Behavioral Competence among Mothers of Infants in the First Year: The Mediational

Role of Maternal Self-Efficacy

Author(s): Douglas M. Teti and Donna M. Gelfand
Source: Child Development , Oct., 1991, Vol. 62, No. 5 (Oct., 1991), pp. 918-929
Published by: Wiley on behalf of the Society for Research in Child Development

Stable URL: https://www.jstor.org/stable/1131143

REFERENCES
Linked references are available on JSTOR for this article:
https://www.jstor.org/stable/1131143?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

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
Behavioral Competence among Mothers of
Infants in the First Year: The Mediational Role
of Maternal Self-Efficacy

Douglas M. Teti
University of Maryland Baltimore County

Donna M. Gelfand

University of Utah

TETI, DOUGLAS M., and GELFAND, DONNA M. Behavioral Competence among Mothers
in the First Year: The Mediational Role of Maternal Self-Efficacy. CHILD DEVELOPME
62, 918-929. This study tests the idea that mothers' self-efficacy beliefs mediate the e
parenting behavior of variables such as depression, perceptions of infant temperame
culty, and social-marital supports. Subjects were 48 clinically depressed and 38 non
mothers observed in interaction with their 3-13-month-old infants (M = 7.35 month
dicted, maternal self-efficacy beliefs related significantly to maternal behavioral com
independent of the effects of other variables. When the effects of self-efficacy were co
parenting competence no longer related significantly to social-marital supports or ma
pression. In addition, maternal self-efficacy correlated significantly with perceptions
difficulty after controlling for family demographic variables. These results suggest that
self-efficacy mediates relations between maternal competence and other psychosocial
and may play a crucial role in determining parenting behavior and infant psychosocia

efficacious individuals tend to persist in

Perceived self-efficacy beliefs concern
judgments of one's ability to perform given
compe-task until success is achieved, whereas
self-inefficacious
tently and effectively in a particular task or individuals give up prema
setting. Bandura (1982, 1986, 1989) hasturely.
iden- In turn, self-efficacy beliefs are en-
tified self-efficacy beliefs as centralhancedto the or decreased, respectively, by suc
understanding of individuals' transactions cess or failure experiences. Performance
with their environments and as a core con- attainments are viewed as having the strong
struct that mediates relations between est impact on self-efficacy beliefs, but othe
knowledge and behavior. For example, sources of information such as vicarious ex-
an
individual may have the knowledgeperiences, needed social persuasion, and emotional
to console a distressed infant, but arousal may be can also be influential (Bandura,
unable to do so because of self-doubt (Bu-
1982, 1986). In the present study, the paren-
genthal, 1987). Since Bandura's (1977, 1982)
tal competence of mothers of young infants
initial formulations, much research was examined
has in- in relation to the mothers'
vestigated the influential role of perceived
feelings of efficacy in the maternal role and
self-efficacy beliefs in guiding behavior, to selected
and maternal, environmental, and
the impact of performance outcomes child on
characteristics.
self-efficacy beliefs. Bandura (1982, 1989)
and Litt (1988) have suggested that the rela- Surprisingly few studies have examined
tion between self-efficacy and performance maternal self-efficacy beliefs as they relate
is best conceptualized as bidirectional. to theSelf-
mother-infant relationship, perhaps

This research was supported by NIMH grant 41474 awarded to the authors. The authors
thank project director Janiece Pompa and research assistants Cathie Fox, Kim Shimoda, Shari
Harbaugh, Heather Walker, and Leslie Poppe. We also thank Donald P. Hartmann and Penny
Jameson for comments on an earlier draft of this paper, and Stanley Feldstein for his statistical
advice. We are indebted to Salt Lake Valley Mental Health Services, participating practitioners,
and John E. Brockert, Director, Bureau of Vital Records, Utah Department of Health, for assis-
tance in subject recruitment. Finally, special thanks are given to the families who participated in
this study. Reprint requests should be addressed to Douglas M. Teti, Department of Psychology,
University of Maryland Baltimore County, Baltimore, MD 21228.

[Child Development, 1991, 62, 918-929. ? 1991 by the Society for Research in Child Development, Inc.
All rights reserved. 0009-3920/91/6205-0003$01.00]

