Clinical Child and Family Psychology Review, Vol. 9, No.

2, June 2006 (Ó 2006)

Parental Control in the Etiology of Anxiety

Natalie Ballash,1 Ovsanna Leyfer,1 Amy F. Buckley,1 and Janet Woodruff-Borden1,2

Anxiety disorders are among the most prevalent forms of adult and childhood psychiatric
disorders, and they are highly familial. However, the mechanisms of transmission remain
unclear. One familial characteristic that may promote the development of anxiety is the
construct of parental control. This paper provides a conceptual overview of the construct of
control in the parenting and anxiety literatures, reviews existing literature on control in
anxious families, and reviews current conceptual models of and developmental approaches to
anxiety. Based on the current empirical literature, an elaborated model that takes child
developmental level into consideration is proposed in order to provide a better understanding
of the role of parental control in the development of anxiety.

KEY WORDS: anxiety; control; parenting; development.

Anxiety disorders are among the most prevalent Orvaschel, and Perrin, 1991; Torgersen, 1983;
forms of adult and childhood psychiatric disorders Weissman et al., 1984). Children of parents with
(Anderson et al., 1987; Bernstein and Brochardt, anxiety disorders are 5–7 times more likely to be
1991; Kashani and Orvaschel, 1990), affecting diagnosed with an anxiety disorder than children of
between 10% and 25% of the population over the parents without an anxiety disorder (Beidel and
course of an individual’s lifetime (Kessler et al., 1994; Turner, 1997; Turner et al., 1987). However, the
Robins et al., 1984). In addition, many anxiety dis- mechanisms for the transmission of anxiety are
orders begin in childhood (Keller et al., 1992; Ost, unclear. Family and twin studies (Hettema et al.,
1987; Stemberger et al., 1995) and may continue into 2001; Kendler et al., 1992; Lenane et al., 1990;
adulthood if left untreated (Cantwell and Baker, Torgerson, 1983) suggest only moderate heritability
1989; Dadds et al., 1999). Due to this early onset, of anxiety disorders (30–40% of the overall vari-
high prevalence rates, and the associated economic ance; Hettema et al., 2001) and thus that the role of
costs, it is crucial to gain an understanding of the environmental factors may have been underesti-
development and prevention of anxiety disorders mated (Tambs and Moum, 1993). Accordingly, it
(Spence, 2001). In order to accomplish this, research appears important to identify psychosocial factors
has largely focused on the transmission of anxiety that put individuals at risk for developing anxiety
from parent to child (e.g., Dumas et al., 1995; disorders.
Hudson and Rapee, 2001; Whaley et al., 1999; Research has relatively recently begun to focus
Woodruff-Borden et al., 2002). on family characteristics that may promote the
Research demonstrates that anxiety tends to development of anxiety. Although the family system
aggregate in families (e.g., Last, Hersen, Kazdin, is complex, research indicates that family relation-
ships differ between anxious and nonanxious families
(e.g., Cobham et al., 1999; Dadds et al., 1996; Stark
1
et al., 1990; Woodruff-Borden et al., 2002). One
Department of Psychological and Brain Sciences, University of
Louisville, Louisville, KY, 40292, USA.
familial characteristic that has consistently been
2
To whom correspondence should be addressed; e-mail: j.woodruff- found to differ and has been emphasized in the
borden@louisville.edu literature is the construct of parental control. Despite

113
1096-4037/06/0600-0113/0 Ó 2006 Springer ScienceþBusiness Media, Inc.
114 Ballash et al.

