they engaged in such antisocial activities
.82 for managing positive affect, .89 for empathic and, if they did, whether they did so only occasion-
efficacy, .88 for academic efficacy, and .86 for ally or often. The reliability coefficient was .86 at the
resistive self-regulatory efficacy. Time 1 assessment and .74 at the Time 2 assessment.
 Affective Self-Efficacy 775

Results Pattern of Influences

Before conducting the analyses, we examined the We tested the posited structural model on the
data for univariate and multivariate outlying cases covariance matrix with the EQS program (Bentler,
using the procedure devised by Tabachnik and Fidell 1995). Participants’ forms of perceived self-efficacy
(1989). Five participants, 2 males and 3 females, were at Time 1 served as predictors of depression,
detected as outliers and eliminated from subsequent delinquent conduct, and prosocial behavior both
analyses. Table 1 presents the means and variances concurrently (Time 1) and 2 years later (Time 2).
for the variables. It also includes the matrix of Because of gender differences, we analyzed the
relations among the forms of perceived self-efficacy structural model by using the multiple-groups
and the three domains of psychosocial functioning: model approach, which estimated simultaneously
depression, delinquent conduct, and prosocial the same pattern of relationships among variables in
behavior. None of the variables presented the two samples of males and females. In this
problems of normality, with skewness ranging approach, equivalence among samples is evaluated
from –.70 to .99 and kurtosis ranging from –.41 by constraints that impose identical estimates for the
to .90. model’s parameters (Byrne, 1994; Scott-Lennox &
ANOVAs of gender variations revealed signifi- Scott-Lennox, 1995). In EQS, the plausibility of these
cant differences on all of the assessed variables. The equality constraints is examined by the Lagrange
degrees of freedom for all of the F values are 1 and multipliers (LM) test (Bentler, 1995). For each of the
462. Females had a stronger sense of academic constraints specified, the LM test provides evidence
efficacy (F 5 27.64, p 5 o.001) and resistive self- that the constraint applies to the populations
regulatory efficacy (F 5 72.64, po.001). They also involved. In the present study the equality con-
showed stronger empathic efficacy (F 5 6.69, po.01) straints were imposed on path coefficients across the
and efficacy to express positive affect (F 5 7.88, gender groups.
po.01) but weaker efficacy to manage negative Figure 2 presents the results of the path analysis
affect (F 5 14.03, po.001). Girls were also more using the multifaceted self-efficacy predictors of
depressed than boys at both Time 1 (F 5 34.72, depression, delinquent conduct, and prosocial beha-
po.001) and Time 2 (F 5 29.53, po.001). However, vior, both concurrently and prospectively. The figure
they were more prosocially oriented at both includes all of the path coefficients that are sig-
Time 1 (F 5 59.41, po.001) and Time 2 (F 5 33.75, nificant beyond the .05 level.
po.001), and less prone to delinquent behavior at Perceived self-efficacy to regulate positive and
Time 1 (F 5 53.15, po.001) and Time 2 (F 5 69.96, negative affect are both accompanied by high
po.001). perceived academic self-efficacy, resistive self-regu-

Table 1
Means, Standard Deviations, and Correlational Matrix for the Affect Regulation and Action-Oriented Forms of Perceived Self-Efficacy, and
Depression, Delinquent Conduct and Prosocial Behavior Measured Concurrently (1) and Longitudinally (2), N 5 459

Variable M SD 1 2 3 4 5 6 7 8 9 10 11

1. Academic efficacy 3.71 .58 F .58nnn .40nnn .30nnn .37nnn .27nnn .49nnn .33nnn .14nn .38nnn .28nnn
2. Regulative efficacy 4.07 .70 F .28nnn .24nnn .36nnn .17nnn .57nnn .31nnn .07 .52nnn .23nnn
3. Empathetic efficacy 3.72 .58 F .44nnn .66nnn .19nnn .24nnn .51nnn .10n .27nnn .41nnn
4. Efficacy in managing 3.19 .67 F .36nnn .44nnn .24nnn .21nnn .33nnn .18nnn .11n
negative emotions
5. Efficacy in expressing 4.13 .61 F .18nnn .22nnn .38nnn .13nn .28nnn .28nnn
positive emotions
6. Depression (1) 30.50 8.57 F .27nnn .03 .57nnn .15nn .02
7. Delinquency (1) 6.20 5.66 F .26nnn .18nnn .58nnn .15nn
8. Prosocial behavior (1) 85.32 14.76 F .03 .25nnn .66nnn
9. Depression (2) 33.00 9.22 F .26nnn .00
10. Delinquency (2) 5.94 5.61 F .22nnn
11. Prosocial behavior (2) 86.48 13.96 F
po.05.
n nn
po.01. po.001.
nnn
776 Bandura et al.