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
 Teti and Gelfand 919

because self-efficacy theory has grown efficacy is the mediating link between
largely out of an experimental paradigm that knowledge and behavior (Bandura, 1982,
has examined changes in self-efficacy as a 1986). Mothers who feel efficacious in the
function of the direct induction of success or parenting role would be expected to be suc-
failure experiences (e.g., Bandura & Wood, cessful in establishing warm and harmoni-
1989), positive and negative moods (e.g., ous relationships with their babies. How-
Kavanagh & Bower, 1985), or performance-ever, mothers who feel inefficacious would
contingent feedback (e.g., Newman & Gold-be expected to have more difficulty. Their
fried, 1987). In these and other instances, handling of their babies may be indecisive,
self-efficacy was examined in regard to insensitive, and awkward because they may
clearly defined laboratory tasks. Here we at- lack the problem-solving skills or the persis-
tempt to examine the role of self-efficacy intence required to establish sensitive interac-
the more ambiguous task of parenting in thetions with their children.
home environment. The present study in-
vestigated the role of maternal self-efficacy Though specific to performance in the
in mediating relations between mothers' be- maternal role, maternal self-efficacy should
havior with infants in the first year of life be sensitive to variations in social-marital
and other factors that might affect the supports, perceptions of infant tempera-
mother-baby interaction, such as mothers' ment, and severity of depression. Self-
social-marital supports, level of depression,efficacy theory suggests that social-marital
and perceptions of infant temperament. supports influence maternal self-efficacy di-
rectly, through social persuasion by inti-
A substantial body of literature has re-
mates and by allowing observation of the
lated quality of mothering to social-marital
parenting behavior of significant support
supports, maternal depression, and percep-
figures. For example, friends and relatives
tions of infant temperament, although the
mechanisms that mediate these relations
may reassure a woman that she is a good
have been unclear. Mothers with low levels
mother, and, by design or accident, they may
demonstrate successful child-care routines
of social support and unhappy marriages
for her. In addition, maternal self-efficacy
have been found to behave less optimally
toward their infants than have mothers with
should be sensitive to mothers' perceptions
of their babies as temperamentally "easy" or
more adequate marital and social supports
"difficult." Goldberg (1977) has speculated
(Cox, Owen, Lewis, & Henderson, 1989;
that maternal feelings of competence and ef-
Crnic, Greenberg, Ragozin, Robinson, &
fectance are likely to be fostered by infants
Basham, 1983; Crockenberg & McCluskey,
who are predictable and manageable, who
1986; Downey & Coyne, 1990; Goldberg &
signal their needs effectively, and who are
Easterbrooks, 1984; Hann, 1989). In addi-
easily soothed. By contrast, low maternal
tion, mothers suffering from depression have
self-efficacy might be expected when moth-
been described as substantially less sensi-
ers must cope with babies who are chroni-
tive and more punitive, negative, and re-
cally fussy, irritable, and difficult to read
jecting toward their infants than have nonde-
(see Cutrona & Troutman, 1986). Finally,
pressed mothers (Cohn, Matias, Tronick,
Bandura (1989) has invoked Bower's
Connell, & Lyons-Ruth, 1986; Field, Healy,
affective-priming theory (Bower, 1983; cited
Goldstein, & Guthertz, 1990; Gelfand &
in Bandura, 1989) to argue that depression
Teti, 1990; Lyons-Ruth, Zoll, Connell, &
predisposes individuals to low self-efficacy
Grunebaum, 1986; Radke-Yarrow, Cum-
and poor performance as a result of the se-
mings, Kuczynski, & Chapman, 1985; Weiss- lective activation of memories of failure ex-
man & Paykel, 1974). Finally, mothers' per-
periences. Thus, depressed women should
ceptions of their infants' temperamental
tend to have lower self-efficacy in the mater-
difficulty have been reported to relate to
nal role than should nondepressed women.
mother-infant interaction, although the na-
ture of this relation has been inconsistent
It is important, however, to emphasize that
maternal self-efficacy and maternal depres-
(Crockenberg, 1986).
sion are separable constructs. Whereas as-
We propose that the influences of mater- sessments of depression such as the Beck
nal depression, social-marital supports, and
Depression Inventory (BDI; Beck, Ward,
perceptions of infant temperament on the Mendelson, Mock, & Erbaugh, 1961) tap
mother-infant relationship are largely indi- into general depressive symptomatology and
rect and mediated by mothers' feelings of dysphoria, maternal self-efficacy is specific
efficacy in the maternal role. This expecta- to women's perceived performance in the
tion is consistent with the view that self- maternal role. Thus, although maternal self-

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
 920 Child Development

efficacy is expected to relate to women's were not included. Most mothers were re-
level of global depression, it is expectedferred
to to the project by their therapists (ex-
have multiple determinants and to have cept three who answered newspaper ads),
clear, straightforward links to maternal be-
and therapists' reported that approximately
havior. 85%-90% of the mothers whom they ap-
In the present study, the mediating roleproached volunteered for the study. Thirty-
eight mothers who were not in therapy of
of maternal self-efficacy was tested using the
any kind were recruited from the same areas
criteria for mediation outlined by Baron and
Kenny (1986). That is, independent vari- of the city as the depressed mothers and
ables should be significantly associated with were located through the State Department
of Vital Statistics birth registry. These moth-
the dependent variable, the mediating vari-
able should show significant relations to ers were sent a description of the project and
both the independent and dependent vari-returned a postcard to the investigators if
they were interested in participating; 15.5%
ables, and statistical control of the mediating
variable should substantially reduce rela-of the mothers who were contacted agreed
tions between the independent and depen-to participate. The nondepressed group had
dent variables. Thus, if maternal self- significantly higher annual family incomes.
efficacy is the final, mediating variable In addition, infant age was significantly
linking parenting behavior (the dependent higher in the nondepressed group, although
variable) to mothers' social-marital supports,the difference in real terms was slight. No
depression, and perceptions of infant tem- other group differences were found for any
perament (the independent variables), then: of the remaining demographic indices.
(a) maternal self-efficacy should relate sig- Procedure
nificantly to perceptions of infant tempera- Two female research assistants who
ment, level of depression, social-marital sup-were unaware of the mothers' psychiatric
ports, and objective assessments of maternalstatus visited mothers in their homes three
behavioral competence; (b) maternal self- times (1.5-2 hours each visit) over a 1-month
efficacy should remain significantly associ- period. Visitors established rapport with
ated with maternal behavioral competence mothers during the first visit by inquiring
after controlling for social-marital supports,about the mothers' recent pregnancy, labor,
level of depression, and perceptions of in- and delivery.
fant temperament; (c) after removing the ef-
fects of maternal self-efficacy, maternal be- Questionnaire measures.-During the
havioral competence should no longer relate three visits, mothers completed a variety of
questionnaires designed to obtain informa-
significantly to social-marital supports, level
of depression, and perceptions of infant tion regarding demographics and maternal
temperament. Thus, maternal self-efficacy psychosocial functioning. Psychosocial mea-
should play a central role in determining sures administered during the first visit in-
maternal competence, mediating any influ- cluded the Beck Depression Inventory
ences of maternal depression, social-marital (BDI; Beck et al., 1961) and the infant fussy-
supports, and infant temperament. difficult subscale from the Infant Character-
istics Questionnaire (Bates, Freeland, &
Method Lounsbury, 1979). The BDI is an exten-
sively used self-report measure of depres-
Subjects sion, and Beck et al. (1981) reported statisti-
Subjects were 86 mothers (including cally48significant relations between BDI
depressed mothers) and their infants whoand clinicians' ratings of depth of de-
scores
were participating in a longitudinal study of in a sample of 226 adult psychiatric
pression
infants of depressed and nondepressed patients. The Infant Characteristics Ques-
mothers (see Table 1 for subject descrip- tionnaire is a brief and behaviorally an-
tions). Mothers were invited to participate choredif measure of perceptions of infant tem-
they had a healthy infant within the first perament
year that has impressive convergence
of life. Ninety-seven percent of the mothers with the Carey Survey of Temperamental
were between 18 and 40 years of age,Characteristics
with (Bates et al., 1979). On the
three mothers between 16 and 17 years of visit, mothers completed a marital
second
age. Infants were 3 to 13 months oldharmony (M = scale adapted from Locke and Wal-
7.35 months). Forty-eight of the mothers lace (1959) and Spanier (1976), the Sense of
were in therapy when recruited and had a
Competence Scale of the Parenting Stress
DSM-III-R diagnosis of depression from Index (PSI; Abidin, 1986), and a Maternal
their therapists. Mothers with bipolar illnessSelf-Efficacy Scale developed for this proj-

