Authoritative Parenting, Psychosocial Maturity, and Academic Success among

Adolescents
Author(s): Laurence Steinberg, Julie D. Elmen and Nina S. Mounts
Source: Child Development , Dec., 1989, Vol. 60, No. 6 (Dec., 1989), pp. 1424-1436
Published by: Wiley on behalf of the Society for Research in Child Development

Stable URL: https://www.jstor.org/stable/1130932

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1130932?seq=1&cid=pdf-
reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

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 [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Society for Research in Child Development and Wiley are collaborating with JSTOR to digitize,
preserve and extend access to Child Development

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
Authoritative Parenting, Psychosocial Maturity,
and Academic Success among Adolescents

Laurence Steinberg
Temple University

Julie D. Elmen
Michael Reese Hospital and Medical Center

Nina S. Mounts

University of Wisconsin-Madison

STEINBERG, LAURENCE; ELMEN, JULIE D.; and MOUNTS, NINA S. Authoritative Parenting, Psy
social Maturity, and Academic Success among Adolescents. CHILD DEVELOPMENT, 1989, 60,1424-1
The over-time relation between 3 aspects of authoritative parenting-acceptance, psycholog
autonomy, and behavioral control-and school achievement was examined in a sample of
10-16 ,-year-olds in order to test the hypothesis that authoritative parenting facilitates, rather
simply accompanies, school success. In addition, the mediating role of youngsters' psychos
maturity was studied. Results indicate that (1) authoritative parenting facilitates adolescents' ac
demic success, (2) each component of authoritativeness studied makes an independent contribut
to achievement, and (3) the positive impact of authoritative parenting on achievement is mediate
least in part through the effects of authoritativeness on the development of a healthy sens
autonomy and, more specifically, a healthy psychological orientation toward work. Adolescents w
describe their parents as treating them warmly, democratically, and firmly are more likely than th
peers to develop positive attitudes toward, and beliefs about, their achievement, and as a co
quence, they are more likely to do better in school.

The present report extends and elabo- parenting is positively correlated with adoles-
rates on a study published in this journalcent
by school performance, whereas authori-
Dornbusch and his colleagues (Dornbusch,
tarian and permissive parenting are nega-
Ritter, Liederman, Roberts, & Fraleigh, 1987)
tively so. More specifically, the authors found
that examined the relation between parenting
that adolescents who describe their parents as
style and adolescent school performance. behaving more democratically, more warmly,
Building on the classic studies of Baumrind
and more encouraging earn higher grades in
(1971, 1973, 1978), Dornbusch and his school
col- than their peers. This finding is consis-
leagues developed several indices designed
tent with an extensive literature linking au-
to capture three prototypic patterns of par-
thoritative parenting practices to children's
enting identified by Baumrind in her earlier
psychosocial competence and well-being, vir-
studies of family interaction and its impact
tually however indexed (see Maccoby & Mar-
on children's competence-the authoritarian
tin, 1983).
pattern, the permissive pattern, and the au-
thoritative pattern. The results of Dornbusch The Dornbusch et al. study is one of a
et al.'s analyses indicated that authoritative
very small number that link parenting prac-

We are grateful to the administrators, teachers, and students of the Madison Unified School
District and to our able staff of data collectors. The work described herein was conducted during the
first author's tenure as a Faculty Scholar under the William T. Grant Foundation's Program in the
Mental Health of Children and was supported as well by grants to the first author from the National
Center on Effective Secondary Schools at the University of Wisconsin-Madison, which is sup-
ported in part by a grant from the Office of Educational Research and Improvement (grant no.
G008690007). Any opinions, findings, and conclusions or recommendations expressed in this publi-
cation are those of the authors and do not necessarily reflect the view of this agency of the U.S.
Department of Education. This paper was prepared at the Temple University Center for Research in
Human Development and Education. Please address correspondence to the first author, Center for
Research in Human Development and Education, Temple University, Philadelphia, PA 19122.

[Child Development, 1989, 60, 1424-1436. ? 1989 by the Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/89/6006-0014$01.00]

