Вы находитесь на странице: 1из 25


discussions, stats, and author profiles for this publication at:


Maternal self-efficacy beliefs,

competence in parenting, and
toddlers' behavior and
developmental status

Article in Infant Mental Health Journal March 2003

DOI: 10.1002/imhj.10048


179 1,262

2 authors:

Priscilla K Coleman Katherine Hildebrandt Karraker

Bowling Green State University West Virginia University


Some of the authors of this publication are also working on these related projects:

Practitioner/Client interactions on termination of pregnancy. A Grounded

Theory study View project

All content following this page was uploaded by Katherine Hildebrandt Karraker on 09 October 2015.

The user has requested enhancement of the downloaded file.


top of AH



Bowling Green State University
West Virginia University

ABSTRACT: This study was designed to examine parenting self-efcacy beliefs as correlates of mothers
competence in parenting toddlers and as predictors of toddlers behavior and development. Sixty-eight
predominantly middle-class mother toddler pairs participated in this study. Mothers completed ques-
tionnaires, toddlers were administered the Bayley Scales of Infant Development (BSID-II), and each dyad
participated in the Crowell Procedure, which is designed to observe parent and toddler behaviors in a
semistructured laboratory context. Although domain-general and domain-specic parenting self-efcacy
beliefs were not associated with parenting competence, domain-specic beliefs were signicantly related
to toddlers scores on the Mental Scale of the BSID-II and several behaviors observed during the Crowell
Procedure (Affection Towards Mother, Avoidance of Mother, Compliance, Enthusiasm, and Negativity).
Implications of the ndings are discussed.

RESUMEN: Este estudio fue disenado con el n de examinar las creencias que se tienen de la autoecacia
en la crianza como correlativos de la capacidad de las madres de criar a sus infantes, as como puntos de
prediccion de la conducta y desarrollo de los propios infantes. Sesenta y ocho pares de madres y sus
infantes, predominantemenmte de clase media, participaron en este estudio. Las madres respondieron
cuestionarios; con los infantes se utilizaron las Escalas Bailey para el Desarrollo Infantil. (BSID-II). Cada
dada participo en el Procedimiento Crowell, el cual esta disenado para observar las conductas de las
madres y de los infantes dentro de un contexto de laboratorio semiestructurado. Aunque las creencias de
la propia ecacia como madres, tanto en los aspectos generales como especcos no estuvieron asociadas
con la capacidad materna de criar, las creencias especcas estuvieron signicativamente relacionadas
con los puntajes de los infantes en la Escala Mental de BSID-II y varios de los comportamientos obser-
vados durante el Procedimiento Crowell (el afecto hacia la madre, el evitar la madre, obediencia, entu-
siasmo y negatividad). Se discuten las implicaciones de estos resultados.

RESUME: Cette etude a ete concue pour examiner les croyances dauto-efcacite de parentage en tant que
correlats de la competence des meres pour le parentage des enfants en bas age et en tant que facteurs de
prediction du comportement et du developpement des enfants en bas age. Soixante huit paires mere-jeune
enfant (issus de la bourgeoisie) ont participe a cette etude. Les meres remplirent des questionnaires, on

Direct correspondence to: Priscilla K. Coleman, Human Development and Family Studies, 16 F Family and Consumer
Sciences Building, Bowling Green State University, Bowling Green, OH 43403; e-mail:pcolema@bgnet.bgsu.edu.

INFANT MENTAL HEALTH JOURNAL, Vol. 24(2), 126 148 (2003)

2003 Michigan Association for Infant Mental Health short
Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/imhj.10048 standard

126 base of drop


top of rh
Maternal Self-Efcacy 127 base of rh

cap height
t passer aux enfants les Echelles de developpement de la petite enfance Bayley (BSID-II), et chaque base of text
dyade participa a la Procedure Crowell, qui est concue pour observer les comportements du parent et de
lenfant dans un contexte de laboratoire semi-structure. Bien que les croyances dauto-efcacite de par-
entage generales et speciques netaient pas associees a la competence de parentage, les croyances
speciques etaient liees de maniere signicative aux resultats des jeunes enfants a lechelle mentale de
BSID-II, ainsi qua plusieurs comportements observes durant la Procedure Crowell (affection envers la
mere, eviter la mere, conformite, enthousiasme, et negativite). Les implications de ces resultats sont

ZUSAMMENFASSUNG: Diese Studie untersucht die Ansicht uber die Fahigkeit zu Elternschaft aus der Sicht
der Eltern als Korrelation der mutterlichen Fahigkeiten ein Kleinkind zu betreuen und als Voraussage-
faktoren fur das Verhalten und die Entwicklung des Kleinkinds. 68 vorwiegend mittelschichtige Mutter
- Kleinkindpaare nahmen an dieser Studie teil. Die Mutter fullten Fragebogen aus, die Kleinkinder wurden
mit der Bayley Skala der kindlichen Entwicklung (BSID-II) untersucht und jedes Paar nahm an der
Cromwell Untersuchung teil, die darauf ausgerichtet ist Eltern- und Kleinkindverhalten in einer semis-
trukturierten Laboruntersuchung festzustellen. Obwohl die Uberzeugungen zur Effektivitat der Fahig-
keiten zu genereller und auf spezielle Bereiche bezogene Elternschaft nicht im Zusammenhang standen
mit der elterlichen Fahigkeit, gab es Bereiche dieser Uberzeugungen, die spezisch mit den Werten der
Kleinkinder auf der Skala: geistige Entwicklung des BSID-II zusammenhingen und auch mit vielen
Verhaltensweisen, die wahrend der Cromwell Untersuchung beobachtet wurden (Zuneigung zur Mutter,
Vermeidung der Mutter, Einverstandnis, Enthusiasmus und Negativitat). Die Bedeutung dieser Ergebnisse
werden besprochen.

* * *
Over the course of the last several decades, a considerable amount of research attention
has been devoted to understanding subjective experiences of parenting. A primary goal of this
effort has been to identify key parental cognitions likely to be associated with successful
personal adjustment to the role and with positive parenting practices. Early work in this area
revealed that women who were able to readily visualize themselves as mothers during preg-
nancy, compared to women who had difculty doing so, experienced more successful post-
partum adjustment, felt more satised with parenting, and exhibited more positive parenting
behavior (Oakley, 1980; Shereshefsky & Yarrow, 1973). Additional longitudinal work has
suggested that visualization of oneself as a mother, beliefs that one possesses the personal
characteristics necessary to become a good parent, and condence related to ones future par-
enting skills measured late in pregnancy are predictive of such self-denitions after birth
(Deutsch, Ruble, Fleming, Brooks-Gunn, & Stangor, 1986) as well as various parent child short
and child outcomes (Heinicke, 1984; Williams et al., 1987). Moreover, maternal perceptions standard

top of rh
128 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
of parenting competence have been found to be more salient predictors of maternal adjustment base of text
than prior experience with infants and a general self-esteem measure (Williams et al., 1987).
Within the broader realm of parenting self-condence literature, research conducted over
the past 15 years has highlighted parenting self-efcacy beliefs as a central correlate of par-
enting behavior, with evidence suggesting that self-efcacy beliefs may mediate the effects of
a number of parent and child variables on the quality of parenting (Bugental, Blue, & Cruzcosa,
1989; Cutrona & Troutman, 1986; Raver & Leadbeater, 1999; Teti & Gelfand, 1991). Parenting
self-efcacy beliefs are broadly dened as parents self-referent estimations of competence in
the parental role. Parenting self-efcacy may also be construed as parents perceived ability to
positively inuence the behavior and development of their children. Implicit in these cognitions
is knowledge of appropriate parenting behaviors as well as condence in ones own ability to
effectively engage in the referent behaviors (Bandura, 1989). Available data from longitudinal
studies suggests that self-efcacy beliefs are relatively stable over time. For example, Schnee-
wind (1995) found parenting self-efcacy beliefs uctuated very little from three to nine months
postpartum among mothers and fathers. Likewise, Gross, Conrad, Fogg, and Wothke (1994)
reported moderately high levels of consistency in self-efcacy beliefs in mothers from the time
their toddlers were 12 to 36 months of age.
Based on the recent wave of parenting self-efcacy research, this study was designed to
explore maternal self-efcacy beliefs, assessed at two different levels of analysis, as correlates
of parenting competence and toddler behavior and developmental status. This objective was
pursued with mother toddler dyads as the participants. The toddler period was selected for
two primary reasons. First, with the tasks associated with effective parenting during toddler-
hood changing dramatically from the infancy period in response to toddlers expanding abili-
ties, parents must develop many new parenting behaviors. According to self-efcacy theory
and related research, acquiring new skills is facilitated by self-efcacy beliefs (Bandura, 1982).
Second, because the salience of self-efcacy beliefs as a predictor of overt behavior tends to
be enhanced under stressful circumstances (Bandura, 1982), the decision was made to observe
parent and child behavior during a rather difcult phase of parenting. By 19 25 months of
age, most toddlers are highly mobile, are actively engaged with the environment, and are testing
the limits imposed on them as they strive for a new level of autonomy from the primary
High maternal self-efcacy beliefs have been found to be related to specic positive par-
enting practices, such as responsive, stimulating and nonpunitive caretaking (Unger & Wau-
dersman, 1985), active maternal coping orientations (Wells-Parker, Miller, & Topping, 1990),
and few maternally perceived child behavior problems (Johnson & Mash, 1989). Low maternal
self-efcacy has been correlated with maternal depression (Cutrona & Troutman, 1986; Teti
& Gelfand, 1991), actual behavior problems in children (Gibaud-Wallson & Waudersman,
1978, cited in Johnston & Mash, 1989), maternal perceptions of child difculty (Johnston &
Mash, 1989), a passive coping style in the parental role (Wells-Parker, Miller, & Topping,
1990), and high levels of stress (Wells-Parker et al.,1990). Self-efcacy has also been found
to be positively associated with specic behavioral tendencies, such as parental efforts to ed-
ucate themselves about parenting (Spoth & Conroy, 1993) and parental efforts to teach their
children injury prevention (Peterson, Farmer, & Kashani, 1990). Building on the expanding
eld of research, linking maternal self-efcacy beliefs to competence in parenting, a few studies
have identied self-efcacy beliefs as mediators of the effects of several historically recognized
correlates of parenting quality including maternal depression, child temperament, and social
support (Bugental et al., 1989; Goodnow & Collins, 1990; Johnston & Mash, 1989; Teti &
Gelfand, 1991). short
Parents who lack a sense of efcacy in their own ability to parent seem to be unable to standard