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
 Teti and Gelfand 921

TABLE 1

DEMOGRAPHIC DATA ON THE DEPRESSED AND NONDEPRESSED GROUPS

GROUP

VARIABLE Depressed (n = 48) Nondepressed (n = 38)

Race (%):
White ................................... 100 94.7
Black ...................................... . 0 2.6
H ispanic.......... ............. .................... 0 2.6
Religion (%):
LDS (Mormon)................................. 68.1 55.6
Catholic ........................................ 6.4 13.9
Protestant .................................. .......... 8.5 8.3
None or other ..................................... 17.1 22.2
Educational attainment (%):
Grades 1-8 ...................................... 4.5 5.1
Grades 9-12 ......................................... 16.7 2.6
High school ..................................... 12.1 13.8
Some college/vocational ...................... 47.0 43.6
College graduate .................................... 15.2 33.3
Graduate/professional .......................... 4.5 2.6
Mean family income ($)............................. 17,837 (17,947) 32,577 (15,184)***
Mean hours per week mother employed. 9.5 (15.19) 8.8 (13.9)
Marital status (%):
Single ........................................... 18.8 13.2
Married/living with partner ................ 81.3 86.8
Mean maternal age (years) ........................ 28.52 (5.83) 28.36 (4.64)
Mean infant age (months) ......................... 6.77 (3.12) 7.97 (1.94)*
Sex of infants (%):
Female ........................... ........... 39.6 44.7
Male ............................................. 60.4 55.3
Birth order status of infant ................ 2.40 2.13
Mean hours per week infants in day-care .41 (2.56) 1.31 (7.42)
NOTE.-Numbers in parentheses are standard deviations.
* Significantly different from the depressed group, p < .05.
*** Significantly different from the depressed group, p < .001.

theoretically
ect by the authors. Locke and Wallace (1959) grounded negative relations
and Spanier (1976) each found their between
measure social supports and neuroticism
of marital harmony to converge with scores on the Eysenck Personality Inven-
clini-
cians' judgments of satisfied and tory. malad-
justed marital dyads. Abidin (1986) has re-
ported an alpha reliability of .74 forOur themeasure of maternal self-efficacy is
Sense of Competence Scale, and the scale
intended to be faithful to Bandura's concept
has been found to discriminate predictably of self-efficacy as being highly situation- or
between mothers of hyperactive children domain-specific, in contrast to other, more
and mothers of normal children (Mash & global parental self-efficacy scales, such as
Johnston, 1983) and between physically those by Abidin (1986) and Gibaud-Wallston
abusive and nonabusive mothers (Mash, and Wandersman (1978). Nine of the 10
Johnston, & Kovitz, 1983). On the third visit,4-point maternal self-efficacy items ad-
mothers completed a questionnaire adapta-dressed mothers' feelings of efficacy in rela-
tion of the Interview Schedule for Social In- tion to specific, delimited domains of infant
teraction (Henderson, Byrne, & Duncan- care, such as soothing the baby, understand-
Jones, 1981), a social support measure that ing what the baby wants, getting the baby
assessed availability and adequacy of moth- to understand mother's wishes, maintaining
ers' attachment figures. Henderson et al. joint attention and interaction with the baby,
(1981) have provided evidence for the con- amusing the baby, knowing what the baby
struct validity of these dimensions in finding enjoys, disengaging from the baby, per-
lower scores among individuals who were forming daily routine tasks (e.g., feeding,
new to a geographical area, and by finding changing, and bathing the baby), and getting

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
922 Child Development
ventory (Caldwell, Heider, & Kaplan, 1966),
the baby to show off for visitors. A final item
tapped mothers' global feelings of efficacy including Emotional and Verbal Responsiv-
in mothering. Item scores were summed ity to of Mother (r = .73, p = .013), Maternal
Involvement (r = .94, p = .001), and Organi-
yield a maternal self-efficacy score. Internal
consistency of the scale was satisfactory, zation of Physical and Temporal Environ-
with a Cronbach's standardized item alpha ment in the home (r = .59, p = .047).
of.79, based on a pilot sample of 29 mothers,
Interrater reliability.-Three of the rat-
and a standardized item alpha of .86 using
ers were female graduate students, who
the present study's sample. As evidence for
trained three undergraduate women as rat-
the concurrent validity of the maternal effi-
ers. Interrater reliability checks between
cacy questionnaire, maternal self-efficacy
trainers and trainee raters were performed
scores were strongly related to the (reverse
periodically for maternal behavior scales in
scored) PSI Sense of Competence Scale in
feeding and free-play contexts on 18 ran-
the present sample (r = -.75, p < .001).
domly selected mother-infant dyads (21% of
the full sample). Using Pearson product-
Behavioral observations.-During the
moment correlations, interrater reliability
second visit, mothers were observed inter-
coefficients were first calculated on the indi-
acting with their infants in two 10-min con-
vidual rating scales, and ranged from .72
texts of feeding and free play with an "Infant
(sensitivity) to .82 (disengagement) in the
Soft Play Set" by Shelcore Toys, which had
feeding context (M = .78) and from .76
three colorful vinyl toys (inflatable bear, bar-
(warmth) to 1.00 (anger) during the free-play
bell rattle, and duck rattle) that mothers
could use to stimulate and interact with their
context (M = .86). Across the feeding and
free-play contexts, scales were significantly
infants. The toys were readily used by all
intercorrelated (sensitivity: r = .66; warmth:
mother-infant dyads. Observations of mater-
r = .75; flatness of affect: r = .58; disen-
nal behavior during feeding and free-play
gagement: r = .48; anger: r = .39, p's <
contexts were coded using rating scales de-
.001). In addition, a standardized item alpha
veloped by Zoll, Lyons-Ruth, and Connellof .86 was obtained based on all 10 behav-
(1984). Scales were based on the work of
ioral ratings (5 per context). Therefore, a sin-
Ainsworth, Blehar, Waters, and Wall (1978)
gle composite score of maternal behavior
and Crittenden (1981). Each rating scale was
was calculated by first adjusting the rating
chosen because of its conceptual relation to
scales as necessary so that higher scores in-
a "good mothering" construct, based on the
dicated more optimal maternal behavior,
findings of Ainsworth and others (Ainsworth
and then summing the ratings across the 10
et al., 1978; Bates, Maslin, & Frankel, 1985;
items. This behavioral composite, which
Egeland & Farber, 1984) that warm, sensi-
was labeled maternal competence, was em-
tive, affectively appropriate mothering fos-
ters secure infant-mother attachment. The
ployed as the unit of analysis for mothers'
behavior. Interrater reliability Pearson r for
five-point rating scales (from less to more of
this composite score was .87 (total n = 18).
the characteristic) were maternal sensitivity
(ability to read and respond appropriately to
infant signals), warmth (level of qualityResults
of
affection toward the infants), flatness of af-
fect (expressionless, impassive, or unemo- The two social support indices and the
marital harmony scores were significantly
tional in facial expressions or tone of voice),
disengagement (disconnectedness from ba- interrelated in the full sample (standardized
bies in terms of body position, pacing,item or alpha = .80). To provide a more stable
control of the interactions), and anger estimate of mothers' quality of support from
(anger/hostility toward babies in verbaliza- intimate support figures, a composite vari-
tions, gestures, or physical assaults). The able of social-marital supports was created
construct validity of these scales was sup- by standardizing and summing the three
variables.
ported by Lyons-Ruth et al. (1986), who
found severity of mothers' depression to re- Hierarchical multiple regression analy-
late to their flatness of affect, hostility, and
sis was performed as a general analytic strat-
intrusiveness during interactions with their egy in the present study to determine rela-
babies. In addition, in an earlier pilot sam- tions between maternal competence and the
ple of eight clinically depressed motherspredictors: maternal depression (BDI),
and their infants, the maternal sensitivity
social-marital supports, infant difficulty, and
scale correlated positively to several logi-
maternal self-efficacy. Table 2 shows that all
cally related dimensions of the HOME In- zero-order correlations between maternal

