International Review of Research in Mental Retardation
Visit the link below to download the full version of this book:
https://medipdf.com/product/international-review-of-research-in-mental-retardati
on/
Click Download Now
PREFACE Xlll
The breadth of this volume required a broad array of reviewers who were both
willing and able to lend their expertise to the field. In this volume, in addition to
the Board of Associate Editors, especially Mike Guralnick and Ted Nettelbeck, I
was particularly fortunate in having the consultation of the following individuals,
listed alphabetically: Grace Baranek, Ivan Brown, Donna Bryant, Paula Caplan,
Rebecca Fewell, Bob Fox, Stephen Greenspan, Ken Keith, Gerald Mahoney, Kofi
Marfo, Karl Newell, Andrew Nocon, Jeannette Rogers-Dulan, Roger Stancliffe,
Wendy Stone, and Keith Widaman. I am grateful to them for the insightful and
incisive reviews that they provided in a timely manner. They are deserving of our
applause, silent though it may be.
LARAINE MASTERS GLIDDEN
Characterizations of the Competence of
Parents of Young Children with Disabilities
CARL J. DUNST
ORELENA HAWKS PUCKETT INSTITUTE
ASHEVILLE, NORTH CAROLINA
AND FAMILY, INFANT, AND PRESCHOOL PROGRAM
WESTERN CAROLINA CENTER
MORGANTON, NORTH CAROLINA
fRACY HUMPHRIES
ORELENA HAWKS PUCKETT INSTITUTE
ASHEV1LLE, NORTH CAROLINA
CAROL M. TRIVETTE
ORELENA HAWKS PUCKETT INSTITUTE
MORGANTON, NORTH CAROLINA
I. INTRODUCTION
The birth of a child with a disability or early diagnosis of a disabling condition
can and often does have a negative impact on family functioning (Crnic, Friedrich,
& Greenberg, 1983). According to Patterson and McCubbin (1983), the family is
fundamentally transformed by these kinds of nonnormative life events, having to
make a host of adaptations and modifications resulting in different emotions and
feelings (tension, conflict, stress, feelings of inadequacy, etc.).
Notwithstanding individual differences in parent and family reactions to the
birth and rearing of a child with a disability (Gabor & Farnham, 1996; Gallimore,
Weisner, Bernheimer, Guthrie, & Nihira, 1993; Hanson & Hanline, 1990; Summers,
Behr, & Turnbull, 1989; TurnbuU, Summers, & Brotherson, 1986), this particular
INTERNATIONALREVIEW OF RESEARCHIN Copyright© 2002 by AcademicPress.
MENTAL RETARDATION,Vol. 25 All rights of reproductionin any form reserved.
0074-7750/02$35.00
2 Carl J. Dunst, Tracy Humphries, and Carol M. Trivette
life event and situation, as well as its accompanying consequences, bring a child
and his/her parents into frequent and often lifelong contact with helping profes-
sionals (Berry & Hardman, 1998; Krauss, 1986; Marshak, Seligman, & Prezant,
1999; Slater, Martinez, & Habersang, 1989). The consequences for parents (as well
as for children) are not always positive (Dunst & Trivette, 1988). For example,
because the "professionals consulted when the client needs help may never [have]
seen the client in a state of general well-being, [they] can therefore have only an
indirect sense of the client's capabilities and strengths. This limited perspective
reinforces the already ingrained tendency for professionals to exercise paternalistic
authority" and judgments (Merton, Merton, & Barber, 1983, p. 21). One domain
of paternalistic judgments is parenting abilities and capabilities (Caplan & Hall-
McCorquodale, 1985a; Hewlett & West, 1998). Professionals often see parents in
negative ways and either implicitly or explicitly blame parents for any number of
their children's problems (Caplan, 2000; Caplan & Hall-McCorquodale, 1985b).