top of rh
Maternal Self-Efcacy 129 base of rh

cap height
put parenting knowledge into action, become preoccupied with themselves, often experience base of text
high levels of emotional arousal, and do not show persistence in parenting (Grusec, Hastings,
& Mammone, 1994). This characterization of parents who exhibit poorly established beliefs in
their own competencies within the domain of parenting is congruent with Banduras general
conceptualization of self-efcacy. Bandura (1982) has proposed that when confronted with
stress, individuals with low estimations of personal efcacy tend to give up easily (presumably
due to failure expectancies), internalize failure, and may experience pronounced anxiety and
depression, as well as diminished role satisfaction. Inefcacious parents tend to feel overly
burdened by their parental duties, frequently becoming immobilized by the emotional and
physical tasks involved. On the opposite end of the spectrum, efcacious parents tend to ex-
perience positive mental health and a sense of personal empowerment, which renders the man-
agement of the varied tasks of parenting less taxing and inherently more satisfying (Bohlin &
Hagekull, 1987; Kwok & Wong, 2000). An efcacious parental outlook should facilitate both
intrinsic interest in parenting and commitment to the process (Bandura, 1995). As noted by
Bandura (1997), personal fulllment and satisfaction are most likely to be achieved through
engagement in challenging activities that individuals feel competent in pursuing.
A few studies have revealed a relationship between low parenting self-efcacy beliefs and
compromised child outcomes in the domains of socio-emotional development (Donovan &
Leavitt, 1985, 1989; Swick & Hassell, 1990) and school achievement (Bandura, Barbaranelli,
Caprara, & Pastorelli, 2001; Elder, Eccles, Ardelt, & Lord, 1995). Nevertheless, as noted by
Coleman and Karraker (1998), the correlates of deciencies in parents self-perceptions of
competence relative to child development remain largely unexamined and represent an impor-
tant goal of this study. The available data do indicate that subjective perceptions of competence
are strong correlates of actual parenting behavior; and parenting behavior is likely to inuence
the emotional and intellectual growth of children in addition to meaningfully affecting the
overall quality of child-rearing environments (for a review, see Strand & Wahler, 1996). Fur-
ther, work by Benedek (1970) and by Pridham and Chang (1992) suggests that mothers with
positive self-evaluations of parenting competence may be freer cognitively and emotionally to
attend to their infants growth, development, and temperament resulting in more adaptive child
Based on the above theoretical and empirical work exploring the link between maternal
self-efcacy and various parenting and child outcomes, it is tempting to assume that these
cognitions are causally related to the outcome measures. However, at this stage in the devel-
opment of the literature, such assumptions of directionality are premature, particularly when
assessments occur at only one point in time as in the current investigation. For instance, the
possibility that particular child behavioral styles attenuate self-efcacy beliefs is as equally
logical as the reverse, and the expression of each variable is surely inuenced by multiple other
factors and processes. Moreover, it is important to remain cognizant of the possibility that
complex, multidirectional relationships among self-efcacy and other variables relevant to
parenting are operative in the development of parenting competence and childrens develop-
ment (Coleman & Karraker, 1998).
Three different formulations of parenting self-efcacy have been described in the literature.
First, the task-specic approach focuses on parents perceptions of their own competence re-
lated to discrete tasks within the domain of parenting (e.g., caring for a child with a fever or
providing particular forms of recreation for ones child; Ballenski & Cook, 1982; Teti &
Gelfand, 1991). The second approach, termed domain-specic, entails combining data gener-
ated from task-specic measures of self-efcacy (Bandura, Babaranelli, Caprara, & Pastorelli,
1996). For example, parents might be asked to rate their perceptions of competence related to short
a number of discrete discipline, emotional nurturance, and physical child care behaviors. Then standard

top of rh
130 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
the efcacy information derived from the individual task-specic items is combined to create base of text
a multidimensional index of domain level parenting self-efcacy. In the third approach, known
as the domain-general method, parenting self-efcacy is viewed as conceptually distinct from
self-efcacy in other domains; however, assessment is based on global competence expectations
that are not linked to particular parenting tasks (Abidin, 1990; Dumka, Stoerzinger, Jackson,
& Roosa, 1996; Wells-Parker et al., 1990).
The dominant approach to measuring parenting self-efcacy has been the use of domain-
general measures. However, as suggested by Bandura (1989), domain-specic measures of
self-efcacy, when compared to more global assessments, are likely to garner more precision
in terms of the associations between self-appraisals and actual behavior. Indeed, a substantial
literature in domains other than parenting attests to the superior predictive validity of task-
specic over global measures of self-efcacy (Beck & Lund, 1981; Earley & Lituchy, 1991;
Lachman & Leff, 1989; Multon, Brown, & Lent, 1991; Pajares & Miller, 1995; Wang &
Richarde, 1988). In addition, Bandura (1989) has noted that specication of causal structures
or the patterning of variables linking self-efcacy beliefs to performance outcomes becomes
realistic only with highly specied self-efcacy beliefs.
With most studies of parenting self-efcacy using only one measurement strategy, very
little information is available pertaining to how assessments of parenting self-efcacy beliefs
based on discrepant conceptualizations may be differentially related to parent and child behav-
ior. Therefore, measures of both domain-specic and domain-general parenting self-efcacy
beliefs were purposely included in this study to enable exploration of possible differences in
the relative strength of association with respect to the two measurement strategies. Although
research outside the parenting area provides evidence supportive of the notion that self-efcacy
beliefs, measured at the domain-specic level, are likely to operate as stronger correlates of
parenting competence and child outcomes, a formal hypothesis relative to this expectation was
not formulated due to the lack of previous data specic to the parenting domain. Nevertheless,
the researchers did expect the domain-specic measure to operate as a more potent predictor
of parent and child behavior than the domain-general index.
Several instruments that measure parenting self-efcacy at the domain-general level are
available (Abidin, 1990; Dumka et al., 1996; Wells-Parker et al., 1990); but domain-specic
parenting self-efcacy measures that incorporate Banduras task-oriented approach are very
scarce in the literature. In fact, an exhaustive review of the existing research revealed only Teti
and Gelfands (1991) Maternal Efcacy Questionnaire, developed primarily for use with par-
ents of infants, and Banduras (1996) unpublished Multidimensional Scales of Perceived Self-
Efcacy, which contain the Parenting Self-Efcacy Scale, designed for use with parents of
school-age children. Therefore, a new domain-specic measure of parenting self-efcacy suit-
able for administration to parents of toddlers was constructed for this study.
Based on the above review of literature, the following hypothesis was tested: parenting
self-efcacy beliefs, measured at the domain-specic and domain-general levels of analysis,
were expected to operate as signicant correlates of maternal competence in the parenting of
toddlers, toddler developmental status, and toddler interactive behavior. Higher self-efcacy
beliefs were expected to be associated with more adaptive mother and child outcomes.