- .20 (-.15)

-.42 (-.26)*

-.05 (-.31)* .41 (.50)

Efficacy .27 (.27) Academic Depression 1 Depression 2
Negative -.07 (
Affect -.08) Efficacy )
08
(-.

)*
.17 5
-.1 -.2

(.13
(.1 9 (-
9) .31)

.08
Self-Regulatory -.34 (-.33) .63 (.37)* Delinquency 2
Delinquency 1
.49 (.33)* Efficacy 3)
(-.1
4)

-.12
(.2

5)
.30

.25

(.2 0)
.28 -.08 (-.1
(.2
7)

Efficacy .60 (.53) Empathic .47 (.51) Prosocial

Positive Prosocial .66 (.59)
Affect Efficacy Behavior 1 Behavior 2

.13 (.12)

Figure 2. Path analysis of the pattern of influences through which perceived self-efficacy to regulate positive and negative emotions acting
in concert with academic, behavioral self-regulatory, and empathic efficacy affect depression, delinquent conduct, and prosocial behavior
concurrently and longitudinally. The first path coefficient on each of the structural links is for males; the second coefficient in brackets is
for females. All the path coefficients are significant beyond the po.05 level except that, for males, empathic self-efficacy is unrelated to
delinquent conduct and depression. These two nonsignificant path coefficients are printed in italic type. The coefficients with an asterisk
on the paths differ significantly across gender.

latory efficacy, and empathic self-efficacy. Perceived pressure for delinquent conduct, and to prosocial
self-efficacy to manage positive affect contributed behavior through empathic self-efficacy.
more strongly to variance in each of the latter Perceived academic, resistive self-regulatory, and
spheres of perceived efficacy. empathic efficacy contribute differentially to depres-
A strong sense of efficacy to regulate negative sion, delinquent conduct, and prosocial behavior
affect is accompanied by low proneness to depres- with some interesting gender differences. Perceived
sion concurrently and distally both directly and academic self-efficacy was accompanied by low
mediationally through academic self-efficacy and concurrent engagement in delinquent activities and
empathic self-efficacy. Perceived self-efficacy to low depression in females but not in males.
regulate negative affect is also accompanied by Adolescents of high perceived resistive self-regula-
concurrent delinquent conduct directly and through tory efficacy exhibited low involvement in delin-
academic self-efficacy, and by distal delinquency quent activities both concurrently and distally.
mediationally through perceived self-efficacy to Perceived empathic self-efficacy contributed to
resist peer pressure for transgressive conduct. The prosocial behavior in males and females both
relation between perceived self-efficacy to regulate concurrently and distally. High perceived empathic
negative affect to concurrent and distal prosocial self-efficacy is associated prospectively with prone-
behavior is entirely mediated through perceived ness to depression in adolescent females but not in
empathic self-efficacy. males. A strong sense of empathic self-efficacy also
Perceived self-efficacy to manage positive affect predicts low engagement in delinquent conduct in
also contributes to variance in the three domains of the longer term.
functioning, but only through its impact on the more Depression, delinquent conduct, and prosocial
action-oriented forms of perceived efficacy. Its behavior showed moderate stability over time, but
mediated link to depression is through its impact more so for depression in females and delinquency
on perceived academic self-efficacy, to delinquent in males, with a similar level of stability across
conduct through perceived self-efficacy to resist peer gender for prosocial behavior. The different forms of
 Affective Self-Efficacy 777