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
 Teti and Gelfand 923

TABLE 2

ZERO-ORDER AND PARTIAL CORRELATIONSa AMONG MATERNAL COMPETENCE AND THE

FOUR PREDICTOR VARIABLES

Maternal Social-Marital Maternal Infant

Self-Efficacy Supports Depression Difficulty

Maternal competence:
Zero-order ............................... .47*** .48*** - .39*** -.30**
Partial ....................................... .44*** .38** -.36** - .20+
Maternal self-efficacy:
Zero-order ....................................48*** - .57*** -.50**
Partial ...........................................44 ** - .54 ** - .39**
Social-marital supports:
Zero-order ...................................- .61 ** - .38***
Partial ........................................- .58*** - 31**
Maternal depression (BDI):
Zero-order ............... ............39***
Partial .......................... ............... .30**
NOTE.-All significance tests are two-tailed.
a Controlling for maternal education, family income, infant
partner).
Sp= .10.
** p < .01.
*** p < .001.

competence and each of the four predictors ried or living with a partner tended to have
were statistically significant and in the ex-higher maternal competence scores (M =
pected direction (positively to maternal 42.16, n = 71) than did mothers without
self-efficacy and social-marital supports and partners (M = 38.89, n = 14), F(1,83) =
negatively to severity of maternal depres- 3.69, p = .058. Maternal competence did not
sion and infant difficulty). As predicted, ma- relate significantly to any other demographic
ternal self-efficacy was significantly and pos- index. Thus, maternal education, family in-
come, infant age, and marital status were
itively related to social-marital supports and
used as covariates in subsequent regression
negatively related to severity of maternal de-
pression and infant difficulty. Table 2 also analyses. With one exception (infant diffi-
shows significant negative correlations be- culty), the correlations among maternal com-
tween social marital supports and maternal petence and all predictor variables remained
depression and infant difficulty and a sig- statistically significant when these covari-
nificant positive association between mater- ates were partialed out (see Table 2).
nal depression and maternally rated infant
Preliminary one-way analyses of vari-
difficulty.
ance were then performed to explore group
In order to determine if any covariates differences on maternal competence and
should be used in subsequent analyses, we each of the four predictors. As expected, the
explored for any simple, zero-order relations depressed mothers as a group had signifi-
between maternal competence and all de- cantly higher BDI scores (M = 23.13, SD =
mographic variables. Correlates of maternal 10.78, range = 3-46) than did the control
competence were maternal education, r = mothers (M = 7.13, SD = 5.17, range =
.32, p = .003, and family income, r = .33, 0-20, F[1,84] = 70.68, p < .0001) and sig-
p = .003. In addition, a positive relation that nificantly lower scores on social-marital sup-
approached significance was found between ports (M = -.91 vs. 1.75, F[1,76] = 33.58,
maternal competence and infant age, r = p < .0001),' maternal self-efficacy (M =
.21, p = .052, and mothers who were mar- 29.63 vs. 33.05, F[1,84] = 12.86, p < .001),

1Analyses were also conducted to examine the possibility that LDS (Mormon) mothers
differed from non-LDS mothers in the amount and range of social-marital supports received. A
one-way analysis of variance revealed no differences in the amount of social-marital supports
reported by LDS and non-LDS mothers. In addition, the variability of social-marital supports
was virtually identical in both groups (LDS mothers: M = .086, SD = 2.57; non-LDS mothers:
M = .393, SD = 2.25, F[1,80] = .55, p = .46).