The participants in this study included sixty-eight mothers and their 19- to 25-month-old tod- short
dlers (33 males, 35 females) residing in, or within 30 miles from, either a moderate size college standard

top of rh
Maternal Self-Efcacy 131 base of rh

cap height
town in rural Virginia (n 42) or a small college town in Tennessee (n 26). The average base of text
toddler age was 21.16 (SD 1.86) months. Participants were primarily identied through
examination of birth announcements listed in the local newspapers from previous years. Sixty-
ve percent of the women initially contacted were willing to participate. Diversity in terms of
ethnicity and socio-economic status was sought; but was unfortunately not obtained due to the
predominantly White, middle-class population of both locales. Specically, 66 of the mother
toddler dyads were White and 2 were Black; and the average gross annual family income was
in the $41,000 to $50,000 range. The average number of years of mothers formal education
was 15 (SD 2.32), with a range extending from 8 to 20. Mothers ages ranged from 20 to
44, with a mean of 31.31 (SD 4.58). With regard to marital status, 88.2% of the mothers
were married for the rst time, 10.3% were married for the second time, and 1.5% were single.
Twenty-three of the mothers (34%) only had one child, 30 mothers (44%) had two children,
and 15 mothers (22%) had three to six children. Twenty-seven mothers (40%) were employed
full-time, 14 mothers (20%) reported working part-time, and 27 mothers (40%) were not em-
ployed outside of the home. Finally, when asked how much previous experience the women
had with children other than their own, two (2.9%) reported none, 15 (23.5%) reported very
little, 25 (37%) indicated a moderate amount, and 24 (35%) reported having had very much

Demographic questionnaire. The subjects were asked to provide some basic background in-
formation pertaining to their age, ethnicity, education level, annual income, employment status,
marital status, number of children, and their childrens child care history.

Domain-specic parental self-efcacy. The Self-Efcacy for Parenting Tasks Index Toddler
Scale (SEPTI -TS) was used to measure domain-specic parenting self-efcacy.
This 53-item scale was developed by the authors of the current study to provide a com-
prehensive index of domain-specic self-efcacy for use with parents of toddlers. The dimen-
sions of parenting represented on the scale were selected based on the most salient dimensions
of toddler caregiver relationships delineated by Zeanah et al. (1997) and originally formulated
by Emde (1989). The seven subdomains along with an example item include the following:
(1) emotional availability When my child needs me, I am able to easily put aside whatever
else I may be doing; (2) nurturance, valuing the child, and empathetic responsiveness My
toddler knows that I understand when his/her feelings are hurt; (3) protection from harm or
injury I am very good about never leaving my child unattended; (4) discipline and limit
setting Setting limits for my toddler is relatively easy for me; (5) play I can always think
of something to play with my child; (6) teaching I believe my toddler learns a great deal
from my efforts to show him/her things; and (7) Instrumental care and establishment of struc-
ture and routines I have been able to establish a daily routine with my toddler that feels
comfortable to both of us. Each of the items are rated on a six-point Likert scale, with possible
responses ranging from Strongly Agree to Strongly Disagree. Total scores have a potential
range of 53 to 318, with higher scores indicative of stronger self-efcacy. The full scale with
all of the items is presented in the appendix.
Although subscale scores were not used in this study, internal consistency reliability co-
efcients were computed to facilitate future development of the scale. Cronbachs alpha co-
efcients for each of the subscales using the current sample were of variable magnitude: Emo-
tional Availability (seven items) .67; Nurturance (eight items) .71; Protection (seven short
items) .53(.63 with item 16 removed); Discipline (nine items) .81; Play (seven items) standard

top of rh
132 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
.92; Teaching (seven items) .73; and Instrumental Care (eight items) .46 (.60 with item base of text
51 removed). Clearly, the Protection and Instrumental Care subscales should be revised prior
to further use. Intercorrelations among the subscale scores ranged from .26 to .71, with a mean
of .43. Cronbachs alpha coefcient for the full scale was found to equal .91. All 53 items were
retained because the results of an item analysis revealed that elimination of any particular items
would not signicantly enhance the internal consistency reliability of the full scale.
A second measure of domain specic self-efcacy was included to provide discriminant
validity evidence for the SEPTI-TS as the tasks related to competent infant care are different
from the tasks related to positive parenting of toddlers. The Maternal Efcacy Questionnaire
(MEQ; Teti & Gelfand, 1991) consists of nine four-point items designed to assess mothers
self-efcacy beliefs in relation to specic areas of infant care, such as soothing the child,
maintaining joint attention and interaction with the child, and performing daily routine tasks.
A tenth item is related to general feelings of competence in parenting: In general, how good
a parent do you feel you are. Scores are summed to derive a total maternal self-efcacy score
and the potential range of scores is from 9 to 36, with higher scores indicative of stronger self-
efcacy beliefs. Cronbachs alpha, an index of internal consistency reliability, was found by
Teti and Gelfand to equal .86 for the MEQ. Concurrent validity evidence is also provided by
the authors, as scores on the MEQ were shown to correlate strongly with the Sense of Com-
petence subscale of the Parenting Stress Index (PSI; Abidin, 1990). Using data derived from
the current sample, Cronbachs alpha was found to equal .83.

Domain-general parenting self-efcacy. To assess domain-general parenting self-efcacy, the

seven-item Efcacy subscale of the Parenting Sense of Competence Scale (PSOC), developed
by Gibaud-Wallson and Wandersman (1978, cited in Johnston & Mash, 1989) was used. Each
item on the PSOC is answered on a six-point scale, with response options ranging from Strongly
Disagree 6 to Strongly Agree 1. With a possible score range spanning from 7 to 42, higher
scores represent stronger self-efcacy. An example of an item on the Efcacy subscale is I
meet my own personal expectations for expertise in caring for my child. With a sample
exceeding 500 subjects, Johnston and Mash (1989) have provided construct validation for the
PSOC and have found the Efcacy subscale to possess sufcient internal consistency reliability
(Cronbachs alpha .76). The developers of the scale have likewise reported adequate internal
consistency reliability (Gibaud-Wallson & Wandersman, 1978, cited in Johnston & Mash,
1989). Data generated with the current sample revealed a Cronbachs alpha coefcient of .81.
Although this scale has primarily been used with parents of older children, the content of the
items is general enough to be appropriate for parents of toddlers.

General self-efcacy. Sherer and Adams (1983) Self-Efcacy Scale (SES), a 17-item mea-
sure, was used as an index of general self-efcacy, unrelated to any particular domain of
functioning. Although not used in any of the primary analyses, this measure was included to
provide the opportunity to assess discriminant validity relative to the newly constructed SEPTI-
TS. The authors report acceptable internal consistency reliability (Conbachs alpha coefcient
.86) and construct validity evidence. The scale was modied from the original ve-point
Likert response option format to a six-point scale to facilitate ease of administration with other
six-point measures used in the current. Cronbachs alpha was found to equal .86 for this slightly
altered version. With a potential range of scores from 17 102, higher scores reect higher
levels of general self-efcacy. An example of an item from this scale is When I decide to do
something, I go right to work on it.
Assessment of toddler and caregiver behaviors. Quality of parenting and various indexes of standard

top of rh
Maternal Self-Efcacy 133 base of rh

cap height
toddlers behavior were assessed using the Crowell Procedure, which was rst described by base of text
Crowell and Feldman (1988) as an adaptation of the Matas, Arend, and Sroufe (1978) tool
use task, and was more recently outlined in detail by Zeanah et al. (1997). Caregivers are
given a detailed visual and verbal orientation beforehand. The procedure consists of nine sep-
arate episodes of varying duration designed to elicit a wide range of caregiver and toddler
behaviors. In the original procedure a telephone is placed in the room and the mother is called
at each transition point with verbal instructions for the next step. However, for this study,
written instructions were provided to the mothers on index cards and there was a clock in the
room to enable them to keep track of the time for each segment. The nine episodes in order of
presentation include the following: (1) unstructured free play with a number of toys, (2) clean-
up, (3) play with bubbles, (4 7) teaching tasks of increasing difculty, (8) brief separation
from the caregiver, and (9) reunion. During free play, the mother and toddler play as they
normally would at home with the aid of a basket of toys containing a doll with a bottle, three
puppets, a doctors kit, two toy telephones, two toy plates and food, three small dollhouse
dolls, and a toy tool kit. The clean-up session consists of requesting the mother to instruct her
child to return the toys to the basket, helping the child if necessary, and having the mother
place the toys outside the room. During the bubble segment the mother is asked to blow bubbles
and to encourage her toddler to pop them. The teaching tasks selected included completion of
a puzzle, placing toy passengers into a toy bus, building a tower of 6 blocks, and operating a
pop-up toy. All of the episodes outlined above were conducted; however, as recommended by
Zeanah et al., only behavior observed during the rst seven episodes (minus the bubbles seg-
ment) were used to assess the teaching-oriented variables of interest. The freeplay, clean-up,
and teaching episodes are designed to provide a relatively unstructured opportunity for the
parent to engage in directing the childs behavior in a way that encourages enjoyment and
learning. The separation and reunion segments provide a brief attachment screening and the
bubbles segment serves to relax the mother and child prior to introduction of the teaching
The testing room was furnished with two adult chairs, a locked cabinet for teaching task
toy storage (the parent was given instructions for opening it), and three large plastic blocks
that could be used as tables or turned on their sides to function as chairs. Free play toys were
kept in a basket and were removed from the room following clean-up.
The Crowell Procedure enables measurement of two broad dimensions of parenting qual-
ity: Supportive Presence, and Quality of Assistance. The Supportive Presence variable is com-
posed of two major criteria (Mother Involvement and Secure Base); as is the Quality of As-
sistance variable (Helping the Child Understand and Complete the Task and Minimal
Assistance). Total scores for each of these two primary global variables are derived using
anchor points ranging from 1 to 7, with higher scores indicative of more positive parenting
behavior. In the current study, scores on these two dimensions of parenting were found to be
highly correlated (r .85); therefore, the Supportive Presence and Quality of Assistance var-
iables were combined to create a composite parenting variable with a potential range extending
from 2 14.
The Crowell Procedure also allows for the assessment of several toddler behaviors rated
on a scale of 1 7: Avoidance of Mother, Affection Towards Mother, Compliance, Enthusiasm
Persistence, Negativity, and Experience of the Session. In each case, higher scores suggest
higher levels of the variable observed. As an index of interrater agreement, Crowell and Feld-
man (1988) reported kappa coefcients of .87 and .94 for the Quality of Assistance and Sup-
portive Presence variables respectively and coefcients ranging from .68 to 1.0 on the toddler
behavior variables. short
Two trained undergraduate students coded maternal and child behaviors observed during standard