The roots of contemporary parent-blaming, or more specifically mother-blaming,
can be traced to the late 1880s (Weiss & Halpern, 1991) and to lay and profes-
sional helpers' attempts to provide guidance and advice to poor families about child
rearing, moral values, and socially acceptable behavior, among other things (e.g.,
Cabot, 1919; Paine, 1880; Richmond, 1899). The consequence of these efforts
often was close scrutiny of parenting behavior, which escalated when profession-
als rather than lay helpers became the major judges of desired parent and family
functioning (see Grubb & Lazerson, 1980). Once this occurred, it took very little
time for parents to be seen by professionals as lacking competence and as a primary
source of their children's problems (e.g., Taft, 1926). "By insisting that families
were weak and could not provide for their own needs without expert interven-
tion, [professionals] ironically undermined whatever confidence and abilities the
families had in their child rearing" (Halpern, 1993, p. 162).
The same kinds of negative attributions and usurpation of parental responsi-
bility have been characteristic of professionals working with families of young
children with disabilities. For example, Kanner (1943), in the now famous paper
in which he coined the term infantile autism, speculated that mothers' poor af-
fective behavior was the root of their children's problems. Parents' psychosocial
functioning continues to be viewed by professionals as a primary source of infant
and early childhood disorders (e.g., Konstantareas & Homatidis, 1983 [cited in
Caplan, 2000]; Trad, 1987; Zelman, 1996), and in instances where parents are
seen as inadequate, professionals often "take over" child-rearing responsibilities
(Zelman, 1996).
Negative characterizations about parents' behavior are not limited to child-
rearing capabilities. Minke and Scott (1995), in a study of professional judgments
about parents' capacities to form parent-professional partnerships, found that pro-
fessionals indicated that parents tended to make poor choices and "could not
be relied upon to act in the best interests of their children" (p. 345), and that
PARENTING COMPETENCE 3
professionals more often than not saw parents as incapable of becoming equal
partners. Of more than 100 competency characterizations made by professionals,
less than 20% were positive in tone.
The magnitude and scope of contemporaneous mother-blaming were made clear
in a review of 125 studies conducted by Caplan and Hall-McCorquodale (1985a).
Their review found that mothers were blamed for 72 different kinds of child psy-
chopathology! The child-related problems mothers supposedly caused included
aggressiveness, minimal brain damage, chronic vomiting, tantrums, bad dreams,
enuresis, hyperactivity, and something called "self-induced television epilepsy."
As pointed out by Caplan and Hall-McCorquodale (1985a), "in no article was the
mother's relationship with her child described as simply healthy, nor was she ever
described only in positive terms" (p. 348). Hewlett and West (1998) went so far
as to say that parents are often the targets of what they call parent-bashing, where
mothers and fathers are ridiculed by professionals rather than provided with the
supports needed to strengthen their competence and carry out parenting responsi-
bilities.
The extent to which parents and families are characterized in negative or positive
ways has recently been the focus of research analyses in both the developmental
disabilities (Helff & Glidden, 1998) and family support (Dunst, in press) fields.
Helff and Glidden (1998) reviewed 60 family adjustment research articles pub-
lished between 1971 and 1993 with a focus on the positive and negative charac-
terizations of families and family functioning. Findings indicated that there were
significant changes in how researchers described and characterized families. Fam-
ilies were described in less negative ways in the 1980s and early 1990s compared
to the 1970s. There was, however, no concomitant increase in positive character-
izations of families. The investigators noted that the social and political climate
of the late 1980s and early 1990s made it less acceptable to describe families in
negative ways, but that research as a whole was implicitly underscored by negative
attributions regarding a family's adjustment to its member with a disability.