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
 Steinberg, Elmen, and Mounts 1425
tices to academic achievement during adoles- coby & Martin, 1983), most empirical studies
cence; most research on parental influences of parental influences on student achieve-
on school performance has focused on youn- ment are constrained by designs that beg the
ger children (see Hess & Holloway, 1984). causal question. As Dornbusch et al. (1987)
One of the most significant aspects of the point out, it is not clear from their study
Dornbusch et al. research is the size and whether poor achievement leads to parental
heterogeneity of their sample and the consis-authoritarianism (or permissiveness) or vice
tency of their results across ethnic, socioeco-
versa. A second aim of the present research,
nomic, and family structure groups. Because therefore, was to examine the over-time re-
the researchers studied nearly 8,000 studentslation between various components of au-
from a variety of backgrounds, they were ablethoritative parenting and school achievement
to examine within-group correlations be- and test the specific hypotheses that accep-
tween parenting practices and school success.tance, psychological autonomy, and behav-
Although there were minor variations from ioral control lead to academic success.
the general trend, the analyses indicated that
in virtually each ethnic, socioeconomic, and Finally, the empirical literature on au-
family structure group, students whose par- thoritative parenting is surprisingly devoid of
studies that examine the psychological pro-
ents were less permissive, less authoritarian,
and more authoritative performed better in cesses that may mediate the impact of par-
school than their peers. enting practices on child and adolescent com-
petence (but see Hill, 1980, for an informative
The compelling consistency of the Dorn- discussion). The Dornbusch et al. study is no
busch et al. findings warrants replication and
exception. In concrete terms, why is it that
extension, on several counts. First, because authoritative parenting is associated with bet-
their indices of parenting style combined sev-
ter school performance? One hypothesis, ex-
eral distinct aspects of parenting practices amined in the present study, is that authorita-
into composite measures, it is impossible totive parenting contributes to the psychosocial
tell whether all, or only certain, features of
development of the adolescent, which in turn
authoritative parenting contribute to aca- facilitates school success. Specifically, we hy-
demic success. Authoritative parenting pothesize
is that authoritative parenting has a
multifaceted. Maccoby and Martin (1983), for positive impact on the development of psy-
example, describe authoritative parenting as chosocial maturity, and, specifically, that ado-
combining both parental responsiveness and lescents who are more psychosocially auton-
parental demandingness. Steinberg (in press) omous than their peers are likely to do better
has suggested that three distinct features
in school. This hypothesis derives mainly
characterize this pattern of parenting: a high
from the work of Greenberger (1982), who
degree of warmth or acceptance, a high de-suggests that differences in psychosocial ma-
gree of psychological autonomy or democ-
turity (a large part of which is rooted in a
racy, and a high degree of behavioral control.
healthy sense of autonomy) may differentiate
While it is tempting to use the "Baumrin- successful from unsuccessful students above
dian" shorthand to refer to the constellationand beyond differences attributable to social
of these characteristics as "authoritative," class
it or academic ability. One reason that au-
is important to ask whether certain aspects tonomy may play an especially important role
of authoritativeness are more predictive in ofinfluencing school performance during ad-
children's competence than others. One pur- olescence is that young people, as they move
pose of the present study, therefore, was to through the school years, are expected to take
"unpack" authoritativeness into its constit-
increasing responsibility for self-manage-
uent components-here, acceptance, psycho-ment. Thus a final aim of this study is to ex-
logical autonomy, and behavioral control-amine the notion that differences in young-
and examine the independent contributionssters' sense of autonomy-here, defined in
of these components to adolescent school per-
terms of their sense of self-reliance, identity,
formance. and self-direction-explain the link between
A second issue concerns the use of a authoritative parenting and academic success.
cross-sectional versus a longitudinal research
Method
design. Questions about the benefits of cer-
tain parenting practices to school perfor- Sample
mance can only be fully examined with longi- The sample for the study is composed of
tudinal data. Although most reviews of the 120 families with a firstborn child between
literature on parent-child relations note that
the ages of 11 and 16. At the time of the first
the relationship is bidirectional (e.g., Mac-
wave of data collection, the average age of the