top of rh
134 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
the Crowell Procedure. Training consisted of viewing a training video produced by the devel- base of text
opers of the procedure, becoming thoroughly acquainted with the coding manual, and coding
the behavior of three sample dyads provided on the training video. The students were also
given ample opportunity to ask questions for clarication on the coding scheme and to share
experiences prior to coding the actual study videos. Cohens kappa statistic was computed for
each of the composite maternal behaviors and the seven toddler behaviors coded using the
training dyads. For the Supportive Presence and Quality of Assistance maternal variables, kappa
coefcients were found to equal 1.0 and .66, respectively, and kappas for the child behaviors
ranged from .66 to 1.0. Based on the levels of interrater agreement achieved with the training
dyads no further training was deemed necessary. With the data collected in Virginia, one master
rater coded behaviors for all 42 mother toddler dyads; whereas the second rater only coded
behaviors for the rst 21 dyads. Cohens kappas were again calculated using 21 of the study
dyads (50% of the data). Kappa coefcients for the Supportive Presence and Quality of As-
sistance maternal variables were equal to .75 and .80, respectively. Kappa coefcients for the
child behavior variables were as follows: Avoidance of Mother .68, Affection Towards Mother
.70, Compliance .78, Enthusiasm .82, Persistence .70, Negativity .75, and Experience of the
Session .68. Because sufcient interrater agreement was obtained, only data coded by the rater
who scored all of the dyads were used for data analysis purposes. Similarly, with the data
collected in Tennessee, two undergraduate students went through a comparable level of train-
ing; however, both students coded 100% of the tapes and the kappa coefcients ranged from
.66 to .82.

Child developmental status. The Mental Scale of the Bayley Scales of Infant Development
(BSID-II; Bayley, 1993) was used to measure the current developmental status of the toddlers.
Items included on the Mental Scale tap into memory, problem-solving, numerical concepts,
generalization, classication, expressive and receptive language, and social skills. Raw scores
are converted to a Mental Development Index (MDI). With the standardization sample, the
mean and standard deviation for MDI scores were equal to 100 and 15, respectively. According
to the test developers (Bayley, 1993), the BSID-II possesses high internal consistency, test-
retest, and interrater reliability. Specically, Cronbachs alpha for the MDI was reported to be
.90 with a sample of 21 month olds, and .92 with a sample of 24 month olds. Further, test-
retest reliability is high across all ages, with an average correlation of .83. Finally, interrater
reliability for the MDI, based on the correlation between scores for 51 children was found to
equal .96. The test developers have also reported substantial validation evidence. For example,
the authors summarize several studies using factor analytic evidence in support of the construct
validity of the scale. They also describe several studies designed to measure the predictive
validity of the MDI, concluding that although scores at the scale level are not predictive of
later intellectual ability, language skills, or achievement, when administered after age 2, scores
do begin to predict later childhood functioning. Finally, as related to discriminant validity, MDI
scores have been found to reliably differentiate between infants at risk and those experiencing
normal development. Moreover, the results obtained with use of the measure are consistent
with trained clinicians impressions of childrens functioning.

Mothers with appropriate age children were mailed a brief description of the study, and ap-
proximately 10 days after the letters were mailed, the women were called, more information
about the study was provided (time required, procedural details, and a general description of short
the purpose and utility of the research), and the women were asked if they were interested in standard

top of rh
Maternal Self-Efcacy 135 base of rh

cap height
participating. Interested mothers were then mailed informed consent forms, which assured them base of text
of the condentiality of their responses, along with a preaddressed and stamped return envelope.
After one copy of the informed consent form was returned, the experimenter contacted each
participant and arranged for a two-hour laboratory assessment during a time of day when
mothers anticipated that their children would be most alert. Two weeks prior to the scheduled
visit, a packet of questionnaires was mailed to each participant with instructions to bring the
completed forms to the scheduled laboratory appointment. To reduce the likelihood of a social
desirability bias in mothers responses to the questionnaires, instructions to the participants
emphasized the difculty of parenting toddlers, the importance of answering the items honestly,
and stressed the fact that all personal identifying information would be detached from their
responses. The questionnaires assessed demographic information, general self-efcacy, and
parenting self-efcacy. Mothers indicated that the time required to complete these measures
did not exceed one hour.
The session began with administration of the Bayley Scales to the toddlers, with mothers
holding their children on their laps throughout the testing. Testing took approximately 45
minutes, and was followed by a brief (5 10-minute) break period during which mothers and
toddlers were encouraged to rest and relax in a room with reading material and toys available.
Observational assessment of parental and child behaviors using the Crowell Procedure followed
the break period and took approximately 30 minutes to complete. The freeplay and the bubbles
segments lasted 5 minutes each and the clean-up and teaching episodes lasted from 3 5 min-
utes, depending on the interest level and cooperation of the toddler. Mothers were instructed
to move on to the next task if it became clear that the child was unable or unwilling to work
on the task at hand after 5 minutes had elapsed. Interested and cooperative toddlers tended to
take at least 3 minutes to complete each task. Both administration of the Bayley Scales and
the Crowell Procedure were videotaped.
Following participation in the study, subjects were graciously thanked, offered the oppor-
tunity to receive a copy of the results of the study, and were given a small commemorative
token of appreciation (a developmentally appropriate toy or an article of clothing for their


Basic Descriptive Statistics and Correlations Among the

Study Variables
Table 1 provides the descriptive statistics for all of the primary study variables. Parenting self-
efcacy beliefs, as measured at both the domain-specic and the domain-general levels, tended
to be very high for the sample. The average participant exhibited high levels of competence in
parenting behaviors as displayed during the Crowell Procedure. The toddlers generally dem-
onstrated positive behavior in interaction with their mothers, as indicated by the group scores
on the seven behavioral variables coded during the Crowell Procedure. Finally, the mean and
standard deviation derived from the MDI scores were nearly equivalent to those achieved with
the standardization sample (Bayley, 1993).
Table 2 provides a zero-order correlational matrix with all of the primary study variables.
As is apparent from examination of the data presented, moderate to strong correlations were
evidenced between each pair of the child behavior variables observed. Bayley MDI scores were
moderately correlated with each of the child behaviors observed during the Crowell Procedure short
as well. Signicant strong correlations were also detected between maternal parenting com- standard

top of rh
136 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
TABLE 1. Descriptive Statistics for Study Variables base of text

Variable Potential Range Observed Range M SD

PSOC-efcacy (domain-general) 7.00 42.00 16.00 42.00 32.61 5.46

SEPTI-TS (domain-specic) 53.00 318.00 210.00 308.00 267.54 22.09
Parenting competence 2.00 14.00 4.00 14.00 11.28 3.06
Toddler enthusiasm 1.00 7.00 2.00 7.00 5.49 .91
Toddler compliance 1.00 7.00 2.00 7.00 5.54 1.16
Toddler affection mother 1.00 7.00 1.00 7.00 4.82 1.52
Toddler persistence 1.00 7.00 2.00 7.00 5.44 1.12
Toddler avoidance of mother 1.00 7.00 1.00 6.00 1.91 1.02
Toddler experience of the session 1.00 7.00 2.00 7.00 5.34 .97
Toddler negativity 1.00 7.00 1.00 7.00 1.76 1.06
MDI (BSID-II, Mental Scale) 55.00 145.00 64.00 128.00 100.21 14.15

petence (combined Supportive Presence and Quality of Assistance variables) and the seven
child behavior variables and MDI scores. The domain-general and the domain-specic mea-
sures of self-efcacy were moderately correlated with each other. Finally, signicant zero-
order correlations were observed between the domain-specic measure of self-efcacy and
MDI scores as well as ve of the behaviors observed during the mother child interactive
session. However, scores on the domain-general measure of self-efcacy were not associated
with any of the child behavior variables.

