Beruflich Dokumente
Kultur Dokumente
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=ucpress.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to The
Elementary School Journal.
http://www.jstor.org
School as a Context of Abstract
by adults in their families, schools, and nomic, and social opportunities due to their
communities for cultivating their academic engagement in high-risk behaviors and ac-
and social aspirations, skills, knowledge, tivities that include violence and vandal-
and commitments is a second essential fac- ism, unprotected sex, abuse of alcohol and
tor in promoting adolescents' developmen- drugs, skipping and failing school, and so
tal success (Erikson, 1968). Both successful on (Carnegie Council, 1989, 1995). And 60%
and problematic development during ado- to 80% of the youth most at risk dwell in
lescence involves, as Erik Erikson said, the disadvantaged inner-city environments
overall configuration of personal and so- (McLaughlin & Irby, 1994). Currently the
cial-historical factors (see Friedman, 1999). challenges to successful adolescent devel-
opment are serious, and millions of U.S.
Promise and Problem young people are at serious risk for unsat-
Although it was once thought that most, if isfying and unproductive lives.
not all, adolescents experienced psycholog-
ical problems as a general condition of their An Intergenerational Perspective
development during the second decade of From a perspective that views adolescents'
life (e.g., Freud, 1936), it is now known that developmental success as grounded in the
this is not the case. For instance, recent re- surrounding social conditions of their de-
search suggests that more than half of all velopment, the fact that a significant mi-
adolescents in the United States are able to nority of adolescents are not doing well in
constructively manage the multiple transi- terms of their academic achievement, social
tions in body, thought and emotion, social and behavioral choices, and mental health
relationships, and school that are character- suggests serious problems not only within
istic of the early to middle adolescent pe- such youth but also across the broader cast
riods without experiencing any major prob- of adults and institutions charged with
lems (Dryfoos, 1990). For many, the early helping them become full members of so-
adolescent years are a time of exciting ex- ciety. This was the view of Erikson (1968)
plorations in identity and the process of be- who, in reflecting on the turbulence of the
coming comfortable with a maturing physi- 1960s in the United States in his book Iden-
cal body; of continued closeness with a tity, Youth,and Crisis, reminded society's el-
nurturing family, deepening mutuality ders that "in looking at the youth of today,
with close friends and opposite-sex peers; one is apt to forget that identity formation,
and of continued engagement with school. while being 'critical' in youth, is really a
The transition into middle level schools in generational issue" (p. 29). His 1968 treatise
particular heralds what is for many adoles- on adolescent identity development was
cents a time of expanding intellectual and meant to draw attention to his contention
social opportunities (Carnegie Council on that adolescents' ability to organize the sig-
Adolescent Development, 1989, 1995). nificant changes they experience during
But if the early adolescent years are no these years into a coherent and positive psy-
longer conceived of as a time of normative chosocial identity was not simply a per-
"storm and stress" for youth, research has sonal project but rather a collective and in-
documented that early adolescence today is tergenerational responsibility of the
a time of more or less quiet distress for mil- adolescent and his or her parents, teachers,
lions of America's young people. For in- and community members.
stance, a distinguished panel of experts re- Whereas Erikson's (1973) life-span the-
cently estimated that somewhere between ory of development clearly defined the life
25% and 50% of all young people in the task of adolescence as the discovery of what
United States between the ages of 10 and 17 one cared to do and who one cared to be,
are at risk for curtailed educational, eco- one of the life tasks he described for adult-
MAY 2000
SCHOOLCONTEXT 445
bigger than oneself, to an ongoing, authen- nities that adults and institutions afford to
tic, and in the best-case scenarios, compas- young people at home, in school, and dur-
sionate, cultural concern (e.g., Erikson, ing the out-of-school hours (Carnegie
1969). When social conditions in a particu- Council, 1989, 1995).
lar place, for a variety of historical reasons,
do not afford to young people the fulfill- Middle Grades Schools
ments they seek for their needs, Erikson In this article, we focus on results from our
saw the seeds of rare moments of genius, own recent research efforts to understand
creativity, and cultural renewal (Erikson, how adults in one social institution, the
1958, 1969), as well as the more common middle grades school, can enhance adoles-
harvest of widespread social problems cents' healthy identity formation and be-
among the elderly and youthful alike (Er- havior during this period. Such healthy de-
ikson, 1950, 1968). velopment is indicated by adolescents'
academic achievement and good conduct in
Positive Youth Development school, positive peer relationships, and
Many researchers interested in adolescent positive self-related processes such as mo-
development today focus, as Erikson once tivation to learn and mental health (Masten
did, on understanding how, through the & Coatsworth, 1998). Middle grades schools
provision or absence of certain social op- have been hailed as one of the most impor-
portunities, adults cultivate or frustrate tant institutions within which to recapture
positive identity formation and behavioral the millions of American youth who, often
development during adolescence (Eccles et because of difficult social conditions, are
al., 1993; Jessor, 1993; Sameroff, Seifer, & probabilistically more likely to show poor
Bartko, 1997). Recently, there has been a achievement and motivation to learn, poor
growing trend in this work toward positive conduct and affiliations with negative
visions of youth development in which the peers, mental health problems, or different
focus is on how, through social systems re- combinations of these problems and others
forms, practitioners can cultivate develop- simultaneously during the early adolescent
mental strengths and assets in young peo- years (Carnegie Council, 1989, 1995; Dry-
ple (e.g., Carnegie Council, 1995; Cowen, foos, 1990).
1991; McLaughlin & Irby, 1994). The many For researchers and practitioners inter-
changes of the early adolescent period ested in middle grades education, the chal-
make it an ideal time for systemic reforms. lenge of cultivating positive youth devel-
Research has documented that current con- opment is the dual challenge of
ditions in many families, schools, and com- understanding, designing, and implement-
munities do not afford adolescents devel- ing schoolwide reform efforts that benefit
opmentally appropriate opportunities for all adolescents during this critical develop-
academic, social, and emotional growth, mental period (e.g., Eccles, Lord, & Roeser,
and also, more hopefully, that when fami- 1996; Midgely & Edelin, 1998) and targeted
lies, schools, and communities do provide intervention and prevention programs that
such opportunities, problems among youth assist those subgroups of adolescents with
are greatly reduced and strengths instilled specific vulnerabilities and needs (Adelman
(Carnegie Council, 1995; Eccles et al., 1993; & Taylor, 1998; Dryfoos, 1998). Such a dual
McLaughlin & Irby, 1994). Thus, the early focus characterized many of the school re-
adolescent years have been aptly labeled "a form efforts during the past century (Dry-
turning point" and a "great transition" in foos, 1994; Hechinger, 1993), and a central
that both promises and problems are poten- assumption of such efforts has been that
tial outcomes of this period and beyond de- school, as a major context of child and ad-
pending on the kind of care and opportu- olescent development, has significant influ-
MAY 2000
SCHOOLCONTEXT 447
"Livesin SchoolContexts"
Adolescents' School
Psychosocial
Functioning SocialContext
School Social-Emotional of
Support of
Support Qualityof
Functioning Functioning Competence Autonomy Relationships
(EGO) (ETHOS)
FIG.3.-Two specific aspects of adolescents' lives in school contexts
"Adolescent
Functioning"
School Social-emotional
Functioning Functioning
Indicators
Psychological Psychological Indicators
to learn
Motivation distressandwell-being
Emotional
(EGO)
time. Our focus here has been on how spe- adolescents (Allport, 1937; Runyan, 1982)
cific instructional, interpersonal, and orga- (see Fig. 5).