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
 1426 Child Development
Measures
participants was 13 years, 1 month. Participat-
ing families were selected via a three-step Parenting practices.-Three aspects of
process, beginning with a large representa-
1985 parenting practices were employed in
tive sample of nearly 900 adolescents en-the present analyses. Adolescents' characteri-
rolled in the Madison (Wisconsin) school dis-zations of their parents' levels of acceptance
trict, who were surveyed in their classrooms,and psychological autonomy were derived
and ending with a subsample of 157 families from acceptance and psychological control
with firstborn adolescents, who were sur- subscales of the revised version of the Child
veyed in their homes in 1985 and again in Report of Parent Behavior Inventory (CRPBI).
1986. Participation rates in each phase of the
The CRPBI is a widely used Likert-scale for-
data collection were adequate: 94% of allmat report of parental disciplinary practices
youngsters attending school on the day of thethat yields separate measures for several
survey participated in the research, 70% aspects
of of the mother-child and father-child
the families contacted by our research staff relationships. As reported in Schwartz, Bar-
agreed to participate in the study in 1985, and
ton-Henry, and Pruzinsky (1985), both the ac-
75% of these families agreed to participate ceptance
in and psychological control subscales
the follow-up 1 year later. Data on achieve- have an alpha coefficient in excess of .80.
ment test scores, which were used as a (Scale scores for psychological control were
covariate in the analyses (see below), were reversed to yield the measure of psychologi-
available for only 120 of the 157 students, lim-
cal autonomy.) For adolescents living with
iting the size of the sample to this maximum. two parents, acceptance and psychological
autonomy scores for mother and father were
As a group, the participating families did
averaged; for adolescents living with one par-
not differ from the eligible nonparticipants on
any of the demographic variables assessed ent, scores for that single relationship were
used.
(socioeconomic status, household composi-
tion, maternal employment status, ethnicity). Adolescents' reports of their parents' use
Comparisons between the study sample and of behavioral control were derived from a
the school district's population, based on de-
checklist concerning 17 areas of family deci-
mographic data provided by the district, indi-
sion making on issues relevant to children in
cate that the school survey sample and family
the age range studied (e.g., curfew, spending
subsample are representative of the districtmoney, leisure activities, completing school
population as a whole. The study sampleassignments).
is For each item, youngsters indi-
evenly divided by sex, predominantly white cated whether their parents dictated how the
(88%), socioeconomically heterogeneous
adolescent should behave, asked the young-
(39% blue collar, 37% white collar, 24% pro-ster's opinion but retained the final say on the
fessional, as determined by parental occupa-matter, or left the decision entirely up to the
tion), from a variety of family structures (64%
youngster. Following both Dornbusch et al.
biologically intact, 21% single-parent, 15%
(1985) and Steinberg (1987), a score repre-
stepfamily), and with a variety of maternal
senting parental permissiveness was cal-
work patterns (58% employed full-time, 31% culated by summing the number of items for
employed part-time, 11% not working). which the child reported having complete de-
Procedure cision-making freedom. Because high scores
reflect lax control, scale scores were reversed
Data on family relations and psychosocial
maturity were collected from the adolescents in analyses examining the effects of behav-
ioral control.
during school and home visits conducted be-
tween April and June 1985 and again 1 year Psychosocial maturity.-Three 10-item
later. Students also provided information on subscales of the autonomy scale of the Psy-
parental occupation and household composi- chosocial Maturity Inventory (Form D)
tion, which were used to index family socio- (Greenberger, Josselson, Knerr, & Knerr,
economic status and family structure, each of 1974) were completed by each adolescent in
which was treated as a three-level variable both 1985 and 1986. The work orientation
(blue collar, white collar, professional; biolog-subscale assesses the adolescent's work skills,
ically intact, single-parent, stepfamily) andaspirations for competent work performance,
used as a covariate in the analyses (see be-and capacity to experience pleasure in work.
low). Data on youngsters' school grades andA sample item from this subscale, reverse
their scores on standardized achievement scored, is "I find it hard to stick to anything
tests (the California Achievement Tests) were that takes a long time to do." The self-reliance
obtained from official school records. subscale taps three related characteristics: the

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
 Steinberg, Elmen, and Mounts 1427
TABLE 1

MEANS AND STANDARD DEVIATIONS OF MAJOR VARIABLES

Variable M SD

GPA 1985 ................................. 2.69 .90

GPA 1986 .................................. 2.69 .90
California achievement test (%) .............. 59.53 24.81
Acceptance ................................. 18.16 2.94
Psychological autonomy ..................... 7.94 2.81
Behavioral control .......................... 10.57 2.75
Psychosocial maturity 1985 .................. 87.52 13.97
Psychosocial maturity 1986 .................. 90.93 11.71
Self-reliance 1986 ........................... 30.75 4.31
Identity 1986 ............................... 31.80 4.68
Work orientation 1986 ....................... 28.27 4.90

absence of excessive dependenceschool on others,

year; in thea middle and elementary
sense of control over one's life, schools,
and initiative.
they are graded quarterly. In the
A sample item from this subscale, reverse
present analyses, all English and math grades
scored, is "When things have gone wrong
(i.e., either four orfor
eight different grades) for a
me, it is usually because of something I
given year were averaged to yield one com-
couldn't do anything about."posite Finally, the
grade-point average for that academic
identity subscale assesses theyear, adolescent's
on a 4-point scale.' Data on grades were
available
sense of self-esteem, concern with for both
life 1985 and 1986.
goals,
internalization of values, and clarity of self-
concept. A sample item, reverse scored, Achievement is "Itest scores.-Each stu-
am not really accepted and liked." The on
dent's scores the verbal and mathematics
sub-
scales have adequate internal California
consistency Achievement Tests were averaged,
and the result was
(the alphas for the work orientation, self-used as a covariate in all
reliance, and identity scales areanalyses.2
.78, .76,In order
and to increase sample vari-
abilityreliability
.81, respectively) and test-retest on this measure, percentile scores
(see Greenberger & Bond, 1976, based on district
for norms, rather than national
details).
Scores on the three subscales are summed to norms, were used. For all fifth and sixth grad-
yield an overall autonomy score. ers, fifth-grade achievement scores were
used; for eighth and ninth graders, eighth-
School grades.-Information on adoles-grade test scores were used.
cents' school grades was obtained from
official school records provided by the school Means and standard deviations for all ma-
district. In order to permit comparability jor variables used in the present analyses
across grade levels, only English and mathe- are presented in Table 1; zero-order correla-
matics grades were used. Students in the tions among these variables are presented in
Madison high schools are graded twice each Table 2.

