From the late 1960s through the 1980s, research on moral judgment flourished, especially
work influenced by Kohlberg’s (1981, 1984) influential cognitive developmental
approach to moral judgment. In addition, there was a marked upsurge in empirical research
on prosocial development from the early to mid-1970s until approximately a
decade ago. Much of the early research on moral judgment included adolescent study
participants; moreover, in recent years there has been an inordinate amount of research
on adolescents’ aggression and antisocial behavior. Nonetheless, as noted by Hoffman
in 1980 and Eisenberg in 1990, studies of the prosocial aspects of moral development
during adolescence have been limited in quantity. Indeed, in 1987 Hill commented that
“capability for relatedness, connectedness, communion, and for what Gilligan has
termed ‘caring morality’ have . . . been little studied” (p. 24). Perhaps the relative
dearth of research on adolescents’ prosocial tendencies is not surprising, given that social
science researchers and the popular press have tended to emphasize the negative
aspects of adolescence, painting a picture of this developmental period as one of emotional
turmoil, hormones, and delinquency (Steinberg & Morris, 2001).
Nonetheless, there is a body of research on adolescent moral reasoning and prosocial
behavior that is informative for researchers and practitioners interested in adolescent
development. In this chapter we review findings on adolescents’ moral reasoning
or attributions and prosocial behaviors and emotional reactions (e.g., empathy and
sympathy). We begin with a brief discussion of some of the reasons why one would expect
morality to continue to develop in adolescence. Next, findings on moral cognitions
(e.g., moral judgment and attributions) are discussed, including those pertaining to
justice-oriented and prosocial issues. Then data on adolescents’ prosocial behavior (including
volunteer and civic activities) and empathy-related responding are reviewed.
Normative development (i.e., age-related changes) and variables related to individual
differences in moral development are considered. Research conducted with children in
late elementary school and high school is emphasized in this chapter, rather than work
with college students. Moreover, more recent findings and trends in conceptual and empirical
work often are highlighted; readers can access earlier reviews for detailed summaries
of prior work (e.g., Hoffman, 1980; Eisenberg, 1990).
155
Work on this chapter was supported by a grant from the National Institutes of Mental Health (1 R01 MH
60838) and the National Institute of Drug Abuse (DA05227).
THE FOUNDATIONS FOR MORAL COGNITIONS AND PROSOCIAL
DEVELOPMENT IN ADOLESCENCE
During the preschool and elementary school years, major advances are evident in moral
judgment and in regard to the frequency of some types of morally relevant behaviors
(e.g., some positive behaviors; Eisenberg & Fabes, 1998; Rest, 1983). Nonetheless, there
are reasons to expect further change in moral cognitions and prosocial tendencies in
adolescence. First, moral judgment and prosocial behaviors such as helping, sharing,
and comforting have been linked both conceptually and empirically with perspectivetaking
skills (Eisenberg, 1986; Kohlberg, 1981, 1984), which continues to develop in
adolescence. For example, it is not until preadolescence (ages 10–12) that individuals
are “aware of the infinite regress (I know that you know that I know that you know, etc.)
characteristic of dyadic relations; that each person is simultaneously aware of his own
and others’ subjective abilities . . . [and begins] to view his own interactions with and
subjective perspectives of others from a third person perspective” (Selman, 1975,
p. 40). Moreover, later in adolescence, the individual may become aware that in taking
another’s perspective, “the mutuality of perspectives includes a view of both self and
other as complex psychological systems of values, beliefs, attitudes, etc. [and the] . . .
further awareness that the mutuality of understanding of each other’s point of view can
take place at different qualitative levels—for example, persons can ‘know’ each other
as acquaintances, friends, closest friends, lovers, etc.” (p. 40). Selman (1980) reported a
linear pattern of change in social perspective taking from childhood to adulthood, including
advances for many individuals from adolescence into adulthood. Given the
conceptual importance of understanding another’s perspective for sympathy, otheroriented
prosocial behaviors, and higher level moral reasoning, advances in perspectivetaking
skills in adolescence would be expected to be associated with further development
of these capabilities during the same period (Colby, Kohlberg, Gibbs, & Lieberman,
1983; Eisenberg, 1986; Kohlberg, 1984).
Similarly, the advances in social problem-solving skills and interpersonal negotiation
skills noted during adolescence (e.g., Berg, 1989; Brion-Meisels & Selman, 1984)
would be expected to contribute to the development of other-oriented social interaction,
as would advances in conceptions of friendship and relationships (Brown & Gilligan,
1992; Rubin, Bukowski, & Parker, 1998; Selman, 1980) and in the ability to make
accurate attributions about others’ motives (Crick & Dodge, 1994; see also Eisenberg,
1986). In addition, changes in conceptions of the self from childhood into adolescence
likely promote moral and prosocial development. In childhood, the self is defined primarily
in terms of nonmoral properties (e.g., bodily properties, material possessions, or
typical behavior); in contrast, by late adolescence, the self is defined in terms of social
and psychological aspects of the self, and morality is the major regulator of social interactions,
whereas belief systems are central to characterizing the psychological self
(see Damon & Hart, 1988; Harter, 1999).
Finally, changes in the quality of moral reasoning and in the likelihood of sympathetic
responding during adolescence that are discussed in this chapter have been conceptually
linked to the development of altruistic tendencies (e.g., Eisenberg, 1986; Eisenberg
& Fabes, 1998). For example, Hoffman (2000) argued that the ability to sympathize
with the distresses of others who are abstract (i.e., are not in the immediate situation)
156 Moral Cognitions and Prosocial Responding in Adolescence
and with the chronic distress of others (including disadvantaged social groups) develops
in late childhood or early adolescence, based on early adolescents’ newfound ability
to view others as having continuing personal identities and life experiences beyond
the immediate situation. This change in sympathy is believed to promote adolescents’
willingness to assist abstract individuals or groups (who are not immediately present).
In brief, during late childhood and adolescence there are significant changes in sociocognitive
skills and affective responses that are believed to foster the development of
moral reasoning and altruistic tendencies (i.e., high-level prosocial responding). Therefore,
adolescence would be expected to be a period of growth for moral and prosocial
dispositions, cognitions, and behaviors.
MORAL REASONING
As mentioned previously, one reason to expect change in moral behavior in adolescence
is that moral reasoning continues to mature during adolescence and into adulthood.
Moral reasoning (or judgment), depending on its conceptualization, reflects the
structure and content of an individual’s reasoning about hypothetical or real-life moral
dilemmas—that is, how an individual justifies his or her moral decisions (Eisenberg,
1986; Kohlberg, 1981; Rest, 1979). In some studies, scores of moral reasoning may reflect
the actual decision made by a person as much or more than its reasoning (e.g., Piaget,
1932/1965).
Time does not permit an in-depth review of the basic findings on the development
of moral reasoning and recent changes in its measurement (see Rest, Narvaez, Bebeau,
& Thoma, 1999a, 1999b; Walker, in press). Rather, focal issues in recent research on
adolescents’ moral judgment are briefly summarized. These include findings on adolescents’
level of moral reasoning and its structure and the relations of moral reasoning
to adolescents’ adjustment, social competence, and risky behaviors, as well as to socialization
correlates.
The Nature of Adolescents’ Moral Reasoning
Justice Reasoning
The type of moral reasoning that has received the most empirical attention is Kohlberg’s
(1981) justice-oriented reasoning. According to Colby et al. (1983), although Stage 2
(individualism, instrumental purpose, and exchange) reasoning is predominant in early
adolescence, at about age 13, and throughout adolescence, Stage 3 (mutual interpersonal
expectations, relationships, and interpersonal conformity reasoning) moral reasoning
is the most common, dominant mode of moral reasoning. In this type of reasoning,
the “right” includes living up to what is expected by people close to you or what
others generally expect of people in your role (as son, wife, etc.). “Being good” is important
and is reflected in having good motives, showing concern for others, and maintaining
mutual relationships through trust, loyalty, respect, and gratitude (Colby et al.,
1983). At this stage, the focus in moral reasoning shifts from self-interest (Stage 2) to
fulfilling others’ expectations and concern with one’s position in others’ eyes, as well as
maintaining positive interpersonal relationships with others. Stage 4 (social system and
Moral Reasoning 157
conscience) reasoning also is used by some adolescents, but generally only infrequently.
Stage 4 reasoning increases with age from early adolescence into adulthood. Stage 4
reasoning emphasizes fulfilling the duties to which you agreed, upholding laws except
in extreme cases in which they conflict with other fixed social duties, and contributing
to the society, group, or institution (Colby et al., 1983).
Recent research is consistent with earlier findings (e.g., Dawson, 2002; Walker,
Gustafson, & Hennig, 2001; Pratt, Arnold, Pratt, & Diessner, 1999; see also Narvaez,
1998) in regard to the nature of adolescents’ moral reasoning. However, it has been argued
that what has been coded as Stage 3 might actually be two different stages (Dawson,
2002) or that there are really only three developmental schemas or levels: personal
interest (Stages 2 and 3), maintaining norms (Stage 4), and postconventional (Stages 5
and 6; Rest, Narvaez, Bebeau, & Thoma, 1999b). Thus, there is currently no consensus
on the nature of moral stages in adolescence.
In recent years there has been a considerable amount of research in which adolescents
have been asked to reason about real-life rather than hypothetical moral dilemmas.
Walker, Pitts, Hennig, and Matsuba (1995) found that there was no significant difference
between 16- to 19-year-olds’ (senior high school students’) moral reasoning
about real-life moral conflicts (coded using Kohlberg’s stages) and 18- to 25-year-olds’
(undergraduates’) reasoning, although 35- to 48-year-olds and 65- to 84-year-olds reasoned
at higher levels than did the two younger groups. Thus, change in moral reasoning
in late adolescence about real-life moral dilemmas (as well as hypothetical dilemmas)
appears to be relatively gradual.
Consistent with Kohlberg’s theory (1981), it also appears that for justice reasoning,
at least as traditionally coded, there is a cycle of consolidation and then transition upward
from stage to stage (with a mix of reasoning—especially higher level reasoning
and one’s modal level of reasoning during the transition). These findings support a
structural model in which moral reasoning during late childhood and adolescence advances
from lower to more mature levels, with periods of apparent disequilibrium between
them (Walker et al., 2001). In addition, it appears that both adolescents and
adults are more likely to use alternative ethical systems (other than Kohlberg’s) such as
religious prescriptions, community norms, professional codes, and care reasoning when
they are in a period of transition between stages (Thoma & Rest, 1999).
Prosocial Reasoning
As was acknowledged by Thoma and Rest (1999), individuals sometimes use moral reasoning
that is not well represented in Kohlberg’s (1981) justice-oriented moral reasoning.
One type of reasoning used by adolescents is care-oriented reasoning (Perry &
McIntire, 1995; Skoe et al., 1999). According to Gilligan (1982), the focus of care reasoning
is on not turning away from others rather than not treating others unfairly (i.e.,
justice concerns). Gilligan’s care reasoning is similar to the stages of prosocial moral
reasoning delineated by Eisenberg (1986), who defines prosocial moral reasoning as
reasoning about moral dilemmas in which one person’s needs or desires conflict with
those of others in a context in which the role of prohibitions, authorities’ dictates, and
formal obligations is minimal.
In the last decade, Eisenberg and colleagues (Eisenberg, Miller, Shell, McNalley, &
Shea, 1991; Eisenberg, Carlo, Murphy, & Van Court, 1995) have followed children
158 Moral Cognitions and Prosocial Responding in Adolescence
through adolescence into early adulthood to delineate the development of their prosocial
moral reasoning. In general, they have found that some self-reflective and internalized
modes of moral reasoning (e.g., reasoning pertaining to role taking; positive or
negative affect based on the consequences of behavioral choices; positive affect related
to living up to internalized values; internalized norm, rule, and law reasoning; generalized
reciprocity) increased in use, whereas stereotypic reasoning (e.g., references to expected
or normative behavior, e.g., “it’s nice to help”) continued to decrease in use from
childhood until the late teens. The linear increases in references to positive affect and
values about consequences and negative affect about consequences was not found until
late adolescence. However, hedonistic reasoning (i.e., reasoning in which the justification
is one’s own desires, e.g., “She wouldn’t help because she would rather go to the
party”), which had decreased from childhood into early midadolescence, increased
modestly in midadolescence and then again in late adolescence, primarily for males.
Moreover, direct reciprocity and approval-oriented reasoning, which had begun to decline
in midadolescence, showed little evidence of declining in the late teens (and even
increased somewhat). Although there was a linear increase in overall level of reasoning
throughout adolescence (see also Eisenberg-Berg, 1979), moral reasoning at age 19 to
20 was not predicted from moral reasoning at earlier points in adolescence, apparently
because of substantial declines in reasoning (due to increases in direct reciprocity and
hedonistic reasoning) for some people and substantial increases in reasoning due to the
use of higher level categories of reasoning for some others. In contrast, there was some
continuity in moral reasoning from age 13 to 14 years to 17 to 18 years.
Somewhat similar findings have been obtained by other researchers using a variety
of methods (Carlo, Eisenberg, & Knight, 1992). For example, in a study of Israeli 12-
to 13-, 14- to 15-, and 16- to 17-year-olds’ self-reported motives for their own volunteering
to help, reports of altruistic motives (i.e., personal willingness to assist without
any expectation of reward or approval, and without reference to compliance) increased
with age (Bar-Tal & Nissim, 1984). Further, in work on attributions about the value of
others’ prosocial actions, investigators have found that from early adolescence into
early adulthood, students increasingly devalue prosocial actions done for self-related
reasons (e.g., tangible rewards, returning a favor), approval, or praise or to avoid criticism
and punishment and increasingly value prosocial actions done out of empathy (see
Eisenberg, 1986, 1990, for reviews).
In one of the few cross-cultural studies on the topic, Boehnke, Silbereisen, Eisenberg,
Reykowski, and Palmonari (1989) examined German, Polish, Italian, and American
elementary, junior, and high school students’ attributions for why story characters
engaged in prosocial actions. Interest in others was a relatively favored motive at all
ages, whereas self-focused motives were chosen infrequently by the preadolescents and
junior and senior high school students. Conformity-related reasons decreased with age
in Italian, German, and Polish samples across grades 6, 9, 10, and 12, whereas taskoriented
reasons (i.e., pragmatic concerns related to the completion of a task) increased.
In another sample of German students in grades 5–6 or 7–9, preference for hedonistic
motives (i.e., motives related to an individual’s feelings of physical well-being
but not other aspects of self-interest) decreased with age, and preference for taskoriented
motives (e.g., “because I know if I helped, the work would get done more
quickly”) increased. For American children in grades 2–3, 5–6, and 7–8, selection of he-
Moral Reasoning 159
donistic motives decreased with age in early adolescence (Boehnke et al., 1989). Thus,
adolescents preferred other-oriented or task-oriented motives for assisting, and conformity
and hedonistic motives were somewhat less preferred with age. The lack of an
age-related change in other-oriented motives may have been due to the format of the
measure.
Gilligan (1982) argued that care-related moral reasoning is somewhat more common
among females, and there is some support for this assertion. In a meta-analysis,
Jaffee and Hyde (2000) found a small sex difference in care-related reasoning (broadly
defined) favoring females (effect size = .28). Of particular interest, this difference was
much larger in adolescents (.53) than in children (.08), university students (.18), or
young adults who were not university students (.33). However, in adolescence, whether
girls score higher than boys in care-related reasoning may depend on the country or culture.
For example, in a study with young adolescents, a sex difference was found in
Canada but not in Norway (Skoe et al., 1999). In the United States, however, the sex
difference favoring females’ higher use of care-oriented reasoning was replicated in a
study of African American seventh graders’ reasoning about dating dilemmas (Weisz
& Black, 2002).