this emphasis, however, the current empirical litera- found evidence for three styles of discipline in her
ture on anxiety and control lacks a unifying concep- assessment of parents and their preschool children.
tual framework that incorporates a developmental An authoritarian style was associated with lack of
analysis of the impact of parental control on the warmth and high control. She found children of
development of child anxiety. The purpose of the these parents to be discontent, withdrawn, and
current review is to provide a conceptual overview of distrustful. An authoritative parenting style was
the construct of control in the parenting and anxiety associated with being controlling and demanding,
literatures, to review existing literature on control in but also warm and receptive. Children of authori-
anxious families, and to review current conceptual tative parents were found to be content, self-reliant,
models of and developmental approaches to anxiety. explorative, and self-controlled. Finally, permissive
Based on the current empirical literature, an elabo- parents were not controlling or demanding, but
rated model that takes child developmental level into were warm. These children were found to be the
consideration is proposed in order to provide a better least self-reliant, explorative, and self-controlled.
understanding of the role of parental control in the Evidence for a fourth category termed stimulative
development of anxiety. parenting was recently empirically supported (Pa-
quette et al., 2000). However, the generalizability of
this finding is questionable. The study was con-
THE CONSTRUCT OF CONTROL
ducted on French–Canadian impoverished fathers,
IN THE PARENTING LITERATURE
and differences found between groups could be ac-
counted for by the demographic variables. Repli-
The parenting literature offers a rich source of
cation of this finding is needed before it is
data regarding control that could be integrated into
considered as a fourth category.
the anxiety literature. Specifically, the parenting lit-
Baumrind’s (1967) original three categories of
erature differentiates between two types of control:
parental discipline styles have received considerable
behavioral and psychological. Behavioral control
attention in the literature. Many studies have
consists of behaviors parents use to manage, regu-
focused on the child outcomes associated with each
late, and supervise their children (Barber, 1996;
type. As Baumrind (1967) originally found, the
Pettit et al., 2001). It is further broken down into
authoritative style appears to be associated with the
two types: discipline and monitoring. Discipline re-
most positive outcomes. Replications of this original
fers to the specific strategies that parents use to
study on preschool children using separate parent
enforce rules and foster values in their children
and child interviews and naturalistic observations
(Cummings et al., 2000). Monitoring is the level of
supported the findings that authoritative parenting
awareness and supervision that parents maintain to
is associated with stability, assertiveness, indepen-
track their child’s activities and whereabouts
dence, achievement, and social responsibility in
(Cummings et al., 2000). Psychological control is
preschool children (Baumrind, 1971; Baumrind and
defined as the attempts parents make to impede
Black, 1967).
their child’s development of independence and
Positive outcomes have also been found in older
autonomy (Barber et al., 1994; Pettit et al., 2001).
children. These studies have all relied upon child
This dichotomous conceptualization enables inves-
report of parental discipline style, therefore it is
tigation of the differential effects of each type of
important to interpret these findings in terms of
control on children. The empirical literature exam-
perceived parental discipline. Authoritative parenting
ining the impact of these types of control on chil-
has been associated with academic success. In high
dren is reviewed next.
school students, an authoritative parenting style was
related to higher GPAs, while authoritarian and
Behavioral Control permissive styles were negatively associated with
GPA (Dornsbusch et al., 1987). Adolescents in 6th,
Discipline 10th, and 12th grades with authoritative parents
demonstrated lower test anxiety than adolescents
Baumrind (1967) first conceptualized a typol- with nonauthoritative parents (Chapell and Overton,
ogy for parenting discipline styles. These categories 1998). In addition, a mediating model has been sup-
seem to consist of the dimensions of warmth and ported for adolescents in a similar age range.
control (Paquette et al., 2000). Baumrind (1967) Authoritative parenting leads to a healthy sense of
Control in Anxiety 115