1 Information on the relative difficulty of each student's courses (i.e., whether the course was
advanced, average, or remedial) was not available for grades prior to ninth grade. Consequently, it
was not possible to adjust students' grade-point averages for this factor.
2 It is not entirely clear whether one should control for scores on standardized tests of achieve-
ment in studies of student GPA. On the one hand, because achievement tests are in some sense an
index of academic performance not unlike grades, partialing out the variance shared between the
two renders the new outcome variable (i.e., GPA with achievement test scores controlled) difficult
to interpret. On the other hand, however, scores derived from measures like the California Achieve-
ment Tests are a closer index of academic ability than is GPA, and because academic ability is likely
to be correlated with both school performance and parenting practices, it is important to ensure that
ability and performance are not confounded in the analyses. Accordingly, all analyses were con-
ducted with and without controlling for student scores on standardized tests of achievement. Al-
though the magnitude of some of the regression coefficients was affected by this procedure, none of
the findings changed substantively (see Tables 3 and 4). For the sake of consistency, all results
presented employed the more conservative approach, with controls for achievement test scores
included.

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
1428 Child Development
TABLE 2

ZERO-ORDER CORRELATIONS AMONG MAJOR VARIABLES

(1) (2) (3) (4) (5)

1. GPA 1985 ....................

2. GPA 1986 ................... .69*
3. California Achievement Test .. .68** .65**
4. Acceptance .................. .21 .34** .21
5. Psychological autonomy ....... .39** .36** .27* .19*
6. Behavioral control ............ .07 .18 .03 .10 -.11
7. Psychosocial maturity 1985 ..... 32** .33** .37** .32** .44**
8. Psychosocial maturity 1986 .... 27* .47** .30** .24* .37**
9. Self-reliance 1986 ............ .27* .41** .29** .09 .32**
10. Identity 1986 ................. .17 .33** .20 .24* .28**
11. Work orientation 1986 .........31** .50** .32** .32** .37**

(6) (7) (8) (9) (10)

7. Psychosocial maturity 1985

8. Psychosocial maturity 1986 ..... 23* .63**
9. Self-reliance 1986 ............ .00 .57** .80**
10. Identity 1986 ................. .28** .51** .85** .53**
11. Work orientation 1986 ........ .23* .56** .86** .53** .60**

NOTE.-N varies from 104 to 119.

* p < .05, two-tailed.
** p < .01, two-tailed.

Plan of Analysis test of the effects of authoritative parenting

Because psychosocial maturity, parentingpractices on changes in school performance
practices, and school performance all vary as
over
a the 1-year period.
function of adolescent age and sex and family
socioeconomic status and composition, all Results
analyses control for the effects of sex, age (in
months), socioeconomic status, and family Over-Time Relation between Parenting
Practices and GPA
structure. In addition, all analyses control for
youngsters' achievement test scores (see foot- In the first set of analyses, adolescents'
note 2). 1986 grade point average was regressed on
the three indices of authoritative parenting
Researchers continue to debate the ap- measured in 1985-acceptance, psychologi-
propriateness of various approaches to the cal autonomy, and behavioral control-while
analysis of longitudinal data in the study of controlling for the mediating effect of the stu-
child development (e.g., Connell & Tanaka, dent's GPA in 1985. The results are displayed
1987). Despite recent advances in structural in Figure 1, which shows all paths significant
equation modeling, it is still generally agreed at p < .10 or better, and in Table 3, which
that the use of multiple regression techniques presents the results of the overall regression
in which one predicts scores on a dependent analysis.
variable at time 2 while controlling for scores
on that same variable at time 1 is an appropri- Not surprisingly, the relation between
ately conservative strategy. Accordingly, the 1985 and 1986 school performance is highly
data were analyzed in a series of path models significant. More important, the results also
employing simultaneous regression analyses, indicate that all three aspects of authoritative
which examined the relations between 1985 parenting lead to increases in school grades.
Specifically, youngsters who in 1985 de-
parenting practices and 1986 school perfor-
mance while controlling for the effects of their parents as granting them greater
scribed
1985 school performance.3 This amounts to a
psychological autonomy and exercising firmer

3 Occasionally, it is suggested that moderate, rather than high, levels of warmth and control are
most predictive of adolescent competence. Accordingly, all analyses examined the curvilinear as
well as linear relations between the three parenting variables and school achievement. In no case
were significant curvilinear effects found.

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
 Ach. Test Scores **
Sex
SES
Family Structure
.127+ Age
Parental
Acceptance
(1985)

.148*

t213**
Parental

Psychological .188* GPA (1985) .363*** GPA (1986)

(1985)

220** R2= .62, F (11,92)= 13.64****

Parental
Behavioral
Control
'1985)
+ p<.10
p<.05
** p<.01
**** p<.(X)1
p**** <.(XX)l

FIG. 1.-Direct and indirect influence of 1985 parenting practices on 1986 GPA

TABLE 3

REGRESSION OF 1986 GPA ON 1985 GPA AND 1985 PARENTING PRACTICES

Independent Variable B SE B Beta T

GPA 1985 .................. .367 .098 .363 3.73***

CAT ...................... .011 .003 .300 3.26**
Sex ......................... -.007 .124 -.004 -.06
Age ....................... -.002 .003 -.041 -.60
SES1 ........................ .072 .161 .034 .45
SES2 ....................... .221 .138 .118 1.60
FAM1 ............... ..... .... .159 .164 .071 .97
FAM2 ..................... -.268 .172 -.107 -1.56
Acceptance ................. .039 .022 .127 1.77+
Psychosocial autonomy ...... .048 .024 .148 1.96*
Behavioral control ........... -.047 .023 .142 2.01*
(Constant) ................. 1.494 .916 1.63

Summary statistics:
Multiple R = .78733
R2 = .61989
Adjusted R2 = .57444
Standard error = .59006
F(11,92) = 13.64****

NOTE.-CAT = California Achievement Test scores; SES1 and SES2 and FAM1
and FAM2 represent effect-coded vectors used to control for the effects of socioeconomic
status and family structure (each a three-level variable).
+ p < .10.
* p < .05.
** p < .01.
*** p < .001.
**** p < .0001.