Similar characterization patterns were found in a research review by Dunst
(in press) of the extent to which early intervention, preschool, elementary, and
secondary school practices were family-centered (Dunst, 1995, 1997). Family-
centeredness is characterized by beliefs and practices that treat families with dig-
nity and respect, recognize and build on family capabilities, promote informed
family choice and decision making, and support family participation in achieving
desired goals and outcomes. Findings as a whole showed that early intervention
and school practices become progressively less family-centered from birth through
high school, indicating that positive family characterizations decreased as children
became older. Patterns of family-centered practices varied as a function of type
of practice as well. Despite generally positive family-professional interactions at
the preschool and early elementary grade levels, mothers (and to a lesser degree
fathers) were viewed as less competent and capable of effective child rearing,
Carl J. Dunst, Tracy Humphries, and Carol M. Trivette
at least without the guidance of professionals. For example, whereas McBride,
Brotherson, Joanning, Whiddon, and Demmitt (1993) found that early interven-
tion practitioners demonstrated positive interpersonal help-giving behavior with
parents, "professionals in this study were not strongly committed to or knowledge-
able about help giving that placed emphasis on building the capacity of families"
(p. 428). This finding implies negative beliefs about existing parenting capabilities
as well as the capacity to become more competent.
Various sources of information suggest that many professionals either implic-
itly or explicitly view parents of young children with delays or disabilities as
lacking in parenting abilities and in need of professional guidance in order to par-
ent effectively (Hankerson, 1984; Minke & Scott, 1995). In some cases, parents
are seen as incapable of becoming competent child rearers (see especially Doll
& Bolger, 2000, pp. 241-242; Minke & Scott, 1995, pp. 344-346). The reasons
for these views are multiple and complex. Regardless of the reasons, the conse-
quences are often negative professional attributions about existing and emerging
parenting competence. Moreover, the logic for these kinds of negative attributions
is often faulty and potentially harmful. For example, Foster, Berger, and McLean
(1981) noted the following in terms of professionals' adoption of parent interven-
tion models based on cultural deficit views (see Zigler & Berman, 1983), which
attempt to explain why children from low socioeconomic backgrounds fare poorly
on any number of child and parent measures in comparison to their middle-income
counterparts. These writers observed:
Many of the children [with disabilities] served by early interventionprograms come from
middle- or upper-incomefamilies.... Models of parent involvementpredicated on a cultural
deficithypothesisare not easily appliedto these families,whose interactionpatterns were the
very standard against which lower-incomefamily styles were compared and found wanting.
(p. 57, emphasisadded)
As Foster et al.'s (1981) observations make clear, professionals sometimes make
leaps of logic, act on that logic, and wind up with practices that are illogical.
II. PURPOSE AND APPROACH
The purpose of this chapter is to review and synthesize research evidence per-
taining to positive and negative characterizations of several different aspects of
parenting capabilities. More specifically, we examine research studies with an
emphasis on professionals' attributions pertaining to the competence of parents
of preschoolers with disabilities or delays. We focus on the preschool period for
several reasons. First, it is a time during which parents spend considerable time
with their children, learning and perfecting parenting roles (Sroufe, 2000). Sec-
ond, the early preschool years are especially important for parent-child bonding,
PARENTING COMPETENCE 5
attachment, and other relationship-building consequences that are influenced by
parenting styles (Davies & Cummings, 1998; Easterbrooks & Lamb, 1979). Third,
considerable effort has been expended in studying interactions between parents and
their preschool-aged children with disabilities, with a focus on the similarities and
differences in parenting styles compared to those used by parents of children with-
out disabilities (Barnard & Kelly, 1990; Doctoroff, 1996; Hodapp, 1995; Kelly &
Barnard, 2000). Fourth, the preschool years are a time when parents are expected
to be competent and where deviations from the "norm" are more likely to be used
to infer incompetence and the need for professional guidance (see Booth, Mitchell,
Barnard, & Spieker, 1989).