In regard to prosocial moral reasoning, Eisenberg et al. (1987) found that a sex difference
in types of reasoning reflecting an other-orientation seemed to emerge in early
adolescence and generally was maintained throughout adolescence for at least some
higher level modes of other-oriented moral reasoning and for the overall level of prosocial
moral reasoning (Eisenberg et al., 1991, 1995). Similarly, Boehnke et al. (1989)
found modest evidence of females providing more other-oriented and less self-interested
reasons for hypothetical story characters’ prosocial actions.
Thus, in general, investigators have found that moral reasoning and attributions regarding
motives for prosocial behavior tend to stabilize or become more other-oriented
and higher level with age during the adolescent years and that females tend to express
more of such reasoning and attributions than do males.
Relations of Higher Level Moral Reasoning to Adolescents’ Adjustment and
Social Competence
Adolescents’ levels (and type) of moral reasoning are important in part because they
relate to differences in their behavior (or attitudes toward various behaviors), including
externalizing problems, prosocial behaviors, adjustment, and risky behaviors.
Moreover, adolescents’ moral reasoning has been linked to their political attitudes and
tolerance of others.
Externalizing Problems
Adolescents’ moral reasoning has been relatively consistently related to their antisocial
behavior. In reviews a decade apart, Nelson, Smith, and Dodd (1990) and Jurkovic
(1980) found that juvenile delinquents use less mature moral reasoning than do their
nondelinquent peers. Additional studies not in those reviews generally are consistent
with their conclusions (e.g., Aleixo & Norris, 2000; Carlo, Koller, & Eisenberg, 1998;
Trevethan & Walker, 1989), although self-reported offending was not related to justicerelated
moral reasoning within a group of convicted young male offenders (Aleixo &
160 Moral Cognitions and Prosocial Responding in Adolescence
Norris, 2000). Similarly, adolescents who score lower on moral judgment are more
aggressive (for boys but not girls; Schonert-Reichl, 1999), hold more positive attitudes
toward violent groups (Sotelo & Sangrador, 1999), and are more likely to perceive intentionally
injurious sport actions as legitimate (Bredemeier, 1985). Acting out preadolescent
and adolescent males are also more accepting in their judgments about others’
aggressive actions (Sanvitale, Saltzstein, & Fish, 1989; see also Berkowitz, Mueller,
Schnell, & Padberg, 1986). Further, gains in moral reasoning due to an intervention
with delinquents have been linked to lower recidivism in adolescents (although the
gains in moral reasoning due to the intervention were not significant; Gibbs, Potter,
Barriga, & Liau, 1996).
Conversely, adolescents who reason at more mature levels are more prosocial and
socially competent. For example, higher level and other-oriented prosocial moral
judgments generally have been positively related to humanitarian political attitudes
(Eisenberg-Berg & Mussen, 1978), as well as self- and other-reported prosocial tendencies
and sympathy across the teen years (Carlo, Eisenberg, & Knight, 1992; Eisenberg
et al., 1991, 1995, 2002). Moreover, measures of moral reasoning tapping a justice
orientation more than a care orientation have been associated with Italian adolescents’
involvement in volunteer activities (Comunian & Gielen, 1995) and with 10- to 13-yearold
Canadian girls’ (but not boys’) prosocial nominations by peers (Schonert-Reichl,
1999), as well as with tolerance of others’ views or lifestyles (Breslin, 1982; Raaijmakers,
Verbogt, & Vollebergh, 1998). In regard to social competence, level of justice-oriented
moral reasoning has been related to pre- and young adolescents’ peer sociometric status
and peer nominations for leadership (girls only). Socially withdrawn behavior or shyness
with peers has been negatively related to the level of justice-related moral reasoning
for boys in early adolescence (Schonert-Reichl, 1999) and for emotionally disturbed
individuals in early and midadolescence (Sigman & Erdynast, 1988; Sigman, Ungerer,
& Russell, 1983). Further, justice-related moral judgment has been linked to higher
level social-problem-solving skills in 14- to 18-year-old inner city youth (Kennedy, Felner,
Cauce, & Primavera, 1988) and with mature ego functioning (inter- and intrapersonal
strategies for coping; Matsuba & Walker, 1998). Mature ego defense mechanisms
at ages 13–14 and 16–18 also have been associated with higher level justice-reasoning
10 to 20 years later (sometimes even when controlling for moral reasoning in adolescence;
Hart & Chmiel, 1992). Thus, although the findings have not always been significant
and the percent of variance accounted for by these relations is generally modest,
adolescents who are more advanced in their moral reasoning appear to be not only
more moral in their behavior but also better adjusted and higher in social competence.
The Relation of Moral Judgments to Adolescents’ Attitudes About Risky Behavior
A special concern in adolescence is with the rise in risky behaviors such as drug use,
sexual activity, and suicide. Although very limited, there is some evidence of links between
cognitions about morality and adolescents’ tendencies to endorse or engage in
such behaviors.
Some investigators have examined the relation of level of moral judgment to adolescents’
risky behavior. There appear to be weak but not very consistent relations between
the two (Berkowitz et al., 1995). For example, in a study of undergraduates,
Hubbs-Tait and Garmon (1995) found that risk taking during sexual intercourse (i.e.,
Moral Reasoning 161
lower likelihood of using condoms) was inversely related with level of justice-related
moral reasoning (on Rest’s, 1979, defining issues test, or DIT). In contrast, in a study
of sexually active teenage girls, Jurs (1984) found no relation between moral reasoning
(also on the DIT) and the responsible use of birth control, getting pregnant, and the decision
to abort (although adolescents reasoning at higher levels were more likely to have
taken a sex education course).
However, whether an individual considers a given risky behavior to be a moral issue
may moderate the relation of moral judgment to risky behaviors. Some investigators
have examined adolescents’ tendencies to categorize risky behaviors as involving
moral, social-conventional, personal, or prudential decisions. Moral judgments involve
categorical and prescriptive judgments of right and wrong about interpersonal issues
such as harm and justice. Social-conventional issues pertain to customs or regulations
intended to ensure social coordination and social organization, such as choices about
modes of dress, table manners, and forms of greeting. Personal choices refer to issues of
private behavior that impinge primarily on the self. For example, within Western culture,
the choice of friends or recreational activities usually is considered a personal
choice, whereas prudential issues involve actual or potential self-harm (but do not involve
others’ welfare; Nucci, Guerra, & Lee, 1991; Tisak, Tisak, & Rogers, 1994).
Many high school students consider the use of legal drugs such as nicotine, caffeine,
and alcohol, as well as premarital sex, as a personal or prudential choice (especially the
latter, if assessed) rather than a behavior that should be controlled by authorities or as
a moral issue (Killen, Leviton, & Cahill, 1991; Kuther & Higgins-D’Alessandro, 2000;
Nucci et al., 1991). In contrast, the use of illegal drugs such as cocaine, crack, and marijuana
is less likely to be viewed as an issue under personal jurisdiction and is more
likely to be viewed as wrong, regardless of authority or laws (Killen et al., 1991). Risky
behaviors are seldom viewed as social-conventional issues (Killen et al., 1991; Nucci et
al., 1991; see Tisak et al., 1994, for data on adolescents’ views on the legitimacy of parents’
attempts to prohibit contact with drug-using friends). Of importance, students
who are higher in the use of drugs are more likely than are their low-using students to
view the drug use as a personal choice (Kuther & Higgins-D’Alessandro, 2000), and
less likely to view it as harmful and a prudentially unacceptable choice (wrong only because
it hurts the self; Nucci et al., 1991). High drug users also are more likely to view
themselves as the only authority with regard to the choice to use drugs and are less
likely than are their peers to view parents or the law as legitimate authorities (Nucci et
al., 1991). Moreover, in one study, adolescents’ views about the nature of decisions regarding
drug use were found to moderate the relation of the level of justice-related
moral reasoning to drug use. When adolescents considered drug use a moral decision,
a higher degree of substance use was related to lower level justice-oriented moral reasoning,
whereas moral reasoning was unrelated to the use of drugs for adolescents who
considered it to be a personal decision (similar findings were not obtained for sexual
behavior or suicide; Kuther & Higgins-D’Alessandro, 2000). Findings such as these
suggest that moral reasoning is related to some risky behaviors, but only for adolescents
who view them as having moral relevance. However, it is not clear whether moral reasoning
actually affects risky behavior or participation in risky behavior affects how
adolescents categorize risky behaviors.
162 Moral Cognitions and Prosocial Responding in Adolescence
Socialization of Moral Reasoning
Parental Influences
Socialization by parents typically has been assigned a circumscribed role in moral development
by cognitive developmental theorists such as Kohlberg (Walker & Hennig,
1999). Thus, it is not surprising that the contributions of parenting to the development
of moral reasoning have not been studied extensively. Some aspects of socialization
that have received the most attention in studies with adolescents are parental moral
reasoning, parental warmth, and aspects of parent-child discussions that might stimulate
perspective taking or autonomous moral thinking.
Based on cognitive developmental theory, one would expect parenting practices that
create cognitive conflict about moral issues to be linked to higher level moral judgment.
Somewhat consistent with this notion, there is evidence that a Socratic style of discussion
(encouraging the child to form opinions and to use reasoning) between parents
and children, combined with other variables (such as parental support) is most conducive
to the development of justice-related moral reasoning in late childhood and adolescence.
Based on studies in which parents and their child discussed hypothetical and
real-life moral dilemmas (sometimes one in the child’s life) and attempted to reach a
consensus, Walker and Hennig (1999) concluded that
parents who engage in cognitively challenging and highly opinionated interactions, who
are hostile, critical, and interfering, and who display poor ego functioning (defensiveness,
rigidity, rationalization, insensitivity, inappropriate emotional expression) provide a context
that hinders children’s opportunities to move toward more mature moral understandings.
In contrast, effective parents are more child-centered and scaffold their child’s development
by eliciting the child’s opinions, drawing out the child’s reasoning with appropriate
probing questions, and checking for understanding; all in the context of emotional support
and attentiveness and with the challenging stimulation of advanced moral reasoning.
(pp. 370, 372)
In Walker’s studies, parent behaviors such as critiquing and directly challenging the
child (especially in a hostile manner), presenting of counterconsiderations, and simply
providing information were not associated with children’s moral growth. Direct challenges
to the child’s reasoning may have been viewed as hostile by the child and, consequently,
may have been counterproductive, whereas simple provision of information
may have been viewed as lecturing (Walker & Hennig, 1999; Walker, Hennig, & Krettenauer,
2000; see Walker & Taylor, 1991). Similar findings in regard to style of interactions
have been found in other studies of preadolescents or adolescents, although
there are some inconsistencies in the literature. Buck, Walsh, and Rothman (1981) examined
the relation of parental behaviors during a discussion with their 10- to 13-yearold
boys’ moral reasoning of how to handle sons’ aggression. Boys with higher moral
reasoning had parents who considered their son’s view and tended to encourage their
son to express his views. Similarly, Holstein (1972) found that parents who encouraged
their children’s participation in discussion and decision making were more likely to
have children who reasoned at relatively high levels. In contrast to Walker and Hennig
Moral Reasoning 163
(1999), Pratt et al. (1999) observed that fathers’ tendencies to extend, challenge, or clarify
the reasoning of their adolescents were positively related to adolescents’ concurrent
moral reasoning and reasoning two years later. Similar findings were not obtained for
mothers, although mothers’ tendencies to consider their children’s perspectives when
recalling socialization encounters were related to higher level moral reasoning at the
2-year follow-up.
Other investigators besides Walker and Hennig (1999) have obtained associations
between parental warmth or involvement and high-level moral reasoning in adolescents
(e.g., Buck et al., 1981; McDevitt, Lennon, & Kopriva, 1991; Palmer & Hollin,
1996; Powers, 1988; Speicher, 1992). In some relevant studies, researchers found relations
between parental nurturance and moral reasoning for one parent or for one group
of children (e.g., age or sex group) but not the other (e.g., Bakken & Romig, 1994; Hart,
1988). Inconsistencies may occur because parental warmth by itself probably is not sufficient
to stimulate higher level moral reasoning. As noted by Hoffman (2000), parental
warmth provides an optimal environment for socialization because children are more
likely to attend to parents and care about pleasing them when the relationship generally
is supportive. Thus, parental warmth may not have a direct effect on children’s
moral reasoning but may moderate the effectiveness of other parental practices in fostering
the growth of moral reasoning. A combination of warmth and other productive
parental practices such as using a Socratic method in discussions and holding high
standards for children may be necessary to foster adolescents’ moral reasoning. In support
of this premise, authoritative parenting (which includes support, demands for appropriate
behavior and control, and practices such as induction) has been linked to
higher level moral judgment in adolescents (Boyes & Allen, 1993; Pratt et al., 1999), although
democratic parenting has not always been associated with adolescents’ moral
reasoning (Speicher, 1992).
Consistent with the relation of authoritative parenting to higher level moral reasoning,
there appears to be a modest relation between parental use of inductions (reasoning)
during discipline and older children and adolescents’ moral judgment (Janssen,
Janssens, & Gerris, 1992), although such relations often vary across parent, social class,
or age group (e.g., Eisikovits & Sagi, 1982; Parikh, 1980; see Eisenberg & Valiente,
2002, for more detail). Moreover, parental emphasis on prosocial behavior has been associated
with higher level prosocial moral reasoning (McDevitt et al., 1991). Further,
agreement between parents in regard to child-rearing practices, attitudes, and values at
age 3 has predicted higher moral reasoning for 14-year-old males (but not females;
Vaughn, Block, & Block, 1988).
In summary, higher level reasoning in adolescence is related to parenting that is supportive
and stimulates adolescents to question and expand on their reasoning, as well
as with an authoritative parenting style (including inductive discipline). However, findings
are limited in number and sometimes inconsistent (e.g., Leahy, 1981; see Eisenberg
& Valiente, 2002). Although infrequently examined, it is possible that the parenting behaviors
and characteristics associated with adolescents’ moral judgment vary across
adolescence. For example, based on her finding that moral judgment was predicted by
reports of comfort with and frequency of family moral and political discussions in later
adolescence and early adulthood, but not earlier adolescence, Speicher (1992) suggested
that the quality of interpersonal family relationships may be more important for
164 Moral Cognitions and Prosocial Responding in Adolescence
the development of moral reasoning in early adolescence, whereas aspects of the family
environment related to cognitive stimulation and perspective taking may be important
at an older age. Currently, there are too few data to test such a prediction. Moreover,
because all the extant data with adolescents are correlational, it is unclear to what
degree parental behaviors actually cause changes in adolescents’ moral reasoning; indeed,
variations in adolescents’ moral reasoning may elicit different parenting styles
and practices (and some other factor such as genetics may affect both parenting and
adolescents’ moral judgment). In addition, it is unclear whether the findings just reviewed
generalize to non-Western countries; little of this work was conducted in non-
Western societies, and there is debate regarding the degree to which systems for coding
moral judgment developed in the United States accurately represent the development
of moral judgment in non-Western, nonindustrialized countries.
Saturday, June 23, 2007
SCHOOLS, ACADEMIC MOTIVATION AND STAGE-ENVIRONMENT FIT
From the time individuals first enter school until they complete their formal schooling,
children and adolescents spend more time in schools than in any other place outside
their homes. Exploring all of the possible ways in which educational institutions influence
motivation and development during adolescence is beyond the scope of a single
chapter. In this chapter I discuss the ways in which schools influence adolescents’ socialemotional
and behavioral development through organizational, social, and instructional
processes ranging from those based in the immediate, proximal relation between
students and the tasks they are asked to perform to the role that principals and the
school boards play in setting school-level and district-level policies, which in turn influence
the social organization of the entire school community. I discuss at length three
examples of the ways in which these multiple organizational levels interact synergistically
to influence adolescent development through their impact on the daily experiences
that adolescents in the United States encounter as they move through the American
school system. The first example focuses on the role of school transitions, the
second on the role of curricular tracking, and the third on extracurricular activities.