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
1430 Child Development

GPA (1985) ****

Ach. Test Scores **
Sex
SES
Family Structure
Age

Parental
Acceptance
(1985)

.209**

Parental Psychosocial
Psychological 327*** Maturity .221** GPA (1986)
Autonomy (1986)
(1985)

R2 = .65, E (12,91) = 14.24****

-.225**

Parental
Behavioral
Control

(1985) + <.05 p<.O5

** p<.01
**** p<001
**** p<.0001

FIG. 2.-Influence of parenting practices on GPA is mediated by adolescent psychosocial maturity

control over their behavior showed greater in- turity was examined in order to test the
creases in grades over the 1-year period than hypothesis that the positive impact of
their peers, and adolescents who described authoritative parenting practices on school
their parents as more accepting tended to do performance is mediated by their effects on
better in school as well.4 In view of the fact youngsters' psychological development. In
that these results hold even after controllingthese analyses, which controlled not only for
for achievement test scores and an array of the set of covariates listed earlier but for 1985
demographic factors correlated with par-GPA as well, students' 1986 GPA was re-
enting practices and school performance, thegressed simultaneously on the three par-
findings present strong evidence of the im- enting dimensions (assessed in 1985) and on
pact of authoritative parenting on school suc- their total score on the autonomy subscale of
cess. Interestingly, whereas the positive the Psychosocial Maturity Inventory. This
impact of psychological autonomy on sub-score represents the sum of students' scores
sequent school performance is both direct on the scales measuring self-reliance, iden-
as well as indirect (through its concurrent re-tity, and work orientation. These results are
lation to school performance), the over-time presented in Figure 2 and Table 4.
effects of parental acceptance and firm control
The results confirm the hypothesis that
on 1986 student performance do not appear to
the impact of authoritative parenting on
be mediated by their relation to 1985 grades.
school success is mediated through its effects
The Mediating Impact of on psychosocial maturity. As expected, given
Psychosocial Maturity the extensive literature documenting the
In the second set of analyses, the mediat- positive contributions of warmth, firm control,
ing impact of youngsters' psychosocial ma-and psychological autonomy to adolescents'

4 Analyses were conducted in order to examine whether the effects of various parenting prac-
tices are moderated by the sex or age of the adolescent. Significant interaction terms were obtained
between sex and each of the three parenting dimensions and between age and both psychological
autonomy and behavioral control. Unfortunately, follow-up analyses were constrained by the small
subsample sizes. Informal inspection of the regression coefficients suggested that the positive im-
pact of parental acceptance appears stronger among girls than boys, whereas the positive impact of
behavioral control is stronger among boys than girls. The positive effects of psychological autonomy
and of behavioral control appear to be stronger among older adolescents than younger ones. These
intuitively sensible findings warrant further research in a larger sample.

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
 Steinberg, Elmen, and Mounts 1431
TABLE 4

REGRESSION OF 1986 GPA ON 1985 PARENTING PRACTICES

AND 1986 PSYCHOSOCIAL MATURITY

Independent Variable B SE B Beta T

GPA 1985 ..................385 .095 .381 4.07****

CAT ......................009 .003 .256 2.85**
Sex ........................ -.039 .119 -.021 -.33
Age ....................... -.003 .003 -.069 -.98
SES1 ..................... .050 .155 .024 .32
SES2 ..................... .124 .137 .066 .90
FAM1 .................... .197 .159 .088 1.24
FAM2 .................... -.266 .166 -.106 -1.60
PSM ..................... .017 .005 .221 2.93**
Acceptance ............... .026 .022 .086 1.22
Psychosocial autonomy ..... .024 .025 .076 .98
Behavioral control .......... .028 .023 .085 1.20
(Constant) ................. .080 1.005 ... .08

Summary statistics:
Multiple R = .80783
R2 = .65259
Adjusted R2 = .60678
Standard error = .56719
F = 14.24****

NOTE.-CAT = California Achievement Test scores; PSM = Psychosocial Maturity

Inventory; SES1 and SES2 and FAM1 and FAM2 represent effect-coded vectors used to
control for the effects of socioeconomic status and family structure (each a three-level
variable).
+ p < .10.
* p < .05.
** p < .01.
*** p < .001.
**** p < .0001.