nizational processes in middle school, Variable-centered, statistically nomo-
based on their fit or mismatch with early thetic analyses are used to explore general
adolescents' implicit psychological needs relations among indicators of psychosocial
for competence, autonomy, and quality re- functioning and middle school experience
lationships, are associated with aspects of across the whole sample of adolescents
their school and social-emotional function- (Magnusson & Bergmann, 1988). Such anal-
ing (e.g., Deci & Ryan, 1985; Eccles et al., yses can yield information on which school
1993). Selected empirical findings bearing practices in general, from adolescents'
on each of these research questions are pre- points of view, seem to promote positive ac-
sented in this article. ademic and social-emotional adjustment,
thus making such practices possible foci of
Two Analytic Methods future schoolwide reform efforts.
Before turning to the substantive findings, Person-centered, statistical idiographic
it is important to mention that we have em- analyses are used to explore subgroups of
ployed two types of statistical analyses to adolescents who show unique configura-
investigate these topics; those of a more tions of psychosocial functioning and mid-
"statistically nomothetic" nature that focus dle school experience (Magnusson & Berg-
on relations among variables across the mann, 1988). Such analyses, because they
whole sample of adolescents, and those of yield information on subgroups of adoles-
a more "statistically idiographic" nature cents with specific assets, vulnerabilities,
that focus on patterns of experiences and and needs, might aid in decisions about
outcomes among particular subgroups of how to target particular educational, health,
MAY 2000
SCHOOLCONTEXT
Variable-centered Person-centered
analysesof the full sample analysesof sub-groups
450 THE ELEMENTARYSCHOOLJOURNAL
such as those measuring motivation to learn all had acceptable reliabilities and were
and academic achievement. In our first set identical across waves of data collection.1
of analyses, we examined how psychologi- Figure 6 presents the results of multiple
cal aspects of early adolescents' evolving regression analyses in which psychological
identities, including their self-perceptions indicators of functioning and adolescents'
of academic competence, valuing of school, demographic characteristics were used to
and feelings of emotional distress in the predict their academic grades, school prob-
prior month, predicted behavioral out- lem behavior, and affiliations with negative
comes such as their achievement, in-school peers during the seventh and eighth
conduct, and affiliations with negative grades.2 These analyses included between
peers. 814 and 945 early adolescents for whom we
had the most complete data at each grade.3
Measures It is important to note that the standardized
The academic competence scale we used regression coefficients that are presented in
Figure 6 are adjusted for (net of) the effect
tapped adolescents' self-perceptions of abil- of particular demographic characteristics of
ity in different school subjects (e.g., How the adolescents and their families that were
good are you at math?) and their ability to also included in these prediction equations.
competently master the academic require- These demographic measures included
ments of school (e.g., How well can you live
educational level of head of household, oc-
up to what your teachers expect of you?).
The valuing-of-school scale measured ado- cupational status of the head of household,
1990 pretax family income, and adolescents'
lescents' perceptions of school as interest-
race and gender.4 Coefficients for these
ing, important, and instrumental in the ful-
fillment of other life goals (e.g., getting a demographic characteristics are left out of
the figures for purposes of clarity. Signifi-
good job later). The emotional distress mea- cant effects of the demographic factors on
sure assessed adolescents' self-reported the outcome measures are described in the
feelings of anger, sadness, or hopelessness text.
during the prior 2-4 weeks. Academic per-
formance was measured by end-of-year, Seventh-Grade Results
teacher-rated grade point average (GPA) in The demographic and psychological
the core academic subjects (e.g., English, variables accounted for 38% of the variance
math, science, social studies). School prob- in adolescents' seventh-grade GPA. Adoles-
lem behaviors were assessed for each school cents who were white, female, or who had
year via adolescents' self-reports. The parents with more education and income
school problem behavior scale was a had higher GPAs than adolescents who
weighted sum of items that assessed were African-American, male, or who had
whether the adolescent had cheated on
parents with less education and income, re-
tests, skipped classes, fought in school, or spectively. In terms of the psychological in-
brought drugs or alcohol to school during dicators, adolescents' self-perceptions of ac-
the school year. Items were weighted in this ademic competence emerged as a
order of severity. The negative peer affilia- significant positive predictor of GPA,
tions scale was based on adolescents' self- whereas emotional distress had a marginal
report of how many of their close friends (p = .055) negative predictive relation with
were involved in problematic behavior later GPA.
within and outside the school setting (e.g., In the prediction of adolescents' engage-
cheating on tests, skipping classes, stealing ment in school problem behavior, we found
something worth more than $50, pressuring that emotional distress predicted increased
others to use drugs, and so on). These scales school problem behavior, whereas self-per-
MAY 2000
SCHOOLCONTEXT 451
Emotional
Distress
-.05t- Academic
.35** GPA
(.38)
Academic
Competence 13t
School
Problem
-.09* Behaviors
Academic
(.13)
Value .234
.14*
Negative
PeerAffiliations
(.12)
of
Beginning of
Beginning Endof
7thGrade 7thGrade 7thGrade
Emotional
Distress - Academic
S-.13"* GPA
.32**" (.42)
Academic
Competence .24**
"
-.10*"*- School
=. Problem
Behaviors
Academic
(.21)
Value .33**
-.23** Negative
PeerAffiliations
(.24)
Endof Endof
8thGrade 8thGrade
FIG.6.-Prediction of achievement, school conduct, and peer relationships during middle school. Standard-
ized regression coefficients are presented. Adjusted R2s are presented in parentheses under each outcome mea-
sure. Demographic characteristics of the adolescent are controlled in these analyses. tp = .055; *p .05; **ps .01.
ceptions of academic competence and val- and male adolescents reported more prob-
uing of school predicted decreased school lem behavior in school than did their white
problem behavior. These effects were found and female counterparts, respectively (p <
after accounting for adolescents' demo- .01).