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
924 Child Development
and maternal competence (M = 39.61 vs. of variance in maternal competence, as did
44.58, F[1,82] = 18.52, p < .001). Depressed the block composed of maternal depression,
and nondepressed groups did not differ insocial-marital supports, and infant difficulty.
their perceptions of their infants' difficultAs predicted, maternal self-efficacy re-
temperaments. To determine whether rela-mained significantly related to maternal
tions between maternal competence and any competence even after the entry of all other
of the predictors differed significantly be- predictors.2 None of the interactive terms in-
tween the two groups, a series of hierarchi- volving maternal self-efficacy was signifi-
cal multiple regression analyses were per- cantly related to maternal competence.
formed using maternal competence as the
dependent variable. In each of these analy- Hierarchical multiple regressions were
ses, the four demographic covariates were also employed to test the hypothesis that,
entered as a block into the regression equa-after statistically controlling for maternal
tion first, followed in separate steps by aself-efficacy, maternal behavioral compe-
dummy variable for group status (1 = de- tence should no longer relate significantly to
pressed, 2 = nondepressed), then one of thesocial-marital supports, maternal depres-
four predictors, and then by a group x pre- sion, and perceptions of infant temperament.
dictor interaction term (i.e., group x depres- Since perceptions of infant difficulty were
sion, group x social-marital supports, group not significantly correlated with maternal
x infant difficulty, and group x maternalcompetence after partialing out maternal ed-
self-efficacy). Because none of these interac- ucation, family income, infant age, and mari-
tive terms related significantly or marginally tal status, only analyses that focused on
to maternal competence, the two groups social-marital supports and maternal depres-
were combined for subsequent analyses. sion are presented. Using maternal compe-
tence as the dependent variable, the four de-
Hierarchical multiple regression analy-mographic covariates were again entered
ses were then performed to test the hypothe-into the regression equation first as a block,
sis that maternal self-efficacy should remainfollowed by maternal self-efficacy, and then
significantly associated with maternal be- by either social-marital supports or maternal
havioral competence after controlling for depression (see Table 4). The demographic
social-marital supports, level of depression,covariates together accounted for a signifi-
and perceptions of infant temperament. Thiscant amount of variance in maternal compe-
strategy enabled an assessment of the tence, as did maternal self-efficacy entered
unique relation between maternal self- on step 5. As predicted, neither social-
efficacy and maternal competence by first marital supports nor maternal depression
partialing out any overlapping variance con- (entered at step 6) significantly related to
tributed by other variables. A series of hier- maternal competence. However, the associa-
archical multiple regression analyses were tion between social-marital supports and
performed with maternal competence as the maternal competence approached signifi-
dependent variable. In each analysis, the cance (p = .055) after maternal self-efficacy
four demographic covariates were entered as was covaried out.2
a block into the regression equation first, fol-
lowed by the entry of social-marital sup- A final set of hierarchical multiple re-
ports, maternal depression, and infant diffi- gressions, employing maternal self-efficacy
culty as a block, and then by maternal as the dependent variable, was performed to
self-efficacy. Finally, using a procedure out- explore whether maternal self-efficacy was
lined by Cohen and Cohen (1983) to exam- sensitive to two-way interactions among ma-
ine the possibility that maternal self-efficacy ternal depression, perceptions of infant dif-
might interact with other variables in pre- ficulty, and social-marital supports (i.e.,
dicting maternal competence, the last entry maternal depression x infant difficulty, ma-
in each of these regressions was a two-way ternal depression x social-marital supports,
interactive term involving maternal self- and infant difficulty x social-marital sup-
efficacy and one of the other predictors (e.g., ports). Three analyses were performed in
maternal self-efficacy x infant difficulty,which the four demographic covariates, ma-
maternal self-efficacy x social-marital sup-ternal competence, and whatever predictor
ports, maternal self-efficacy x maternal de- that was not involved in the interaction were
pression). Table 3 summarizes these analy- entered as a block into the regression equa-
ses. The four demographic covariates tion first, followed at separate steps by the
together accounted for a significant amount
two predictors involved in the interaction to

2 Results were virtually identical when the four demographic covariates were omitted from
analyses.

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
 Teti and Gelfand 925

TABLE 3

HIERARCHICAL MULTIPLE REGRESSION ANALYSES ASSESSING THE UNIQUE RELATION BETWE

COMPETENCE (the Dependent Variable) AND MATERNAL SELF-EFFICACY

Step and
Order of Entry Multiple R R2 Increase df F Change p Beta
1-4:
Maternal education ...................... .31
Family income........................... .40 .16 4,69 3.22 .017 .19
Infant age ...................................... .13
Marital status .............................. .17
5-7:
Maternal depression (BDI) ......... -.34
Infant difficulty .......................... .55 .15 3,66 4.70 .005 -.12
Social-marital supports ................ .24
8:
Maternal self-efficacy ................ .60 .05 1,65 5.23 .026 .31

NOTE.-Variables at steps 1-4 and 5-7 were entered as a block. No two-way in

self-efficacy with maternal depression, infant difficulty, or social-marital suppor
associated with maternal competence.

ficulty scores
be tested, and finally by the two-way interac-or high BDI scores
tive term itself. Only the maternal depres- scores had interm
infant difficulty
ternal was
sion x infant difficulty interaction self-efficacy
sig- scores (Y = 2
nificantly related to maternal 20.63, respectively).
self-efficacy (r2 As these valu
change = .04, F change [1,64] = 5.42,
cate, p =
the interaction was account
.023, beta at entry = -.88). To
the examine this
larger difference in predicted
interaction further, predictedself-efficacy
maternal scores
self-between depressed
efficacy scores were calculated from
mothers with the
highbeta
versus low infant diffi-
culty scores
weights supplied by this regression than between nondepressed
analysis
(following Cohen & Cohen, 1983). Predicted
mothers with high versus low infant diffi-
maternal self-efficacy scores were
culty scores. lowest
among mothers who reported high BDI and
infant difficulty scores (Y = 13.01) and high-
Discussion
est among mothers with low BDI and low
These
infant difficulty scores (Y = 29.38). results support the premise tha
Mothers
who had either low BDI and high infant
maternal dif-
self-efficacy is a central mediat

TABLE 4

HIERARCHICAL MULTIPLE REGRESSION ANALYSES ASSESSING THE UNIQUE RELATIONS BETWE

COMPETENCE (the Dependent Variable), MATERNAL DEPRESSION AND SOCIAL-MARITAL SU
CONTROLLING FOR MATERNAL SELF-EFFICACY

Step and
Order of Entry Multiple R R2 Increase df F Change p Beta
1-4:
Maternal education ............. .31
Family income ................. .40 .16 4,69 3.22 .017 .19
Infant age............................. .13
M arital status ....................... .17
5:
Maternal self-efficacy ......... .56 .16 1,68 16.06 <.001 .43
6:
Maternal depression ........ .58 .02 1,67 1.74 .19 -.17
6:
Social-marital supports ....... .60 .04 1,67 3.83 .055 .23

NOTE.-Variables at steps 1-4 were entered as a block. Perceptions of infan

related to maternal competence at step 6. This analysis is not presented, how
competence were not significantly associated after maternal education, fami
were partialed out (see Table 2).