Few of these processes have been studied in countries other than the United States. I
assume similar processes are true in other countries, but this remains to be demonstrated
empirically.
A DEVELOPMENTAL VIEW OF THE IMPACT OF SCHOOLS
ON DEVELOPMENT
Understanding the impact of schools on adolescent development requires a conceptual
framework for thinking simultaneously about schools as contexts in which development
takes place and about the changing developmental needs of students as they move
through the school system. My colleagues and I have been working on such a framework
for the last 20 years. In the late 1980s Carol Midgley and I proposed our model of stageenvironment
fit to guide research on the impact of school transitions on adolescent development
(see Eccles & Midgley, 1989; Eccles et al., 1993). We argued that individuals
have changing emotional, cognitive, and social needs and personal goals as they mature.
Drawing on ideas related to person-environment fit and self-determination theory
(e.g., Deci & Ryan, 1985), we argued that schools need to change in developmentally
125
appropriate ways if they are to provide the kind of social context that will continue to
motivate students’ interest and engagement as the students mature. To the extent that
this does not happen, we predicted that students would disengage first psychologically
and then physically from school as they matured into and through adolescence. This
should be particularly true as the adolescents acquired more incentives and more power
to control their own behavior. I say more about both of these psychological perspectives
on the impact of classroom experiences later.
More recently, Robert Roeser and I (see Eccles & Roeser, 1999) proposed a framework
for thinking about school influences that dissected the school context into a series
of hierarchically ordered, interdependent levels of organization beginning at the most
basic level of the classroom and then moving up in complexity to the school as an organizational
system embedded in a larger cultural system. In adopting this heuristic, we
assumed that (a) schools are systems characterized by multiple levels of regulatory processes
(organizational, social, and instructional in nature); (b) these processes are interrelated
across levels of analysis; (c) such processes are usually dynamic in nature,
sometimes being worked out each day between the various social actors (e.g., teachers
and students); (d) these processes change as children move through different school levels
(elementary, middle, and high school); and (e) these processes regulate children’s and
adolescents’ cognitive, social-emotional, and behavioral development. In this chapter
I focus on the interface between these theoretical frameworks. I begin with a summary
of Eccles and Roeser’s multilevel description of school contexts.
AN ECOLOGICAL VIEW OF SCHOOLS AND THEIR IMPACT ON
DEVELOPMENT DURING ADOLESCENCE
From the location of the school within macroregulatory systems characterized by national,
state, and school district laws and educational policies to the miniregulatory systems
that involve the minute-to-minute interactions between teachers and individual
students, schools are a system of complex, multilevel, regulatory processes. Eccles and
Roeser (1999) described these different levels of the school environment in terms of
their hierarchical ordering—moving from the student in a classroom, to the school
building itself, then to the school district, and finally to the larger communities in which
school districts are located. Within each of these levels, we discussed those beliefs and
practices that affect students’ experiences on a daily basis. At the classroom level, we
focused attention on teacher beliefs and instructional practices, teacher-student relationships,
the nature and design of tasks and instruction, and the nature and structure
of classroom activities and groups. At the level of the school building, we focused attention
on organizational climate and such schoolwide practices as academic tracking,
school start time, and the provision of extracurricular activities. At the level of the
school district, we focused on the between-school grade configurations that create particular
school-transition experiences for students. Finally at the level of schools embedded
in larger social systems, we discussed such issues as school resources, as well as
the linkages of schools with parents and with the labor market.
Eccles and Roeser (1999) further assumed that in any given school setting these multilevel
processes are interwoven with one another. Relations between different levels of
126 Schools, Academic Motivation, and Stage-Environment Fit
organization in the school may be complementary or contradictory and may influence
students either directly or indirectly. For instance, a principal may decide that all of his
or her teachers should use a particular practice such as cooperative learning. However,
the impact of such a decision on the daily experiences of students depends on how well
this practice is actually implemented at the classroom level. If done well, students
should be seen working successfully in groups on complex, authentic problems. Such a
well-implemented school policy is likely to produce gains in self-esteem, interethnic relationships,
and achievement among students, especially those of low ability or status
(Slavin, 1990). In contrast, if done poorly, chaos can result, leading to far less positive
outcomes at the student level. How such a schoolwide instructional policy is implemented
depends on many factors including the morale within the school, the relationships
between the principal and the teachers, the teachers’ understanding and endorsement
of the new instructional practice, the way in which the policy change was
decided upon, the provision of adequate in-service training, the provision of adequate
supports for implementation of new strategies, and the students’ willingness to go
along with the new practice. Recent debates about the likely impact of national standards
testing provide another example of the complex ways in which a new policy—this
time a state- or national-level policy—can affect the daily experiences of teachers and
students in the classroom and in the school building.
Eccles and Roeser (1999) also assumed that the processes associated with the different
levels of school interacting dynamically with each other, rather than static resources
or characteristics of the curriculum, teachers, or school per se, influence adolescents’
development. In addition, adolescents’ own constructions of meaning and interpretations
of events within the school environment are critical mediators between school
characteristics and students’ feelings, beliefs, and behavior.
Finally, in keeping with the stage-environment perspective proposed by Eccles and
Midgely (1989), Eccles and Roeser (1999) assumed that these different school-related
processes change across the course of children’s and adolescents’ development as they
progress through elementary, middle, and high school. That is, not only are children
and adolescents developing, but so too is the whole nature of the schools that they attend.
For example, the organizational, social, and instructional processes in schools
change as children move from elementary to middle school. Eccles and Midgley argued
that these changes are often associated with declines in many adolescents’ motivation
and behavior. Understanding the interaction of different school features with the developmental
needs of adolescents is critical to understanding the role of schooling in
young people’s development (see Eccles & Midgley, 1989). In the next sections I discuss
those characteristics of each level likely to be most important for understanding the impact
of schools on adolescent development. I also discuss how school characteristics at
each level may also influence group differences in adolescent development, paying particular
attention to gender and ethnic group differences within the United States.
LEVEL 1: CLASSROOMS
The most immediate educational environment to the student is the classroom. This is
also the level that has received the most attention from educational psychologists. In
Level 1: Classrooms 127
this section I review what we know about teacher beliefs, classroom climate, the nature
of the academic work itself, and experiences of racial/ethnic discrimination.
Teacher Beliefs
Teacher beliefs have received much attention in educational psychology. In this section
I focus on two types of beliefs: Teachers’ general sense of their own teaching efficacy
and teachers’ expectations for specific students in their class.
Teachers’ General Sense of Efficacy
When teachers hold high general expectations for student achievement and students
perceive these expectations, students learn more, experience a greater sense of self-worth
and competence as learners, feel more connected to their teacher and their school, and
resist involvement in problem behaviors (Eccles et al. 1993; Lee & Smith, 2001; Roeser,
Eccles, & Sameroff, 1998; Rutter, 1983; Weinstein, 1989). Similarly, teachers who feel
they are able to reach even the most difficult students and who believe in their ability to
affect students’ lives communicate such positive expectations and beliefs to their students.
Thus, a high sense of general teacher efficacy can enhance students’ own confidence
in their ability to master academic material, thereby promoting effort investment
and achievement as well as a positive emotional relationship with their teacher and
greater engagement in school as a social institution (Ashton, 1985; Midgley, Feldlaufer,
& Eccles, 1989b). Alternatively, teachers who have low confidence in their teaching efficacy
often engage in behaviors that reinforce feelings of incompetence and alienation
in their students, increasing the likelihood that their students will develop learned helpless
responses to failure in the classroom, depressive affect, anger, and disengagement
(see Cole, 1991; Roeser, Eccles, & Freedman-Doan, 1999). Lee and Smith (2001) stressed
this aspect of teachers’ general beliefs as a critical component for secondary school reform
(see also Jackson & Davis, 2000).
As I discuss in more detail later, the prevalence of teachers with a low sense of personal
teaching efficacy is higher in junior high and middle schools than in elementary
schools and higher in schools that serve high proportions of ethnic minority and poor
adolescents than in schools that serve more affluent and higher achieving adolescents
(Darling-Hammond, 1997; Eccles, Wigfield, & Schiefele, 1998). This fact alone provides
a possible explanation for both average levels of declining school engagement
during early to middle adolescence and for social class and ethnic group differences in
school engagement.
Differential Teacher Expectations
Equally important are the differential expectations teachers often hold for various individuals
within the same classroom and the differential treatments that sometimes accompany
these expectations. Beginning with the work by Rosenthal (1969), many researchers
have shown that undermining teacher-expectancy effects depend on how
teachers structure activities differently, as well as interact differently with, high- and lowexpectancy
students and on how the students perceive these differences (Brophy, 1985;
Cooper, 1979; Eccles & Wigfield, 1985; Weinstein, 1989). Most concerns have been
128 Schools, Academic Motivation, and Stage-Environment Fit
raised over behaviors that create a self-fulfilling prophecy by undermining the learning
and well-being of those students for whom the teachers hold the lowest expectations.
Much work on teacher expectancy effects has focused on differential treatment related
to gender, race/ethnic group, and/or social class. Most of this work has documented
the small but fairly consistent undermining effects of low teacher expectations
on girls (for math and science), on minority children (for all subject areas), and on children
from lower-social-class family backgrounds (again for all subject areas) (see Eccles
& Wigfield, 1985; Ferguson, 1998; Jussim, Eccles, & Madon, 1996; Valencia, 1991).
In addition, Jussim et al. (1996) found that even though these effects are typically quite
small, young women, African American adolescents, and students from poorer homes
are more subject to both the positive and negative teacher expectancy effects than are
other students.
Researchers such as Steele (1992) have linked this form of differential treatment,
particularly for African American students, to school disengagement and disidentification
(the separation of one’s self-esteem from all forms of school-related feedback).
Steele argued that African American students become aware of the fact that teachers
and other adults have negative stereotypes of African Americans’ academic abilities.
This awareness (labeled stereotype threat by Steele and his colleagues) increases their
anxieties, which in turn lead them to disidentify with the school context to protect their
self-esteem. It is interesting that recent studies using the same theoretical notions and
experimental techniques have shown that Asian students believe that teachers and
adults expect them to perform very well and that this belief leads Asian students to perform
better on tests when their ethnic identity is made salient (Shih, Pittinsky, & Ambady,
1999). Thus, the psychological processes associated with stereotype threat can
either undermine or facilitate performance on standardized tests depending on the nature
of commonly held stereotypes about the intellectual strengths and weaknesses of
different social groups.
Classroom Climate
Classroom climate refers to the more general character of the classroom and teacherstudent
relationships within the classroom. In this section I focus on the following aspects
of classroom climate: Teacher-student relationships, classroom management, and
motivational climate.
Teacher-Student Relationships
Teacher-student relationships are a key component of classroom climate: High-quality
teacher-student relationships facilitate academic motivation, school engagement, academic
success, self-esteem, and more general socioemotional well-being (Deci & Ryan,
1985; Eccles et al., 1998; Goodenow, 1993; Midgley et al., 1989b; Roeser, Midgley, &
Urdan, 1996). Teachers who trust, care about, and are respectful of students provide
the social-emotional support that students need to approach, engage, and persist on
academic learning tasks and to develop positive achievement-related self-perceptions
and values. Feeling emotionally supported is one of the most important characteristics
of contexts that support positive development. Correlational studies with adoles-
Level 1: Classrooms 129
cents show that students’ perceptions of caring teachers enhance their feelings of selfesteem,
school belonging, and positive affect in school (Roeser & Eccles, 1998; Roeser
et al., 1996).
Declines in both adolescents’ perception of emotional support from their teachers
and in the adolescents’ sense of belonging in their classrooms are quite common as adolescents
move from elementary school into secondary schools (Eccles et al., 1998). This
shift is particularly troublesome in our highly mobile society in which teachers represent
one of the last stable sources of nonparental role models for adolescents. In addition
to teaching, teachers in mobile societies such as the United States can provide
guidance and assistance when social-emotional or academic problems arise. This role
is especially important for promoting developmental competence when conditions in
the family and neighborhood cannot or do not provide such supports (Eccles, Lord, &
Roeser, 1996; Simmons & Blyth, 1987).
Classroom Management
Work related to classroom management has focused on two general issues: orderliness/
predictability and control/autonomy. With regard to orderliness and predictability, the
evidence is quite clear: Student achievement and conduct are enhanced when teachers
establish smoothly running and efficient procedures for monitoring student progress,
providing feedback, enforcing accountability for work completion, and organizing
group activities (e.g., Eccles et al., 1998; Pintrich & Schunk, 1996). Unfortunately, such
conditions are often absent, particularly in highly stressed and underfunded schools
with inexperienced teachers (Darling-Hammond, 1997).
In addition, research on international comparisons of instruction suggest that
American teachers are often more lax in their classroom management and provide less
systematic and rigorous control over the instructional sequences (Stevenson & Stigler,
1992). Furthermore, this research suggests that these differences in teachers’ controlrelated
practices could be a partial explanation for the relatively poor performance of
many American youth on international standardized tests of math and science achievement
(Schmidt, McKnight, & Raizen, 1997).
Motivational Climate
Several teams of researchers have suggested that teachers engage in a wide range of behaviors
that create a pervasive motivational climate in the classroom. For example,
Rosenholtz and Simpson (1984) suggested a cluster of general teaching practices (e.g.,
individualized vs. whole-group instruction, ability grouping practices, and publicness
of feedback) that should affect motivation because these practices make ability differences
in classrooms especially salient to students (see Mac Iver, 1988). They assumed
that these practices affect the motivation of all students by increasing the salience of extrinsic
motivators and ego-focused learning goals, leading to greater incidence of social
comparison behaviors and increased perception of ability as an entity state rather than
an incremental condition. All of these changes reduce the quality of students’ motivation
and learning. The magnitude of the negative consequences of these shifts, however,
should be greatest for low-performing students: As these students become more aware
of their relative low standing, they are likely to adopt a variety of ego-protective strategies
that unfortunately undermine learning and mastery (Covington, 1992).
130 Schools, Academic Motivation, and Stage-Environment Fit
More recently, researchers interested in goal theory have proposed a similar set of
classroom characteristics (Ames, 1992; E. M. Anderman, & Maehr, 1994; Maehr &
Midgley, 1996; Pintrich & Schunk, 1996; Roeser, Midgley, & Maehr, 1994). Goal theorists
propose two major achievement goal systems: mastery-oriented goals and performanceoriented
goals. Students with mastery-oriented goals focus on learning the material
and on their own improvement over time. Students with performance-oriented goals
focus on doing better than other students in their class. Goal theorists further argue
that a mastery orientation sustains school engagement and achievement better than
does a performance orientation (see Ames, 1992; Maehr & Midgley, 1996; Midgley,
2002). Evidence is quite strong for the first prediction and more mixed for the second:
The desire to do better than others often has positive rather than negative consequences,
whereas the fear of failing (performance-avoidance goal orientation) undermines school
performance (see Midgley, 2002). Finally, these theorists suggest that the publicness of
feedback, particularly social comparative feedback, and a classroom focus on competition
between students undermine mastery motivation and increase performance motivation.
The school-reform work of Midgley, Maehr, and their colleagues has shown
that social reform efforts to reduce these types of classroom practices, particularly
those associated with performance feedback, social comparative grading systems, and
ego-focused, competitive motivational strategies have positive consequences for adolescents’
academic motivation, persistence on difficult learning tasks, and socioemotional
development (e.g., Maehr & Midgley, 1996).