perceived self-efficacy made unique contributions to factor, with a significant chi square, w2(90, N 5
variance in the latter spheres of functioning after 459) 5 134.88, po.001, NNFI 5 .97, CFI 5 .98, and
controlling for the prior levels of these outcomes. RMSEA 5 .033.
The refined model, which includes the significant The Akaike information (AIC) index is particu-
nonspecified relation between empathic self-efficacy larly well suited for comparing the adequacy of non-
and depression, provided an excellent fit to the nested models fitted to the same correlational
empirical data as shown by different goodness-of-fit matrix. The lower the AIC index, the better is the
indexes. These tests yielded a nonsignificant chi goodness of fit. In this comparison, the AIC index
square, w2(80, N 5 459) 5 98.41, a non-normed fit confirmed a better fit for the posited causal structure
index (NNFI) of .99, a comparative fit index (CFI) of (–45) than the model conferring causal primacy to
.99, and a root mean square error of approximation the more behaviorally oriented forms of perceived
(RMSEA) of .023. For the CFI and NNFI indexes, the self-efficacy (–29). The model reversing the causal
closer the value is to 1 the better is the model fit. For ordering by making prior prosocial behavior, delin-
the RMSEA index, the closer the value is to 0, the quent behavior, and depression the primary causal
better is the fit of the conceptual model to the factors (54) provided the poorest fit to the empirical
empirical data. data.
The model accounted for 27% and 30% of the
variance in depression for females at Time 1 and
Discussion
Time 2, respectively, and 26% and 26% of the
variance for males at the two time points. The model The findings of this research provide a good
accounted for 34% and 54% of the variance at Times empirical fit to the structural model specifying
1 and 2 delinquent conduct in males, and 35% and how perceived affective self-regulatory efficacy
23% of the variance in Times 1 and 2 delinquent operates in concert with action-oriented perceived
conduct in females. For prosocial behavior, the self-efficacy in governing adaptation in diverse
model explained 23% and 53% of the variance for spheres of functioning. There is a notable pattern
males at Times 1 and 2, and 26% and 43% of the of gender differences in self-appraisals of efficacy.
variance for females at Times 1 and 2. Compared with adolescent males, females manifest
a stronger sense of efficacy to manage academic
activities, to rebuff peer pressure for transgressive
Alternative Models
behavior, to experience empathy for another’s
Although the refined model provided an excellent feelings and experiences, and to express positive
fit to the empirical data, alternative plausible models affectively in their interpersonal relationships. How-
were also tested. One alternative model presumes ever, adolescent females doubt their efficacy to
that quality of psychosocial functioning shapes manage negative affective states. These differential
efficacy beliefs. It reverses the direction of causation patterns of perceived self-efficacy are accompanied
with Time 1 depression, delinquent conduct, and by different styles of adaptation. Compared with
prosocial behavior influencing efficacy beliefs at males, females are more prosocial in their behavior,
Time 1 and the three modes of functioning at Time 2. less prone to delinquent conduct, and more prone to
This trimmed model, which includes only the paths depression.
found to be significant, provided a much poorer fit Although efficacy beliefs differed as a function of
to the data. It yielded a significant chi square, w2(90, gender, the causal structures were essentially the
N 5 459) 5 233.61, po.001, and fared less well on the same for both groups. In accord with prediction, a
other goodness-of-fit indexes, with NNFI 5 .90, strong sense of efficacy to manage one’s positive and
CFI 5 .93, and RMSEA 5 .059. negative emotional life contributes to perceived self-
A second plausible model conferred causal efficacy to take charge of one’s academic activities, to
primacy to perceived academic, resistive self-reg- ward off peer pressures for transgressive behavior,
ulatory, and empathic efficacy. In this structural and to feel empathy for the experiences of others.
model the latter behaviorally oriented efficacy Theorizing and research on human affect is
beliefs affect the three spheres of functioning both heavily oriented toward the detrimental psychoso-
directly and indirectly through their impact on cial effects of disregulation of negative emotional
positive and negative affect self-regulatory efficacy. states. The recent years have witnessed a shift in our
This trimmed structural model provided a better fit discipline toward the contribution of positive factors
to the data on the indexes than the model assigning to human self-development, adaptation, and change
causal primacy to prior behavior as the predominant (Bandura, 2001; Seligman & Csikszentmihalyi, 2000).
778 Bandura et al.