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
926 Child Development

of relations between mothers' competence

they have available to establish sensitive,
with their infants and factors such as mater-
harmonious relationships with their chil-
nal perceptions of infant difficulty, maternal dren. By contrast, women who develop feel-
depression, and social-marital supports. This ings of self-doubt in response to infant diffi-
conclusion is based on three interrelated culty may be expected to withdraw from the
findings: First, maternal self-efficacy corre- challenges of a difficult baby and to be less
lated significantly with maternal compe- creative and resourceful in fostering suc-
tence, perceptions of infant difficulty, cessful mother-infant exchanges than would
social-marital supports, and maternalmore de- efficacious mothers. Thus, maternal
pression after controlling for selected demo-
self-efficacy appears to be a crucial media-
graphic variables. Second, maternal self-tor of relations between perceptions of in-
efficacy was the factor most directlyfant andtemperament and maternal behavior,
unambiguously related to parenting behav-and central to an understanding of the
"goodness-of-fit" (Thomas & Chess, 1977)
ior. Maternal self-efficacy remained signifi-
cantly associated with maternal competence
between infants and their caregivers.
after statistically controlling for all other pre-
dictors. In addition, after first controlling forWe also note that, like many other stud-
maternal self-efficacy, neither maternalies involving infant temperament (see Ler-
de-
ner & Lerner, 1986), the present study em-
pression nor social-marital supports related
significantly to maternal competence. Third,
ployed a measure of maternal perceptions of
infant
although perceptions of infant difficulty did temperament rather than a more "ob-
jective" assessment of infant temperament
not relate significantly to maternal compe-
tence when demographic variables wereinvolving
sta- trained observers. As Crockenberg
(1986), Lamb and Bornstein (1987), Rothbart
tistically controlled, they were significantly
(1981), and others have observed, the inher-
associated with maternal self-efficacy, which
ent problems with such a measure include
in turn was related to maternal competence.
the possibility that mothers' perceptions of
Collectively, these results satisfy the condi-
infant
tions set forth by Baron and Kenny (1986) for difficulty are related to characteristics
of the parent. This disadvantage is weighed
identifying variables that putatively mediate
relations between "independent" or pre-against potential advantages, such as the
dictor variables and a "dependent" or crite-
mothers' thorough familiarity with their chil-
rion variable. dren. Further, Bates et al. (1979) have ar-
gued that parents' perceptions of infant tem-
The significant relation between per- perament are important in their own right
ceptions of infant difficulty and maternal in the prediction of child developmental
self-efficacy is consistent with Goldberg's outcomes. They reported low levels of
(1977) hypotheses regarding mothers' feel- agreement between parental scores on the
ings of "effectance" and their infants' abili-
Infant Characteristics Questionnaire, used
ties to signal their needs effectively. Fur- in the present study, and independent ob-
ther, the potential role of maternal server ratings of the same children, and also
self-efficacy in mediating the effects of per-
found "easy" infants to be associated with
ceptions of infant difficulty on maternal be-extraverted, gregarious mothers. Despite
havior is especially relevant to empirical ef-
these concerns, our results nevertheless sug-
forts to establish links between variations in
gest that maternal perceptions of infant dif-
mothers' perceptions of infant temperamentficulty influenced mothers' behavioral com-
and maternal behavior. Crockenberg (1986) petence only to the extent that they affected
has noted that the research literature shows their feelings of efficacy in the maternal role.
no clear consensus on just how infant diffi- Of strong interest would be an examination
culty influences maternal behavior. In theof maternal self-efficacy as a potential medi-
present study, the impact of mothers' per-
ator of relations between more objective
ceptions of infant temperamental difficultymeasures of infant difficulty and maternal
on maternal competence appeared to de- behavior.
pend on the mothers' feelings of efficacy.
Thus, a difficult, irritable baby may be ex- The significant negative relation be-
pected to have a negative effect on maternal
tween maternal depression and maternal
behavior only to the degree that infant tem-
self-efficacy is consistent with Kavanagh and
perament reduces a woman's feelings of ef-
Bower's (1985) experimental work involving
ficacy as a mother. Some mothers with diffi-
the effects of positive and negative mood on
cult babies but who have a strong sense of
self-efficacy beliefs, and extends their find-
maternal self-efficacy may be expected toings to a clinical population. The present
use whatever personal and social resources
findings also support Bandura's (1982, 1989)