The work on understanding group differences in achievement and achievement
choices is another example of an attempt to identify a broad set of classroom characteristics
related to motivation. The work on girls and math is one example of this approach.
There are sex differences in adolescents’ preference for different types of learning
contexts that likely interact with subject area to produce sex differences in interest
in different subject areas (Eccles, 1989; Hoffmann & Haeussler, 1995). Females appear
to respond more positively to math and science instruction if it is taught in a cooperative
or individualized manner rather than a competitive manner, if it is taught from an
applied or person-centered perspective rather than a theoretical or abstract perspective,
if it is taught using a hands-on approach rather than a book-learning approach,
and if the teacher avoids sexism in its many subtle forms. The reason given for these effects
is the fit between the teaching style, the instructional focus, and females’ values,
goals, motivational orientations, and learning styles. The few relevant studies support
this hypothesis (Eccles & Harold, 1993; Hoffmann & Haeussler, 1995). If such classroom
practices are more prevalent in one subject area (e.g., physical science or math)
than another (e.g., biological or social science), one would expect sex differences in motivation
to study these subject areas. In addition, however, math and physical science
do not have to be taught in these ways; more girl-friendly instructional approaches can
be used. When they are, girls, as well as boys, are more likely to continue taking courses
in these fields and to consider working in these fields when they become adults.
The girl-friendly classroom conclusion is a good example of person-environment fit.
Many investigators have suggested that students are maximally motivated to learn in
situations that fit well with their interests, current skill level, and psychological needs,
so that the material is challenging, interesting, and meaningful (e.g., Csikszentmihalyi,
Rathunde, & Whalen, 1993; Eccles et al., 1993; Krapp, Hidi, & Renninger, 1992). Vari-
Level 1: Classrooms 131
ations on this theme include aptitude by treatment interactions and theories stressing
cultural match or mismatch as one explanation for group differences in school achievement
and activity choices (e.g., Fordham & Ogbu, 1986; Suarrez-Orozco & Suarrez-
Orozco, 1995; Valencia, 1991). For example, Valencia (1991) concluded that a mismatch
of both the values of the school and the materials being taught contributed to
the poor performance and high dropout rates among Latino youth in the high school
they studied. Deyhle and LeCompte (1999) made a similar argument in their discussion
of the poor performance of Native American youth in traditional middle school contexts.
The misfit between the needs of young adolescents and the nature of junior high
school environments is another example of these person-environment fit dynamics.
The Nature of Academic Work
Academic work is at the heart of the school experience. Two aspects of academic tasks
are important: the content of the curriculum and the design of instruction. The nature
of academic content has an important impact on students’ attention, interest, and cognitive
effort. Long ago, Dewey (1902/1990) proposed that academic work that is meaningful
to the historical and developmental reality of students’ experiences will promote
sustained attention, high investment of cognitive and affective resources in learning,
and strong identification with educational goals and aims. In general, research supports
this hypothesis: Content that provides meaningful exploration is critical given
that boredom in school, low interest, and perceived irrelevance of the curriculum are
associated with poor attention, diminished achievement, disengagement, and alienation
from school (e.g., Finn, 1989; Jackson & Davis, 2000; Larson & Richards, 1989).
Curricula that represent the voices, images, and historical experiences of traditionally
underrepresented groups are also important (Valencia, 1991). Choosing materials that
provide an appropriate level of challenge for a given class, designing learning activities
that require diverse cognitive operations (e.g., opinion, following routines, memory,
comprehension), structuring lessons so that they build on each other in a systematic
fashion, using multiple representations of a given problem, and explicitly teaching students
strategies that assist in learning are but a few of the design features that scaffold
learning and promote effort investment, interest in learning, and achievement (Blumenfeld,
1992; Deci & Ryan, 1985; Eccles et al., 1998).
Unfortunately, American secondary schools have problems providing each of these
types of educational experiences. Recent work by Larson and his colleagues has documented
the fact that adolescents are bored most of the time that they are in secondary
school classrooms (Larson, 2000; Larson & Richards, 1989). Culturally meaningful
learning experiences are rare in many American secondary schools (Fine, 1991; Valencia,
1991). The disconnection of traditional curricula from the experiences of these
groups can explain the alienation of some group members from the educational process,
sometimes eventuating in school dropout (Fine, 1991; Sheets & Hollins, 1999).
Appropriately designed tasks that adequately scaffold learning are also rare in many
inner-city and poor schools (Darling-Hammond, 1997).
In addition, from a developmental perspective, there is evidence that the nature of
academic work too often does not change over time in ways that are concurrent with
the increasing cognitive sophistication, diverse life experiences, and identity needs of
132 Schools, Academic Motivation, and Stage-Environment Fit
adolescents as they move from the elementary into the secondary school years (Carnegie
Council, 1989; Lee & Smith, 2001). As one indication of this, middle school students
report the highest rates of boredom when doing schoolwork, especially passive work
(e.g., listening to lectures) and in particular classes such as social studies, math, and science
(Larson & Richards, 1989). There is also evidence that the content of the curriculum
taught in schools does not broaden to incorporate either important health or social
issues that become increasingly salient as adolescents move through puberty and
deal with the identity explorations associated with adolescence (Carnegie Council,
1989). Further, academic work sometimes becomes less, rather than more, complex in
terms of the cognitive demands as adolescents move from elementary to junior high
school (Eccles et al., 1998). It may be that declines in some adolescents’ motivation during
the transition to secondary school in part reflect academic work that lacks challenge
and meaning commensurate with adolescents’ cognitive and emotional needs (Eccles
& Midgley, 1989). Recent efforts at middle school reform support this hypothesis: motivation
is maintained when middle schools and junior high schools introduce more
challenging and meaningful academic work (Jackson & Davis, 2000). I discuss this in
more detail later.
Experiences of Racial-Ethnic Discrimination
Researchers interested in the relatively poor academic performance of adolescents
from some ethnic/racial groups have suggested another classroom-based experience as
critical for adolescent development, namely, experiences of racial/ethnic discrimination
(Essed, 1990; Feagin, 1992; Fordham & Ogbu, 1986; Garcia Coll et al., 1996; Rosenbaum,
Kulieke, & Rubinowitz, 1988; Ruggiero & Taylor, 1995; Taylor, Casten, Flickinger,
Roberts, & Fulmore, 1994; Wong, Eccles, & Sameroff, in press). Two types of
discrimination have been discussed: (a) anticipation of future discrimination in the labor
market, which might be seen as undermining the long-term benefits of education
(Fordham & Ogbu, 1986), and (b) the impact of daily experiences of discrimination on
one’s mental health and academic motivation (Essed, 1990; Wong et al., in press). Both
types are likely to influence adolescent development, but research on these issues is in
its infancy. Wong et al. (in press) found that anticipated future discrimination leads to
increases in African American youth’s motivation to do well in school, which in turn
leads to increases in academic performance. In this sample, anticipated future discrimination
appeared to motivate the youth to do their very best so that they would be maximally
equipped to deal with future discrimination. In contrast, daily experiences of
racial discrimination from their peers and teachers led to declines in school engagement
and confidence in one’s academic competence and grades, along with increases in
depression and anger.
Level 1: Summary
The studies of classroom-level influences suggest that development is optimized when
students are provided with challenging tasks in a mastery-oriented environment that
also provides good emotional and cognitive support, meaningful material to learn and
master, and sufficient support for their own autonomy and initiative. Connell and Well-
Level 1: Classrooms 133
born (1991), as well as Deci and Ryan (1985), suggested that humans have three basic
needs: to feel competent, to feel socially attached, and to have autonomous control in
their lives. Further, they hypothesized that individuals develop best in contexts that
provide opportunities for each of these needs to be met. Clearly, the types of classroom
characteristics that emerge as important for both socioemotional and intellectual development
would provide such opportunities.
LEVEL 2: SCHOOL BUILDINGS
Schools are formal organizations and, as such, have characteristics and features that
are superordinate to classroom characteristics. These aspects of the whole school environment
should impact on adolescents’ intellectual, social-emotional, and behavioral
development. Important school-level organizational features include school climate
and sense of community (Goodenow, 1993; Rutter, 1983) and the relationships among
the students themselves. School organizational features also include such schoolwide
practices as curricular tracking, start and stop times, and the availability of extracurricular
activities.
General Social Climate
Researchers have become interested in the social climate of the entire school. These researchers
suggest that schools vary in the climate and general expectations regarding
student potential and that such variations affect the development of both teachers and
students in very fundamental ways (e.g., Bandura, 1994; Bryk, Lee, & Holland, 1993;
Mac Iver, Reuman, & Main, 1995; Rosenbaum et al., 1988; Rutter, Maughan, Mortimore,
& Ouston, 1979). For example, in their analysis of higher achievement in Catholic
schools, Bryk et al. (1993) discussed how the culture within Catholic schools is fundamentally
different from the culture within most public schools in ways that positively
affect the motivation of students, parents, and teachers. This culture (school climate)
values academics, has high expectations that all students can learn, and affirms the belief
that the business of school is learning. Similarly, Lee and Smith (2001) showed that
between-school differences in teachers’ sense of their own personal efficacy as well as
their confidence in the general ability of the teachers at their school to teach all students
accounted, in part, for between-school differences in adolescents’ high school performance
and motivation. Finally, Bandura (1994) documented between-school differences
in the general level of teachers’ personal efficacy beliefs and argued that these differences
translate into teaching practices that undermine the motivation of many
students and teachers in the school.
Maehr, Midgley, and their colleagues argued that just as classroom practices give rise
to certain achievement goals, so too do schools through particular policies and practices.
A school-level emphasis on different achievement goals creates a schoolwide psychological
environment that affects students’ academic beliefs, affects, and behaviors
(e.g., Maehr & Midgley, 1996; Roeser et al., 1996). For example, schools’ use of public
honor rolls and assemblies for the highest achieving students, class rankings on report
cards, differential curricular offerings for students of various ability levels, and so on
134 Schools, Academic Motivation, and Stage-Environment Fit
all emphasize relative ability, competition, and social comparison in the school and create
a school-level ability rather than mastery/task focus. In contrast, through the recognition
of academic effort and improvement, rewards for different competencies that extend
to all students, and through practices that emphasize learning and task mastery
(block scheduling, interdisciplinary curricular teams, cooperative learning), schools
can promote a school-level focus on discovery, effort and improvement, and academic
mastery. The academic goal focus of a school also has important implications for students’
mental health. In a series of studies, Roeser and Eccles found that students’ belief
that their school is ability-focused leads to declines in students’ educational values,
achievement, and self-esteem and increases in their anger, depressive symptoms, and
school truancy as they move from seventh to eighth grade (Roeser & Eccles, 1998;
Roeser et al., 1998). Fiqueira-McDonough (1986) reported similar findings in a study
of two high schools that were similar in intake characteristics and achievement outcomes
but differed in their academic orientation and rates of delinquent behavior. The
high school characterized by a greater emphasis on competition and high grades (ability
orientation) had higher delinquency rates, and the students’ grades were a major
correlate of students’ involvement in delinquent behavior (low grades predicted increased
delinquent behavior). In contrast, in the school that had more diverse goals and
greater interest in non-academic needs, school attachment (valuing of school, liking
teachers) was greater on average, and those students with high school attachment engaged
in the least delinquent activity.
One final note on school-level academic goal emphases: They are strongly correlated
with adolescents’ perceptions of the school’s social climate. Adolescents who perceive
a task orientation in their school also report that their teachers are friendly, caring, and
respectful. These factors in turn predict an increased sense of belonging in school among
adolescents (see also Goodenow, 1993). In contrast, perceptions of a schoolwide ability
orientation are negatively correlated with adolescents’ perceptions of caring teachers
(Roeser et al., 1996). From the adolescents’ perspective, a deemphasis on comparison
and competition and an emphasis on effort and improvement are intertwined with
their view of caring teachers.
Academic Tracks and Curricular Differentiation
Another important school-level feature relates to academic tracks or curriculum differentiation
policies. These terms refer to the regularities in the ways in which schools
structure the learning experiences for different types of students (Oakes, Gamoran, &
Page, 1992). The practice of providing different educational experiences for students of
different ability levels is widespread in American schools. Tracking takes different
forms at different grade levels. It includes within-class ability grouping for different
subject matters or between-class ability grouping in which different types of students
are assigned to different teachers. Within-classroom ability grouping for reading and
math is quite common in elementary school. In secondary school, between-class tracking
is more widespread and is often linked to the sequencing of specific courses for students
bound for different post–secondary school trajectories (e.g., the college prep,
general, or vocational tracks). Differentiated curricular experiences for students of different
ability levels influence school experiences in two major ways: First, tracking de-
Level 2: School Buildings 135
termines the quality and kinds of instruction each student receives (Rosenbaum, 1976,
1980; Oakes et al., 1992), and second, it determines exposure to different peers and
thus, to a certain degree, the nature of social relationships that youth form in school
(Fuligni, Eccles, & Barber, 1995).
Despite years of research on the impact of tracking practices, few strong and definitive
answers have emerged (see Fuligni et al., 1995; Gamoran & Mare, 1989; Kulik &
Kulik, 1987; Slavin, 1990). The results vary depending on the outcome assessed, the
group studied, the length of the study, the control groups used for comparison, and the
specific nature of the context in which these practices are manifest. The best justification
for these practices, derived from a person-environment fit perspective, is the belief
that students are more motivated to learn if the material is adapted to their current
competence level. There is some evidence to support this view for students placed in
high ability and gifted classrooms, high within-class ability groups, and college tracks
(Dreeben & Barr, 1988; Fuligni et al., 1995; Gamoran & Mare, 1989; Kulik & Kulik,
1987; Pallas, Entwisle, Alexander, & Stluka, 1994).
The results for adolescents placed in low-ability and noncollege tracks are usually
inconsistent with this hypothesis. By and large, the effects found for this group of students
are negative (Dreeben & Barr, 1988; Pallas et al., 1994; Rosenbaum, 1976, 1980;
Rosenbaum et al., 1988; Vanfossen, Jones, & Spade, 1987). Low-track placement predicts
poor attitudes toward school, feelings of incompetence, and problem behaviors
both within school (nonattendance, crime, misconduct) and in the broader community
(drug use, arrests); it also predicts lower educational attainments (Oakes et al., 1992).
These negative effects reflect the fact that students placed in the lower tracks are often
provided with inferior educational experience and support.
Ability grouping also has an impact on students’ peer groups: Between-classroom
ability grouping and curricular tracking increase the extent of contact among adolescents
with similar levels of achievement and engagement with school. For those doing
poorly in school, tracking is likely to facilitate friendships among students who are similarly
alienated from school and are more likely to engage in risky or delinquent behaviors
(Dryfoos, 1990). Dishion, McCord, and Poulin (1999) showed experimentally how
such collecting of alienated adolescents increases their involvement in problem behaviors.
This collecting of adolescents with poor achievement or adjustment histories also
places additional discipline burdens on the teachers who teach these classes (Oakes et
al., 1992), making such classes unpopular with the teachers as well as the students and
decreasing the likelihood that the teachers with the most experience will allow themselves
to be assigned to these classes.