graphic characteristics. African-American In the prediction of adolescents' affilia-
452 THE ELEMENTARYSCHOOLJOURNAL
tions with negative peers, we found that behaviors and with negative peers is con-
emotional distress and poor valuing of sistent with other research on "school bond-
school were associated with increased in- ing" that finds both males and females who
volvement with such peers. We also found are committed to school engage in less
that negative peer affiliations were more problem behavior generally (Finn, 1989;
common among males than females (p < Hawkins, Doueck, & Lishner, 1988). Our re-
.01). sults support the notion that adolescents
who find school interesting, important, and
Eighth-Grade Results instrumental for attaining other life goals
As one can see in the bottom portion of are less likely to cheat, to skip classes, to
Figure 6, an identical pattern of relations be- fight in school, or to gravitate toward peer
tween these psychological and behavioral groups that engage in and reinforce such
indicators of functioning was found at maladaptive behaviors (Dishion, Patterson,
eighth grade. We also note that an identical Stoolmiller, & Skinner, 1991). We also
pattern of demographic effects on these out- found, perhaps not surprisingly, that ado-
comes was found at eighth grade. Thus, it lescents who experience frequent feelings of
appears that adolescents' demographic sadness or anger are more likely to act out
characteristics, their motivation to learn, in these ways in school and to gravitate to-
and their emotional distress are related to ward negative peer groups.
their academic achievement, in-school con-
duct, and affiliation with problematic peers Longitudinal Relations
in similar ways during the middle and later We next looked at the relations among
middle school years. Finally, the pattern of a subset of psychological and behavioral in-
results in Figure 6 was replicated when we dicators of functioning over time. We were
substituted parents' ratings of their child's interested in the direction of effects among
emotional distress for the adolescent self- these measures. For instance, we wondered
report measures of distress. if motivation to learn and academic
That adolescents' self-perceptions of ac- achievement protect adolescents against
ademic competence predicted higher sub- feeling emotionally distressed at a later
sequent achievement corroborates other time. Alternatively, we wondered if emo-
educational research that shows reciprocal tional distress served as a risk factor for
relations among self-perceived academic subsequent motivational and achievement
competence and actual performance (Ban- difficulties in school. We chose to focus on
dura, 1993; Eccles, 1983). When adolescents the cross-time relations between adoles-
perceive themselves as academically com- cents' motivation to learn, mental health,
petent and able to master school-related and academic grades because we had three
tasks, they get higher grades, and when temporally lagged measures of these con-
they achieve good grades, they feel more structs in our study. Thus, this set of mea-
competent academically. The negative re- sures provided the best opportunity to ex-
lation of emotional distress and academic plore causal relations over time.
grades corroborates other mental health re- To examine these direction-of-effect is-
search that shows emotional distress can sues, we regressed adolescents' end of sev-
impair children's and adolescents' ability to enth-grade GPA on measures of motivation
effectively learn in school (Blechman, and emotional distress that were collected
McEnroe, Carella, & Audette, 1986; Nolen- at the beginning of seventh grade. We then
Hoeksema, Girgus, & Seligman, 1986; Tes- regressed measures of academic compe-
iny, Lefkowitz, & Gordon, 1980). tence, valuing of school, and emotional dis-
That adolescents' valuing of school pre- tress that were collected at the end of eighth
dicted less involvement in school problem grade on the seventh-grade motivation, dis-
MAY 2000
SCHOOLCONTEXT 453
tress, and achievement measures. Thus, two motivation and achievement over time. We
sets of longitudinal relations were exam- also found evidence for the opposite causal
ined: the prediction of grades from previ- path: adolescents who felt academically
ous motivation and distress and the predic- competent, valued school, and received
tion of motivation and distress from good marks at some point in seventh grade
previous grades (holding constant prior were less likely to feel emotionally dis-
levels of motivation and distress). A sche- tressed at the end of eighth grade. Such re-
matic representing the significant cross- sults further suggest the positive effects that
time relations we found is presented in Fig- adolescents' motivation to learn and aca-
ure 7. Demographic factors were once again demic achievement can have on their men-
accounted for in these statistical analyses. tal health over time.
Such relations are not represented in Figure
7 for purposes of economy and clarity. The Summary
full set of standardized regression coeffi- Variable-centered results suggested
cients is presented in Table 1. some of the psychological reasons why ad-
Results revealed a pattern of reciprocal olescents show academic and social success
relations between adolescents' academic or problems during these years. These rea-
competence beliefs, feelings of emotional sons included whether adolescents re-
distress, and academic grades over time. ported feeling competent at learning,
Adolescents who felt more emotional dis- whether they found value in what they are
tress at the beginning of seventh grade asked to learn at school, and how they felt
showed lower grades 1 year later and lower about their lives in general. It is important
self-perceptions of academic competence 2 for teachers to realize that students' motives
years later. Such results further document for particular actions in the classroom are
the negative effect that feelings of emotional often not mysterious: their motives are of-
distress can have on adolescents' academic ten precisely the kinds of beliefs, values,
Emotional Emotional
Distress Distress
Grades
Academic Academic
Competence Competence
School School
Value Value
Outcomes
NOTE.-N = 945.
+p = .055.
*p .05.
**p .01.
and emotions we have been examining countered. Social motivation can be har-
here. That is, adolescents' decisions to en- nessed through collaborative problem solv-
gage in learning or not in the classroom de- ing and group work that is carefully
pend in some measure on whether they feel designed.
able to meet the challenges presented them,
whether they see purpose and value in Patterns of School and
classroom activities, and whether they feel Social-Emotional Functioning
safe and cared for by others in the setting. Given these general relations among vari-
The challenge for middle school teachers is ables at the level of the whole sample, we
really not a question of getting students mo- next turned our attention toward identify-
tivated or not but rather, in part, getting ing subgroups of adolescents who might
them to be motivated to learn rather than show different patterns of school and so-
motivated to protect themselves from situ- cial-emotional functioning. For example,
ations they perceive as threatening to their we wondered if we could identify a group
self, meaningless, or somehow threatening of adolescents who were sad or angry but
to their social image. Teachers' task in this who nonetheless remained engaged in
respect is to shape adolescents' motivation school and stayed away from bad peer in-
to engage in fruitful intellectual activities fluences, or if we could identify a group of
and to harness social motivation for aca- adolescents who seemed free of emotional
demic purposes. Teachers can enhance distress but who nonetheless felt psycho-
achievement-related motivation in the logically disengaged from learning in
classroom by reinforcing adolescents' per- school. Person-centered, cluster analytic
ceptions of competence through scaffolded techniques were employed toward these
skill development and feedback, by framing ends. Such techniques assess the interde-
for them the value and purposes of activi- pendence among variables within persons
ties, and by providing emotional support and thereby classify persons into relatively
and encouragement during the learning homogeneous groups based on their simi-
process, especially when difficulties are en- larity across a series of measures (Magnus-
MAY 2000
SCHOOLCONTEXT 455
1.00
0.80
0.60
S0.40"
0.20
0.00 Academic Competence
rl Academic Value
, -0.20 * Mental Health
-0.40
-0.60
-0.80
-1.00
Positive Poor School Poor Mental Multiple
Adjustment Value Health Problems
(N = 418) (N = 147) (N = 152) (N = 324)
School Functioning
Social-Emotional
Functioning
Negative Peers -0.27 a 0.00 b 0.00 b 0.27 c F (3, 1029) = 17.47**
Youth ReportSelf Esteem 0.49 a 0.17 b -0.31 c -0.56
d
F (3, 1035) = 91.77**
ParentReportedDistress a -0.06 b 0.02 b
-0.27 0.35 c F (3, 1031) = 24.39**
FIG.8.-Emergent patterns of school and social-emotional functioning at seventh grade. Student Newman-
Keuls comparisons were used to test group differences. Different superscripts for a particular variable across
clusters indicate a significant mean difference at the p s .05 level.