psychological and social functioning (see vailing theories of socialization, the analyses
Maccoby & Martin, 1983), each of the par-leave open the possibility that youngsters
enting dimensions assessed contributes inde-who are relatively more emotionally mature
pendently to youngsters' scores on the mea- merely provoke more positive behavior from
sure of psychosocial maturity. And, not their parents. If this were true, one could very
surprisingly, psychosocial maturity is itself reasonably suggest that the observed relation
significantly related to school success (seebetween authoritative parenting and GPA is
Greenberger, 1982). As Figure 2 shows, how-actually due to the impact of psychosocial ma-
ever, when psychosocial maturity is includedturity both on youngsters' performance in
in the model, the direct paths between 1986 school and their parents' performance at
school performance and 1985 parental ac-home. In view of debates over the direction of
ceptance, psychological autonomy, and firm socialization effects, especially in studies of
control are reduced to nonsignificance. Ap- the effects of firm control (e.g., Lewis, 1981),
parently, school success is enhanced byit is important to examine the alternative
authoritative parenting at least in part be- causal explanation.
cause authoritative parenting contributes to
youngsters' psychosocial development. To this end, a model was examined that
included both 1985 and 1986 scores on the
Psychosocial Maturity and Authoritative Psychosocial Maturity Inventory. In these
Parenting: Direction of Effects analyses, the central question of interest was
The previous analyses indicate that au-whether 1985 parenting practices had a sig-
thoritative parenting has a positive impact onnificant effect on 1986 psychosocial maturity
psychosocial maturity, which in turn is posi- after taking into account 1985 psychosocial
tively correlated with school performance. Al-maturity. The results are presented in Figure
though the ostensible causal chain is both3 and Table 5. As was the case in previous
intuitively sensible and consistent with pre-figures, only the significant paths are drawn.

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
1432 Child Development

GPA (1985)****
Ach. Test Scores**
SES
Sex
Family Sturcture
Age

-.168+

Psychosocial .340**** Parental Psychosocial

Maturity Psychological .152*
(1985) Autonomy Maturity .313"** . GPA (1986)
(1986)
(1985)

-.204*

I2 R2 = .67, E(
.295*** Parental270***
Behavioral .270**
Control
(1985)

Parental
Acceptance
(1985) + p<.10
p<.05
**p<.O1
**** p<.(XX)1

FIc. 3.-Authoritative parenting practices influence psychosocial maturity, which in turn influences
GPA.

influence, it is interesting to ask whether the

The analyses suggest that the relation be-
tween authoritative parenting and psychoso- positive impact of authoritative parenting on
adolescent school success is mediated more
cial maturity is reciprocal, although different
parenting dimensions show different patterns through certain aspects of psychosocial func-
tioning than through others. The analyses pre-
of relations. Parental acceptance appears to be
more a correlate of youngsters' psychosocialsented in Figure 4 and Table 6 address this
question. In these analyses, a model that in-
maturity than a contributor to it. Firm control
appears to be more a contributor to psychoso- cludes separate scores on the scales tapping
cial maturity than a correlate of it. And psy-self-reliance, work orientation, and identity,
chological autonomy appears to be both in
a place of the global index of psychosocial
correlate and antecedent of youngsters' psy-
maturity, was examined.
chosocial maturity. Overall, however, the
The analyses help to clarify the psycho-
analyses support the hypothesis that authori-
logical pathways through which authoritative
tative parenting enhances youngsters' psycho-
parenting enhances adolescents' grades in
social development, which in turn contributes
school. Specifically, the results show that all
to their success in school. In particular, the
three aspects of authoritative parenting may
results point to the benefits of a parent-child
enhance youngsters' work orientation, which
relationship characterized by the combination in turn contributes to school success. Al-
of psychological autonomy and firm behav-
ioral control. though self-reliance is also related to school
success, its development is not directly re-
The Relative Importance of Self-Reliance, lated to the parenting practices studied (al-
Work Orientation, and Identity though it may be affected indirectly, as the
The measure of psychosocial maturity figure indicates). And although firm control is
employed in this study is an aggregate of positively related to adolescents' identity de-
youngsters' scores on three related, but con- velopment, scores on the measure of identity
ceptually distinct, dimensions: self-reliance, are not directly related to GPA (although
work orientation, and identity. Given theoret- identity development may contribute to GPA
ical concerns about the mechanism through indirectly, through its impact on work orienta-
which authoritative parenting may exert its tion or self-reliance). In any event, it appears

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
 Steinberg, Elmen, and Mounts 1433
TABLE 5

REGRESSION OF 1986 GPA ON 1985 PARENTING PRACTICES

AND 1985 AND 1986 PSYCHOSOCIAL MATURITY

Independent Variable B SE B Beta T

GPA 1985 ................. .390 .093 .386 4.17***

CAT ....................... .010 .003 .276 3.09**
Sex ....................... -.050 .118 -.028 -.43
Age ...................... -.002 .003 -.062 -.89
SES1 ..................... .074 .154 .035 .48
SES2 .................... .101 .136 .054 .74
FAM1 .................... .160 .158 .072 1.01
FAM2 .................... -.270 .163 -.108 -1.65
Acceptance ................ .034 .022 .119 1.58
Psychosocial autonomy ..... .031 .025 .095 1.24
Behavioral control .......... .019 .023 .057 .80
PSM 1985 ................ -.011 .005 -.168 -1.87+
PSM 1986 ................. .024 .006 .313 3.50***
(Constant) ................ .185 .993 ... .19
Summary statistics:
Multiple R = .81580
R2 = .66553
Adjusted R2 = .61722
Standard error = .55962
F(13,90) = 13.76****