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
 Teti and Gelfand 927

hypothesis that depression predisposes indi- incidental induction of basic problem sol
viduals to low self-efficacy. Especially strik- ing skills.
ing was the finding that maternal depression
had no relation to maternal competence in- The present study has several limita-
dependent of maternal self-efficacy. As with tions that must be addressed. First, it is pos-
infant difficulty, maternal depression related
sible that the relations obtained are in part
to maternal competence only to the extent the result of shared method variance; that
that it potentially affected maternal self- is, our measures of severity of depression,
efficacy. Thus, whereas maternal self- social-marital supports, infant difficulty, and
efficacy may have been influenced by level maternal self-efficacy were obtained directly
of depression, the latter's association withfrom maternal reports. While this concern
maternal behavior appeared to be depen-cannot be minimized, it appears that shared
dent on maternal self-efficacy. The present method variance alone cannot account for
findings thus support the premise that ma- the obtained results. To begin, the maternal
ternal self-efficacy is specific to women'scompetence ratings were made by indepen-
perceived performance in the maternal role dent observers, so not all data took the form
and is a viable construct in its own right. of maternal reports. In addition, maternal
Indeed, maternal self-efficacy was extremely self-efficacy was the most robust predictor of
low among depressed mothers who per- maternal competence. This relation re-
ceived their infants as difficult but substan- mained statistically significant after partial-
tially higher among depressed mothers who ing out the effects of all other variables,
perceived their infants as less difficult. whereas maternal competence was no
These findings suggest that some depressedlonger significantly associated with maternal
mothers may nevertheless develop adequatedepression and social-marital supports when
feelings of efficacy in the maternal role, es-maternal self-efficacy was partialed out. We
pecially if they perceive their babies as hav- would not have expected these differential
ing "easy" temperaments. The present study correlational patterns if the findings ob-
did not determine whether or not easy in-tained were due primarily to shared method
fants of depressed mothers are less at risk variance. A second limitation of the present
than difficult infants. As an easy infant's be- study was that our observations of maternal
havioral repertoire develops, that child may behavior were relatively brief, and despite
become more likely to be perceived as a the fact that corresponding dimensions of
source of stress by a demoralized mother,maternal behavior in the feeding and free-
with negative consequences for the relation- play observation contexts were significantly
ship. Greater focus needs to be given to the intercorrelated, our samples of maternal be-
role of infant temperament in shaping de-havior may not have been truly representa-
pressed and nondepressed mothers' mater-tive. Studies of the psychosocial correlates
nal self-efficacy and mothering skills. of mother-infant interaction have varied
widely with regard to duration of observa-
The significant relation between social-tions, ranging from 3-min sessions of face-
marital supports and maternal self-efficacy to-face interactions (Cohn et al., 1986; Field
supports Bandura's (1982, 1986) premise et al., 1990) to several hours of unstructured
that self-efficacy is sensitive to social persua-activities (Crockenberg & McCluskey, 1986;
sion and modeling influences provided byRadke-Yarrow et al., 1985). The two 10-min
significant others. Social-marital supportsobservational periods used in the present
may also have influenced maternal self- study were viewed as behavioral probes, de-
efficacy indirectly via its relation to maternalsigned to examine affective dimensions of
depression. It is interesting that, althoughmaternal behavior in specific contexts. As a
the significant relation between social- group, our depressed mothers were rated as
marital supports and maternal competenceless competent than nondepressed mothers,
was eliminated when maternal self-efficacya finding that is in line with other studies
was statistically controlled, the relation nev- of depressed mothers that employed either
ertheless approached significance (p = very short (e.g., Cohn et al., 1986) or very
.055). Conceivably, maternal self-efficacylong observational periods (e.g., Radke-
may be only one of several mechanisms Yarrow et al., 1985). Third, our sample was
through which social-marital supports po- heavily skewed with LDS (Mormon) moth-
tentially influence mothers' behavioral com- ers, which might have affected the results.
petence. As Rutter (1987) has discussed, fu- There is no reason to believe, however, that
ture research may wish to examine how psychosocial correlates of mothering among
social-marital supports influence quality ofpredominantly middle-class LDS mothers
mothering directly, perhaps via the direct or should be different from other middle-class

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
928 Child Development

samples. In the present study, there was Bandura,

no A., & Wood, R. (1989). Effect of per-
difference between LDS and non-LDS ceived controllability and performance stan-
mothers in the amount and variation of dards on self-regulation of complex decision
social-marital supports. In addition, the sig-
making. Journal of Personality and Social
nificant relations obtained in the present Psychology, 56, 805-814.
study are consistent with studies of Baron,
non- R. M., & Kenny, D. A. (1986). The
LDS depressed and nondepressed mothers moderator-mediator variable distinction in so-
(e.g., Cohn et al., 1986; Cox et al., 1989; cial psychological research: Conceptual, stra-
Crockenberg & McCluskey, 1986; Crnic et tegic, and statistical considerations. Journal
al., 1983; Field et al., 1990; Goldberg & of Personality and Social Psychology, 51,
Easterbrooks, 1984; Lyons-Ruth et al., 1986; 1173-1182.
Weissman & Paykel, 1974), which reinforces Bates, J. E., Freeland, C. A., & Lounsbury, M. L
our confidence in the external validity of the (1979). Measurement of infant difficultness.
present findings. Fourth, the present study Child Development, 50, 794-803.
is limited by its use of a contemporaneousBates, J. E., Maslin, C. A., & Frankel, K. A. (1985
data set, and longitudinal data are clearly Attachment security, mother-child interac-
necessary to establish causal links that may tion, and temperament as predictors of behav-
exist among variables. Indeed, maternal ior problem ratings at age three years. In I.
self-efficacy may both influence mothers' be- Bretherton & E. Waters (Eds.), Growing
havior toward their babies and be influenced points of attachment theory and research (pp.
by their experiences with their babies. We 167-193). Monographs of the Society for Re-
view self-efficacy as a dynamic construct that
search in Child Development, 50(1-2, Serial
shapes and is shaped by behavior, knowl- No. 209).
edge, and perceptions. Beck, A. T., Ward, C. H., Mendelson, M., Mock,
J., & Erbaugh, J. (1961). An inventory for mea-
Future research might longitudinally in-
vestigate the possible impact of maternal suring depression. Archives of General Psy-
chiatry, 4, 561-571.
self-efficacy on developmental outcomes in
Bugenthal, D. B. (1987, April). Caregiver attribu-
infants and young children, perhaps with re-
gard to infants' quality of attachment and tions as moderators of child effects. Paper
children's compliance behavior. Indeed, our presented at the meeting of the Society
for Research in Child Development, Balti-
findings suggest that maternal self-efficacy
more, MD.
may be a critical determinant of infant risk
Caldwell, B. M., Heider, J., & Kaplan, B. (1966).
status insofar as it affects the quality of moth-
ers' behavior toward their children. Addi- The Inventory of Home Stimulation. (Avail-
able from Center for Early Development and
tional research might explore the temporal
Education, 814 Sherman, Little Rock, AR
stability of maternal self-efficacy, perhaps by
77202.)
examining the sensitivity of maternal self-
Cohen, J., & Cohen, P. (1983). Applied regression/
efficacy to longitudinal changes in maternal
depression, child behavior, and maternal correlation analysis for the behavioral sci-
ences (2d ed.). Hillsdale, NJ: Erlbaum.
competence.
Cohn, J. F., Matias, R., Tronick, E. Z., Connell,
References D., & Lyons-Ruth, K. (1986). Face-to-face in-
teractions of depressed mothers and their in-
Abidin, R. R. (1986). Parenting Stress Index. Char-
fants. In E. Z. Tronick & T. Field (Eds.), Ma-
lottesville, VA: Pediatric Psychology Press.
ternal depression and infant disturbance:
Ainsworth, M. D. S., Blehar, M. C., Waters, NewE.,directions for child development (W.
& Wall, S. (1978). Patterns of attachment:Damon,
A editor-in-chief) (pp. 31-45). San
psychological study of the Strange Situation.
Francisco: Jossey-Bass.
Hillsdale, NJ: Erlbaum. Cox, M. J., Owen, M. T., Lewis, J. M., & Hender-
Bandura, A. (1977). Self-efficacy: Toward a unify-
son, V. K. (1989). Marriages, adult adjustment,
ing theory of behavior change. Psychological
and early parenting. Child Development, 60,
Review, 84, 191-215. 1015-1024.
Bandura, A. (1982). Self-efficacy mechanisms in
Crittenden, P. M. (1981). Abusing, neglecting,
human agency. American Psychologist, problematic
37, and adequate dyads: Differenti-
122-147. ating by patterns of interaction. Merrill-
Palmer Quarterly, 27, 201-218.
Bandura, A. (1986). Social foundations of thought
Crnic, K. A., Greenberg, M. T., Ragozin, A. S.
and action: A social-cognitive theory. Engle-
wood Cliffs, NJ: Prentice-Hall. Robinson, N. M., & Basham, R. B. (1983). Ef-
Bandura, A. (1989). Regulation of cognitive fects
pro- of stress and social support on mothers
cesses through perceived self-efficacy. Devel-
and premature and full-term infants. Child
opmental Psychology, 25, 729-735. Development, 54, 209-217.