Our interest in parenting competence is both empirical and practical. Empiri-
cally, we have conducted research investigating the sources of variations in parent-
ing styles (Dunst & Trivette, 1988), the impact different parenting styles have on
child behavior and development (Dunst, Trivette, & Jodry, 1996), and how differ-
ent intervention practices strengthen or attenuate a sense of parenting competence
(Dunst, 1999). Practically, we are concerned with a trend that is inconsistent with
research evidence indicating that parenting behavior is important for producing op-
timal child development outcomes (Mahoney, Boyce, Fewell, Spiker, & Wheeden,
1998; NICHD Early Child Care Research Network, 1999). This trend was articu-
lated in the following way by Hewlett and West (1998):
[A]n unintended consequence of [government's] well-meaning [attempts] to save our children
was the emergence of a new class of professionals--social workers, therapists, foster care
providers, family court lawyers--who have a vested interest in taking over parental function.
Bureaucracies everywhere have a remorseless drive to expand--to widen their client base. If chil-
dren are the clients, parents can quite easily become the adversaries--the people who threaten to
take business a w a y . . . [T]he "so-called helping professions" have functioned to undermine the
confidence of families, unwittingly creating a vicious cycle of dependency on external sources
of expertise. (pp. 109-110)
To the extent that usurpation of parenting roles in fact occurs as a function of
professionals blaming parents and assuming parenting responsibilities, parents'
sense of confidence and competence would more likely be attenuated.
The approach and method used to ascertain the extent to which parents of young
children with disabilities are seen as competent or incompetent, as well as capable
or incapable of becoming more competent, were similar to those used by Helff and
Glidden (1998) to determine whether research on adjustment of families rearing
children with disabilities had become less negative and more positive. We reviewed
58 articles published between 1970 and 2000, organized by 5-year blocks. Selection
criteria and study characteristics are described in detail later in the chapter. Suffice
it to say at this point that we focused on articles that reported the findings of
researchers who examined the parenting styles and interactional behavior used
with children from birth to 5 years of age with disabilities or developmental delays
6 Carl J. Dunst, Tracy Humphries, and Carol M. Trivette
because these kinds of studies were more likely to be ones for which we could
make judgments about different aspects of parenting competence.
III. CONTRASTING MODELS OF
PARENTING COMPETENCE
Our review and analysis were guided by a framework for distinguishing between
the characteristics of deficit and competence models of performance (Cicchetti,
Rappaport, Sandier, & Weissberg, 2000; Cowen, 1985; Dunst, Trivette, & Thomp-
son, 1994; Rappaport, 1981; Trivette, Dunst, Hamby, & LaPointe, 1996; Zautra
& Sandier, 1983). Those who operate from a deficit-based model view people as
lacking abilities, deficient in some aspect of functioning, or having identifiable dis-
orders or problems. Interventions devised from this perspective focus on correcting
people's weaknesses or problems (Howard, 1982; Pransky, 1991). Those who hold
deficit-based perspectives of parents view them as lacking in parenting abilities
and as sources or causes of children's developmental problems or behavior aberra-
tions (see Berlin, 1974; Booth et al., 1989). Parents are also seen by professionals
as lacking desired parenting behavior or using inappropriate parenting styles and
as being in need of professional interventions in order to become capable parents
(Hornby, 1995; Howard, 1978, 1982; Kirkman, 1993; Webster-Stratton, 1990).
In extreme cases, parents are viewed as incapable of becoming even minimally
competent, and parenting roles and functions are assumed by professionals (see
Lee & Guck, 2000, for a description of this set of circumstances).
In contrast, those who use competence models view people as having existing
capabilities as well as the capacity to become more competent when provided
opportunities to display abilities. Interventions based on this model afford people
experiences that strengthen and build on their existing competence and that use their
strengths to learn new skills (Bond, 1982; Danish & D'Augelli, 1980; Drake, 1995;
Judge, 1998; Rappaport, 1981, 1987; Seeman, 1989; Zautra & Sandler, 1983).
Professionals who have competence-based perspectives of parents view all parents
as having existing capabilities as well as the capacity to parent more effectively
(Judge, 1998; Powell, Batsche, Ferro, Fox, & Dunlap, 1997; Stoneman, 1985).