Concerns have also been raised about the way students get placed in different classes
and how difficult it is for students to change tracks once initial placements have been
made. These issues are important both early in a child’s school career (e.g., Entwisle &
Alexander, 1993) and later in adolescence, when course placement is linked directly to
the kinds of educational options that are available to the student after high school. Minority
youth, particularly African American and Latino boys, are more likely to be assigned
to low-ability classes and non-college-bound curricular tracks than are other
groups; furthermore, careful assessment of the placements has shown that many of
these youth were incorrectly assigned to these classes (Dornbusch, 1994; Oakes et al.,
136 Schools, Academic Motivation, and Stage-Environment Fit
1992). The consequences of such misassignment are great. It has long-term consequences
for students’ ability to go to college once they complete secondary school.
children and adolescents spend more time in schools than in any other place outside
their homes. Exploring all of the possible ways in which educational institutions influence
motivation and development during adolescence is beyond the scope of a single
chapter. In this chapter I discuss the ways in which schools influence adolescents’ socialemotional
and behavioral development through organizational, social, and instructional
processes ranging from those based in the immediate, proximal relation between
students and the tasks they are asked to perform to the role that principals and the
school boards play in setting school-level and district-level policies, which in turn influence
the social organization of the entire school community. I discuss at length three
examples of the ways in which these multiple organizational levels interact synergistically
to influence adolescent development through their impact on the daily experiences
that adolescents in the United States encounter as they move through the American
school system. The first example focuses on the role of school transitions, the
second on the role of curricular tracking, and the third on extracurricular activities.
Few of these processes have been studied in countries other than the United States. I
assume similar processes are true in other countries, but this remains to be demonstrated
empirically.
A DEVELOPMENTAL VIEW OF THE IMPACT OF SCHOOLS
ON DEVELOPMENT
Understanding the impact of schools on adolescent development requires a conceptual
framework for thinking simultaneously about schools as contexts in which development
takes place and about the changing developmental needs of students as they move
through the school system. My colleagues and I have been working on such a framework
for the last 20 years. In the late 1980s Carol Midgley and I proposed our model of stageenvironment
fit to guide research on the impact of school transitions on adolescent development
(see Eccles & Midgley, 1989; Eccles et al., 1993). We argued that individuals
have changing emotional, cognitive, and social needs and personal goals as they mature.
Drawing on ideas related to person-environment fit and self-determination theory
(e.g., Deci & Ryan, 1985), we argued that schools need to change in developmentally
125
appropriate ways if they are to provide the kind of social context that will continue to
motivate students’ interest and engagement as the students mature. To the extent that
this does not happen, we predicted that students would disengage first psychologically
and then physically from school as they matured into and through adolescence. This
should be particularly true as the adolescents acquired more incentives and more power
to control their own behavior. I say more about both of these psychological perspectives
on the impact of classroom experiences later.
More recently, Robert Roeser and I (see Eccles & Roeser, 1999) proposed a framework
for thinking about school influences that dissected the school context into a series
of hierarchically ordered, interdependent levels of organization beginning at the most
basic level of the classroom and then moving up in complexity to the school as an organizational
system embedded in a larger cultural system. In adopting this heuristic, we
assumed that (a) schools are systems characterized by multiple levels of regulatory processes
(organizational, social, and instructional in nature); (b) these processes are interrelated
across levels of analysis; (c) such processes are usually dynamic in nature,
sometimes being worked out each day between the various social actors (e.g., teachers
and students); (d) these processes change as children move through different school levels
(elementary, middle, and high school); and (e) these processes regulate children’s and
adolescents’ cognitive, social-emotional, and behavioral development. In this chapter
I focus on the interface between these theoretical frameworks. I begin with a summary
of Eccles and Roeser’s multilevel description of school contexts.
AN ECOLOGICAL VIEW OF SCHOOLS AND THEIR IMPACT ON
DEVELOPMENT DURING ADOLESCENCE
From the location of the school within macroregulatory systems characterized by national,
state, and school district laws and educational policies to the miniregulatory systems
that involve the minute-to-minute interactions between teachers and individual
students, schools are a system of complex, multilevel, regulatory processes. Eccles and
Roeser (1999) described these different levels of the school environment in terms of
their hierarchical ordering—moving from the student in a classroom, to the school
building itself, then to the school district, and finally to the larger communities in which
school districts are located. Within each of these levels, we discussed those beliefs and
practices that affect students’ experiences on a daily basis. At the classroom level, we
focused attention on teacher beliefs and instructional practices, teacher-student relationships,
the nature and design of tasks and instruction, and the nature and structure
of classroom activities and groups. At the level of the school building, we focused attention
on organizational climate and such schoolwide practices as academic tracking,
school start time, and the provision of extracurricular activities. At the level of the
school district, we focused on the between-school grade configurations that create particular
school-transition experiences for students. Finally at the level of schools embedded
in larger social systems, we discussed such issues as school resources, as well as
the linkages of schools with parents and with the labor market.
Eccles and Roeser (1999) further assumed that in any given school setting these multilevel
processes are interwoven with one another. Relations between different levels of
126 Schools, Academic Motivation, and Stage-Environment Fit
organization in the school may be complementary or contradictory and may influence
students either directly or indirectly. For instance, a principal may decide that all of his
or her teachers should use a particular practice such as cooperative learning. However,
the impact of such a decision on the daily experiences of students depends on how well
this practice is actually implemented at the classroom level. If done well, students
should be seen working successfully in groups on complex, authentic problems. Such a
well-implemented school policy is likely to produce gains in self-esteem, interethnic relationships,
and achievement among students, especially those of low ability or status
(Slavin, 1990). In contrast, if done poorly, chaos can result, leading to far less positive
outcomes at the student level. How such a schoolwide instructional policy is implemented
depends on many factors including the morale within the school, the relationships
between the principal and the teachers, the teachers’ understanding and endorsement
of the new instructional practice, the way in which the policy change was
decided upon, the provision of adequate in-service training, the provision of adequate
supports for implementation of new strategies, and the students’ willingness to go
along with the new practice. Recent debates about the likely impact of national standards
testing provide another example of the complex ways in which a new policy—this
time a state- or national-level policy—can affect the daily experiences of teachers and
students in the classroom and in the school building.
Eccles and Roeser (1999) also assumed that the processes associated with the different
levels of school interacting dynamically with each other, rather than static resources
or characteristics of the curriculum, teachers, or school per se, influence adolescents’
development. In addition, adolescents’ own constructions of meaning and interpretations
of events within the school environment are critical mediators between school
characteristics and students’ feelings, beliefs, and behavior.
Finally, in keeping with the stage-environment perspective proposed by Eccles and
Midgely (1989), Eccles and Roeser (1999) assumed that these different school-related
processes change across the course of children’s and adolescents’ development as they
progress through elementary, middle, and high school. That is, not only are children
and adolescents developing, but so too is the whole nature of the schools that they attend.
For example, the organizational, social, and instructional processes in schools
change as children move from elementary to middle school. Eccles and Midgley argued
that these changes are often associated with declines in many adolescents’ motivation
and behavior. Understanding the interaction of different school features with the developmental
needs of adolescents is critical to understanding the role of schooling in
young people’s development (see Eccles & Midgley, 1989). In the next sections I discuss
those characteristics of each level likely to be most important for understanding the impact
of schools on adolescent development. I also discuss how school characteristics at
each level may also influence group differences in adolescent development, paying particular
attention to gender and ethnic group differences within the United States.
LEVEL 1: CLASSROOMS
The most immediate educational environment to the student is the classroom. This is
also the level that has received the most attention from educational psychologists. In
Level 1: Classrooms 127
this section I review what we know about teacher beliefs, classroom climate, the nature
of the academic work itself, and experiences of racial/ethnic discrimination.
Teacher Beliefs
Teacher beliefs have received much attention in educational psychology. In this section
I focus on two types of beliefs: Teachers’ general sense of their own teaching efficacy
and teachers’ expectations for specific students in their class.
Teachers’ General Sense of Efficacy
When teachers hold high general expectations for student achievement and students
perceive these expectations, students learn more, experience a greater sense of self-worth
and competence as learners, feel more connected to their teacher and their school, and
resist involvement in problem behaviors (Eccles et al. 1993; Lee & Smith, 2001; Roeser,
Eccles, & Sameroff, 1998; Rutter, 1983; Weinstein, 1989). Similarly, teachers who feel
they are able to reach even the most difficult students and who believe in their ability to
affect students’ lives communicate such positive expectations and beliefs to their students.
Thus, a high sense of general teacher efficacy can enhance students’ own confidence
in their ability to master academic material, thereby promoting effort investment
and achievement as well as a positive emotional relationship with their teacher and
greater engagement in school as a social institution (Ashton, 1985; Midgley, Feldlaufer,
& Eccles, 1989b). Alternatively, teachers who have low confidence in their teaching efficacy
often engage in behaviors that reinforce feelings of incompetence and alienation
in their students, increasing the likelihood that their students will develop learned helpless
responses to failure in the classroom, depressive affect, anger, and disengagement
(see Cole, 1991; Roeser, Eccles, & Freedman-Doan, 1999). Lee and Smith (2001) stressed
this aspect of teachers’ general beliefs as a critical component for secondary school reform
(see also Jackson & Davis, 2000).
As I discuss in more detail later, the prevalence of teachers with a low sense of personal
teaching efficacy is higher in junior high and middle schools than in elementary
schools and higher in schools that serve high proportions of ethnic minority and poor
adolescents than in schools that serve more affluent and higher achieving adolescents
(Darling-Hammond, 1997; Eccles, Wigfield, & Schiefele, 1998). This fact alone provides
a possible explanation for both average levels of declining school engagement
during early to middle adolescence and for social class and ethnic group differences in
school engagement.
Differential Teacher Expectations
Equally important are the differential expectations teachers often hold for various individuals
within the same classroom and the differential treatments that sometimes accompany
these expectations. Beginning with the work by Rosenthal (1969), many researchers
have shown that undermining teacher-expectancy effects depend on how
teachers structure activities differently, as well as interact differently with, high- and lowexpectancy
students and on how the students perceive these differences (Brophy, 1985;
Cooper, 1979; Eccles & Wigfield, 1985; Weinstein, 1989). Most concerns have been
128 Schools, Academic Motivation, and Stage-Environment Fit
raised over behaviors that create a self-fulfilling prophecy by undermining the learning
and well-being of those students for whom the teachers hold the lowest expectations.
Much work on teacher expectancy effects has focused on differential treatment related
to gender, race/ethnic group, and/or social class. Most of this work has documented
the small but fairly consistent undermining effects of low teacher expectations
on girls (for math and science), on minority children (for all subject areas), and on children
from lower-social-class family backgrounds (again for all subject areas) (see Eccles
& Wigfield, 1985; Ferguson, 1998; Jussim, Eccles, & Madon, 1996; Valencia, 1991).
In addition, Jussim et al. (1996) found that even though these effects are typically quite
small, young women, African American adolescents, and students from poorer homes
are more subject to both the positive and negative teacher expectancy effects than are
other students.
Researchers such as Steele (1992) have linked this form of differential treatment,
particularly for African American students, to school disengagement and disidentification
(the separation of one’s self-esteem from all forms of school-related feedback).
Steele argued that African American students become aware of the fact that teachers
and other adults have negative stereotypes of African Americans’ academic abilities.
This awareness (labeled stereotype threat by Steele and his colleagues) increases their
anxieties, which in turn lead them to disidentify with the school context to protect their
self-esteem. It is interesting that recent studies using the same theoretical notions and
experimental techniques have shown that Asian students believe that teachers and
adults expect them to perform very well and that this belief leads Asian students to perform
better on tests when their ethnic identity is made salient (Shih, Pittinsky, & Ambady,
1999). Thus, the psychological processes associated with stereotype threat can
either undermine or facilitate performance on standardized tests depending on the nature
of commonly held stereotypes about the intellectual strengths and weaknesses of
different social groups.
Classroom Climate
Classroom climate refers to the more general character of the classroom and teacherstudent
relationships within the classroom. In this section I focus on the following aspects
of classroom climate: Teacher-student relationships, classroom management, and
motivational climate.
Teacher-Student Relationships
Teacher-student relationships are a key component of classroom climate: High-quality
teacher-student relationships facilitate academic motivation, school engagement, academic
success, self-esteem, and more general socioemotional well-being (Deci & Ryan,
1985; Eccles et al., 1998; Goodenow, 1993; Midgley et al., 1989b; Roeser, Midgley, &
Urdan, 1996). Teachers who trust, care about, and are respectful of students provide
the social-emotional support that students need to approach, engage, and persist on
academic learning tasks and to develop positive achievement-related self-perceptions
and values. Feeling emotionally supported is one of the most important characteristics
of contexts that support positive development. Correlational studies with adoles-
Level 1: Classrooms 129
cents show that students’ perceptions of caring teachers enhance their feelings of selfesteem,
school belonging, and positive affect in school (Roeser & Eccles, 1998; Roeser
et al., 1996).
Declines in both adolescents’ perception of emotional support from their teachers
and in the adolescents’ sense of belonging in their classrooms are quite common as adolescents
move from elementary school into secondary schools (Eccles et al., 1998). This
shift is particularly troublesome in our highly mobile society in which teachers represent
one of the last stable sources of nonparental role models for adolescents. In addition
to teaching, teachers in mobile societies such as the United States can provide
guidance and assistance when social-emotional or academic problems arise. This role
is especially important for promoting developmental competence when conditions in
the family and neighborhood cannot or do not provide such supports (Eccles, Lord, &
Roeser, 1996; Simmons & Blyth, 1987).
Classroom Management
Work related to classroom management has focused on two general issues: orderliness/
predictability and control/autonomy. With regard to orderliness and predictability, the
evidence is quite clear: Student achievement and conduct are enhanced when teachers
establish smoothly running and efficient procedures for monitoring student progress,
providing feedback, enforcing accountability for work completion, and organizing
group activities (e.g., Eccles et al., 1998; Pintrich & Schunk, 1996). Unfortunately, such
conditions are often absent, particularly in highly stressed and underfunded schools
with inexperienced teachers (Darling-Hammond, 1997).
In addition, research on international comparisons of instruction suggest that
American teachers are often more lax in their classroom management and provide less
systematic and rigorous control over the instructional sequences (Stevenson & Stigler,
1992). Furthermore, this research suggests that these differences in teachers’ controlrelated
practices could be a partial explanation for the relatively poor performance of
many American youth on international standardized tests of math and science achievement
(Schmidt, McKnight, & Raizen, 1997).
Motivational Climate
Several teams of researchers have suggested that teachers engage in a wide range of behaviors
that create a pervasive motivational climate in the classroom. For example,
Rosenholtz and Simpson (1984) suggested a cluster of general teaching practices (e.g.,
individualized vs. whole-group instruction, ability grouping practices, and publicness
of feedback) that should affect motivation because these practices make ability differences
in classrooms especially salient to students (see Mac Iver, 1988). They assumed
that these practices affect the motivation of all students by increasing the salience of extrinsic
motivators and ego-focused learning goals, leading to greater incidence of social
comparison behaviors and increased perception of ability as an entity state rather than
an incremental condition. All of these changes reduce the quality of students’ motivation
and learning. The magnitude of the negative consequences of these shifts, however,
should be greatest for low-performing students: As these students become more aware
of their relative low standing, they are likely to adopt a variety of ego-protective strategies
that unfortunately undermine learning and mastery (Covington, 1992).
130 Schools, Academic Motivation, and Stage-Environment Fit
More recently, researchers interested in goal theory have proposed a similar set of
classroom characteristics (Ames, 1992; E. M. Anderman, & Maehr, 1994; Maehr &
Midgley, 1996; Pintrich & Schunk, 1996; Roeser, Midgley, & Maehr, 1994). Goal theorists
propose two major achievement goal systems: mastery-oriented goals and performanceoriented
goals. Students with mastery-oriented goals focus on learning the material
and on their own improvement over time. Students with performance-oriented goals
focus on doing better than other students in their class. Goal theorists further argue
that a mastery orientation sustains school engagement and achievement better than
does a performance orientation (see Ames, 1992; Maehr & Midgley, 1996; Midgley,
2002). Evidence is quite strong for the first prediction and more mixed for the second:
The desire to do better than others often has positive rather than negative consequences,
whereas the fear of failing (performance-avoidance goal orientation) undermines school
performance (see Midgley, 2002). Finally, these theorists suggest that the publicness of
feedback, particularly social comparative feedback, and a classroom focus on competition
between students undermine mastery motivation and increase performance motivation.