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms
 Teti and Gelfand 929

Litt, M. D.
Crockenberg, S. B. (1986). Are temperamental (1988). Cognitive mediators of stressful
dif-
experience: Self-efficacy and perceived con-
ferences in babies associated with predict-
trol.
able differences in care-giving? In J. V. Cognitive Therapy and Research, 12,
Ler-
ner & R. M. Lerner (Eds.), Temperament 241-260.
and
social interaction in infants and Locke, H., & Wallace, K. (1959). Short marital-
children:
New directions for child development adjustment
(W. and prediction tests: Their reli-
Damon, editor-in-chief) (pp. 53-73). San ability and validity. Marriage and Family Liv-
Francisco: Jossey-Bass. ing, 21, 251-255.
Crockenberg, S., & McCluskey, K. (1986). Lyons-Ruth, K., Zoll, D., Connell, D., & Grune-
Changes in maternal behavior during the baum, H. U. (1986). The depressed mother
baby's first year of life. Child Development, and her one-year-old infant: Environment, in-
57, 746-753. teraction, attachment, and infant develop-
Cutrona, C. E., & Troutman, B. R. (1986). Social ment. In E. Z. Tronick & T. Field (Eds.), Ma-
support, infant temperament, and parenting ternal depression and infant disturbance:
self-efficacy: A mediational model of postpar- New directions for child development (W.
tum depression. Child Development, 57, Damon, editor-in-chief) (pp. 61-82). San
1507-1518. Francisco: Jossey-Bass.
Downey, G., & Coyne, J. C. (1990). Children of Mash, E. J., & Johnson, C. (1983). Parental percep-
depressed parents: An integrative review. tions of child behavior problems, parenting
Psychological Bulletin, 108, 50-76. self-esteem, and mothers' reported stress in
Egeland, B., & Farber, E. A. (1984). Infant-mother younger and older hyperactive and normal
attachment: Factors related to its develop- children. Journal of Consulting and Clinical
ment and changes over time. Child Develop- Psychology, 51, 86-99.
ment, 55, 753-771. Mash, E. J., Johnston, C., & Kovitz, K. (1983). A
Field, T., Healy, B., Goldstein, S., & Guthertz, M. comparison of the mother-child interactions
(1990). Behavior-state matching and syn- of physically abused and non-abused children
chrony in mother-infant interactions of non- during play and task situations. Journal of
depressed versus depressed dyads. Develop- Clinical Child Psychology, 12, 337-346.
mental Psychology, 26, 7-14. Newman, C., & Goldfried, M. R. (1987). Disabus-
Gelfand, D. M., & Teti, D. M. (1990). The effects ing, negative self-efficacy expectations via ex-
of maternal depression on children. Clinical periences, feedback, and discrediting. Cogni-
Psychology Review, 10, 329-353. tive Therapy and Research, 11, 401-417.
Gibaud-Wallston, J., & Wandersman, L. P. (1978, Radke-Yarrow, M., Cummings, E. M., Kuczynski,
September). Development and utility of the L., & Chapman, M. (1985). Patterns of attach-
Parenting Sense of Competence Scale. Paper ment in two- and three-year-olds in normal
presented at the meeting of the American families and families with parental depres-
Psychological Association, Toronto. sion. Child Development, 56, 884-893.
Goldberg, S. (1977). Social competence in in- Rothbart, M. K. (1981). Measurement of tempera-
fancy: A model of parent-infant interaction. ment in infancy. Child Development, 52,
Merrill-Palmer Quarterly, 23, 163-177. 569-578.
Goldberg, W., & Easterbrooks, M. A. (1984). Role Rutter, M. (1987). Psychosocial resilience and pro
of marital quality in toddler development. De- tective mechanisms. American Journal of Or-
velopmental Psychology, 20, 504-514. thopsychiatry, 57, 316-331.
Hann, D. M. (1989). A systems conceptualization Spanier, G. B. (1976). Measuring dyadic adjust-
of the quality of mother-infant interaction. ment: New scales for assessing the quality of
Infant Behavior and Development, 12, marriage and similar dyads. Journal of Mar-
251-263. riage and the Family, 38, 15-28.
Henderson, S., Byrne, D. G., & Duncan-Jones, Thomas,P. A., & Chess, S. (1977). Temperament and
(1981). Neurosis and the social environment. development. New York: Brunner/Mazel.
Sydney: Academic Press Australia. Weissman, M. M., & Paykel, E. D. (1974). The
Kavanagh, D. J., & Bower, G. H. (1985). Mood depressed woman: A study of social relation-
and self-efficacy: Impact of joy and sadness ships. Chicago: University of Chicago Press.
on perceived capabilities. Cognitive TherapyZoll, D. A., Lyons-Ruth, K., & Connell, D. (1984).
and Research, 9, 507-525. Infants at psychiatric risk: Maternal behav-
Lamb, M. E., & Bornstein, M. H. (1987). Develop- ior, depression, and family history. Paper
ment in infancy. New York: Random House. presented at the 92nd Annual Convention of
Lerner, J. V., & Lerner, R. M. (Eds.). (1986). Tem- the American Psychological Association, To-
perament and social interaction in infants ronto.

and children. San Francisco: Jossey-Bass.

This content downloaded from

146.50.98.82 on Fri, 20 Oct 2023 12:45:00 +00:00
All use subject to https://about.jstor.org/terms