Lack of expression of parenting competence is not viewed as a deficit or weakness,
but rather as the lack of opportunity to express competence (Dunst, Trivette, &
LaPointe, 1992). In such cases, professionals provide opportunities or facilitate
parent participation in settings that encourage competence expression and adoption
of a wider range of parenting behavior (Bond, 1982; Bradley, Rock, Whiteside,
Caldwell, & Brisby, 1991; Lojkasek, Goldberg, Marcovitch, & MacGregor, 1990;
McCollum & Yates, 1994; Stoneman, 1985).
The contrasting features of deficit and competence models of parenting abilities
were used in this chapter to (a) conduct a systematic analysis of parent-child
interaction studies to ascertain if there have been changes in characterizations of
PARENTING COMPETENCE 7
parenting competencies during the past 30 years and (b) compare and contrast
selective parenting studies that illustrate deficit and competence perspectives of
child rearing.
IV. LITERATURE ANALYSIS
The last three decades of the 20th century were busy ones in terms of research
studies and literature reviews on parenting styles and behavior used with children
with (e.g., Hadadian, 1996; Hanzlik, 1989; Kelly & Barnard, 2000; Marfo, 1988,
1991) and without (e.g., Bell, 1974; Bornstein, 1995; DeChateau, 1987; Hess,
1981; Hoffman, Gandelman, & Schiffman, 1982; Maccoby & Martin, 1983) dis-
abilities and developmental delays. The research literature on interactional styles
of parents of children with disabilities or delays constituted the focus of our anal-
ysis and categorization. We reviewed parenting studies over a 30-year period to
determine whether researchers have characterized different aspects of parenting
competence as deficit-based or competence-based and whether there have been
changes in patterns of characterizations over time. We conducted both quantitative
and qualitative analyses of the research literature. First, we rated selected features
of 58 articles to ascertain researchers' perspectives of parenting competence, and
then we conducted several different kinds of statistical analyses to determine pat-
terns of characterizations of parenting abilities. Second, we abstracted researcher
attributions of selected studies to highlight deficit-based and competence-based
characterizations of parenting abilities and competence.
A. Approach
1. METHOD
To ensure that a representative sample of studies was included in our review,
articles were gathered from several relevant fields, including early intervention,
early childhood special education, child development, and developmental psychol-
ogy. For the purposes of our analysis, the three-decade review and analysis was
divided into six 5-year blocks. Articles were located through a PsycINFO database
search, literature reviews, and reference lists of several pertinent publications. The
PsycINFO database was searched for the targeted years using the search terms
parent competence, parent child interaction, parent behavior, parent interactions,
and parenting styles to locate studies that examined parents' behavior with their
children. Delimiting terms included mother or father and sample characteristics
included mental retardation, disabilities, cerebral palsy, autism, handicapped,
developmentally delayed, Down syndrome, prematurity, visually impaired, blind,
deaf hearing impaired, mentally retarded, developmental risk, etc.
For all six time periods, we first examined the articles to determine their suitabil-
ity for our review. In addition to the content criteria regarding parenting interactions
8 Carl J. Dunst, Tracy Humphries, and Carol M. Trivette
and styles, the articles had to contain research findings and had to be either em-
pirical investigations or case studies. Studies also had to have focused on children
with, or at risk for, developmental or physical disabilities and whose ages were
between birth and 5 years.
Because it was possible to locate more than 10 articles that met all of our criteria
for all but one of the 5-year time periods, sampling was concluded after 10 articles
were identified, even for those time periods in which more articles were available.
There was a shortage of articles meeting the criteria for the 1970 to 1974 time
period, so the search was expanded to include the year 1969 and was concluded
after 8 articles had been identified for that time period.
The final sample of studies included in our analysis was considered representa-
tive of most, if not all, of the articles that would meet our criteria because of the
comprehensive nature of the identification process. The sample of articles included
in our review came from 21 different journals and 8 edited books. The Appendix
includes the complete citations for the articles included in our analysis.