The school-reform work of Midgley, Maehr, and their colleagues has shown
that social reform efforts to reduce these types of classroom practices, particularly
those associated with performance feedback, social comparative grading systems, and
ego-focused, competitive motivational strategies have positive consequences for adolescents’
academic motivation, persistence on difficult learning tasks, and socioemotional
development (e.g., Maehr & Midgley, 1996).
The work on understanding group differences in achievement and achievement
choices is another example of an attempt to identify a broad set of classroom characteristics
related to motivation. The work on girls and math is one example of this approach.
There are sex differences in adolescents’ preference for different types of learning
contexts that likely interact with subject area to produce sex differences in interest
in different subject areas (Eccles, 1989; Hoffmann & Haeussler, 1995). Females appear
to respond more positively to math and science instruction if it is taught in a cooperative
or individualized manner rather than a competitive manner, if it is taught from an
applied or person-centered perspective rather than a theoretical or abstract perspective,
if it is taught using a hands-on approach rather than a book-learning approach,
and if the teacher avoids sexism in its many subtle forms. The reason given for these effects
is the fit between the teaching style, the instructional focus, and females’ values,
goals, motivational orientations, and learning styles. The few relevant studies support
this hypothesis (Eccles & Harold, 1993; Hoffmann & Haeussler, 1995). If such classroom
practices are more prevalent in one subject area (e.g., physical science or math)
than another (e.g., biological or social science), one would expect sex differences in motivation
to study these subject areas. In addition, however, math and physical science
do not have to be taught in these ways; more girl-friendly instructional approaches can
be used. When they are, girls, as well as boys, are more likely to continue taking courses
in these fields and to consider working in these fields when they become adults.
The girl-friendly classroom conclusion is a good example of person-environment fit.
Many investigators have suggested that students are maximally motivated to learn in
situations that fit well with their interests, current skill level, and psychological needs,
so that the material is challenging, interesting, and meaningful (e.g., Csikszentmihalyi,
Rathunde, & Whalen, 1993; Eccles et al., 1993; Krapp, Hidi, & Renninger, 1992). Vari-
Level 1: Classrooms 131
ations on this theme include aptitude by treatment interactions and theories stressing
cultural match or mismatch as one explanation for group differences in school achievement
and activity choices (e.g., Fordham & Ogbu, 1986; Suarrez-Orozco & Suarrez-
Orozco, 1995; Valencia, 1991). For example, Valencia (1991) concluded that a mismatch
of both the values of the school and the materials being taught contributed to
the poor performance and high dropout rates among Latino youth in the high school
they studied. Deyhle and LeCompte (1999) made a similar argument in their discussion
of the poor performance of Native American youth in traditional middle school contexts.
The misfit between the needs of young adolescents and the nature of junior high
school environments is another example of these person-environment fit dynamics.
The Nature of Academic Work
Academic work is at the heart of the school experience. Two aspects of academic tasks
are important: the content of the curriculum and the design of instruction. The nature
of academic content has an important impact on students’ attention, interest, and cognitive
effort. Long ago, Dewey (1902/1990) proposed that academic work that is meaningful
to the historical and developmental reality of students’ experiences will promote
sustained attention, high investment of cognitive and affective resources in learning,
and strong identification with educational goals and aims. In general, research supports
this hypothesis: Content that provides meaningful exploration is critical given
that boredom in school, low interest, and perceived irrelevance of the curriculum are
associated with poor attention, diminished achievement, disengagement, and alienation
from school (e.g., Finn, 1989; Jackson & Davis, 2000; Larson & Richards, 1989).
Curricula that represent the voices, images, and historical experiences of traditionally
underrepresented groups are also important (Valencia, 1991). Choosing materials that
provide an appropriate level of challenge for a given class, designing learning activities
that require diverse cognitive operations (e.g., opinion, following routines, memory,
comprehension), structuring lessons so that they build on each other in a systematic
fashion, using multiple representations of a given problem, and explicitly teaching students
strategies that assist in learning are but a few of the design features that scaffold
learning and promote effort investment, interest in learning, and achievement (Blumenfeld,
1992; Deci & Ryan, 1985; Eccles et al., 1998).
Unfortunately, American secondary schools have problems providing each of these
types of educational experiences. Recent work by Larson and his colleagues has documented
the fact that adolescents are bored most of the time that they are in secondary
school classrooms (Larson, 2000; Larson & Richards, 1989). Culturally meaningful
learning experiences are rare in many American secondary schools (Fine, 1991; Valencia,
1991). The disconnection of traditional curricula from the experiences of these
groups can explain the alienation of some group members from the educational process,
sometimes eventuating in school dropout (Fine, 1991; Sheets & Hollins, 1999).
Appropriately designed tasks that adequately scaffold learning are also rare in many
inner-city and poor schools (Darling-Hammond, 1997).
In addition, from a developmental perspective, there is evidence that the nature of
academic work too often does not change over time in ways that are concurrent with
the increasing cognitive sophistication, diverse life experiences, and identity needs of
132 Schools, Academic Motivation, and Stage-Environment Fit
adolescents as they move from the elementary into the secondary school years (Carnegie
Council, 1989; Lee & Smith, 2001). As one indication of this, middle school students
report the highest rates of boredom when doing schoolwork, especially passive work
(e.g., listening to lectures) and in particular classes such as social studies, math, and science
(Larson & Richards, 1989). There is also evidence that the content of the curriculum
taught in schools does not broaden to incorporate either important health or social
issues that become increasingly salient as adolescents move through puberty and
deal with the identity explorations associated with adolescence (Carnegie Council,
1989). Further, academic work sometimes becomes less, rather than more, complex in
terms of the cognitive demands as adolescents move from elementary to junior high
school (Eccles et al., 1998). It may be that declines in some adolescents’ motivation during
the transition to secondary school in part reflect academic work that lacks challenge
and meaning commensurate with adolescents’ cognitive and emotional needs (Eccles
& Midgley, 1989). Recent efforts at middle school reform support this hypothesis: motivation
is maintained when middle schools and junior high schools introduce more
challenging and meaningful academic work (Jackson & Davis, 2000). I discuss this in
more detail later.
Experiences of Racial-Ethnic Discrimination
Researchers interested in the relatively poor academic performance of adolescents
from some ethnic/racial groups have suggested another classroom-based experience as
critical for adolescent development, namely, experiences of racial/ethnic discrimination
(Essed, 1990; Feagin, 1992; Fordham & Ogbu, 1986; Garcia Coll et al., 1996; Rosenbaum,
Kulieke, & Rubinowitz, 1988; Ruggiero & Taylor, 1995; Taylor, Casten, Flickinger,
Roberts, & Fulmore, 1994; Wong, Eccles, & Sameroff, in press). Two types of
discrimination have been discussed: (a) anticipation of future discrimination in the labor
market, which might be seen as undermining the long-term benefits of education
(Fordham & Ogbu, 1986), and (b) the impact of daily experiences of discrimination on
one’s mental health and academic motivation (Essed, 1990; Wong et al., in press). Both
types are likely to influence adolescent development, but research on these issues is in
its infancy. Wong et al. (in press) found that anticipated future discrimination leads to
increases in African American youth’s motivation to do well in school, which in turn
leads to increases in academic performance. In this sample, anticipated future discrimination
appeared to motivate the youth to do their very best so that they would be maximally
equipped to deal with future discrimination. In contrast, daily experiences of
racial discrimination from their peers and teachers led to declines in school engagement
and confidence in one’s academic competence and grades, along with increases in
depression and anger.
Level 1: Summary
The studies of classroom-level influences suggest that development is optimized when
students are provided with challenging tasks in a mastery-oriented environment that
also provides good emotional and cognitive support, meaningful material to learn and
master, and sufficient support for their own autonomy and initiative. Connell and Well-
Level 1: Classrooms 133
born (1991), as well as Deci and Ryan (1985), suggested that humans have three basic
needs: to feel competent, to feel socially attached, and to have autonomous control in
their lives. Further, they hypothesized that individuals develop best in contexts that
provide opportunities for each of these needs to be met. Clearly, the types of classroom
characteristics that emerge as important for both socioemotional and intellectual development
would provide such opportunities.
LEVEL 2: SCHOOL BUILDINGS
Schools are formal organizations and, as such, have characteristics and features that
are superordinate to classroom characteristics. These aspects of the whole school environment
should impact on adolescents’ intellectual, social-emotional, and behavioral
development. Important school-level organizational features include school climate
and sense of community (Goodenow, 1993; Rutter, 1983) and the relationships among
the students themselves. School organizational features also include such schoolwide
practices as curricular tracking, start and stop times, and the availability of extracurricular
activities.
General Social Climate
Researchers have become interested in the social climate of the entire school. These researchers
suggest that schools vary in the climate and general expectations regarding
student potential and that such variations affect the development of both teachers and
students in very fundamental ways (e.g., Bandura, 1994; Bryk, Lee, & Holland, 1993;
Mac Iver, Reuman, & Main, 1995; Rosenbaum et al., 1988; Rutter, Maughan, Mortimore,
& Ouston, 1979). For example, in their analysis of higher achievement in Catholic
schools, Bryk et al. (1993) discussed how the culture within Catholic schools is fundamentally
different from the culture within most public schools in ways that positively
affect the motivation of students, parents, and teachers. This culture (school climate)
values academics, has high expectations that all students can learn, and affirms the belief
that the business of school is learning. Similarly, Lee and Smith (2001) showed that
between-school differences in teachers’ sense of their own personal efficacy as well as
their confidence in the general ability of the teachers at their school to teach all students
accounted, in part, for between-school differences in adolescents’ high school performance
and motivation. Finally, Bandura (1994) documented between-school differences
in the general level of teachers’ personal efficacy beliefs and argued that these differences
translate into teaching practices that undermine the motivation of many
students and teachers in the school.
Maehr, Midgley, and their colleagues argued that just as classroom practices give rise
to certain achievement goals, so too do schools through particular policies and practices.
A school-level emphasis on different achievement goals creates a schoolwide psychological
environment that affects students’ academic beliefs, affects, and behaviors
(e.g., Maehr & Midgley, 1996; Roeser et al., 1996). For example, schools’ use of public
honor rolls and assemblies for the highest achieving students, class rankings on report
cards, differential curricular offerings for students of various ability levels, and so on
134 Schools, Academic Motivation, and Stage-Environment Fit
all emphasize relative ability, competition, and social comparison in the school and create
a school-level ability rather than mastery/task focus. In contrast, through the recognition
of academic effort and improvement, rewards for different competencies that extend
to all students, and through practices that emphasize learning and task mastery
(block scheduling, interdisciplinary curricular teams, cooperative learning), schools
can promote a school-level focus on discovery, effort and improvement, and academic
mastery. The academic goal focus of a school also has important implications for students’
mental health. In a series of studies, Roeser and Eccles found that students’ belief
that their school is ability-focused leads to declines in students’ educational values,
achievement, and self-esteem and increases in their anger, depressive symptoms, and
school truancy as they move from seventh to eighth grade (Roeser & Eccles, 1998;
Roeser et al., 1998). Fiqueira-McDonough (1986) reported similar findings in a study
of two high schools that were similar in intake characteristics and achievement outcomes
but differed in their academic orientation and rates of delinquent behavior. The
high school characterized by a greater emphasis on competition and high grades (ability
orientation) had higher delinquency rates, and the students’ grades were a major
correlate of students’ involvement in delinquent behavior (low grades predicted increased
delinquent behavior). In contrast, in the school that had more diverse goals and
greater interest in non-academic needs, school attachment (valuing of school, liking
teachers) was greater on average, and those students with high school attachment engaged
in the least delinquent activity.
One final note on school-level academic goal emphases: They are strongly correlated
with adolescents’ perceptions of the school’s social climate. Adolescents who perceive
a task orientation in their school also report that their teachers are friendly, caring, and
respectful. These factors in turn predict an increased sense of belonging in school among
adolescents (see also Goodenow, 1993). In contrast, perceptions of a schoolwide ability
orientation are negatively correlated with adolescents’ perceptions of caring teachers
(Roeser et al., 1996). From the adolescents’ perspective, a deemphasis on comparison
and competition and an emphasis on effort and improvement are intertwined with
their view of caring teachers.
Academic Tracks and Curricular Differentiation
Another important school-level feature relates to academic tracks or curriculum differentiation
policies. These terms refer to the regularities in the ways in which schools
structure the learning experiences for different types of students (Oakes, Gamoran, &
Page, 1992). The practice of providing different educational experiences for students of
different ability levels is widespread in American schools. Tracking takes different
forms at different grade levels. It includes within-class ability grouping for different
subject matters or between-class ability grouping in which different types of students
are assigned to different teachers. Within-classroom ability grouping for reading and
math is quite common in elementary school. In secondary school, between-class tracking
is more widespread and is often linked to the sequencing of specific courses for students
bound for different post–secondary school trajectories (e.g., the college prep,
general, or vocational tracks). Differentiated curricular experiences for students of different
ability levels influence school experiences in two major ways: First, tracking de-
Level 2: School Buildings 135
termines the quality and kinds of instruction each student receives (Rosenbaum, 1976,
1980; Oakes et al., 1992), and second, it determines exposure to different peers and
thus, to a certain degree, the nature of social relationships that youth form in school
(Fuligni, Eccles, & Barber, 1995).
Despite years of research on the impact of tracking practices, few strong and definitive
answers have emerged (see Fuligni et al., 1995; Gamoran & Mare, 1989; Kulik &
Kulik, 1987; Slavin, 1990). The results vary depending on the outcome assessed, the
group studied, the length of the study, the control groups used for comparison, and the
specific nature of the context in which these practices are manifest. The best justification
for these practices, derived from a person-environment fit perspective, is the belief
that students are more motivated to learn if the material is adapted to their current
competence level. There is some evidence to support this view for students placed in
high ability and gifted classrooms, high within-class ability groups, and college tracks
(Dreeben & Barr, 1988; Fuligni et al., 1995; Gamoran & Mare, 1989; Kulik & Kulik,
1987; Pallas, Entwisle, Alexander, & Stluka, 1994).
The results for adolescents placed in low-ability and noncollege tracks are usually
inconsistent with this hypothesis. By and large, the effects found for this group of students
are negative (Dreeben & Barr, 1988; Pallas et al., 1994; Rosenbaum, 1976, 1980;
Rosenbaum et al., 1988; Vanfossen, Jones, & Spade, 1987). Low-track placement predicts
poor attitudes toward school, feelings of incompetence, and problem behaviors
both within school (nonattendance, crime, misconduct) and in the broader community
(drug use, arrests); it also predicts lower educational attainments (Oakes et al., 1992).
These negative effects reflect the fact that students placed in the lower tracks are often
provided with inferior educational experience and support.
Ability grouping also has an impact on students’ peer groups: Between-classroom
ability grouping and curricular tracking increase the extent of contact among adolescents
with similar levels of achievement and engagement with school. For those doing
poorly in school, tracking is likely to facilitate friendships among students who are similarly
alienated from school and are more likely to engage in risky or delinquent behaviors
(Dryfoos, 1990). Dishion, McCord, and Poulin (1999) showed experimentally how
such collecting of alienated adolescents increases their involvement in problem behaviors.
This collecting of adolescents with poor achievement or adjustment histories also
places additional discipline burdens on the teachers who teach these classes (Oakes et
al., 1992), making such classes unpopular with the teachers as well as the students and
decreasing the likelihood that the teachers with the most experience will allow themselves
to be assigned to these classes.