In the present research, perceived self-efficacy to heavily on fear-based control. They sought to
express positive affect in interpersonal transactions discourage their sons’ aggressive conduct by em-
is generally a stronger contributor to beliefs that one phasizing the external punishment it would bring on
can manage academic, transgressive, and empathic them. In contrast, the parents of prosocial sons
aspects of one’s life than is perceived self-efficacy to cultivated empathic-based control. They portrayed
regulate negative affect. the consequences of aggressive conduct in terms of
Except for the direct effect of perceived self- the injury and suffering it brings to others. In
efficacy to regulate negative affect on depression, handling problems of misconduct, parental sociali-
affective self-regulatory efficacy was related to the zation practices that direct attention to the suffering
domains of functioning indirectly through its impact inflicted on others foster development of empathic
on behaviorally oriented aspects of perceived self- perspective taking and prosocial behavior (Bandura
efficacy. In the mediated paths of influence, per- & Walters, 1959; Hoffman, 2001; Mussen & Eisen-
ceived capability to manage positive and negative berg, 2001).
affect is associated with low engagement in delin- The structural analysis revealed an interesting
quent activities concurrently and lessened proneness nonspecified relation between perceived empathic
to depression in girls through its impact on academic self-efficacy and longer term depression. This was
self-efficacy. These enabling and protective develop- true for adolescent females but not for males.
mental benefits of perceived academic self-efficacy Empathic self-efficacy in females increases vulner-
replicate functional relations obtained at younger ability to depression over time. To the extent that
ages (Bandura et al., 1996b, Bandura, Barbaranelli, many of the experiences involve perturbing aspects,
Caprara, Patorelli, et al., 2001; Bandura et al., 1999). personalizing the distresses of others can take an
This evidence adds to the generalizability of the emotional toll on empathizers. It is widely assumed
functional properties of this belief system. Similarly, that empathic arousal motivates prosocial behavior
the impact of perceived efficacy to manage positive to reduce one’s own vicarious distress. However, it is
and negative affect on delinquency and prosocial not uncommon for people to avoid empathic distress
behavior is entirely mediated through perceived by disengaging themselves psychologically and
empathic and behavioral self-regulatory efficacy. physically from the suffering of others (Bandura,
Perceived self-efficacy to rebuff peer pressures 1999b, in press; Bandura & Rosenthal, 1966).
for transgressive behavior is accompanied by In social cognitive theory (Bandura, 1999b) being
low proneness to depression concurrently and low empathetic can serve a proactive intrapersonal and
delinquent conduct both concurrently and long- prosocial function as well. People form self-concep-
itudinally. When examined within the context of tions embodying self-evaluative standards of social
familial relationships, perceived self-regulatory effi- obligation. They act in accordance with their
cacy deters involvement in delinquent activities after personal standards to preserve their self-respect.
controlling for both prior delinquent conduct and A vulnerability based on being empathetic presents
quality of parental communication (Caprara et al., the challenge of how to moderate the personaliza-
2002). Adolescents who are assured in their efficacy tion of other people’s distress and suffering to
to manage peer pressure stay clear of delinquent minimize impairing personal anguish without be-
activities and freely discuss with their parents the coming emotionally indifferent to the plight of
predicaments they face outside the home. others (Maslach, 1982).
Perceived empathic self-efficacy functioned as a The question remains of how much of the gender
significant mediator in each of the forms of adapta- difference in empathic efficacy is the product of
tion. Adolescents with a high sense of efficacy to evolutionary endowment, differential socialization,
involve themselves in the emotional lives of others sociostructural role prescription, or dynamic inter-
were more prosocial in their relationships and action among these factors. Males and females do
refrained from delinquent conduct. These findings not differ in their ability to recognize affect in others,
are in accord with prediction. Research comparing but females generally display stronger vicarious
the early familial management practices of adjudi- arousal (Hoffman, 1977). Personal distress over
cated delinquents with those of prosocial adoles- another’s adverse experiences, as indexed by auto-
cents in the same milieu shed some light on the nomic arousal, predicts proneness to helpfulness in
development of empathic self-efficacy and its role as girls but not in boys (Fabes, Eisenberg, & Eisenbud,
a restrainer of aggression (Bandura & Walters, 1959). 1993). For the socialization of emotion, males are
In their early socialization practices, parents of sons traditionally socialized to be instrumental and emo-
who adopted aggressive styles of behavior relied tionally stoic, and females are socialized to be
 Affective Self-Efficacy 779