son & Bergmann, 1988). In this case, we graph portion of Figure 8. The first pattern
clustered adolescents into subgroups based of school and social-emotional functioning
on their pattern of self-reported academic was labeled "Positive Adjustment." Ado-
competence, academic value, and emo- lescents in this group were characterized by
tional distress during the beginning of sev-
positive academic motivation and positive
enth grade. For the remainder of this article, mental health (indicated by low emotional
we reverse the "emotional distress" scale
distress) and comprised 40% of the sample.
and call it "mental health" for purposes of
The second group of adolescents showed
clarity in figures and tables.
relatively poor valuing of school but rela-
Describing the Clusters tively positive perceptions of their aca-
The four clusters that emerged and were demic competence and mental health
retained for analyses are depicted in the (called "Poor Academic Value Group").
456 THE ELEMENTARYSCHOOLJOURNAL
The third group was just the opposite. La- likely to act out in school. The other three
beled "Poor Mental Health Group," these groups did not differ from each other on
adolescents showed positive school moti- any of these measures. This latter finding
vation but experienced relatively poorer supports a commonsensical but generally
mental health than each of the other three unacknowledged notion: different sub-
groups. Finally, a "Multiple Problems groups of adolescents arrive at their aca-
Group" emerged that included 31% of the demic success in different ways-some are
adolescents in our sample. This latter group happy and some sad; some are genuinely
of adolescents reported feeling the least ac- engaged in the subject matter and some are
ademically competent, saw the least value bored; but all of those who achieved good
in school, and reported the second-highest grades also reported a sense of competence
level of emotional distress among the four and efficacy in the academic domain (Ban-
groups. We examined the gender and eth- dura, 1993; Eccles, 1983).
nic composition of these groups and found
two noteworthy things: first, white males Mental Health
were overrepresented in the poor value Next, we compared the groups on the
group, and second, white females and Af- two additional indicators of social-emo-
rican-American males were overrepre- tional functioning. These results are also
sented in the poor mental health group. presented in the table portion at the bottom
of Figure 8. More differentiation among the
Concurrent Validity of Clusters
groups was found on these measures than
As a means of "validating" the clusters, on the school behavior measures. All of the
we compared the groups on a series of in- groups, for instance, differed in terms of ad-
dicators that were not used in creating the olescents' self-reported self-esteem. Adoles-
clusters (Aldenderfer & Blashfield, 1984). cents in the positively adjusted group re-
These indicators included adolescents' aca- ported the highest levels of esteem,
demic grades, the percentage of adolescents followed sequentially by the other three
in each group whose parents reported they groups. In addition, parents of adolescents
had failed or done very poorly in a class characterized by low academic value or
during the past year, and adolescents' self- poor mental health also reported poorer
reported problem behavior in school and af- mental health in their children compared to
filiations with negative peers. We also com- the parents of positively adjusted adoles-
pared the groups on two indicators of cents. Adolescents characterized by multi-
emotional functioning that were not used to ple problems were rated by their parents as
form the groups: adolescents' self-reported having the poorest mental health of any of
self-esteem and parents' ratings of their ad- the adolescents. Finally, we found that the
olescents' mental health. well-adjusted adolescents were the least
likely to affiliate with negative peers,
School Behavior whereas the multiple-risk youth were the
As one can see in the table portion at the most likely to affiliate with peers who
bottom of Figure 8, analyses of variance and skipped school, stole things, and so on. In
post-hoc mean comparisons among all pos- sum, the four subgroups of adolescents,
sible pairs of groups revealed that only the each showing a unique pattern of motiva-
multiple problems group differed from the tion and mental health, differed in theoret-
others in terms of their school-related be- ically and substantively meaningful ways
haviors. Compared to each of the other on the validation measures of achievement,
groups, this group of adolescents got conduct, peer relationships, and mental
poorer grades, were more likely to do health. Thus, examining how patterns of
poorly in or fail a class, and were more psychological processes configure with pat-
MAY 2000
SCHOOLCONTEXT 457
terns of manifest behavior among different processes, intrapsychic and social in nature.
subgroups of adolescents is one useful way Considering the multiproblem youth in
of studying what Erikson called psychoso- particular, our results support the idea that
cial functioning and others now call devel- rather than designing unique interventions
opmental competence, resilience, and mal- for each unique developmental problem, a
adjustment during adolescence (Masten & single set of interventions focused on build-
Coatsworth, 1998). ing academic skills and motivation and
teaching adolescents strategies to deal with
Predictive Validity of Clusters emotional, academic, and social setbacks
The next series of analyses compared would potentially address a diversity of
these four groups, formed at the beginning problems including academic failure, tru-
of seventh grade, on their academic moti- ancy, peer problems, and so on (e.g., Haw-
vation and mental health 2 years later. We kins et al., 1988). Designing single interven-
wanted to see if the profiles found at sev- tions aimed at these kinds of outcomes
enth grade remained stable over time. Re- could eventuate in significant cost savings
sults showed the same pattern of differ- and ease of implementation in the next gen-
ences in motivation and mental health eration of school-based programs. The re-
among the groups at eighth grade as was sults for the multiple problems group also
found at seventh grade. In short, the clus- highlight the fact that for many "high-risk"
ters remained fairly stable over time, adolescents, school reforms are not going to
though the magnitude of the between- be enough to help them get back on track
group differences in motivation and mental academically and school-linked services are
health diminished over time (see top por- an important adjunct to such reform (Adel-
tion of Fig. 9). man & Taylor, 1998).