autonomy and orientation toward work, which then children (Pettit et al., 1993). Finally, in a sample of
leads to academic success (Steinberg et al., 1989). 4th and 5th grade boys, an interaction was found
A variety of other positive outcomes have been between parent-reported harsh discipline and tea-
found for the authoritative discipline style. For cher-reported aggression. Children’s fearfulness and
instance, it has been associated with advanced moral depression were assessed by self-report. Harsh disci-
reasoning in 6th, 10th, and 12th graders, and 1st year pline was assessed using a 6-item scale, consisting of
college students (Boyes and Allen, 1993; Chapell and descriptions of physically aggressive behaviors par-
Overton, 1998). Using a longitudinal methodology, ents may have used in conflicts with their children.
Baumrind (1991) assessed children at 4, 9, and For moderately to highly fearful children, harsh dis-
15-years-old and found that adolescents of authori- cipline predicted high levels of aggression. For chil-
tative parents were less likely to have problems with dren having low fear, this relationship did not exist.
drug use. Conversely, authoritarian parenting has In addition, harsh discipline was related to children’s
been related to negative outcomes, such as increased self-reported depression in highly fearful boys
self-consciousness and avoidance in the classroom, (Colder et al., 1997).
sports, and amongst peers, in 4th, 5th, 7th, and 8th Comparison is difficult among existing studies
grade students (Yee and Flanagan, 1985). because discipline is frequently defined differently.
Despite the empirical success with Baumrind’s The dimensions of warmth and control are common
typologies, many researchers have defined discipline to each of them, but the parental categories differ
differently, using a wide variety of measures to assess greatly in these studies. Many measures have been
discipline styles. With respect to academic success, an identified as reliable and valid for the study of disci-
ambitious longitudinal project was conducted. pline (Locke and Prinz, 2002). One particular psy-
Maternal behavior was assessed during a teaching chometrically sound measure assesses Baumrind’s
task and a hypothetical discipline situation completed (1967) typology (PPQ; Robinson et al., 1995).
when the child was 4-years-old. Parenting style was Empirical evidence supports the importance of par-
found to impact school performance at ages 4, 5/6, enting discipline styles. Considering that several reli-
and 12. Specifically, parental direct control tactics led able and valid instruments assessing parenting
to lower school performance (Hess and McDevitt, discipline have been developed, continued research
1984). on these categories is needed.
The influence of discipline on aggression in
children has been commonly studied as well. Parents’ Monitoring
discipline style was found to impact preschool chil-
dren’s behavior (Hart et al., 1992). The discipline Parental monitoring has been studied in relation
style was assessed by means of disciplinary style to many psychosocial variables, such as academic
interview administered to a parent and observation of achievement and adjustment, and is central to
parent–child interactions on a playground. Children behavior change (see Dishion and McMahon,
of parents with a less power-assertive style, i.e., a 1998, for a review). In a sample of 11–14-year-old
parenting style characterized by less coercion to elicit children, children and teachers rated parents’
certain behaviors (e.g., less physical punishment, behavioral, intellectual, and personal involvement in
withdrawal of privileges), demonstrated fewer dis- their children’s school (Grolnick and Slowiaczek,
ruptive behaviors during a naturalistic playground 1994). Behavioral involvement was rated by children
observation. These children were also more preferred and teachers, and it included such activities as having
by their peers. Child behaviors mediated the rela- met the child’s teacher, attending parent–teacher
tionship between maternal discipline and peer status. conferences, participation in school events, and
In addition, a significant age difference was found. attending open houses. Intellectual involvement
Older preschool children of less power-assertive included activities fostering the child’s intellectual
mothers demonstrated more prosocial behavior than development, such as going to libraries and museums,
did older children of more power-assertive mothers, reading books and newspapers, and others. Personal
but this difference was not found in younger children. involvement was rated by children only and included
In a longitudinal study, using home-based the involvement subscale items from the Parenting
observations of parent–child interactions, teacher Context Questionnaire and the items from the Parent
and parent-reported child aggression was predicted Involvement Measure. The first questionnaire defined
by negative-coercive parenting style, in 5-year-old involvement as ‘‘the dedication of psychological
116 Ballash et al.