emotionally expressive and nurturing responsive. welfare of others provides many occasions for
Differences in the socialization of emotion can create dejection over their sorrowful predicaments. Gender
differential proclivity for empathic arousal. The differences in the management of despondency is
findings of research conducted within the frame- still another process by which empathic efficacy may
work of sociological theory (Thoits, 1989) and social increase vulnerability to depression. Males tend to
cognitive theory (Bandura, 1986; Bussey & Bandura, work their way out of a despondent mood by
1999) document the social construction of affect and immersing themselves in activities, whereas females
its expressive display. Systematic research into are more inclined to dwell on their despondency,
gender differences in frequency and intensity of which sustains or exacerbates it (Nolen-Hoeksema,
affective experiences and expressive display styles 1990).
shed light on the notion that affective interpersonal Although the predictive relations were studied
commonality may be a contributing factor. prospectively, self-report data present certain limita-
Knowledge of the mechanisms governing em- tions. Note, however, that self-beliefs are subjective
pathic response provides another possible explana- phenomena that are necessarily accessible through
tion for gender difference in capability for vicarious self-report. Some of the developmental outcomes,
self-arousal. Evidence indicates that personalizing especially prosocial behavior and delinquent con-
the emotional experiences of others is more vicar- duct, are socially measurable behavioral phenom-
iously arousing than simply viewing events from ena, but many delinquent activities remain hidden
their perspective. Thus, for example, observers react from the public. In prior research, prosocial behavior
more emotionally to the sight of a person in pain if and antisocial conduct were assessed by multiple
they imagine how they themselves would feel under methods (self-report, sociometric ratings, behavior
the circumstances than if they imagine how the other observations) and by multiple sources (self, peer,
person might feel (Stotland, 1969). Studies of the teachers, parents). Self-reports of prosocial behavior
development of empathic understanding corrobo- and transgressive conduct correlated with assess-
rate the importance of personalization (Hughes, ments by these diverse methods and sources
Tingle, & Sawin, 1981). Young children who focus (Bandura et al., 1996b; Caprara & Pastorelli, 1993).
on their own emotional reactions to the plight of These findings lend support to the reliability of self-
others gain better understanding of others’ emotions reports of behavioral patterns.
than if they focus on how others might feel. Conceptions of the impact of positive and
The ability to visualize oneself undergoing the negative affect on human functioning are often
experiences to which others are being subjected framed in terms of direct effects. Positive affect
seems to be a critical factor in the empathic process promotes beneficial outcomes; negative affect breeds
(Bandura, 1986; Stotland, 1969). The imaginal self- dysfunctional outcomes. However, the variability of
arousal is facilitated by revivifying similar experi- behavioral effects requires further theoretical speci-
ences. The emotional expressions of others serve as fication. Theories of affect regulation focus on factors
retrieval cues for observers to reactivate similar that may mediate the relation between affect and
emotions that they have experienced. If females have behavioral outcomes (Gross, 1998; Larsen, 2000). As
more emotional experiences to draw on, they will previously noted, beliefs in personal efficacy to
exhibit stronger empathic responsiveness. Given that regulate positive and negative affective states oper-
females are more prone to depression than males ate in concert with action-oriented efficacy beliefs on
(Culbertson, 1997; Nolen-Hoeksema, 1990), they are diverse forms of adaptation both concurrently and
likely to have greater commonality of despondency longitudinally.
for revivification. To further complicate affect reg- Although control is more central to negative affect
ulation, females tend to have a lower sense of and expression is more central to positive affect in
efficacy to manage negative affect than do males. socializing emotion, effective social functioning
The combined effect of higher vicarious emotional requires discriminative regulation of expressions of
activation and a low sense of efficacy to manage affect. There are times and circumstances when
negative affect increases personal vulnerability to expression of negative affect, such as indignation
others’ emotional distress. over unjust practices, has positive functional value
As this and other studies reveal, compared with and indiscriminate displays of positive affect are
males, females are more involved prosocially in socially out of place.
relationships as expressed in being helpful, sharing, The obtained structural pattern of influences
consoling, and cooperative (Bandura, Barbaranelli, verifies that complex human adaptations are gov-
Caprara, Patorelli, et al., 2001). High concern for the erned by multiple forms of perceived efficacy
780 Bandura et al.