Results of the person-centered analyses
Summary also suggested that academic motivation is
Our person-centered results support the a key factor in understanding how some
idea that psychological and behavioral in- early adolescents who experience signifi-
dicators of risk, including poor motivation cant emotional distress can nonetheless stay
to learn, poor mental health, poor grades on track academically. In our study, youth
and academic failure, school misconduct, in the poor mental health group continued
and affiliations with negative peers all tend to get good grades, follow the rules in
to co-occur in what is a substantial minority school, and avoid affiliations with negative
of adolescents (Carnegie Council, 1989; peers despite feeling sad, angry, and un-
Dryfoos, 1990). These problems character- happy. Both the variable- and person-cen-
ized about one-third of our sample (e.g., tered results support the idea that feeling
Multiple Problems Group). By contrast, we academically competent and valuing school
also found that positive motivation and were two intrapsychic resources that may
mental health, positive achievement and have helped such youth overcome other life
school conduct, and abstention from affili- adversities that often threaten the attain-
ating with bad crowds co-occurred in a sub- ment of a good education. Understanding
stantial minority of positively adjusted more about how and why such students
youth. Such well-adjusted youth comprised manifest this pattern, as well as the kinds of
40% of our sample. school experiences that might promote such
That both problems and assets co-occur resilience is a fruitful avenue of future re-
among subgroups of early adolescents sug- search and one we discuss more fully below
gests that a constellation of favorable and (Luthar, 1995). At this point, it seems plau-
unfavorable developmental outcomes is sible to conjecture that educators who assist
likely linked to a common set of underlying their adolescent students in developing
458 THE ELEMENTARYSCHOOLJOURNAL
their motivation to learn as well as their ac- dle schools is a topic we turn to in the next
ademic skills and knowledge are providing section where we further explore and dis-
them with intrapsychic resources that can cuss how middle grades educators, through
protect them against social-emotional prob- particular policies and practices, may be
lems contemporaneously and in the future. able to promote academic, social, and emo-
Perhaps the best mental health intervention tional health and well-being in all students
teachers can implement in middle schools as well as counteract problems in these ar-
is good teaching. eas among particular subgroups of stu-
Finally, our person-centered results sup- dents.
port the idea that sometimes poor valuing
of school is a marker for broadband diffi- Adolescent Functioning and the
culties (e.g., the multiple problems group), Perceived School Context
and sometimes low school valuing is an in- The second focus of our research has been
dicator that students who are otherwise do- concerned with describing and understand-
ing well are just bored with their schooling ing how adolescents' experiences in middle
(e.g., the poor academic value group). Poor school are related to the quality of their ac-
valuing of school has been linked to a host ademic and social-emotional functioning
of problems that become more prevalent in within and over time. To understand how
adolescence, including the "major negative adolescents' psychosocial functioning is re-
outcomes" described by Dryfoos (1990): lated to their experiences in school, we
dropping out of school, drug and alcohol drew on contemporary theories of human
use and abuse, involvement in delinquent motivation. In particular, like others, we
activity, and pregnancy. Our results sug- have found it useful to conceptualize the
gest that knowing whether an adolescent middle school from the adolescents' per-
feels "bonded to school" is not enough to spective in terms of how well aspects of the
indicate whether he or she is at risk for school ethos provide a "fit" with their im-
other behavior problems-it depends on plicit psychological needs for competence,
whether poor school values are accompa- autonomy, and quality relationships with
nied by low perceived academic compe- others (Connell & Wellborn, 1991; Deci &
tence and frequent feelings of emotional Ryan, 1985; Eccles et al., 1993). The theo-
distress, among other things (e.g., socioeco- retical argument is straightforward: to the
nomic factors; see Roeser et al., 1998, for a extent that adolescents perceive teachers
description of demographic characteristics and school staff as providing them with op-
of this group). portunities to develop their academic and
In our discussions of middle school re- social competencies, to exercise autono-
forms, we must be balanced in attending to mous control over aspects of their learning,
the needs of high-risk students as well as and to feel cared for and supported during
those who, like these adolescents showing learning, adolescents' perceptions of their
poor valuing of school, show acceptable but academic competence, their valuing of
not optimal functioning in school. Indeed, school, and their emotional well-being
if we do so, we often find that the educa- should all be enhanced. Such psychological
tional changes that could promote optimal resources, as we have already described, in
adjustment in most youth also could serve turn promote achievement, good conduct,
as interventions for high-risk students. Af- and positive peer relationships (for review,
ter all, we do not just want high-risk stu- see Eccles, Wigfield, & Schiefele, 1998). In
dents to be ready to learn or for students to contrast, to the extent that adolescents per-
get high grades, we also want all students ceive teachers and staff creating environ-
to develop a love of learning. How to pro- ments that potentiate feelings of self-con-
mote a valuing and love of learning in mid- sciousness and incompetence, control, and
MAY 2000
SCHOOLCONTEXT 459
interpersonal anonymity or disrespect, then cents with opportunities that fulfill their
we hypothesize that adolescents will feel needs in healthy and productive ways.
less motivated to learn and more unhappy Consistent with the idea that it is in the
and will be more likely to manifest aca- "fit" or "mismatch" between personal
demic and social problems (Connell & Well- needs and social opportunities that devel-
born, 1991; Deci, Vallerand, Pelletier, & opmental outcomes are forged (Eccles &
Ryan, 1991; Eccles & Midgley, 1989). Midgley, 1989), we have assumed that ad-
This motivational perspective is similar olescents, when in school, implicitly attend
to the configural approach of Erikson, who to questions such as: (1) How do students
was concerned with how the psychological become successful in this school and am I
construct of identity, in some more or less successful here? (associated with the need
adaptive arrangement with prevailing so- for competence); (2) What kinds of oppor-
cial circumstances, produced behavioral tunities for self-expression, choice, and de-
outcomes. Erikson (1968) too focused on ba- cision making are there for me in this
sic needs and environmental affordances. school? (associated with the need for auton-
He believed that the basic needs and life omy); and (3) Do I feel cared for and re-
tasks that characterized infancy and child- spected as a person in this school by teach-
hood were later reworked during adoles- ers, staff, and other students? (associated
cence-the second time in relation to adult with the need for quality relationships).
roles, responsibilities, and social relation- Like others, we use adolescents' percep-
tions to assess the school social context. We
ships rather than childhood ones. For Er-
ikson (1968), the previous developmental developed survey measures of adolescents'
issues that were reworked during adoles- perceptions of various dimensions of school
cence revolved around trust versus mis- life that we assumed were related to either
the fulfillment or the frustration of their im-
trust; autonomy and initiative versus
shame, doubt, and guilt; and competence plicit psychological needs. We then exam-
versus inferiority. Issues of interpersonal ined how these perceived school features
were associated with psychological indica-
trust were reworked in terms of adoles-
tors of adolescents' school and social-emo-
cents' development of faith in particular in-
tional functioning over time. This reflects
stitutions and adult role models; issues of
our interest in active, meaning-making stu-
personal autonomy were reworked in terms dents whose perceptions of the school en-
of adolescents' commitments to particular
vironment, beliefs, values, and emotions
ideologies and activities; and issues of aca- mediate the influence of the actual school
demic and social competence were re-
environment on their academic and social
worked in relation to occupational and so-
behavior (Ford, 1992; Ryan & Grolnick,
cial roles toward which adolescents aspired
1986). A summary of the school perception
(Erikson, 1968). Erikson's basic themes in measures is presented in the appendix.