Concerns have also been raised about the way students get placed in different classes
and how difficult it is for students to change tracks once initial placements have been
made. These issues are important both early in a child’s school career (e.g., Entwisle &
Alexander, 1993) and later in adolescence, when course placement is linked directly to
the kinds of educational options that are available to the student after high school. Minority
youth, particularly African American and Latino boys, are more likely to be assigned
to low-ability classes and non-college-bound curricular tracks than are other
groups; furthermore, careful assessment of the placements has shown that many of
these youth were incorrectly assigned to these classes (Dornbusch, 1994; Oakes et al.,
136 Schools, Academic Motivation, and Stage-Environment Fit
1992). The consequences of such misassignment are great. It has long-term consequences
for students’ ability to go to college once they complete secondary school.
Adolescents’ socialization
Adolescents’ socialization into adulthood and related self-development were conceptualized
in terms of four mechanisms: First, it was assumed that the age-graded developmental
tasks, role transitions, and institutional tracks define an opportunity space
that channels young people’s future-oriented motivation, thinking, and behavior. Second,
the kinds of motives and personal goals adolescents construct, and the ways in
which they explore, plan, construct strategies, and enter into commitments were assumed
to be responsible for the ways in which adolescents direct their future development
and select their developmental environments. Third, as a consequence of their efforts,
adolescents attain outcomes, either successes or failures, which requires them to
adjust their previous efforts in terms of goal reconstruction, coping, and the use of self-
110 Socialization and Self-Development
protective causal attributions. Finally, after ending up in a particular social position
and related life situation, adolescents construct reflections and tell stories about who
they are.
Socialization and Self-Development
A review of previous research suggested, first, that adolescents have relatively detailed
conceptions of their age-graded developmental environments (i.e., the timing and sequential
structure of the transitions and tracks they are facing in the future). It was
therefore no surprise that such age-graded structures channel adolescents’ personal
goals and interests: Young people’s future hopes and interests were found to focus typically
on the major developmental tasks of their own age period. Young people also
continuously reconstruct their personal goals to match with the specific stages of a particular
transition through which they are going, as well as the institutional tracks in
which they are involved.
Both the personal goals adolescents have and the cognitive strategies they deploy,
which were assumed to be the major mechanisms in the selection process, were found
to contribute to the developmental trajectories they face subsequently, as well as how
well they are able to deal with the related challenges and demands. Clear evidence was
found from longitudinal studies that adolescents’ motives and personal goals predict
how their lives will proceed in educational, occupational, and family-related trajectories.
Similarly, the kinds of plans and strategies that adolescents apply have consequences
for their success in dealing with major challenges at school, at work, and also
in interpersonal relationships. However, adolescents become interested in forthcoming
developmental tasks and transitions as they grow older, and the tools they have for
dealing with these demands and challenges develop rapidly during early adolescence in
particular. Although the majority of adolescents deploy adaptive strategies, such as optimistic
and task-focused patterns, some of them deploy avoidant strategies as a way to
deal with a fear of failure or anxiety.
This review showed also that adolescents whose goals focus on major age-graded developmental
tasks have higher well-being than do those who have other kinds of goals,
perhaps because such goals help them to deal with the major demands and challenges
they are facing. Although it has been assumed that thinking of self-related issues is a
part of adolescents’ lives, strong evidence was found that self-focused, existential type
of goals are detrimental to young people’s well-being. Moreover, the deployment of
adaptive strategies led not only to higher levels of success in academic and interpersonal
domains of life but also, in the longer run, to higher well-being.
There is also considerable evidence that parents and their adolescent children share
similar kinds of goals concerning the adolescent’s future. Moreover, positive and authoritative
parenting is associated not only with adolescents’ high level of interest in
major developmental tasks, such as education and occupation, but also with adolescents’
use of adaptive strategies, particularly in achievement contexts.
After adolescents have received feedback about the outcomes of their efforts to deal
with the major developmental challenges and demands, they have to adjust their previous
efforts in terms of coping, reconstruction of goals, and making causal attributions.
Conclusions 111
Surprisingly, much less research has been conducted on the antecedents and consequences
of this adjustment compared with those of selection processes. Research on
coping showed that problem-focused coping and engagement coping are associated
with higher levels of psychological adjustment, whereas emotion-focused coping seems
to lead to maladjustment. However, there was little evidence that coping has clear
consequences for individual success in dealing with particular kinds of tasks. Authoritative
and positive parenting was shown to be associated with problem-focused coping,
whereas more negative parenting is related to emotion-focused coping.
Very little research has been carried out on how adolescents reconstruct their personal
goals based on their previous successes and failures. The few studies suggest that
adolescents reconstruct their goals on the basis of the feedback they receive concerning
goal attainment and that such goal reconstruction contributes to their well-being.
Similarly, little research has been carried out on the role that causal attributions have
in the situations in which young people have had problems in dealing with previous demands.
The few studies that exist show that problems in dealing with major transitions
decrease the use of self-protective causal attributions, which then leads to an increase
in depressive symptoms. Dysfunctional causal attributions also lower adolescents’ active
engagement in school activities, and subsequently their academic achievement.
It was also assumed that entrance into certain roles and social positions has consequences
for the identities or self-concepts that adolescents construct. Although the
studies suggest that younger adolescents more frequently report less developed identity
statuses than do older adolescents, relatively little is know about the developmental antecedents
or consequences of these developments.
By contrast, we know much, on a descriptive level, about how self-concept and
self-esteem develop during adolescence. However, some of the recent findings have
challenged previous theories by suggesting a more dynamic view according to which
adolescents’ self-concepts fluctuate significantly and follow, in many cases, individual
developmental trajectories. This fluctuation has been found to reflect many changes in
the individual’s development environments, such as school transitions, grades, and a
variety of stressful life events.
Socialization in Place: Different Developmental Environments
In this chapter adolescent socialization and self-development were conceptualized in
terms of four mechanisms that are responsible for the transaction between the developing
adolescent, on the one hand, and his or her age-graded sociocultural environment,
on the other. It can also be assumed that the substantial amount of variation
across societies and cultures in the developmental environments in which adolescents
grow up (Brown, Larson, & Saraswathi, 2002; Hurrelmann, 1994) channels their subsequent
development in many ways. One key factor that contributes to this variation is
education: There are many differences in the educational systems that are reflected in
adolescents’ thinking and lives across the world (Hurrelmann, 1994; Nurmi, Seginer, et
al., 1995; Scnabel, Alfeld, Eccles, Köller, & Baumert, 2002). For example, in many European
countries and the United States, streaming in education based on academic
achievement begins relatively early (Hurrelmann, 1994), which also influences adolescents’
subsequent opportunities. In some other societies, such as Scandinavian coun-
112 Socialization and Self-Development
tries, adolescents receive comprehensive education until the age of fifteen without any
streaming (Nurmi & Siurala, 1994). These differences in educational transition then
cooccur with those of occupational life. For example, a large proportion of young
British youths leave school and enter the labor force at the age of 16, which is very different
compared to countries that aim at long education for a whole cohort, such as the
United States and Scandinavian countries.
There are also many cross-national differences in the transitions related to interpersonal
life, such as the age of first marriage and the patterns of starting family life, that
influence adolescents’ socialization into adulthood in many ways (cf. Martínez, de
Miguel, & Fernández, 1994; Roe, Bjurström, & Förnäs, 1994). One further factor along
which developmental environments vary is the relative importance of parents and peers
in adolescents’ lives. Although peers are an important part of adolescents lives in most
parts of the world, in some contexts (e.g., in India and in Arab countries) peer groups
play a relatively minor role, particularly for girls (Brown et al., 2002).
Besides cross-national differences, adolescents’ developmental environments vary
also within societies along many factors, such as ethnicity and socioeconomic status.
Young people who come from diverse backgrounds face different opportunity structures,
age-related normative demands, and standards and are provided different role
models and parental tutoring. Such differences then have consequences for the ways in
which adolescents direct their lives in the domains of education, occupation, and interpersonal
relations; for how they adjust to the outcomes of their efforts; and for the
kinds of reflections they construct about themselves during the transition from adolescence
to adulthood. For example, in the United States the percentage of adolescents
who had completed high school and the percentage of those who have a bachelor degree
vary substantially according to ethnicity: Whites show the highest percentages,
followed by Blacks, while Latinos show the lowest educational attainment (Kerckhoff,
2002). Such differences are important because they are also reflected in an individual’s
occupational career and problems with it, such as unemployment.
Another important factor that influences the challenges, opportunities, and standards
that adolescents face is the socioeconomic status of their family, which consists
of a number of interrelated variables, such as family income and values, parental education,
and membership in particular subcultures and communities. For example, family
income is a foremost factor in differentiating the paths taken through the transition
from adolescence to adulthood in many countries (Mortimer & Larson, 2002). As adolescents
have relatively detailed conceptions of their age-graded developmental environments
(Crockett & Bingham, 2000), they are probably also conscious of how their
social background will influence them. The impact of socioeconomic status is reflected
not only in the opportunity structure but also in parents’ values and aspirations, which
have been shown to influence adolescents’ subsequent life paths (Hogan, 1985).
Because the key assumption in this chapter is that adolescents’ age-graded developmental
environments provide a basis for the ways in which adolescents direct their development
and adjust to developmental outcomes, this variation in adolescents’ environments
across and within countries can be expected to be reflected in many ways in
the channeling, selection, adjustment, and reflection processes. For example, Scnabel
et al. (2002) showed that academic achievement was predictive of adolescents’ career
decisions both in Germany and the United States, whereas social background influ-
Conclusions 113
ences were more pronounced in Germany. Moreover, Nurmi, Seginer, et al. (1995)
found that due to earlier and shorter educational transitions, Australian adolescents
showed higher levels of exploration and commitments, both in the domain of future education
and work, compared with their Israeli and Finnish counterparts. They also expected
their goals and hopes related to future education and work to be realized earlier
in their lives than did young Finns and Israelis.
Overall, these results are important because they suggest that the different environments
in which adolescents grow up produce substantial amounts of variation in their
subsequent life paths. Consequently, one must be careful in making generalizations
from results found in one sociocultural context to other environments. However, this
variation provides researchers with an interesting option to examine the extent to which
their theories and findings generalize across different developmental environments.
Socialization in Time: Historical Changes
It is only during the past 100 years that adolescence emerged as an independent and extended
life period, mainly due to the extended period of education (Hurrelman, 1994).
Moreover, adolescence, as well as how it is defined by society and culture, shows continuous
change. Herdandez (1997) summarized the trends in the United States during
the past 150 years. According to him, the major changes in the developmental context
of children and adolescents during the century before the second world war included
the shift to nonfarm work by fathers, a drastic constriction of family size, and enormous
increases in educational attainments. After the half century that followed, the key
changes have included the increase of labor force participation by mothers, the rise of
single parenthood, and a large decline and then substantial rise in child poverty. Although
the timing of these historical changes has varied from one country to another,
the general patterns are more or less the same in industrialized countries. Moreover,
some of the recent changes in developing countries resemble the changes that happened
in industrialized societies several decades ago (Brown et al., 2002). The importance of
these analyses for adolescent research is that they help us to understand that how things
appear in young people’s lives at a given moment is not a consequence of unchangeable
general laws but rather is influenced by many historical and societal developments.
A few recent trends also modify adolescent development. First, gender differences
in adolescents’ thinking and interests seem to be changing in industrialized countries.
Although girls continue to be more interested in future family and human relationships
and boys in material aspects of life, comparisons of research findings across the past 30
years suggest that girls’ interests in education and occupation began to exceed those of
boys (Nurmi, 2001). These results accord well with the statistics that in many countries
the proportion of girls in higher education exceeds that for boys. However, in less industrialized
countries the experiences and opportunities of adolescents boys and girls
have remained markedly different (Brown & Larson, 2002).
Another still-continuing change is a move from rural areas to urban environments.
It has been suggested that living in urban versus rural living environments, along with
related differences in the opportunity structures, is reflected in young people’s motivation
and thinking in many ways. For example, Nurmi et al. (1994) found that adolescents’
exploration and commitments related to education and occupation increased
114 Socialization and Self-Development
with age in urban environments but not in rural environments. This difference was suggested
to be due to the fact that rural environments provide less educational choices
than do urban contexts.
The third developmental trend that influences adolescents’ lives in most parts of the
world is globalization (Brown & Larson, 2002). Besides a move from rural to urban environments
and increasing length of education, globalization refers also to the important
role of a uniform youth culture as reflected in standard elements of dress, music
taste, and entertainment. This development is closely connected to the increasing importance
of new information technologies and worldwide media business.
One recent change in adolescents’ lives in industrialized countries consists of an increase
in so-called turbulences in the transition into job markets. Many young people
start their occupational careers in jobs that both they and their employer expect to be
temporary. Although it has been suggested that this trend is due to the educational system
of the United States (Kerckhoff, 2002), a similar trend is evident in many European
countries. The problem of this development is that it may also postpone other transitions
during young adulthood, such as gaining independence from parents and starting
one’s own family.
The final important recent change in adolescent life is the increase in divorce and
single parenthood. If this trend continues, it may lead to many changes in adolescent
socialization. For example, as most of the single parents are women, increasing amounts
of adolescents are living in a situation in which they lack advice and support from their
fathers (Jenkins Tucker et al., 2001). This may cause particular problems in adolescent
socialization into the adult world, particularly for boys.
Methodological Implications
This chapter summarized the results of research about adolescents’ socialization and
self-development. Unfortunately, many parts of the literature review concluded that the
research included serious methodological limitations.
One typical feature of the reviewed studies was that the direction of influence was
presupposed on the basis of cross-sectional findings. This problem was particularly true
for the research that focused on the role of parent and peer relations in adolescents development.
Consequently, there is a need to enhance the quality of data when examining
any developmental mechanisms. One way to do this is to use cross-lagged longitudinal
data in which the same variables are repeatedly measured across time. This means
that the time for easy solutions to conduct adolescent research is over: Cross-sectional
procedures in the examination of developmental processes are in many cases wastes of
time and money.
Another way to test the direction of effects is to use intervention studies. This approach
has not been typical in the research on adolescents socialization, perhaps
because it is not clear what should be targeted in interventions. In the case of problem
behaviors, such as criminality and drug abuse, this, of course, is not a problem.
One assumption in this chapter is that adolescent development consists of interactions
between the developing individual and his or her changing age-graded environments.
The major idea is that age-graded environments channel adolescents motives
and interests, and feedback concerning their efforts in dealing with a variety of transi-
Conclusions 115
tions and challenges leads to the adjustment of previous strategies and goals and the
formation of reflections concerning oneself. Examination of such mechanisms requires
at least two features of the data.
First, as the processes included in socialization might be assumed to change rapidly,
there is a need for intensive measurements, such as every half year or even less. The
problem with traditional longitudinal studies is, namely, that they may not be intensive
enough to reach the critical developmental changes. Examination of developmental
processes should be preceded by theoretical analysis of the time range during which
major developmental processes take place, and this time range should then be applied
to define the length of the time intervals between the measurements (Aunola, Leskinen,
Onatsu-Arvilommi, & Nurmi, 2002).
The second requirement for the successful examination of processes such as socialization
is that studies focus on periods of adolescent development during which key developmental
processes take place. It might be assumed that the times when adolescents
are facing some major transitions in their lives are such important periods. During
early adolescence such transitions are typically scheduled by the individual’s age. During
late adolescence, they may appear more independently from age, which may require
less traditional research designs. One additional aspect of such studies on critical transitions
is that if a group of individuals is followed only across a particular transition, the
phenomena under focus can be measured intensively.
In sum, the results reviewed in this chapter indicate that we know much about how
age-graded sociocultural environments channel adolescent development, as well as
about the mechanisms by which adolescents select their developmental environments.