Articles were prepared for evaluation by removing all identifying information,
which included the article source, title, authors' names, all citations, and any
reference to specific years. The articles were randomly ordered and assigned an
identification number for the evaluation. Consistent with the review and analysis
conducted by Helff and Glidden (1998), we were primarily interested in the char-
acterizations of parenting competence rather than the methodological soundness
of the articles, so only the abstract, introduction, and discussion sections were used
in the analysis.
Prior to the conduct of the evaluation of the selected articles, two raters were
extensively trained by one of the coauthors using similarly prepared articles that
were not included in the review. During training, the raters were familiarized with
the coding categories and their corresponding definitions and were provided with
numerous examples and opportunities to practice coding articles. After the raters
had established a satisfactory understanding of the coding guidelines and had
reached an acceptable level of interrater reliability (Cohen's kappa coefficients
greater than .80), they began coding the articles analyzed for this chapter. Raters
were not informed of the purpose of the review and the fact that the data were
being collected for the preparation of this chapter.
2. CATEGORIES OF PARENTING COMPETENCE
Three dimensions of researcher attributions regarding parenting capabilities
were assessed, with each dimension having a positive-oriented and a negative-
oriented indicator. The three dimensions were parenting competence (competent
vs incompetent), parenting enjoyment (positivity vs negativity), and paradigmatic
orientation (deficit-based vs strengths-based). Differences between the respective
pairs of indicators were used to establish the relative positive or negative tone of
researcher attributions.
PARENTING COMPETENCE 9
In addition to the three major dimensions of researcher attributions, we as-
sessed the extent to which researchers deemed parents as having the capacity to
become competent and the extent to which researchers perceived parents as need-
ing professional interventions in order to become more competent. A difference
between the need for help and the capacity to become competent was used as a
measure of positive or negative attributions regarding parenting abilities; higher
difference scores for the need for help were considered a negative researcher attri-
bution, whereas higher difference scores for parenting capacity were considered a
positive attribution.
a. Parenting Competence. The extent to which researchers characterized parent-
ing competence as positive or negative was ascertained by examining researcher
descriptions of parenting capabilities, skills, or performance. Positive attributions
of parenting competence included descriptions of positive parenting performance,
appropriate parent-child interactions, and effective parenting skills. Negative at-
tributions of parenting competence focused on descriptions regarding parents' lack
of capabilities, skills, or performance and included descriptions of poor parenting
abilities, problematic parent-child relationships, inadequate parenting skills, and
similar characterizations.
b. Parenting Enjoyment. The extent to which researchers characterized parents
as liking or disliking interactions with their children was analyzed in terms of
judgments and attributions about the positivity or negativity of parents' affect.
Positive enjoyment attributions included descriptions of parents' positive feelings,
enjoyment of parenting, feelings of satisfaction, and fulfillment. Statements re-
garding parents' feelings of worthiness or pride were also included in the category
of positive affect. Negative attributions of affect included descriptions of parents'
negative feelings toward themselves or their children, dissatisfaction with par-
enting, feelings of stress, frustration, or guilt, or a sense of being overwhelmed
with parenting roles. Statements regarding parents' feelings of ineffectiveness or
incompetence were also coded as negative affect attributions.
c. Paradigmatic Orientation. The extent to which researchers adopted contrast-
ing paradigms for characterizing parenting behavior was assessed in terms of
whether descriptions were strengths-based or deficit-based. Descriptions were con-
sidered strengths-based if researchers discussed parenting from the perspective of
the skills that parents had and could use to increase or improve their interactions
or skills. Strengths-based categorizations also included descriptions of parents as
positive or empowered agents in their child's life, positive role models, advocates
for their children, or allies in intervention. Deficit-based descriptions included
those that characterized parenting in terms of deficits that parents had or that
were in need of amelioration. Descriptions were also judged deficit-based if they
characterized parents as having parenting difficulties or problems, as being sources
of their child's difficulties, as unable to effectively intervene with their children
without guidance, or as unable to effectively parent their children.