adolescent identity development were simi- Table 2 presents a conceptual summary
lar to the kinds of implicit, basic psycholog- of a series of variable-centered, multiple re-
ical needs motivational theorists focus on
gression analyses in which we examined
today in studying successful adolescent de- the multivariate effects of different school
velopment: the need for trusting and caring perceptions on changes in adolescents' mo-
relationships, the need for autonomous self- tivational beliefs and mental health over
expression, choice, and decision making; time. In Table 2, plus signs (+) indicate
and the need for challenge and competence school perceptions that were linked with
development (Deci & Ryan, 1985; Eccles et changes in outcomes that were develop-
al., 1993). In both approaches, adults are mentally healthy over time, whereas minus
seen as responsible for providing adoles- signs (-) indicate school experiences that
460 THE ELEMENTARYSCHOOLJOURNAL
Change in Outcomes:
Seventh to Eighth Grade
Support of competence:
Positive teacher regard
Emphasis on relative ability
Emphasis on mastery and effort
Support of autonomy:
Meaningful curricula
Class autonomy provisions
Quality of relationships:
Teacher emotional support
Negative treatment by race
Negative treatment by gender
MAY 2000
SCHOOLCONTEXT 461
middle schools as (1) defining academic the most frequent (though still relatively in-
success and supporting the development of frequent) experiences of mistreatment
competence in a noncomparative, noncom- based on their race and gender at school by
petitive way; (2) supporting autonomy teachers and other school staff. In addition,
through provision of meaningful curricula; compared to adolescents in the other
and (3) having caring, respectful teachers groups, these youth felt that their teachers
also showed improved academic, social, had the lowest regard for them academi-
and emotional functioning over time. cally, they perceived their academic work
as the least meaningful, they reported the
Patterns of Adolescent Functioning fewest opportunities for autonomy in the
and Perceived School Contexts classroom, and they felt the least emotional
The final set of findings we present here is support from teachers. In short, these alien-
concerned with how perceived school ex- ated youth reported the most alienating
periences during eighth grade differed school social ethos of any of the adolescents
among the four groups of adolescents de- in our sample. Not surprisingly, we also
scribed earlier (see Fig. 8). That is, we found that disproportionate numbers of
wanted to go beyond general relations these youth with multiple problems were in
among these variables at the aggregate sam- what could be considered low-track or re-
ple level to see if we could link specific medial classes in school based on parents'
school experiences with subgroups of ado-
reports of their course-taking patterns
lescents who were characterized by specific
(Oakes, Gamoran, & Page, 1992).
patterns of assets and vulnerabilities. Recall Although the negative psychological
that the patterns of motivation and mental state of these youth may have colored their
health that emerged during seventh grade
remained constant over the 2 years of the perceptions of school, we believe these re-
sults have substantive meaning. Other re-
study and still differentiated these four search has documented that low-achieving
groups of adolescents from one another in
terms of their school and social-emotional youth (most of whom are placed in low-
ability classes) often experience classrooms
functioning during the eighth grade. that are characterized by poor opportuni-
Figure 9 presents the clusters at eighth ties to learn, low expectations, unimagina-
grade, as well as group differences and sta- tive curricula, and teacher control (Fine,
tistical comparisons between these groups
on the school perception variables. Note 1991; Kagan, 1990; Oakes et al., 1992). Al-
that the school perception and the motiva- though grouping alienated youth together
tional and mental health measures depicted in low-track classes presents a challenge to
in Figure 9 were collected at the same time, teachers, and although misconduct and
during the end of adolescents' eighth-grade poor motivation on the part of such stu-
school year in middle school. Three find- dents can no doubt influence teachers' de-
cisions about curricula and their use of con-
ings are worth highlighting here.
First, perhaps not surprisingly, adoles- trolling practices in the classroom (Oakes et
cents characterized by "multiple problems" al., 1992; Skinner & Belmont, 1993), it also
(e.g., poor motivation to learn, poor mental seems plausible that the kinds of experi-
health, and poor academic and social be- ences reported by these students in school
havior) also reported the most consistently deepen their motivational, emotional, and
negative, developmentally inappropriate behavioral difficulties. That new teachers
school environment of any of the adoles- are often assigned to the low-track classes
cents. For instance, these youth reported the that serve such students, assignments that
most competition and greatest emphasis on are among the toughest in any school, likely
relative ability in their schools, as well as exacerbates unproductive cycles of student
462 THE ELEMENTARYSCHOOLJOURNAL
1.00
0.80
0.60
0.40
?
-e 0.20
0
E 0.00 ~. AcademicCompetence
0 AcademicValue
U EmotionalFunctioning
S-0.20a
-0.40
-0.60
-0.80
-1.00
Positive Poor School Poor Mental Multiple
Adjustment Value Health Problems
(N = 418) (N = 147) (N = 152) (N = 324)
SchoolPerceptions F-Values
c
PositiveTeacherRegard 4 09 a 3.85 b 3.96 ab 3.50 [ (3, 1031) 24 81**
FIG. 9.-Patterns of functioning at eighth grade and perceptions of school. Student Newman-Keuls com-
parisons were used to test group differences. Different superscripts for a particular variable across clusters
indicate a significant mean difference at the p < .05 level.
failure and teacher frustration (Oakes et al., earlier, African-American males and white
1992). females were also overrepresented in this
A second set of interesting findings from group. Why do these two groups feel sin-
these analyses concerned the poor mental gled out in a negative way by teachers and
health group: those adolescents who contin- school staff? Could such (perceived) mis-
ued to be engaged and do well in school treatment contribute to their high distress,
despite emotional difficulties. These adoles- especially given their commitment to
cents reported relatively more frequent ex- school? Researchers are only beginning to
periences of mistreatment by teachers or explore the effects of differential treatment
other school staff predicated on their race by race and gender on adolescent function-
or gender than did the first two groups of ing (e.g., Fine, 1991; Sadker & Sadker, 1994;
adolescents. Interestingly enough, as noted Tatum, 1997), but our results suggest that
MAY 2000
SCHOOLCONTEXT 463
MAY 2000
SCHOOLCONTEXT 465
and learning truly are, in the best-case sce- ing and competition with other students
nario, a series of "assisted performances" in (Maehr & Midgley, 1991, 1996). Educators
which the more expert assist the less expert need to ask themselves if rewarding the
to develop their skills and understandings highest-achieving students provides "role
along a fruitful line of inquiry and practice models" and thus "motivates" other stu-
(Dewey, 1902/1990; Tharp & Gallimore, dents to do well. We are unconvinced that
1988), then effective schools for early ado- it does and believe that during the self-con-
lescents necessitate effective work environ- scious early adolescent period, such prac-
ments for middle grades teachers, effective tices likely turn off rather than turn on
management environments for middle many adolescents to the process of learning
grades principals, and so on up the educa- (Eccles & Midgley, 1989; Maehr & Midgley,
tional administration hierarchy (Maehr & 1991).