involving self-management of cognitive, motiva- Bandura, A. (1986). Social foundations of thought and action:
tional, and affective aspects of functioning operating A social cognitive theory. Englewood Cliffs, NJ: Prentice-
in concert. The verified codetermining patterns of Hall.
self-regulatory efficacy in the structural model vary Bandura, A. (1991). Self-regulation of motivation through
for different forms of adaptation. The patterned anticipatory and self-reactive mechanisms. In R. A.
Dienstbier (Ed.), Perspectives on motivation: Nebraska
multicausality further underscores the dynamic
symposium on motivation (Vol. 38, pp. 69–164). Lincoln:
interplay of efficacy beliefs in the regulation of
University of Nebraska Press.
socioemotional functioning. Of particular interest Bandura, A. (1992). Social cognitive theory of social
are relations reflecting dynamic processes such as referencing. In S. Feinman (Ed.), Social referencing and
perceived empathic self-efficacy increasing vulner- the social construction of reality in infancy (pp. 175–208).
ability to depression but curtailing socially injurious New York: Plenum.
aggression. Such findings underscore the explana- Bandura, A. (Ed.). (1995). Self-efficacy in changing societies.
tory and predictive value of a multifaceted, self- Cambridge, England: Cambridge University Press.
system perspective on personal dispositions. Multi- Bandura, A. (1997). Self-efficacy: The exercise of control. New
faceted causality requires multidimensional assess- York: Freeman.
ment and structural analysis to disentagle direct and Bandura, A. (1999a). A social cognitive theory of person-
ality. In L. Pervin & O. John (Eds.), Handbook of
mediated paths of influence.
personality (2nd ed., pp. 154–196). New York: Guilford.
Emotional experiences are heavily embedded in
Bandura, A. (1999b). Moral disengagement in the perpe-
interpersonal transactions. In maneuvering through tration of inhumanities. Personality and Social Psychology
emotionally arousing situations, people have to take Review, 3, 193–209.
charge of their inner emotional life and regulate their Bandura, A. (2001). Social cognitive theory: An agentic
expressive behavior and strategically manage their perspective. Annual Review of Psychology, 52, 1–26.
modes of adaptation. Those who believe they can Bandura, A. (2002). Social cognitive theory in cultural
exercise some measure of control over their emo- context. Journal of Applied Psychology: An International
tional life are more successful in their self-regulatory Review, 51, 269–290.
efforts than individuals who believe they are at the Bandura, A. (in press). Reflexive empathy: On predicting
more than has ever been observed. Behavioral and Brain
mercy of their emotional states (Bandura, 1997,
Sciences.
1999a; Sanderson, Rapee, & Barlow, 1989). Research
Bandura, A., Barbaranelli, C., Capara, G. V., & Pastorelli, C.
has also shown that perceived self-regulatory effi- (1996a). Mechanisms of moral disengagement in the
cacy mediates the effects of affective states on health exercise of moral agency. Journal of Personality and Social
and addictive behavior. The findings of the present Psychology, 71, 364–374.
research further corroborate the mediational func- Bandura, A., Barbaranelli, C., Caprara, G. V., & Pastorelli,
tion of different configurations of perceived action- C. (1996b). Multifaceted impact of self-efficacy beliefs
oriented efficacy in diverse spheres of functioning. on academic functioning. Child Development, 67,
Explanatory and predictive generalizability affords 1206–1222.
theoretical utility. Bandura, A., Barbaranelli, C., Caprara, G. V., & Pastorelli,
C. (2001). Self-efficacy beliefs as shapers of children’s
aspirations and career trajectories. Child Development, 72,
187–206.
References Bandura, A., Barbaranelli, C., Caprara, G. V., Pastorelli, C.,
& Regalia, C. (2001). Sociocognitive self-regulatory
Achenbach, T. M. (1991). Manual for the Youth Self-Report mechanisms governing transgressive behavior. Journal
and 1991 Profile. Burlington: University of Vermont. of Personality and Social Psychology, 80, 125–135.
Achenbach, T. M., Howell, C. T., McConaughy, S. H., & Bandura, A., Pastorelli, C., Barbaranelli, C., & Caprara,
Stanger, C. (1995a). Six-year predictors of problems in a G. V. (1999). Self-efficacy pathways to childhood
national sample of children and youth: I. Cross- depression. Journal of Personality and Social Psychology,
informant syndromes. Journal of American Academy Child 76, 258–269.
and Adolescence Psychiatry, 34, 336–347. Bandura, A., & Rosenthal, T. L. (1966). Vicarious classical
Achenbach, T. M., Howell, C. T., McConaughy, S. H., & conditioning as a function of arousal level. Journal of
Stanger, C. (1995b). Six-year predictors of problems in a Personality and Social Psychology, 3, 54–62.
national sample of children and youth. II. Signs of Bandura, A., & Walters, R. H. (1959). Adolescent aggression.
disturbance. Journal of American Academy Child and New York: Ronald Press.
Adolescence Psychiatry, 34, 488–498. Bentler, P. M. (1995). EQS structural equations program
Achenbach, T. M., & McCounaghy, S. H. (1996). Relation manual. Encino, CA: Multivariate Software.
between DSM IV and empirically based assessment. Bower, G. H. (1992). How might emotions affect learning?
School Psychology Review, 25, 329–344. In S. A. Christianson (Eds.), The handbook of emotion and
 Affective Self-Efficacy 781