10 Carl J. Dunst, Tracy Humphries, and Carol M. Trivette
d. Parenting Capacity. The extent to which researchers characterized parents as
having the capacity to become competent was assessed in terms of descriptions that
reflected the belief that parents were capable of acquiring or learning new skills or
benefiting from parent-child interactions. The capacity to become competent also
included descriptions that suggested ways parents could improve their parenting
skills or interactions with their children. Parents were described as lacking the
capacity to become competent if they were characterized as unable to or incapable
of improving their parenting skills.
e. Professional Interventions. The extent to which researchers described par-
ents as needing professional guidance, training, and advice was assessed in terms
of descriptions indicating that professional interventions were needed to improve
parenting abilities. The need for professional assistance in order for parents to
become competent was assessed in terms of descriptions which indicated that
parenting competence or capability would be limited or impossible without pro-
fessional guidance. The ability to parent effectively without professional guidance
or training was assessed in terms of descriptions indicating that participatory op-
portunities (Dunst, 2000) other than professional interventions could strengthen
parenting competence and confidence.
The various competence indicators were judged individually on a 5-point scale
for determining the extent to which researcher descriptions provided the following:
1 = little or no evidence, 3 = implied evidence, or 5 -----explicitly stated evidence
for the presence of the attribution or characteristic. Abstracts or paragraphs that
clearly did not include any discussion or evidence whatsoever regarding researcher
attributions about the presence of the parenting characteristic were considered
noncodable and were not included in the analyses.
Each rater scored 40 articles. An overlap of 20 randomly selected articles was
used to calculate interrater reliability. Interrater agreement within 1 point on the
5-point scale averaged 95% across coding categories for the 20 articles.
Each article was coded paragraph by paragraph, including the abstract, which
was judged in the same manner as a paragraph. The abstracts for more than half
the articles were not codable and were excluded from further consideration. The
number of paragraphs per 5-year block varied from 75 to 140, with earlier stud-
ies having fewer paragraphs than more recently published articles and chapters.
Conclusion sections of articles included just over half (54%) of the total number
of paragraphs ( N = 615). A 6-Year Block x Type of Paragraph (Introduction vs
Conclusion) X2 found that the proportions of introduction and conclusion section
paragraphs included in the analyses were relatively similar across time blocks,
X 2 = 10.40, d f = 5, p > .05.
3. DATA SUMMARY AND ANALYSIS
We calculated several different measures of parenting competence to assess
researcher characterizations of parenting abilities and to produce converging
PARENTING COMPETENCE 11
evidence that supported or disconfirmed changes in patterns of characterizations.
First, where paragraphs were the unit of analysis, the ratings for individual para-
graphs were used as the dependent measures. Second, where individual stud-
ies were the unit of analysis, an average score for all codable paragraphs was
used as the dependent measure since different articles had different numbers of
paragraphs for both the introduction and conclusion sections. Third, the relative
positivity or negativity of each paragraph or article was established by calcu-
lating a balance score (positive rating minus negative rating for each parenting
capability dimension). Balance scores were calculated for the individual ratings
in the paragraph analyses and for the average paragraph ratings for the article
analyses.
A second set of dependent measures was calculated to provide another index
that judged characterizations of the presence or absence of research attributions
regarding the different dimensions of parenting abilities. A score of 3 or higher
on the 5-point rating scale was used as an index of the presence of each parenting
characteristic, and a proportion measure was calculated for the paragraph, arti-
cle, and balance score analyses. The proportion measures provide direct indices
of the percentage of paragraphs or articles that were characterized as deficit- or
competence-based. The difference (balance) between the proportion of positively
and negatively characterized paragraphs and articles provides a measure of the
overall tone of researcher attributions.
Multivariate repeated-measures ANOVAs were conducted with time block as
the between factor and both paragraph type (introductions and discussions) and
the parenting competence measures as the within factors to establish overall main
and interaction effects for the six sets of dependent measures. Separate sets of
analyses were conducted with the paragraph and article ratings and proportion
scores as the dependent measures. Since we were primarily interested in whether
there were changes in patterns of researcher attributions, we included multivariate
and univariate tests of linear trends for both positive and negative characterizations
in each analysis.