Midgley, 1996; Talbert & McLaughlin, Second, once management routines
1999). have been established in the classroom, our
Educators' phenomenological experi- results suggest the importance of teachers'
ences of their work environment in terms of ability to find ways to include students'
the quality of leadership, resources, and the voices in the learning process and to pro-
student body are crucial factors that affect vide them with some choice and control
their teaching and, thereby, their adolescent over their learning. Supporting autonomy
students' experience of schooling and qual- does not mean having no structure, just as
ity of functioning (Lee, Bryk, & Smith, structure in the classroom should not ex-
1993). Thus, linking the study of adoles- clude provisions for student autonomy.
cents' experience, motivation, and behavior Teachers need to ask how their students can
in school with the study of their teachers' reasonably participate, with teacher sup-
experience, motivation, and behavior at port, in the design of tasks or in the choice
school is clearly needed to understand how of topics. Perhaps students should be in-
to create middle schools that work, middle cluded in designing the kinds of manage-
schools in which successful adolescent and ment routines that govern effective class-
adult development are seen and treated as room life. Our results suggest that such
interdependent. It seems clear that creating practices will bear fruit, though teachers'
professional work environments where ability to give away some authority in the
teachers feel supported by other profession- classroom certainly rests first and foremost
als and school leaders in relation to their on their ability to establish that authority.
own needs for competence, autonomy, and Third, teachers need to be open to the
quality relationships is essential to their de- interests and prior knowledge of their stu-
cisions to create these conditions for stu- dents so they can craft lessons that touch
dents (Deci et al., 1991; Maehr & Midgley, meaningful life events, experiences, and
1996; Tharp & Gallimore, 1988). questions that occupy adolescents in their
With that said, our results suggest sev- everyday lives during this exciting period
eral things teachers and school leaders in the life course (Dewey, 1902/1990). We
might consider in relation to their everyday have encountered middle grades teachers
practice as a means of enhancing positive who do this by surveying students at the
academic, social, and emotional develop- outset of units and who, after gathering
ment among youth and redressing prob- such data, use it in the design of their cur-
lems in high-risk youth. First, our results riculum and instruction for that unit in
highlight the need to contemplate implicit small but significant ways.
and explicit school practices, including Fourth, many of the Carnegie Council's
feedback and recognition practices, that em- (1989, 1995) recommendations on making
phasize students' relative academic stand- schools smaller through organizational ar-
MAY 2000
SCHOOLCONTEXT 467
grades unprepared for what lies ahead of Everyoneis challengedto do their very best.
them. A fundamental transformation of the (1 = not at all true, 3 = somewhat true, 5 =
MAY 2000
SCHOOLCONTEXT 469
Teachers/counselors discourage you from tak- $5,000 to 16 = more than $75,000, with each scale
ing certain classes because of your sex? value representing a $10,000 incremental range.
You are disciplined more harshly by teachers
than kids of the opposite sex?
(1 = never, 3 = a couple of times a month, 5 =
every day) References
lutionin healthand socialservicesfor children, Finn, J. (1989). Withdrawing from school. Review
youth, and families. San Francisco: Jossey- of Educational Research,59, 117-142.
Bass. Ford,M. E. (1992).Motivatinghumans:Goals,emo-
Dryfoos, J. (1998).Safepassage:Makingit through tions, and personalagency beliefs.Newbury
adolescence in a riskysociety.New York:Cam- Park, CA: Sage.
bridge University Press. Freud, A. (1936). Theego and the mechanismsof
Eccles, J. S. (1983). Expectancies, values and ac- defense. London: Hogarth.
ademic behaviors. In J. T. Spence (Ed.), The Friedman,L.J. (1999).Identity'sarchitect:
A biog-
developmentof achievementmotivation(pp. raphyof ErikH. Erikson.New York:Scrib-
283-331). Greenwich, CT: JAI. ner's.
Good, T. L., & Weinstein, R. S. (1986). Schools
Eccles, J. S., Early, D., Fraser, K., Belansky, E., &
McCarthy, K. (1997). The relation of connec- make a difference: Evidence, criticisms, and
tion, regulation, and support for autonomy new directions. American Psychologist, 41,
in the context of family, school, and peer 1090-1097.
group to successful adolescent development. Goodenow, C. (1992). Strengthening the links be-
Journalof AdolescentResearch,12, 263-286. tween educational psychology and the study
Eccles, J. S., Lord, S., & Roeser, R.W. (1996). of social contexts. EducationalPsychologist,27,
Round holes, square pegs, rocky roads, and 177-196.
sore feet: A discussion of stage-environment Hawkins, J. D., Doueck, H. J., & Lishner, D. M.
fit theory applied to families and school. In (1988). Changing teaching practices in main-
D. Cicchetti & S. L. Toth (Eds.), Rochester stream classrooms to improve bonding and
symposiumon developmental psychopathology: behavior of low achievers. American Educa-
Vol. 7. Adolescence:Opportunitiesand chal- tionalResearch
Journal,25, 31-50.
lenges (pp. 47-92). Rochester, NY: University Hechinger, J. (1993). Schools for adolescents: An
of Rochester Press. historic dilemma. In R. Takanishi (Ed.), Ad-
Eccles, J. S., & Midgley, C. (1989). Stage-environ- olescencein the 1990's:Risk and opportunity
ment fit: Developmentally appropriate class- (pp. 64-81). New York: Teachers College
rooms for young adolescents. In C. Ames & Press.
R. Ames (Eds.),Research on motivationin edu- Jessor, R. (1993). Successful adolescent develop-
cation: Goalsand cognitions (Vol. 3, pp. 13-44). ment among youth in high-risk settings.
New York: Academic Press. AmericanPsychologist,48, 117-126.
Eccles, J. S., Midgley, C., Wigfield, A., Buchanan, Kagan, D. M. (1990). How schools alienate stu-
C. M., Reuman, D., Flanagan, C., & Mac Iver, dents at risk: A model for examining proxi-
D. (1993). Development during adolescence: mal classroom variables. EducationalPsychol-
The impact of stage-environment fit on ad- ogist, 25, 105-125.
olescents' experiences in schools and fami- Lee, V. E., Bryk, A. S., & Smith, J. B. (1993). The
lies. American Psychologist, 48, 90-101. organization of effective secondary schools.