resources in the context of positive affect’’ (p. 241) parental involvement in children’s activities and
and included items such as the parent knowing what behaviors, were positively associated with academic
happens to the child in school and what the child achievement and self-reliance, and negatively related
does in school. The second questionnaire included to drug use, which were all subsequently related to
such behaviors as knowing where the child is and peer group association (Brown et al., 1993). A review
keeping track of how the child does in school. A of the literature indicated that co-occurring individ-
mediating model was supported in which mother’s uality and connectedness in family relationships
behavioral and intellectual involvement led to chil- contributed to adolescents’ identity exploration and
dren’s perceived competence and understanding of role-taking skills (Grotevant and Cooper, 1986).
control, i.e., understanding of how their behaviors There is much less research in the area of mon-
affect outcomes, which then predicted school perfor- itoring, as compared to discipline. In fact, the
mance. School performance was also indirectly research leaves many questions to be answered
affected by father’s behavioral involvement through regarding the relationship of parental monitoring to
perceived competence. Thus, parental monitoring child outcome. The main limitation of this literature
was associated with positive outcomes. is that the methods of assessment of parental moni-
For highly active 4th and 5th grade children, toring have been highly inconsistent. Further,
poor parental monitoring assessed by a parent-report although there is clearly an established link between
questionnaire of parental supervision of children’s low levels of monitoring and child externalizing
behaviors and activities, was associated with aggres- behaviors (Dishion and McMahon, 1998), the rela-
sion. A more complicated relationship between tion between parental monitoring (i.e., high levels of
monitoring and self-reported depressive symptoms monitoring) and child internalizing behaviors (viz.,
was noted. For moderately fearful children, a curvi- anxiety) remains unclear. The extent to which sig-
linear relationship was found. Both extremely high nificantly high levels of parental monitoring affect
and low levels of parental involvement were associ- child internalizing behaviors/outcome developmen-
ated with depression. For highly fearful children, tally deserves further attention.
depressive symptoms increased as parental involve- One important contribution of this literature is
ment increased (Colder et al., 1997). its attention to age differences. Changes in parents’
Parental monitoring has received partial support and children’s views of monitoring with age have
as a protective factor against adolescents’ behavior been studied. For instance, older children have
problems. For instance, a longitudinal study of reported more permissive parental treatment than
10-year-old boys demonstrated that poor parental younger children (Steinberg, 1987). In addition,
monitoring at age 10 was predictive of subsequent junior high school students have described their
involvement with antisocial peers at age 12 (Dishion family environments as more participatory and
et al., 1991). Parents’ monitoring scores were based allowing more autonomy than elementary school
on children’s report of how their activities and students (Yee and Flanagan, 1985). Finally, a longi-
behaviors were supervised by the parents at home, the tudinal study indicated that between the ages of 13
parents’ report about the number of hours spent with and 15 the beliefs that personal issues are under
the children during the 24 h prior to the time of the parental authority decreased in African American
assessment, as well as an interviewer’s rating of how adolescents. These beliefs were assessed by presenting
well a particular child was supervised by his or her the adolescents with a set of 24 hypothetical situa-
parents. However, this relationship was no longer tions and inquiring about their opinions on parental
significant when continuity of friendships was taken authority in each of these situations. The adolescents
into account (Dishion et al., 1991). Other research also changed their perception of parental authority in
supported a relationship between adolescents’ practical, friendship, and personal issues, viewing it
involvement in family decision-making, as reported as less restrictive (Smetana, 2000). These findings
by either the adolescent or both parent and adoles- address important developments in both children and
cent, and susceptibility to peer pressure and deviance. parents’ perceptions of monitoring.
However, this relationship also became insignificant
once family structure was considered (Dornsbusch Psychological Control
et al., 1985; Steinberg, 1987). High-school students’
perceptions of parental monitoring, as measured by a Psychological control and behavioral control
self-report questionnaire assessing the degree of have been empirically demonstrated to be distinct
Control in Anxiety 117