Discriminant Validity Evidence for the SEPTI-TS

Total scores on the SEPTI-TS, the domain-specic measure developed for use in this study,
were found to be signicantly correlated with scores on the PSOC-Efcacy subscale, r(66)
.48, p .001. A stronger correlation between scores on the SEPTI-TS and the PSOC was not
expected (and presumably not obtained), because the PSOC is a domain-general measure.
Scores on the SEPTI-TS were related to total scores on the MEQ, an index of domain-specic
self-efcacy designed for parents of infants, r(66) .27, p .028. More congruence was not
expected here, because the tasks associated with competent parenting of infants versus toddlers
vary pretty dramatically. There was also a signicant association between SEPTI-TS scores
and the one domain-general item on the MEQ, r(66) .27, p .031. Finally, a signicant
correlation was observed between SEPTI-TS scores and the SES, a measure of general self-
efcacy, r(66) .32, p .009.

Preliminary Exploratory Analyses with Demographic Data

A number of preliminary exploratory analyses were conducted using correlation coefcients
to determine whether or not any of the demographic variables were related systematically to
any of the primary study variables, necessitating the introduction of possible covariates into
the analyses conducted to test the hypotheses. Maternal age was found to be inversely correlated
with the domain-general measure of parenting self-efcacy (PSOC scores), r(67) .37,
p .01. Therefore, maternal age was introduced as a covariate in all of the analyses conducted
to test the hypothesis involving domain-general parenting self-efcacy. None of the remaining
demographic items (maternal education, annual income, maternal employment, marital status,
number of children, child age, nor child gender) were signicantly related to any of the primary short
study variables. standard
IMHJ (Wiley)
TABLE 2. Zero-Order Correlations Among Study Variables

2 3 4 5 6 7 8 9 10 11

1. PSOC-efcacy (domain-general) .48*** .01 .06 .03 .02 .06 .03 .04 .08 .00
2. SEPTI-TS (domain-specic) .13 .24* .27* .27* .33** .13 .25* .21 .24*
3. Parenting competence .31** .52*** .60*** .54** .54*** .55*** .50*** .44***
4. MDI (BSID-II mental) .49*** .48*** .40*** .42*** .50*** .54*** .34***

5. Toddler enthusiasm .82*** .65*** .84*** .75** .89*** .34**
6. Toddler compliance .60*** .83*** .80*** .77*** .73***
7. Toddler affection .51*** .46*** .56*** .36**
8. Toddler presistence .80** .83*** .70***
9. Toddler avoidance .77*** .77***
10. Toddler experience of the session .69***
11. Toddler negativity

Note. * p .05; ** p .01, *** p .001.

base of text
cap height

base of rh
top of rh

top of rh
138 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
With 42 subjects tested in one geographical location and 26 subjects tested in a second base of text
location, several analyses were conducted prior to combining the data to make sure that the
samples did not differ signicantly in terms of any of the demographic variables. No differences
were observed relative to maternal age, child age, gender of the child, maternal education,
family income, marital status, race, or maternal employment.

Parenting Self-Efcacy Beliefs as Predictors of Parenting

To test the rst part of the hypothesis, which predicted that domain-specic and domain-general
parenting self-efcacy beliefs would operate as signicant predictors of maternal competence
in the parenting of toddlers, a multiple regression analysis was conducted with the composite
parenting competence variable from the Crowell Procedure operating as the criterion. The two
measures of parenting self-efcacy operated as the independent variables with maternal
age entered initially as a covariate. The analysis was not signicant, F(3, 62) .79, p .05,
r 2 .04.

Parenting Self-Efcacy Beliefs as Predictors of Child

Developmental Status and Behavior
A series of regression analyses were conducted to test the second part of the hypothesis that
postulated that domain-specic and domain-general parenting self-efcacy beliefs would op-
erate as signicant predictors of toddler developmental status (Bayley MDI scores) and behav-
ioral outcomes (seven Crowell Procedure behaviors). In every analysis, the two measures of
self-efcacy operated as the predictors and were entered after maternal age was controlled. The
rst test employed Bayley MDI scores as the criterion and it was signicant, F(3, 62) 2.46,
p .05, r 2 .11. The details of this analysis are provided in Table 3. High scores on the
domain-specic measure of self-efcacy (SEPTI-TS) signicantly predicted high scores on the
Bayley MDI scores; however, the domain-general index of self-efcacy (PSOC) was not found
to be a signicant predictor. Five of the seven regression analyses conducted to examine the
strength of the two measures of self-efcacy as predictors of various toddler behaviors were
signicant. The results of the analyses were as follows: Affection Towards Mother,
F(3, 62) 3.21, p .05, r 2 .13; Avoidance of Mother, F(3, 62) 3.09, p .05, r 2 .13;
Compliance, F(3, 62) 2.90, p .05, r 2 .12; Enthusiasm, F(3, 62) 2.63, p .05, r 2
.11; Negativity, F(3, 62) 2.80, p .05, r 2 .12; Persistence, F(3, 62) 1.19, p .05,
r 2 .05; and Experience of the session, F(3, 62) 1.46, p .05, r 2 .07. The details of
these seven analyses are provided in Table 3. High scores on the domain-specic measure of
self-efcacy (SEPTI-TS) signicantly predicted high levels of child Affection Towards Mother,
Compliance, and Enthusiasm in addition to low levels of Avoidance of Mother and Negativity;
however, scores on the domain-general index of self-efcacy (PSOC) did not signicant predict
scores on any of these variables.
In a follow-up analysis with the seven child behaviors entered into a regression analysis
as predictors of domain-specic self-efcacy, the combination of variables explained 19%
of the variance in mothers self-efcacy beliefs and the model approached signicance,
F(7, 60) 2.02, p .067.

Parenting self-efcacy beliefs are well established in the available literature as a central cor- short
relate of parenting behaviors likely to be related to positive child-rearing environments that are standard

top of rh
Maternal Self-Efcacy 139 base of rh

cap height
TABLE 3. Regression Analyses Using Domain-Specic and Domain-General Parenting Self-Efcacy base of text
as Predictors of Child Variables with Maternal Age Entered as a Covariate

Dependent Variable Predictors Beta T Sig T Change in r2

Bayley MDI scores Maternal age .04 .31 .755 .00

SEPTIS-TS .37 2.65 .010 .07
PSOC-efcacy .23 1.59 .117 .03
Enthusiasm Maternal age .03 .27 .798 .00
SEPTIS-TS .39 2.79 .007 .09
PSOC-efcacy .19 1.29 .202 .02
Compliance Maternal age .02 .15 .889 .00
SEPTIS-TS .40 2.94 .005 .09
PSOC-efcacy .23 1.59 .117 .03
Affection towards mother Maternal age .04 .35 .774 .00
SEPTIS-TS .44 4.21 .019 .12
PSOC-efcacy .16 1.09 .282 .10
Persistence Maternal age .12 .89 .377 .01
SEPTIS-TS .24 1.64 .105 .04
PSOC-efcacy .11 .75 .459 .01
Avoidance of mother Maternal age .03 .26 .795 .00
SEPTIS-TS .41 3.01 .004 .09
PSOC-efcacy .15 1.90 .094 .04
Experience of the session Maternal age .04 .29 .776 .00
SEPTIS-TS .28 1.98 .052 .06
PSOC-efcacy .08 .543 .589 .00
Negativity Maternal age .12 .94 .354 .01
SEPTIS-TS .38 2.75 .008 .10
PSOC-efcacy .15 1.04 .303 .01

conducive to maximizing childrens development. This study entailed an exploration of par-

enting self-efcacy beliefs as correlates of parenting competence, toddlers overt behavior in
a laboratory setting, and toddlers performance on the Mental Scale of the Bayley Scales of
Infant Development (BSID-II).
The data revealed that domain-specic parenting self-efcacy beliefs were signicantly
related to six of the child variables (Avoidance of Mother, Affection Towards Mother, Com-
pliance, Enthusiasm, and Negativity as assessed during the Crowell Procedure and BSID-II
Mental Scale scores). Parenting self-efcacy beliefs measured at the domain-specic level were
not, however, related to two toddler behavior variables coded during the Crowell Procedure
(Persistence and Experience of the Session). Domain-general self-efcacy beliefs were not
found to be signicantly associated with any of the child variables. Perhaps even stronger
relations between self-efcacy beliefs and the various child behaviors would have emerged if
the mother toddler dyads were observed when the children were tired or understimulated and
their behavior more variable. Under such circumstances, when toddlers are more prone to
exhibiting oppositional behavior, more active parental effort would be required to elicit com-
pliance and persistence. Simply expanding the length of the observational period would in-
crease the probability of demanding child behavior arising. With greater challenges to mothers
attempts to keep their children focused and on task, the relations between self-efcacy beliefs short
and the toddler behaviors may have become more pronounced. As indicated previously, the standard