There are also data depicting how adolescents reflect themselves as a consequence of
these adventures. However, less is known about how adolescents try to adjust their previous
efforts as a means to deal with negative feedback and failures. Moreover, the fact
that only a few cross-lagged longitudinal studies have been conducted on the role that
parents and peers play in adolescent socialization and self-development limits our possibilities
to understand the processes taking place in these interpersonal settings.
in terms of four mechanisms: First, it was assumed that the age-graded developmental
tasks, role transitions, and institutional tracks define an opportunity space
that channels young people’s future-oriented motivation, thinking, and behavior. Second,
the kinds of motives and personal goals adolescents construct, and the ways in
which they explore, plan, construct strategies, and enter into commitments were assumed
to be responsible for the ways in which adolescents direct their future development
and select their developmental environments. Third, as a consequence of their efforts,
adolescents attain outcomes, either successes or failures, which requires them to
adjust their previous efforts in terms of goal reconstruction, coping, and the use of self-
110 Socialization and Self-Development
protective causal attributions. Finally, after ending up in a particular social position
and related life situation, adolescents construct reflections and tell stories about who
they are.
Socialization and Self-Development
A review of previous research suggested, first, that adolescents have relatively detailed
conceptions of their age-graded developmental environments (i.e., the timing and sequential
structure of the transitions and tracks they are facing in the future). It was
therefore no surprise that such age-graded structures channel adolescents’ personal
goals and interests: Young people’s future hopes and interests were found to focus typically
on the major developmental tasks of their own age period. Young people also
continuously reconstruct their personal goals to match with the specific stages of a particular
transition through which they are going, as well as the institutional tracks in
which they are involved.
Both the personal goals adolescents have and the cognitive strategies they deploy,
which were assumed to be the major mechanisms in the selection process, were found
to contribute to the developmental trajectories they face subsequently, as well as how
well they are able to deal with the related challenges and demands. Clear evidence was
found from longitudinal studies that adolescents’ motives and personal goals predict
how their lives will proceed in educational, occupational, and family-related trajectories.
Similarly, the kinds of plans and strategies that adolescents apply have consequences
for their success in dealing with major challenges at school, at work, and also
in interpersonal relationships. However, adolescents become interested in forthcoming
developmental tasks and transitions as they grow older, and the tools they have for
dealing with these demands and challenges develop rapidly during early adolescence in
particular. Although the majority of adolescents deploy adaptive strategies, such as optimistic
and task-focused patterns, some of them deploy avoidant strategies as a way to
deal with a fear of failure or anxiety.
This review showed also that adolescents whose goals focus on major age-graded developmental
tasks have higher well-being than do those who have other kinds of goals,
perhaps because such goals help them to deal with the major demands and challenges
they are facing. Although it has been assumed that thinking of self-related issues is a
part of adolescents’ lives, strong evidence was found that self-focused, existential type
of goals are detrimental to young people’s well-being. Moreover, the deployment of
adaptive strategies led not only to higher levels of success in academic and interpersonal
domains of life but also, in the longer run, to higher well-being.
There is also considerable evidence that parents and their adolescent children share
similar kinds of goals concerning the adolescent’s future. Moreover, positive and authoritative
parenting is associated not only with adolescents’ high level of interest in
major developmental tasks, such as education and occupation, but also with adolescents’
use of adaptive strategies, particularly in achievement contexts.
After adolescents have received feedback about the outcomes of their efforts to deal
with the major developmental challenges and demands, they have to adjust their previous
efforts in terms of coping, reconstruction of goals, and making causal attributions.
Conclusions 111
Surprisingly, much less research has been conducted on the antecedents and consequences
of this adjustment compared with those of selection processes. Research on
coping showed that problem-focused coping and engagement coping are associated
with higher levels of psychological adjustment, whereas emotion-focused coping seems
to lead to maladjustment. However, there was little evidence that coping has clear
consequences for individual success in dealing with particular kinds of tasks. Authoritative
and positive parenting was shown to be associated with problem-focused coping,
whereas more negative parenting is related to emotion-focused coping.
Very little research has been carried out on how adolescents reconstruct their personal
goals based on their previous successes and failures. The few studies suggest that
adolescents reconstruct their goals on the basis of the feedback they receive concerning
goal attainment and that such goal reconstruction contributes to their well-being.
Similarly, little research has been carried out on the role that causal attributions have
in the situations in which young people have had problems in dealing with previous demands.
The few studies that exist show that problems in dealing with major transitions
decrease the use of self-protective causal attributions, which then leads to an increase
in depressive symptoms. Dysfunctional causal attributions also lower adolescents’ active
engagement in school activities, and subsequently their academic achievement.
It was also assumed that entrance into certain roles and social positions has consequences
for the identities or self-concepts that adolescents construct. Although the
studies suggest that younger adolescents more frequently report less developed identity
statuses than do older adolescents, relatively little is know about the developmental antecedents
or consequences of these developments.
By contrast, we know much, on a descriptive level, about how self-concept and
self-esteem develop during adolescence. However, some of the recent findings have
challenged previous theories by suggesting a more dynamic view according to which
adolescents’ self-concepts fluctuate significantly and follow, in many cases, individual
developmental trajectories. This fluctuation has been found to reflect many changes in
the individual’s development environments, such as school transitions, grades, and a
variety of stressful life events.
Socialization in Place: Different Developmental Environments
In this chapter adolescent socialization and self-development were conceptualized in
terms of four mechanisms that are responsible for the transaction between the developing
adolescent, on the one hand, and his or her age-graded sociocultural environment,
on the other. It can also be assumed that the substantial amount of variation
across societies and cultures in the developmental environments in which adolescents
grow up (Brown, Larson, & Saraswathi, 2002; Hurrelmann, 1994) channels their subsequent
development in many ways. One key factor that contributes to this variation is
education: There are many differences in the educational systems that are reflected in
adolescents’ thinking and lives across the world (Hurrelmann, 1994; Nurmi, Seginer, et
al., 1995; Scnabel, Alfeld, Eccles, Köller, & Baumert, 2002). For example, in many European
countries and the United States, streaming in education based on academic
achievement begins relatively early (Hurrelmann, 1994), which also influences adolescents’
subsequent opportunities. In some other societies, such as Scandinavian coun-
112 Socialization and Self-Development
tries, adolescents receive comprehensive education until the age of fifteen without any
streaming (Nurmi & Siurala, 1994). These differences in educational transition then
cooccur with those of occupational life. For example, a large proportion of young
British youths leave school and enter the labor force at the age of 16, which is very different
compared to countries that aim at long education for a whole cohort, such as the
United States and Scandinavian countries.
There are also many cross-national differences in the transitions related to interpersonal
life, such as the age of first marriage and the patterns of starting family life, that
influence adolescents’ socialization into adulthood in many ways (cf. Martínez, de
Miguel, & Fernández, 1994; Roe, Bjurström, & Förnäs, 1994). One further factor along
which developmental environments vary is the relative importance of parents and peers
in adolescents’ lives. Although peers are an important part of adolescents lives in most
parts of the world, in some contexts (e.g., in India and in Arab countries) peer groups
play a relatively minor role, particularly for girls (Brown et al., 2002).
Besides cross-national differences, adolescents’ developmental environments vary
also within societies along many factors, such as ethnicity and socioeconomic status.
Young people who come from diverse backgrounds face different opportunity structures,
age-related normative demands, and standards and are provided different role
models and parental tutoring. Such differences then have consequences for the ways in
which adolescents direct their lives in the domains of education, occupation, and interpersonal
relations; for how they adjust to the outcomes of their efforts; and for the
kinds of reflections they construct about themselves during the transition from adolescence
to adulthood. For example, in the United States the percentage of adolescents
who had completed high school and the percentage of those who have a bachelor degree
vary substantially according to ethnicity: Whites show the highest percentages,
followed by Blacks, while Latinos show the lowest educational attainment (Kerckhoff,
2002). Such differences are important because they are also reflected in an individual’s
occupational career and problems with it, such as unemployment.
Another important factor that influences the challenges, opportunities, and standards
that adolescents face is the socioeconomic status of their family, which consists
of a number of interrelated variables, such as family income and values, parental education,
and membership in particular subcultures and communities. For example, family
income is a foremost factor in differentiating the paths taken through the transition
from adolescence to adulthood in many countries (Mortimer & Larson, 2002). As adolescents
have relatively detailed conceptions of their age-graded developmental environments
(Crockett & Bingham, 2000), they are probably also conscious of how their
social background will influence them. The impact of socioeconomic status is reflected
not only in the opportunity structure but also in parents’ values and aspirations, which
have been shown to influence adolescents’ subsequent life paths (Hogan, 1985).
Because the key assumption in this chapter is that adolescents’ age-graded developmental
environments provide a basis for the ways in which adolescents direct their development
and adjust to developmental outcomes, this variation in adolescents’ environments
across and within countries can be expected to be reflected in many ways in
the channeling, selection, adjustment, and reflection processes. For example, Scnabel
et al. (2002) showed that academic achievement was predictive of adolescents’ career
decisions both in Germany and the United States, whereas social background influ-
Conclusions 113
ences were more pronounced in Germany. Moreover, Nurmi, Seginer, et al. (1995)
found that due to earlier and shorter educational transitions, Australian adolescents
showed higher levels of exploration and commitments, both in the domain of future education
and work, compared with their Israeli and Finnish counterparts. They also expected
their goals and hopes related to future education and work to be realized earlier
in their lives than did young Finns and Israelis.
Overall, these results are important because they suggest that the different environments
in which adolescents grow up produce substantial amounts of variation in their
subsequent life paths. Consequently, one must be careful in making generalizations
from results found in one sociocultural context to other environments. However, this
variation provides researchers with an interesting option to examine the extent to which
their theories and findings generalize across different developmental environments.
Socialization in Time: Historical Changes
It is only during the past 100 years that adolescence emerged as an independent and extended
life period, mainly due to the extended period of education (Hurrelman, 1994).
Moreover, adolescence, as well as how it is defined by society and culture, shows continuous
change. Herdandez (1997) summarized the trends in the United States during
the past 150 years. According to him, the major changes in the developmental context
of children and adolescents during the century before the second world war included
the shift to nonfarm work by fathers, a drastic constriction of family size, and enormous
increases in educational attainments. After the half century that followed, the key
changes have included the increase of labor force participation by mothers, the rise of
single parenthood, and a large decline and then substantial rise in child poverty. Although
the timing of these historical changes has varied from one country to another,
the general patterns are more or less the same in industrialized countries. Moreover,
some of the recent changes in developing countries resemble the changes that happened
in industrialized societies several decades ago (Brown et al., 2002). The importance of
these analyses for adolescent research is that they help us to understand that how things
appear in young people’s lives at a given moment is not a consequence of unchangeable
general laws but rather is influenced by many historical and societal developments.
A few recent trends also modify adolescent development. First, gender differences
in adolescents’ thinking and interests seem to be changing in industrialized countries.
Although girls continue to be more interested in future family and human relationships
and boys in material aspects of life, comparisons of research findings across the past 30
years suggest that girls’ interests in education and occupation began to exceed those of
boys (Nurmi, 2001). These results accord well with the statistics that in many countries
the proportion of girls in higher education exceeds that for boys. However, in less industrialized
countries the experiences and opportunities of adolescents boys and girls
have remained markedly different (Brown & Larson, 2002).
Another still-continuing change is a move from rural areas to urban environments.
It has been suggested that living in urban versus rural living environments, along with
related differences in the opportunity structures, is reflected in young people’s motivation
and thinking in many ways. For example, Nurmi et al. (1994) found that adolescents’
exploration and commitments related to education and occupation increased
114 Socialization and Self-Development
with age in urban environments but not in rural environments. This difference was suggested
to be due to the fact that rural environments provide less educational choices
than do urban contexts.
The third developmental trend that influences adolescents’ lives in most parts of the
world is globalization (Brown & Larson, 2002). Besides a move from rural to urban environments
and increasing length of education, globalization refers also to the important
role of a uniform youth culture as reflected in standard elements of dress, music
taste, and entertainment. This development is closely connected to the increasing importance
of new information technologies and worldwide media business.
One recent change in adolescents’ lives in industrialized countries consists of an increase
in so-called turbulences in the transition into job markets. Many young people
start their occupational careers in jobs that both they and their employer expect to be
temporary. Although it has been suggested that this trend is due to the educational system
of the United States (Kerckhoff, 2002), a similar trend is evident in many European
countries. The problem of this development is that it may also postpone other transitions
during young adulthood, such as gaining independence from parents and starting
one’s own family.
The final important recent change in adolescent life is the increase in divorce and
single parenthood. If this trend continues, it may lead to many changes in adolescent
socialization. For example, as most of the single parents are women, increasing amounts
of adolescents are living in a situation in which they lack advice and support from their
fathers (Jenkins Tucker et al., 2001). This may cause particular problems in adolescent
socialization into the adult world, particularly for boys.
Methodological Implications
This chapter summarized the results of research about adolescents’ socialization and
self-development. Unfortunately, many parts of the literature review concluded that the
research included serious methodological limitations.
One typical feature of the reviewed studies was that the direction of influence was
presupposed on the basis of cross-sectional findings. This problem was particularly true
for the research that focused on the role of parent and peer relations in adolescents development.
Consequently, there is a need to enhance the quality of data when examining
any developmental mechanisms. One way to do this is to use cross-lagged longitudinal
data in which the same variables are repeatedly measured across time. This means
that the time for easy solutions to conduct adolescent research is over: Cross-sectional
procedures in the examination of developmental processes are in many cases wastes of
time and money.
Another way to test the direction of effects is to use intervention studies. This approach
has not been typical in the research on adolescents socialization, perhaps
because it is not clear what should be targeted in interventions. In the case of problem
behaviors, such as criminality and drug abuse, this, of course, is not a problem.
One assumption in this chapter is that adolescent development consists of interactions
between the developing individual and his or her changing age-graded environments.
The major idea is that age-graded environments channel adolescents motives
and interests, and feedback concerning their efforts in dealing with a variety of transi-
Conclusions 115
tions and challenges leads to the adjustment of previous strategies and goals and the
formation of reflections concerning oneself. Examination of such mechanisms requires
at least two features of the data.
First, as the processes included in socialization might be assumed to change rapidly,
there is a need for intensive measurements, such as every half year or even less. The
problem with traditional longitudinal studies is, namely, that they may not be intensive
enough to reach the critical developmental changes. Examination of developmental
processes should be preceded by theoretical analysis of the time range during which
major developmental processes take place, and this time range should then be applied
to define the length of the time intervals between the measurements (Aunola, Leskinen,
Onatsu-Arvilommi, & Nurmi, 2002).
The second requirement for the successful examination of processes such as socialization
is that studies focus on periods of adolescent development during which key developmental
processes take place. It might be assumed that the times when adolescents
are facing some major transitions in their lives are such important periods. During
early adolescence such transitions are typically scheduled by the individual’s age. During
late adolescence, they may appear more independently from age, which may require
less traditional research designs. One additional aspect of such studies on critical transitions
is that if a group of individuals is followed only across a particular transition, the
phenomena under focus can be measured intensively.
In sum, the results reviewed in this chapter indicate that we know much about how
age-graded sociocultural environments channel adolescent development, as well as
about the mechanisms by which adolescents select their developmental environments.
There are also data depicting how adolescents reflect themselves as a consequence of
these adventures. However, less is known about how adolescents try to adjust their previous
efforts as a means to deal with negative feedback and failures. Moreover, the fact
that only a few cross-lagged longitudinal studies have been conducted on the role that
parents and peers play in adolescent socialization and self-development limits our possibilities
to understand the processes taking place in these interpersonal settings.
Subscribe to:
Posts (Atom)