B. Results
1. PARAGRAPH SCORES
The multivariate repeated-measures ANOVA with paragraphs as the unit of anal-
ysis, paragraph type and year as research factors, and the three sets of parenting
competence ratings as the dependent measures showed that parents were charac-
terized more negatively overall, F(1,603) = 148.77, p < .0001, but that charac-
terization patterns changed across blocks of years, F(5,603) = 4.69, p < .001.
Characterization patterns did not differ as a function of paragraph type (introduc-
tions vs discussions) for either positive or negative attributions by researchers in
either the multivariate or univariate analyses.
12 Carl J. Dunst, Tracy Humphries, and Carol M. Trivette
Multivariate tests of linear trends showed an overall increase in positive char-
acterizations, F(1,603) = 14.35, p < .0001, and an overall decrease in negative
characterizations, F(1,603) = 3.90, p < .05, from 1970 to 2000. Further analy-
ses showed that positive characterizations of parenting competence, F(1,603) =
11.69, p < .001, parenting enjoyment, F(1,603) = 5.18, p < .05, and strengths-
based descriptions, F(1,603) = 10.19, p < .0001, increased linearly across the
blocks of years. Linear decreases in negative characterizations were limited to the
negative enjoyment measure, F(1,603) = 6.11, p < .01.
The same multivariate ANOVA with the presence or absence of each charac-
teristic as the dependent measure produced identical patterns of findings. Figure 1
shows these results in terms of the percentage of articles that were characterized
as having positive and negative attributions by researchers regarding the different
dimensions of parenting abilities. The multivariate test of linear trends showed
an overall increase in positive characterizations across time blocks, F(1,603) =
10.31, p < .001, and univariate linear increases in positive characterizations of
parenting competence, F ( 1 , 6 0 3 ) = 8.59, p < .001, parenting positive affect,
F(1,603) = 5.53, p < .05, and strengths-based descriptions, F(1,603) = 5.76,
p < .05. The multivariate test for changes in negative characterizations was non-
significant, F(1,603) = 1.39, p > .05, but there was a significant univariate de-
crease in negative characterizations of parenting affect across time blocks,
F(1,603) = 4.71, p < .05. Neither the multivariate nor the univariate analysis
findings were influenced by type of paragraph (introductions vs discussions).
Examination of Fig. 1 shows that despite linear increases in the percentage of
paragraphs in which parents were characterized in positive terms, overall character-
izations tended to be more negative for the three sets of measures for all but one
parenting measure (enjoyment) at all but one time block. Pairwise comparisons for
the parenting competence measures showed that parents were characterized more
negatively at all six time periods, Fs(1,603) = 5.15 to 28.34, ps < .05 to .001.
Parents were also described in more deficit-based terms at all six time periods,
Fs(1,603) = 7.68 to 48.87, ps < .05 to .001. Pairwise contrasts of the parent-
ing enjoyment measures showed that researchers made more negative attributions
about parents, Fs(1,603) = 8.26 to 24.48, ps < .01 to .001, at all but two time
periods (1985 to 1989 and 1995 to 2000). Exactly the same pairwise comparison
results were produced using the paragraph ratings as were produced using the
dependent measures.
The multivariate repeated-measures ANOVA with paragraphs as the unit of anal-
ysis and the ratings of parents' capacity to become competent and the need for pro-
fessional interventions as the dependent measures produced a significant within-
factor effect, F(1,603) = 5.47, p < .05. Ratings for the need for professional
interventions were significantly higher than ratings for parents' capacity to be-
come competent. A marginally significant linear trend x within factor interaction,
F(1,603) = 2.86, p < . 10, indicated that the within-factor differences varied ac-
cording to blocks of years. Figure 2 shows the data in terms of the percentage