top of rh
140 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
association between self-efcacy and overt behavior tends to be most salient under stressful base of text
Clearly, many variables in addition to self-efcacy are at play in explaining the behavioral
and developmental variability observed in this investigation. Nevertheless, when the six (out
of eight) signicant associations are considered, an average of 12% of the variance in the child
variables was explained by the variability in domain-specic self-efcacy beliefs, rendering
the associations rather intriguing. If mothers with high self-efcacy tend to have children, who
are more advanced developmentally, are more likely to interact positively, and are less inclined
to exhibit negative behavior, then intervention studies designed to enhance self-efcacy beliefs
could be designed to help resolve the directionality problem. In the event that a causal rela-
tionship is revealed, programs designed to elevate mothers feelings of efcacy may enhance
childrens competencies, while reducing the prevalence of behavior problems. Recent inter-
vention studies do suggest that maternal self-efcacy beliefs are potentially alterable through
education and training (e.g., Gross, Fogg, & Tucker, 1995; Tucker, Gross, Delaney, & Lap-
porte, 1998).
Although signicant associations were observed between self-efcacy and a number of
the child outcomes, signicant relationships were not observed between parenting self-efcacy
beliefs (measured at both the domain-specic and domain-general levels) and the composite
parenting variable coded during the Crowell Procedure. These results become more compre-
hensible when various methodologic issues are considered. For example, stronger relationships
between mothers self-efcacy beliefs and the quality of parenting behaviors might have been
detected if a microanalytic approach to coding mothers interactive behaviors had been adopted.
Instead of coding variables such as encouraging and supporting the childs efforts, which
require the rater to decide if the mother generally responds to the childs discrete actions and
efforts to solve the task by showing strong positive affect or some verbal praise or reinforce-
ment, the frequencies of more highly specied behaviors such as positive or negative contin-
gent verbalizations, changes in tone of voice, or physical gestures could have been analyzed.
As noted by Kochanska (1990), previous attempts to demonstrate signicant associations
between parenting beliefs and actual child-rearing behaviors have been met with mixed results.
More reliable correlations between beliefs and actions tend to emerge in the literature when
there is a close match between the self-report items and the particular behaviors of interest. If
the behavioral indicators of parenting competence had been more closely associated with the
specic dimensions of self-efcacy of interest, stronger associations may very well have been
revealed. Additional research in this area should strive for a closer match between the aspects
of self-efcacy measured and the actual behaviors coded.
Many of the parenting behaviors that may be differentially impacted by parents feelings
of efcacy in the role are perhaps difcult to measure based on a relatively brief and structured
excerpt of dyadic behavior in a laboratory setting, such as in the Crowell Procedure. The
laboratory context, complete with interesting toys and activities, may not have been conducive
to evoking toddlers typical at-home behaviors, which are likely to elicit the critical forms of
parenting behavior inuenced by self-efcacy beliefs. A number of mothers expressed surprise
with how cooperative their toddlers were, and there was very little toddler behavioral varia-
bility, with the exception of that added by a few irritable toddlers. All of the Crowell Procedure
behaviors were moderately to highly intercorrelated and they were all signicantly related to
toddlers Mental Scale scores on the Bayley, suggesting that the toddlers generally displayed
qualitatively uniform behavior throughout the session. Toddlers in good spirits tended to be
affectionate, cooperative, enthusiastic, etc., exhibiting positive behavior for the duration of the
Crowell Procedure; whereas toddlers who were more negativistic tended to be so during the short
entire observation period. This nding of seemingly nonorthogonal child behaviors is consistent standard

top of rh
Maternal Self-Efcacy 141 base of rh

cap height
with previously published work employing the Crowell Procedure (Crowell & Feldman, 1988). base of text
However, the tendency toward a halo effect may not have been evidenced with observations
of children and mothers under more naturalistic, less novel circumstances. For example, Teti
and Gelfands (1991) study of the relationship between maternal self-efcacy beliefs and be-
havioral competence involved assessment of parenting behavior based on three in-home ob-
servations that took place on different days over a one-month period. Further research on the
relationship between parenting self-efcacy and parenting competence under less articial con-
ditions and involving the sampling of behavior over a more extended time frame seems merited.
The mean was very high on the composite assessment of parenting competence. Mothers
are undoubtedly more inclined to engage in appropriate stimulating and nurturing behavior,
when there are few distractions, time is limited, and they are aware of being videotaped. The
mothers were informed that the focus of the study was on individual differences in toddler
behavior as well as parenting strategies used during this difcult time in parenting, rendering
an element of exaggerated performance likely.
Previous studies revealing correlations between maternal self-efcacy beliefs and parenting
behavior have typically been done with either demographically diverse samples (e.g., Gross et
al., 1994) or samples that include both mothers who are clinically depressed and nondepressed
(e.g., Teti & Gelfand, 1991). Testing for associations between self-efcacy beliefs and parent
and child outcomes with such heterogeneous samples obviously has the advantage of maxi-
mized variance.
In this study, self-efcacy beliefs were found to be signicantly related to the majority of
the child variables despite the limited variability observed across nearly all measures (self-
efcacy beliefs, socio-demographic characteristics, parenting competence, and child interactive
behaviors). Most of the mothers were White, married, well-educated, middle-class women,
who had a considerable amount of previous experience with children and felt highly efcacious
in their roles as parents of toddlers. There was also rather low variability relative to the measure
of parenting competence and the child variables, presumably due in part to the brevity of the
observation period and the rather inviting nature of the laboratory context. In future samples,
an effort should be made to enhance variability by recruiting a demographically diverse sample,
locating women with more variable feelings of efcacy (perhaps by studying only rst-time
or adolescent mothers), and/or testing in more naturalistic or stress-evoking situations over
time. As noted above, with efforts such as these, self-efcacy beliefs will in all likelihood be
found to operate as even stronger correlates of parent and toddler outcomes. Moreover, in-
creasing variability within samples of sufcient size will provide the opportunity to identify
certain groups of parents or children, dened in terms of psychologic, socio-demographic, or
other parent or child individual difference factors (e.g., age, gender, personality), for whom
self-efcacy beliefs are more meaningfully related to adaptive outcomes. For example, in ad-
vantaged child-rearing contexts, where there are two parents in the home, nances are stable,
mothers are well-educated and psychologically stable, there are numerous positive factors likely
to enhance parenting and childrens development, rendering parents self-efcacy beliefs a
potentially less salient correlate of outcomes. On the other hand, when parents are faced with
adverse living conditions, feelings of efcacy could very well be the central variable that denes
parents abilities to cope effectively within various environmental constraints.
A signicant positive relationship was observed between parenting competence and tod-
dlers Mental Scale scores on the BSID-II. This nding is consistent with previous research
linking various indicators of parenting competence with toddlers cognitive performance (Born-
stein & Tamis-LeMonda, 1989; MacPhee Ramey, & Yeates, 1984). Parenting competence was
positively associated with the adaptive Crowell Procedure toddler behaviors (Affection To- short
wards Mother, Compliance, Enthusiasm, Persistence, and Experience of the Session), and was standard

top of rh
142 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
negatively associated with undesirable toddler behaviors coded during the Crowell Procedure base of text
(Avoidance of Mother and Negativity). The link between parenting quality and child behaviors
is well established in the existing literature (Crowell & Feldman, 1988; Strand & Wahler,
Bornstein and Tamis-LeMonda (1989) have suggested that various parental factors such
as responsiveness may instill children with feelings of self-efcacy, which in turn, may foster
their desire to acquire information and bolster their performances on cognitive tasks. Maternal
responsiveness or sensitivity may provide essential mechanisms for the transmission of high
parenting self-efcacy beliefs into parental behavioral competence in addition to promoting
achievement enhancing efcacious feelings in children. However, multidimensional constructs,
such as sensitivity or responsiveness, are difcult to dene in terms of specic behaviors, which
are most likely to be inuenced by task-related parenting self-efcacy beliefs. Most of the
available research designed to examine the relationship between parenting self-efcacy and
parenting competence has tended to focus on global indexes of parenting quality and the extent
to which highly efcacious parents do actually excel in more specied areas remains largely
unexamined, particularly with mothers of toddlers. What seems to be needed is more empirical
work designed to examine how self-efcacy beliefs are associated with overt parent and child
behaviors through their link to more complex parental factors, such as responsiveness, sensi-
tivity, empathy, the capacity to emotionally connect to others, interest or motivation in par-
enting, a sense of responsibility associated with the role, long-range parenting goals or aspi-
rations for ones child, stress associated with parenting, etc. If subsequent studies include
parenting self-efcacy beliefs, several of these more complex internal-type variables, and dis-
crete measures of parenting competence and child outcomes, a more comprehensive, explan-
atory model is likely to emerge.
Research pertaining to self-efcacy beliefs tends to be restricted to either the domain or
the general levels of analysis. As a result, very little is known about how assessments of self-
efcacy beliefs, based on discrepant conceptualizations, may be differentially related to be-
havior. The results of this study revealed signicant moderate to low correlations between
scores on the domain-specic self-efcacy measure and the domain-general and general self-
efcacy measures. A few other isolated studies pertaining to domains of behavior other than
parenting have found moderate correlations between general and more task-oriented self-ef-
cacy measures, implying that generalized self-efcacy indexes may operate as an informational
source used to estimate domain-specic self-efcacy or conversely that task-specic beliefs
converge to produce more generalized efcacy beliefs (Watt & Martin, 1994; Woodruff &
Cashman, 1993).
As described earlier, Bandura (1989) has suggested that domain-specic measures of self-
efcacy, when compared to global assessments, are likely to be more precise in terms of
predicting actual behavior. The results of the present study provided data consistent with this
proposition as domain-specic self-efcacy beliefs were found to be stronger correlates than
domain-general self-efcacy beliefs relative to toddlers developmental status as measured by
the Bayley Mental Scale and several toddler interactive behaviors coded during the Crowell
Although extensive psychometric analysis of the SEPTI-TS was not possible given the
relatively small sample size, preliminary internal consistency reliability evidence and the cor-
relations with the SES (a general measure of self-efcacy), the PSOC (a domain general mea-
sure), and the MEQ (a domain-specic measure tied to parenting during infancy) were prom-
ising. Approximately 10% of the variance in SEPTI-TS scores could be explained by variability
in general self-efcacy, and nearly one quarter of the variance in SEPTI-TS scores represented short
shared variance with the domain-general measure. Finally, there was some common variance standard