Eccles, J. S., Wigfield, A., & Schiefele, U. (1998). In L. Darling-Hammond (Ed.), Review of re-
Motivation to succeed. In W. Damon (Ed.), search in education (Vol. 19, pp. 171-267).
N. Eisenberg (Series Ed.), Handbookof child Washington, DC: American Educational Re-
search Association.
psychology(5th ed.): Vol. 3. Social,emotional, Lee, V. E.,
and personalitydevelopment (pp. 1017-1095). Croninger, R. G., Linn, E., & Chen, X.
New York: Wiley. (1996). The culture of sexual harassment in
Erikson, E. H. (1950). Childhoodand society. New secondary schools. American EducationalRe-
York: Norton. searchJournal,33, 383-417.
Erikson, E. H. (1958). Young man Luther. New Lewin, K. (1936). Principlesof a topologicalpsy-
York: Norton. chology. New York: McGraw-Hill.
Luthar, S. S. (1995). Social competence in the
Erikson, E. H. (1968). Identity, youth and crisis. school setting: Prospective cross-domain as-
New York: Norton. sociations among inner-city teens. Child De-
Erikson, E. H. (1969). Ghandi's truth. New York:
velopment,66, 416-429.
Norton.
Maehr, M. L., & Midgley, C. (1991). Enhancing
Erikson, E. H. (1970). Autobiographic notes on student motivation: A school-wide ap-
the identity crisis. Daedalus, 99, 730-759. proach. EducationalPsychologist,26, 399-427.
Erikson, E. H. (1973). Dimensions of a new identity. Maehr, M. L., & Midgley, C. (1996). Transforming
New York: Norton. schoolculturesto enhancestudentmotivation
Fine, M. (1991). Framingdropouts:Notes on the and learning. Boulder, CO: Westview.
politicsof an urbanpublichighschool.Albany, Magnusson, D., & Bergmann, L. R. (1988). Indi-
NY: SUNY Press. vidual and variable-based approaches to lon-
MAY 2000
SCHOOLCONTEXT 471
gitudinal research on early risk factors. In M. gress: Research findings and policy implica-
Rutter (Ed.), Studies of psychosocial risk: The tions. Child Development,54, 1-29.
power of longitudinal data (pp. 45-61). New Ryan, R. M., & Grolnick, W. (1986). Origins and
York: Cambridge University Press. pawns in the classroom: Self-report and pro-
Masten, A. S., & Coatsworth, J. D. (1998). The de- jective assessments of individual differences
velopment of competence in favorable and in children's perceptions. Journalof Personal-
unfavorable environments: Lessons from re- ity and Social Psychology, 50, 550-558.
search on successful children. American Psy- Sadker, M., & Sadker, D. (1994). Failing at fair-
chologist, 53, 205-220. ness: How our schools cheat girls. New York:
McLaughlin, M. W., & Irby, M. A. (1994). Urban Simon & Schuster.
sanctuaries: Neighborhood organizations Sameroff, A. J. (1983). Developmental systems:
that keep hope alive. Phi Delta Kappan, 76, Contexts and evolution. In W. Kessen (Ed.),
300-306. P. H. Mussen (Series Ed.), Handbookof child
Mead, M. (1970). Cultureand commitment:A study psychology:Vol. 1. History, theory, and meth-
of the generation gap. Garden City, NY: Nat- ods (pp. 237-294). New York: Wiley.
ural History Press. Sameroff, A. J., Seifer, R., & Bartko, W. T. (1997).
Midgley, C., & Edelin, K. C. (1998). Middle Environmental perspectives on adaptation
school reform and early adolescent well-be- during childhood and adolescence. In S. S.
ing: The good news and the bad. Educational Luthar, J. A. Burack, D. Cicchetti, & J. R.
Psychologist, 33, 195-206. Weisz (Eds.), Developmentalpsychopathology:
Nam, C. B., & Powers, M. G. (1983). The socioeco- Perspectives on adjustment, risk, and disorder
nomic approachto status measurement:With a (pp. 507-526). New York: Cambridge Uni-
guide to occupationaland socioeconomicstatus versity Press.
scores. Houston: Cap & Gown. Skinner, E. A., & Belmont, M. J. (1993). Motiva-
Noddings, N. (1992). The challenge to care in tion in the classroom: Reciprocal effects of
schools. New York: Teachers College Press. teacher behavior and student engagement
Nolen-Hoeksema, S., Girgus, J. S., & Seligman, across the school year. Journalof Educational
M. E. P. (1986). Learned helplessness in chil- Psychology, 85, 571-581.
dren: A longitudinal study of depression, Talbert, J. E., & McLaughlin, M. W. (1999). As-
achievement, and explanatory style. Journal sessing the school environment: Embedded
of Personality and Social Psychology, 51, 435- contexts and bottom-up research strategies.
442. In American Psychological Association (Ed.),
Oakes, J., Gamoran, A., & Page, R. N. (1992). Measuring environment across the life span:
Curriculum differentiation: Opportunities, Emerging methodsand concepts (pp. 197-227).
outcomes, and meanings. In P. Jackson (Ed.), Washington, DC: American Psychological
Handbookof researchon curriculum (pp. 570- Association.
608). New York: Macmillan. Tatum, B. D. (1997). Why are all the blackkids sit-
Roeser, R. W., & Eccles, J. S. (1998). Adolescents' ting togetherin the cafeteria?New York: Basic.
perceptions of middle school: Relation to Tesiny, E. P., Lefkowitz, M. M., & Gordon, N. H.
longitudinal changes in academic and psy- (1980). Childhood depression, locus of con-
chological adjustment. Journalof Researchon trol, and school achievement. Journalof Edu-
Adolescence,8, 123-158. cational Psychology, 72, 506-510.
Roeser, R. W., Eccles, J. S., & Sameroff, A. J. Tharp, R. G., & Gallimore, R. (1988). Rousing
(1998). Academic and emotional functioning minds to life: Teaching, learning, and schooling
in early adolescence: Longitudinal relations, in social context. New York: Cambridge Uni-
patterns, and prediction by experience in versity Press.
middle school. Developmentand Psychopathol- Wong, C., & Eccles, J. S. (1996, April). The effects
ogy, 10, 321-352. of perceived racial discrimination on African-
Runyan, W. M. (1982). Life histories and psychobi- American students' motivation and school
ography: Explorations in theory and method. achievement. Paper presented at the annual
New York: Oxford University Press. meeting of the American Educational Re-
Rutter, M. (1983). School effects on pupil pro- search Association, New York.