memory: research and theory (pp. 3–31). Hillsdale, NJ: Folkman, S., & Moskowitz, J. T. (2000). Positive affect and
Erlbaum. the other side of coping. American Psychologist, 55,
Bussey, K., & Bandura, A. (1999). Social cognitive theory of 647–654.
gender development and differentiation. Psychological Forgas, J. P., Bower, G. H., & Moylan, S. J. (1990). Praise or
Review, 106, 676–713. blame? Affective influences on attributions for achieve-
Byrne, B. (1994). Testing the factorial validity, replication, ment. Journal of Personality and Social Psychology, 59,
and invariance of a measuring instrument: A paradig- 809–819.
matic application based on the Maslach Burnout Fredrickson, B. L. (1998). What good are positive emo-
Inventory. Multivariate Behavioral Research, 29, 289–311. tions? Review of General Psychology, 2, 300–319.
Caprara, G. V. (Ed.). (2001). La Valutazione dell’autoefficacia Furstenberg, F. F., Eccles, J., Elder, G. H., Jr., Cook, T., &
[Self-efficacy assessment]. Trento, Italy: Edizioni Sameroff, A. (1999). Adolescent development in urban
Erickson. communities: How families manage risk and opportunity.
Caprara, G. V. (2002). Personality: Filling the gap between Chicago: University of Chicago Press.
basic processes and molar functioning. In C. von Graber, J. A., Brooks-Gunn, J., & Petersen, A. C. (1996).
Hofsten & L. Bäckman (Eds.), Psychology at the turn of Transitions through adolescence: Interpersonal domains and
the millennium: Vol. 2. Social, development, and clinical context. Mahwah, NJ: Erlbaum.
perspectives (pp. 201–224). Brighton, England: Psychol- Gross, J. J. (1998). The emerging field of emotion
ogy Press. regulation: An integrative review. Review of General
Caprara, G. V., & Pastorelli, C. (1993). Early emotional Psychology, 2, 271–299.
instability, prosocial behaviour, and aggression: Some Gross, J. J., & Munoz, R. F. (1995). Emotion regulation and
methodological aspects. European Journal of Personality, mental health. Clinical Psychology: Science and Practice, 2,
7, 19–36. 151–164.
Caprara, G. V., Regalia, C., & Bandura, A. (2002). Long- Hinshaw, S. P. (1992). Externalizing behavior problems and
itudinal impact of perceived self-regulatory efficacy on academic underachievement in childhood and adoles-
violent conduct. European Psychologist, 7, 63–69. cence: Causal relationships and underlying mechan-
Caprara, G. V., Scabini, E. Barbaranelli, C., Pastorelli, C. isms. Psychological Bulletin, 111, 127–155.
Regalia, C., & Bandura, A. (1998). Impact of adolescents’ Hoffman, M. L. (1977). Sex differences in empathy and
perceived self-regulatory efficacy on familial commu- related behaviors. Psychological Bulletin, 84, 712–720.
nication and antisocial conduct. European Psychologist, 3, Hoffman, M. L. (2001). Toward a comprehensive empathy-
125–132. based theory of prosocial moral development. In A. C.
Carstensen, L. L. (1992). Selectivity theory: Social activity Bohart & D. J. Stipek (Eds.), Constructive and destructive
in life-span context. In K. W. Schaie (Ed.), Annual Review behavior (pp. 61–86). Washington, DC: American Psy-
of Gerontology and Geriatrics (Vol. 11, pp. 195–217). New chological Association.
York: Springer. Hughes, R., Jr., Tingle, B. A., & Swain, D. B. (1981).
Culbertson, F. M. (1997). Depression and gender: An Development of empathic understanding in children.
international review. American Psychologist, 52, 25–31. Child Development, 52, 122–128.
Dishion, T. J. (1990). Peer context of child and adolescent Isen, A. M. (1987). Positive affect, cognitive processes, and
troublesome behavior. In P. Leone (Ed.), Understanding social behavior. Advances in Experimental Social Psychol-
troubled and troublesome youth (pp. 128–135). Beverly ogy, 20, 203–253.
Hills, CA: Sage. Jessor, R., Donovan, J. E., & Costa, F. M. (1991). Beyond
Eccles, J. S., & Midgley, C. (1989). State-environment fit: adolescence: Problem behavior and young adult development.
Developmentally appropriate classrooms for young Cambridge, England: Cambridge University Press.
adolescents. In R. Ames & C. Ames (Eds.), Research on Kaplan, R. M., Atkins, C. J., & Reinsch, S. (1984). Specific
motivation in education, Vol. 3: Goals and cognitions efficacy expectations mediate exercise compliance in
(pp. 139–186). New York: Academic. patients with COPD. Health Psychology, 3, 223–242.
Englis, B. G., Vaughan, K. B., & Lanzetta, J. T. (1982). Kavanagh, D. J. (1983). Mood and self-efficacy. Unpublished
Conditioning of counter-empathetic emotional re- doctoral dissertation, Stanford University.
sponses. Journal of Experimental Social Psychology, 18, Kavanagh, D. J., & Bower, G. H. (1985). Mood and self-
375–391. efficacy: Impact of joy and sadness on perceived
Fabes, R. A., Eisenberg, N., & Eisenbud, L. (1993). capabilities. Cognitive Therapy and Research, 9, 507–525.
Behavioral and physiological correlates of children’s Lanzetta, J. T., & Englis, B. G. (1989). Expectations of
reactions to others’ distress. Developmental Psychology, cooperation and competition and their effects on
29, 655–663. observers’ vicarious emotional responses. Journal of
Feshbach, S., & Feshbach, N. D. (1986). Aggression and Personality and Social Psychology, 56, 543–554.
altruism: A personality perspective. In C. Zahn-Waxler, Larsen, R. J. (2000). Toward a science of mood regulation.
E. M. Cummings, & R. Iannotti (Eds.), Altruism and Psychological Inquiry, 11, 129–141.
aggression: Biological and social origins (pp. 189–217). Love, S. Q., Ollendick, T. H., Johnson, C., & Schlezinger, S.
Cambridge, England: Cambridge University Press. E. (1985). A preliminary report of the prediction of
782 Bandura et al.