top of rh
Maternal Self-Efcacy 143 base of rh

cap height
(.07%) between scores on the domain-specic measure designed for administration to parents base of text
of infants and scores on the SEPTI-TS, providing preliminary discriminant validity evidence.
With further renement, it seems very likely that the SEPTI-TS, which is based on a concep-
tualization of the construct in a manner that is consistent with the Bandurian tradition and taps
into parents efcacy beliefs relative to a wide range of parenting behaviors, is likely to provide
a useful instrument for further work.
The present study, like most of the existing parenting self-efcacy research, investigated
the construct concurrently with various indexes of parenting quality and child characteristics,
precluding the conrmation of directional effects. Subsequent prospective research efforts,
coupled with more intervention studies as discussed earlier, should help decipher the temporal
relations between self-efcacy beliefs and various dimensions of parenting and child behavior.
Further, as noted by Coleman and Karraker (1998), research pertaining to the continuities and
discontinuities in the evolution of parental self-efcacy beliefs across all of the years of active
parenting is needed as this literature base matures.
Parenting self-efcacy is a relatively new research area possessing great promise for re-
solving many ambiguities related to individual differences in adapting to parenting. However,
several avenues for further study of parenting self-efcacy remain virtually unexamined. For
example, are there particular parental personality variables or environmental situations that
tend to predispose individuals to high or low parenting self-efcacy? How resistant to change
are parental self-efcacy beliefs in the face of nonnormative events such as the presence of a
lengthy physical or mental illness or divorce? Do mothers tend to have higher self-efcacy
beliefs than fathers, because they tend to have more direct care experiences and their roles are
currently perhaps more well-dened than those of fathers? Do mothers and fathers self-
efcacy beliefs tend to differ depending on the referent tasks? Are there life-enhancing benets
that extend beyond parenting for those fortunate enough to experience high parental self-
efcacy beliefs? What are the mental health implications for parents who raise their children,
without ever having achieved a sense of personal competence in the role?
Highly self-efcacious parents apparently experience personal empowerment associated
with parenting that relates to successful management and performance of the numerous req-
uisite tasks associated with the role. Moreover, high self-efcacy is related to the potential for
strong intrinsic interest in parenting activities and deep personal involvement in the parenting
process (Bandura, 1995). Alternatively, parents with low self-efcacy tend to be overwhelmed
by their parental responsibilities, often feeling excessively burdened by the daily and long-term
demands of parenthood. Although these general trends have been well-established in the avail-
able literature on the topic, the direct and indirect relations between parents perceptions of
competence and child social, affective, and cognitive development remain fertile ground for
further investigative efforts with more socio-demographically diverse parents.



Emotional Availability
1. Even when I have had an unusually distressing day, I think my child knows I am
available to meet his or her emotional needs.
2. I believe that I adequately meet my childs needs to feel secure and accepted. short
3. When my child needs me, I am able to easily put aside whatever else I may be doing. standard

top of rh
144 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
4. I nd it difcult to always be available to provide my child with the comfort he/she base of text
needs in dealing with the many frustrations and fears that toddlers face each day.*
5. Providing physical comfort for my child is easy for me.
6. I am usually willing to stop what Im doing and cuddle my child when he/she seems
to need affection.
7. I am often too preoccupied with my own problems to keep up with my childs chang-
ing emotions.*

Nurturance/Valuing/ Empathetic Responsiveness

8. I am able to sense when my child is starting to become distressed.
9. My toddler knows that I understand when his/her feelings are hurt.
10. I think my child knows by my behavior how much I really adore him/her.
11. My child feels very loved by me.
12. I think I am tolerant and understanding when my child displays negative emotions.
13. I nd it very distressing when my child isnt in a good mood.*
14. I denitely fulll my parental duties when it comes to providing emotional support
for my child.
15. When my child has a problem, he/she knows I will want to help.

16. Providing a safe, hazard-free environment for my child is very difcult for me.*
17. I feel comfortable with my ability to react well should an emergency arise in which
my childs physical well-being is in danger.
18. When I leave my child in someone elses care, I make sure that the substitute care
provider will be capable of protecting my child from harm.
19. I have my home arranged to prevent as many accidents as possible with my toddler.
20. I am very good about never leaving my child unattended.
21. I always make sure I can see my child in order to make sure he/she does not get hurt.
22. I have difculty determining what is and is not safe for my child to do.*

Discipline/Limit Setting
23. Disciplining my child does not seem to be coming as naturally to me as other parts
of parenting.*
24. I have trouble getting my child to listen to me.*
25. Other parents seem to have more success with setting limits for their children than I
do with my child.*
26. Setting limits for my toddler is relatively easy for me.
27. When my toddler tests the limits that I have set up, I nd myself becoming extremely
discouraged.* short

top of rh
Maternal Self-Efcacy 145 base of rh

cap height
28. Telling my child no when safety isnt the issue is hard for me.* base of text
29. I allow my child enough freedom to actively explore the environment.

30. I can always think of something to play with my child.
31. I am a fun playmate for my toddler.
32. I nd it hard to loosen up and just play with my child.*
33. I am able to get actively involved in playing with my child.
34. Playing is a part of my relationship with my child that I have very little difculty with.
35. I really need to learn how to just have fun with my child.*
36. I think I spend an appropriate amount of time just playing with my child.

37. I believe my toddler learns a great deal from my efforts to show him/her things.
38. Assisting my child with learning to talk and understand words is a part of parenting
that I leave to others.*
39. Sitting down regularly with my child to read or do some other one-on-one activity is
not difcult for me.
40. I am probably not that great at teaching my child about the world.*
41. I have some difculty guring out the appropriate level of instruction when Im trying
to explain something to my child.*
42. Helping my child learn colors, names of objects, etc. is not one of my strongest points.*
43. My child learns more from me than anyone else in his/her life.
44. I easily nd opportunities to point out things about the world during my daily inter-
actions with my child.
45. Although I would like to help my child learn more about his/her surroundings, this is
an area of parenting that I do not feel well-equipped for.*

Instrumental Care/ Structure/ Routines

46. I have been able to establish a daily routine with my toddler that feels comfortable to
both of us.
47. I am able to provide my child with a comfortable amount of daily structure.
48. I have been successful in getting my child to eat on a fairly regular schedule.
49. I feel like I have no control over my childs sleep habits.*
50. I am not very good at getting my child to stick to a regular daily schedule.*
51. Although I have tried to train my child to eat well, my efforts have been met with
very little success.*
52. I dont seem to be able to establish a regular bed time routine with my child.*
53. I have worked out a fairly regular morning routine with my toddler.
Note: *Indicates item is reverse scored. short

top of rh
146 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
REFERENCES base of text