NOTE.-CAT = California Achievement Test; PSM = Psychosocial Maturity In-

ventory; SES1 and SES2 and FAM1 and FAM2 represent effect-coded vectors used to
control for the effects of socioeconomic status and family structure (each a three-level
variable).
+p < .10.
* p < .05.
** p < .01.
*** p < .001.
**** p < .0001.

from these analyses that the most direct psy-warmly, democratically, and firmly are more
chosocial mechanism through which authori- likely than their peers to develop positive
tative parenting may encourage school suc-attitudes toward, and beliefs about, their
cess is through the effects of such parentingachievement, and as a consequence, they are
practices on youngsters' work orientation. more likely to do better in school. The corrob-
oration of Dornbusch et al.'s (1987) cross-
Discussion sectional findings with longitudinal data indi-
cates that parental authoritativeness is not
The findings of the present study both
merely a response to youngsters' demon-
corroborate and extend those of previous re-
strated school success and calls into question
search. Specifically, the results indicate that
the notion that parental authoritativeness
authoritative parenting likely facilitates ado-
merely follows from, rather than precedes, the
lescents' academic success; that all three com-
ponents of authoritativeness studied here-development of competence in children (e.g.,
Lewis, 1981).
parental acceptance, psychological autonomy,
and behavioral control-make independent
It is important to keep in mind, of course,
contributions to school achievement; and that
that the information on parenting practices
the positive impact of authoritative parenting was obtained from adolescents and not
on school success is mediated in part through through objective observations of parent-chil
the effects of authoritativeness on the devel- interaction. Thus, we can only say with cer-
opment of a healthy sense of autonomy and, tainty that adolescents who feel that their par
more specifically, on the development of ents are accepting, democratic, and firm out
a healthy psychological orientation toward perform their peers in school. Because the
work. Put in concrete terms, adolescents who information on school performance does
describe their parents as treating them come from official school records, however,

This content downloaded from

31.205.11fff:ffff:ffff:ffff on Thu, 01 Jan 1976 12:34:56 UTC
All use subject to https://about.jstor.org/terms
 GPA (1985)****
Ach. Test Scores**
S1986 Sex
SES
Age
Parental Self Family Structure
Acceptance Reliance

.208** .348***

Parental .147* Work .198*

Psychological Orientation GPA(1986)
Autonomy

.331***
R2 = .67, E(14,89)= 12.81***

-.230** .304***

Parental

Behavioral
Control .193** Identity
* + p<.10
p<.05
** p<.01
**** p<.001
**** p<0001

FIG. 4.-Influence of parenting practices on GPA is mediated primarily by adolescent work orien-
tation.

TABLE 6

REGRESSION OF 1986 GPA ON 1985 PARENTING PRACTICES AND 1986

SCORES ON WORK ORIENTATION, SELF-RELIANCE, AND IDENTITY

Independent Variable B SE B Beta T

GPA 1985 .................. .361 .094 .357 3.85***
CAT ...................... .008 .003 .241 2.71**
Sex ........................ -.033 .120 -.018 -.28
Age ....................... -.003 .003 -.083 -1.17
SES1 ..................... .074 .154 .035 .48
SES2 ..................... .154 .136 .082 1.13
FAM 1 ..................... .184 .157 .082 1.17
FAM2 .................... -.292 .165 -.117 -1.77+
Acceptance ................. .025 .022 .081 1.14
Psychosocial autonomy ..... .022 .024 .068 .88
Behavioral control .......... .033 .023 .109 1.43
Self-reliance ............... .030 .018 .142 1.70
Work orientation ........... .036 .016 .198 2.24*
Identity ................... -.012 .016 -.061 -.71
(Constant) .................. .280 .995 ... .28

Summary statistics:
Multiple R = .81752
R2 = .66834
Adjusted R2 = .61617
Standard error = .56038
F(14,89) = 12.81****

NOTE.-CAT = California Achievement Test scores; SES1 and SES2 and FAM1
and FAM2 represent effect-coded vectors used to control for the effects of socioeconomic
status and family structure (each a three-level variable).
+ p < .10.
* p < .05.
** p < .01.
*** p < .001.
**** p < .0001.