bulimic behavior: A social learning analysis. Bulletin induced via inhalation of 5.5% carbon dioxide-enriched
of the Society of Psychologists in Addictive Behavior, 4, air. Archives of General Psychiatry, 46, 157–162.
93–101. Schneider, J. A., O’Leary, A., & Agras, W. S. (1987). The role
Maslach, C. (1982). BurnoutFThe cost of caring. Englewood of perceived self-efficacy in recovery from bulimia:
Cliffs, NJ: Prentice-Hall. A preliminary examination. Behaviour Research and
Mayer, J. D., & Salovey, P. (1997). What is emotional Therapy, 25, 429–432.
intelligence? In P. Salovey & D. Sluyter (Eds.), Emotional Schunk, D. H. (1989). Self-efficacy and cognitive skill
development and emotional intelligence: Implications for learning. In C. Ames & R. Ames (Eds.), Research on
educators (pp. 3–31). New York: Basic Books. motivation in education. Vol. 3: Goals and cognitions
McHugo, G. J., Smith, C. A., & Lanzetta, J. T. (1982). The (pp. 13–44). San Diego, CA: Academic.
structure of self-reports of emotional responses to film Scott-Lennox, J. A., & Scott-Lennox, R. D. (1995). Sex-race
segments. Motivation and Emotion, 6, 365–385. differences in social support and depression in older
Miller, P. A., & Eisenberg, N. (1988). The relation of low-income adults. In R. H. Hoyle (Ed.), Structural
empathy to aggressive and externalizing/antisocial equation modeling: Concepts, issues, and applications
behavior. Psychological Bulletin, 103, 324–244. (pp. 199–216). Thousand Oaks, CA: Sage.
Mussen, P., & Eisenberg, N. (2001). Prosocial development Seligman, M. E. P., & Csikszentmihalyi, M. (2000). Positive
in context. In A. C. Bohart & D. J. Stipek (Eds.), psychology. American Psychologist, 55, 5–14.
Constructive and destructive behavior (pp. 103–126). Staub, E. (1971). The learning and unlearning of aggres-
Washington, DC: American Psychological Association. sion: The role of anxiety, empathy, efficacy, and
Nolen-Hoeksema, S. (1990). Sex differences in depression. prosocial values. In J. Singer (Ed.), The control of
Stanford, CA: Stanford University Press. aggression and violence: Cognitive and physiological factors.
Nolen-Hoeksema, S. (1991). Responses to depression and New York: Academic Press.
their effects on the duration of depressive episodes. Stotland, E. (1969). Exploratory investigations of empathy.
Journal of Abnormal Psychology, 100, 569–582. In L. Berkowitz (Ed.), Advances in experimental social
Pajares, F., & Valiante, G. (1997). The predictive and psychology (Vol. 4, pp. 271–314). New York: Academic
mediational roles of the writing self-efficacy beliefs of Press.
elementary students. Journal of Educational Research, 6, Tabachnik, B. G., & Fidell, L. S. (1989). Using multivariate
353–360. statistics. New York: Harper Collins.
Pastorelli, C., Caprara, G. V., Barbaranelli, C., Rola, J., Teti, D. M., & Gelfand, D. M. (1991). Behavioral compe-
Rozsa, S., & Bandura, A. (2001). Structure of children’s tence among mothers of infants in the first year: The
perceived self-efficacy: A cross-national study. European mediational role of maternal self-efficacy. Child Develop-
Journal of Psychological Assessment, 17, 87–97. ment, 62, 918–929.
Patterson, G. R., Capaldi, D., & Bank, L. (1991). An early Thoits, P. A. (1989). The sociology of emotions. Annual
starter model for predicting delinquency. In D. Pepler & Review of Sociology, 15, 317–342.
K. H. Rubin (Eds.), The development and treatment of Watson, D., & Tellegen, A. (1985). Toward a consensual
childhood aggression (pp. 139–168). Hillsdale, NJ: structure of mood. Psychological Bulletin, 98, 219–235.
Erlbaum. Weissman M. M., Sholomskas, D., Pottenger, M., Prushoff,
Radloff, L. S. (1977). The CES–D Scale: A self-report B. A., & Locke, B. Z. ( 1977). Assessing depressive
depression scale for research in the general population. symptoms in five psychiatric populations: A validation
Applied Psychological Measurement, 1, 385–401. study. American Journal of Epidemiology, 106, 203–214.
Russell, J. A., & Carroll, J. M. (1999). On the bipolarity of Williams, S. L. (1995). Self-efficacy, anxiety, and phobic
positive and negative affect. Psychological Bulletin, 125, disorders. In J. E. Maddux (Ed.), Self-efficacy, adaptation,
3–30. and adjustment: Theory, research and application (pp. 69–
Saarni, C. (1999). The development of emotional competence. 107). New York: Plenum.
New York: Guilford. Zimmerman, B. J., Bandura, A., & Martinez-Pons, M.
Salovey, P., & Birnbaum, D. (1989). Influence of mood on (1992). Self-motivation for academic attainment: The
health-relevant cognitions. Journal of Personality and role of self-efficacy beliefs and personal goal-setting.
Social Psychology, 57, 539–551. American Educational Research Journal, 29, 663–676.
Sanderson, W. C., Rapee, R. M., & Barlow, D. H. (1989). The
influence of an illusion of control on panic attacks