Abidin, R.R. (1990). Parenting Stress Index manual (4th ed.). Charlottesville, VA: Pediatric Psychology
Ballenski, C.B., & Cook, A.S. (1982). Mothers perceptions of their competence in managing selected
parenting tasks. Family Relations, 31, 489494.
Bandura, A. (1982). Self-efcacy in human agency. American Psychologist, 37, 122147.
Bandura, A. (1989). Regulation of cognitive processes through perceived self-efcacy. Developmental
Psychology, 25, 729 735.
Bandura, A. (1995). Exercise of personal and collective efcacy in changing societies. In A. Bandura
(Ed.), Self-efcacy in changing societies (pp. 145). New York: Cambridge University Press.
Bandura, A. (1996). Multidimensional scales of perceived self-efcacy. Unpublished.
Bandura, A. (1997). Self-efcacy: The exercise of control. New York: W. H. Freeman.
Bandura, A., Babaranelli, C., Caprara, G.V., & Pastorelli, C. (1996). Multifaceted impact of self-efcacy
on academic functioning. Child Development, 67, 12061222.
Bandura, A., Babaranelli, C., Caprara, G.V., & Pastorelli, C. (2001). Self-efcacy beliefs as shapers of
childrens aspirations and career trajectories. Child Development, 72, 187206.
Bates, J.E., Freeland, C.A., & Lounsbury, M.L. (1979). Measurement of infant difcultness. Child De-
velopment, 50, 784 803.
Bayley, N. (1993). Manual for the Bayley Scales of Infant Development. New York: The Psychological
Beck, K.H., & Lund, A.K. (1981). The effects of health threat seriousness and personal efcacy upon
intentions and behavior. Journal of Applied Social Psychology, 11, 401415.
Benedek, T. (1970). Parenthood during the life cycle. In E.J. Anthony and T. Benedek (Eds.), Parenthood:
Its psychology and psychopathology (pp. 185206). Boston: Little, Brown.
Bohlin, G., & Hagekull, B. (1987). Good mothering: Maternal attitudes and motherinfant interaction.
Infant Mental Health Journal, 8, 352 363.
Bornstein, M.H., & Tamis-LeMonda, C.S. (1989). Maternal responsiveness and cognitive development
in children. In M.H. Bornstein (Ed.), Maternal responsiveness: Characteristics and consequences
(pp. 49 61). San Francisco: Josey-Bass.
Bugental, D.B., Blue, J., & Cruzcosa, M. (1989). Perceived control over caregiving outcomes: Implica-
tions for child abuse. Developmental Psychology, 25, 532539.
Coleman, P.K., & Karraker, K.H. (1998). Self-efcacy and parenting quality: Findings and future appli-
cations. Developmental Review, 18, 47 85.
Crowell, J.A., & Feldman, S.S. (1988). Mothers internal models of relationships and childrens behav-
ioral and developmental status. Child Development, 59, 12731285.
Cutrona, C., & Troutman, B. (1986). Social support, infant temperament, and parenting self-efcacy: A
mediational model of postpartum depression. Child Development, 57, 15071518.
Deutsch, F.M. , Ruble, D.N., Fleming, A., Brooks-Gunn, J., & Stangor, C. (1986, April). Becoming a
mother: Information seeking and self-denitional processes. Paper presented at the meeting of the
Eastern Psychological Association. New York, NY.
Donovan, W.L., & Leavitt, L.A. (1985). Simulating conditions of learned helplessness: The effects of
interventions and attributions. Child Development, 56, 594603.
Donovan, W.L., & Leavitt, L.A. (1989). Maternal self-efcacy and infant attachment: Integrating phys-
iology, perceptions, and behavior. Child Development, 60, 460472.
Dumka, L.E., Stoerzinger, H.D., Jackson, K.M., & Roosa, M.W. (1996). Examination of the cross-cultural
and cross-language equivalence of the Parenting Self-Agency Measure. Family Relations, 45, 216 short
222. standard

top of rh
Maternal Self-Efcacy 147 base of rh

cap height
Earley, P.C., & Lituchy, T.R. (1991). Delineating goal and efcacy effects: A test of three models. Journal base of text
of Applied Psychology, 76, 81 98.
Elder, G.H., Eccles, J.S., Ardelt, M., & Lord, S. (1995). Inner city parents under economic pressure:
Perspectives on the strategies of parenting. Journal of Marriage and the Family, 57, 771784.
Emde, R.N. (1989). The infants relationship experience: Developmental and clinical aspects. In A.J.
Sameroff & R.N. Emde (Eds.), Relationship disturbances in early childhood (pp. 3351). New York:
Basic Books.
Goodnow, J.J., & Collins, W.A. (1990). Development according to parents: The nature, sources, and
consequences of parents ideas. Hillsdale, NJ: Lawrence Erlbaum.
Gross, D., Conrad, B., Fogg, L., & Wothke, W. (1994). A Longitudinal model of maternal self-efcacy,
depression, and difcult temperament during toddlerhood. Research in Nursing and Health, 17, 207
Gross, D., Fogg, L., & Tucker, S. (1995). The efcacy of parent training for promoting positive parent
toddler relationships. Research in Nursing and Health, 18, 489499.
Grusec, J.E., Hastings, P., & Mammone, N. (1994). Parenting cognitions and relationship schemas. In
J.G. Smetana (Ed.), Beliefs about parenting: Origins and developmental implications (pp. 519).
San Francisco: Jossey-Bass.
Heinicke, C.M. (1984). Impact of prebirth parent personality and marital functioning on family devel-
opment; A framework and suggestions for further study. Developmental Psychology, 26, 1044
Johnston, C., & Mash, E.J. (1989). A measure of parenting satisfaction and efcacy. Journal of Clinical
Child Psychology, 18, 167 175.
Kochanska, G. (1990). Maternal beliefs as long-term predictors of motherchild interaction and report.
Child Development, 61, 1934 1943.
Kwok, S., & Wong, D. (2000). Mental health of parents with young children in Hong Kong: The roles
of parenting stress and parenting self-efcacy. Child and Family Social Work, 5, 5765.
Lachman. M.E., & Leff, R. (1989). Perceived control and intellectual functioning in the elderly: A 5-
year longitudinal study. Developmental Psychology, 25, 722728.
MacPhee, D., Ramey, C.T., & Yeates, K.O. (1984). Home environment and early cognitive development:
Implications for intervention. In A.W. Gottfried (Ed.), Home environment and early cognitive de-
velopment (pp. 343 369).Orlando, FL: Academic Press Inc.
Matas, L., Arend, R.A., & Sroufe, L.A. (1978). Continuity of adaptation in the second year: The rela-
tionship between quality of attachment and later competence. Child Development, 49, 547556.
Multon, K.D., Brown, S.D., & Lent, R.W. (1991). Relation of self-efcacy beliefs to academic outcomes:
A meta-analytic investigation. Journal of Counseling Psychology, 38, 3038.
Oakley, A. (1980). Women conned: Toward a sociology of childbirth. New York: Stocken Books.
Pajares, F., & Miller, M.D. (1995). Mathematics self-efcacy and mathematics performances: The need
for specicity of assessment. Journal of Counseling Psychology, 42, 190198.
Peterson, L., Farmer, J., & Kashani, J. (1990). Parental injury prevention endeavors: A function of health
beliefs? Health Psychology, 9, 177 191.
Pridham, K.F., & Chang, A.S. (1992). Transition to being the mother of a new infant in the rst 3 months:
Maternal problem solving and self-appraisals. Journal of Advanced Nursing, 17, 204216.
Raver, C.C., & Leadbeater, B.J. (1999). Mothering under pressure: Environment, child, and dyadic cor-
relates of maternal self-efcacy among low-income women. Journal of Family Psychology, 13, 523
Schneewind, K. A. (1995). Impact of family processes on control beliefs. In A. Bandura (Ed.), Self- short
efcacy in changing societies (pp. 1 45). New York: Cambridge University Press. standard

top of rh
148 P.K. Coleman and K. Hildebrandt Karraker base of rh

cap height
Sherer, M., & Adams, C. (1983). Construct validity of the Self-Efcacy Scale. Psychological Reports, base of text
53, 899 902.
Shereshefsky, P., & Yarrow, L. (1973). Psychological aspects of a rst pregnancy and early postnatal
adaptation. New York: Raven Press.
Spoth, R., & Conroy, S. (1993). Survey of prevention-relevant beliefs and efforts to enhance parenting
skills among rural parents. The Journal of Rural Health, 9, 227239.
Strand, P.S., & Wahler, R.G. (1996). Predicting maladaptive parenting: Role of maternal object relations.
Journal of Clinical Child Psychology, 25, 4351.
Swick, K.J., & Hassell, T. (1990). Parental efcacy and the development of social competence in young
children. Journal of Instructional Psychology, 17, 2432.
Teti, D.M., & Gelfand, D.M. (1991). Behavioral competence among mothers of infants in the rst year:
The mediational role of maternal self-efcacy. Child Development, 62, 918929.
Tucker, S., Gross, D., Fogg, L., Delaney, K., & Lapporte, R. (1998). The long-term efcacy of a behav-
ioral parent training intervention for families with 2-year-olds. Research in Nursing & Health, 21,
Unger, D.G., & Wandersman, L.P. (1985). Social support and adolescent mothers: Action research con-
tributions to theory and application. Journal of Social Issues, 41, 2945.
Williams, T.M., Joy, L.A., Travis, L., Gotowiec, A., Blum-Steele, M., Aiken, L.S., Painter, S.L., &
Davidson, S.M. (1987). Transition to motherhood: A longitudinal study. Infant Mental Health Jour-
nal, 8, 251 265.
Wang, A.Y., & Richarde, R.S. (1988). Global versus task-specic measures of self-efcacy. The Psy-
chological Record, 38, 533 541.
Watt, S.E., & Martin, P.R. (1994). Effect of general self-efcacy expectancies on performance attribu-
tions. Psychological Reports, 75, 951 961.
Wells-Parker, E.., Miller, D.I., & Topping, S. (1990). Development of control of outcome scales and self-
efcacy scales for women in four life roles. Journal of Personality Assessment, 54, 564575.
Woodruff, S., & Cashman, J.F. (1993). Task, domain, and general self-efcacy: A reexamination of the
Self-Efcacy Scale. Psychological Reports, 72, 423432.
Zeanah, C.H., Boris, N.W., Heller, S.S., Hinshaw-Fuselier, S., Larrieu, J.A., Lewis, M., Palomino, R.,
Rovaris, M., & Valliere, J. (1997). Relationship assessment in infant mental health. Infant Mental
Health Journal, 18, 182 197.

View publication stats