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
 Steinberg, Elmen, and Mounts 1435
we can rule out the possibility that the stu- practices have similar benefits in other popu-
dents who describe their parents in more lations.
positive light are simply more likely to de-
scribe their own behavior more positively as
well. References
The findings concerning the psychosocial Baumrind, D. (1971). Current patterns of parental
mediators of the relation between authorita-
authority. Developmental Psychology Mono-
tive parenting and academic performance are graphs, 4, 1-103.
especially interesting in light of controversies Baumrind, D. (1973). The development of instru-
over the putative contribution of self-esteem mental competence through socialization. In
to youngsters' academic success (see Harter, A. Pick (Ed.), Minnesota symposium on child
1983). Despite widespread public belief psychology (Vol. 7, pp. 3-46). Minneapolis:
about the benefits of raising adolescents' self- University of Minnesota Press.
esteem as a means of enhancing their school Baumrind, D. (1978). Parental disciplinary patterns
performance, no systematic research has cor- and social competence in children. Youth and
roborated this notion. The present study sug- Society, 9, 239-276.
gests that the relation between youngsters' Connell, J., & Tanaka, J. (Eds.). (1987). Structural
self-esteem (as indexed via the identity mea- equation modeling. [Special section of issue].
sure) and their school performance is, at best, Child Development, 58(1).
indirect. In contrast, it appears that changing Dornbusch, S., Carlsmith, J., Bushwall, S., Ritter,
adolescents' motivation to work hard and
P., Leiderman, P., Hastorf, A., & Gross, R.
strive for success (tapped here by the work (1985). Single parents, extended households,
orientation measure) holds the most promise and the control of adolescents. Child Develop-
for psychological interventions designed to ment, 56, 326-341.
enhance school performance among adoles- Dornsbusch, S., Ritter, P., Liederman, P., Roberts,
cents.
D., & Fraleigh, M. (1987). The relation of par-
Whether interventions designed to en-enting style to adolescent school performance.
hance schQol performance should includeChild Development, 58, 1244-1257.
attempts to modify parenting practices is Greenberger,
a E. (1982). Education and the acquisi-
different matter, however. An important tion of psychosocial maturity. In D. McClel-
limitation of the present study concerns its land (Ed.), The development of social maturity
sample. Unlike the Dornbusch et al. (1987) (pp. 155-189). New York: Irvington.
effort, the present research focused on an Greenberger, E., & Bond, L. (1976). Technical
ethnically homogeneous group of adoles- manual for the Psychosocial Maturity Inven-
cents, and it is important to exercise caution tory. Unpublished manuscript, Program in So-
in generalizing the findings. Virtually all of cial Ecology, University of California, Irvine.
the participants were white, and a majority Greenberger, E., Josselson, R., Knerr, C., & Knerr,
were from middle-class or professional back- B. (1974). The measurement and structure of
grounds. To date, the benefits of parental au- psychosocial maturity. Journal of Youth and
thoritativeness have been most consistently Adolescence, 4, 127-143.
demonstrated in studies of this specific popu- Harter, S. (1983). Developmental perspectives on
lation (Steinberg, in press); even in the Dorn- the self-system. In E. M. Hetherington (Ed.),
busch study, it was found that the positive P. H. Mussen (Series Ed.), Handbook of child
effects of authoritativeness were greatest for psychology: Vol. 4. Socialization, personality,
white youth. Although there is certainly room and social development (pp. 275-385). New
for more systematic study of authoritative par- York: Wiley.
enting and child development in samples Hess, R., & Holloway, S. (1984). Family and school
of middle-class, white youngsters, it would as educational institutions. In R. Parke (Ed.),
seem far more important, given the changing Review of child development research (Vol. 7,
demography of American adolescents (Wet- pp. 179-222). Chicago: University of Chicago
zel, 1987), to see whether patterns of social- Press.

ization effects observed in white, middle- Hill, J. (1980). The family. In M. Johnson (Ed.),
class samples hold true in other groups. One Toward adolescence: The middle school years.
must therefore resist the temptation to con- (Seventy-ninth yearbook of the National Soci-
clude on the basis of these data that large- ety for the Study of Education) (pp. 32-55).
scale efforts should be implemented in order Chicago: University of Chicago Press.
to increase parental authoritativeness. Al-Lewis, C. (1981). The effects of parental firm con-
though such a strategy may hold promise trol. Psychological Bulletin, 90, 547-563.
among white, middle-class families, it re- Maccoby, E., & Martin, J. (1983). Socialization in
mains to be seen whether the same parenting the context of the family: Parent-child interac-

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms
1436 Child Development
tiQn. In E. M. Hetherington (Ed.), P. H. Mus- social peer pressure. Child Development, 58,
sen (Series Ed.), Handbook of child psychol- 269-275.
ogy: Vol. 4. Socialization, personality, andSteinberg, L. (in press). Interdependency in the
social development (pp. 1-101). New York: family: Autonomy, conflict, and harmony in the
Wiley. parent-adolescent relationship. In S. Feldman
Schwartz, J., Barton-Henry, M., & Pruzinsky, T. & G. Elliot (Eds.), At the threshold: The de-
(1985). Assessing childrearing behaviors: A veloping adolescent. Cambridge, MA: Harvard
comparison of ratings made by mother, father, University Press.
child, and sibling on the CRPBI. Child Devel- Wetzel, J. (1987). American youth: A statistical
opment, 56, 462-479. snapshot. New York: William T. Grant Foun-
Steinberg, L. (1987). Single parents, stepparents, dation Commission on Work, Family, and Citi-
and the susceptibility of adolescents to anti- zenship.

This content downloaded from

31.205.114.218 on Wed, 30 Nov 2022 19:12:40 UTC
All use subject to https://about.jstor.org/terms