constructs and to play independent roles in child teacher reports of the child’s overt aggression
outcomes (Barber et al., 1994; Smetana and Daddis, (Hart et al., 1998). In terms of internalizing out-
2002). There is a great deal of research demonstrating comes, perceived psychological control was found to
that behavioral control is more consistently linked be positively related to self-reported shyness in col-
with child externalizing outcomes, while psychologi- lege students (Eastburg and Johnson, 1990) and
cal control is associated with internalizing problems negatively related to self-esteem in 7th, 8th, and 9th
(Barber, 1996; Barber et al., 1994). graders (Litovsky and Dusek, 1985). In addition,
Barber (1996) conducted three separate cross- psychological control, as assessed by child report on
sectional studies to support this finding. In 5th, 8th, items related to parental guilt induction in an attempt
and 10th grade students, parent-reported child to influence child behaviors and behavioral observa-
depression was predicted by child-reported parental tion of a 3-min parent–child interaction, was found to
psychological control, as measured by a 6-item be a mediating factor between family conflict and
questionnaire assessing the child’s perception of adolescents’ (11–14-years-old) internalizing problems
parental behaviors and interaction patterns. The (Fauber et al., 1990).
questionnaire included such items as [the parent]’’is Another study evaluated this issue using a lon-
always trying to change me’’, ‘‘wants to control gitudinal methodology (Pettit et al., 2001). Internal-
whatever I do’’, and ‘‘is less friendly with me if I do izing and externalizing problems were assessed at
not see things his (her) way’’. Increased levels of ages 8, 10, and 13/14 using mother, teacher, and
psychological control were associated with increased adolescent reports. The questions on the degree of
depression. Conversely, parent-reported child delin- parental involvement in children’s lives and psycho-
quency was predicted by child-reported parental logical control were modeled after Brown et al.
behavioral control, which was measured by a 5-item (1993), Dishion et al. (1991), and Barber (1996). At
child report scale that included questions about all ages, high levels of parental monitoring were
parental knowledge regarding who the child’s friends associated with fewer delinquency problems, while
are and how the child spends his or her free time. the relationship between parental psychological con-
Higher levels of behavioral control were associated trol and child outcome was more complicated. In
with lower delinquency. Psychological control was girls and teens low in preadolescent delinquent
only found to add a small amount of additional problems, high levels of psychological control were
variance in mother–son dyads. Barber (1996) at- associated with increased delinquency. In girls and
tempted to replicate these findings in a subsample teens high in preadolescent anxiety and depression,
using behavioral observation of parental control and increased psychological control was related to
assessing child outcome one year later at 12 years of increased anxiety and depression.
age. He found a similar relationship between psy- The relationship between psychological control
chological control and depression in mother–daugh- and internalizing versus externalizing outcomes is
ter dyads. Counter to previous findings, behavioral unclear, but high levels of psychological control are
control was uniquely related to depression for father– consistently associated with negative outcomes. Age
son dyads. The previous association between behav- may potentially impact the link between psychologi-
ioral control and delinquency was replicated in mo- cal control and outcome and add to the complexity of
ther–child dyads. Barber’s third study was a this relationship. The relationship between psycho-
replication of study 1 with 5th and 8th grade students logical control and delinquency was stronger for 5th
only. He supported those findings, with the exception grade than 8th grade children. An opposite effect was
that a relationship between psychological control and found for depression; the relationship with psycho-
delinquency was also found (Barber, 1996). Thus, it logical control was stronger for 8th graders than 5th
appears that both types of control may be related to graders (Barber, 1996).
both externalizing and internalizing child problems.
Due to the inconsistency of these findings, further Limitations and Conclusions
replication is needed.
Psychological control has been found to be The literature on parental psychological control
related to externalizing child outcomes in preschool appears to be better established than that of behav-
age children. Parent-reported psychological control ioral control. However, it shares limitations of the
(assessed by a set of items similar to the ones parenting literature in general. Many of which are
described by Barber, 1996) was positively related to related to measurement. For instance, reporting
118 Ballash et al.

biases are important to consider. Whether the in which behaviors occur. Parenting behaviors are
parenting style is reported by child, parent, or both or affected most by a child’s changing age, while par-
whether it is objectively observed influences the enting style should remain more consistent (Locke
interpretations that can be made regarding findings. and Prinz, 2002). This is especially relevant for dis-
Reports by children gauge their perceptions of their cipline, because as children develop, different disci-
parents’ style, which may be most meaningful, but pline strategies become more or less appropriate and
this is limited by age. Younger children are not able useful (Baumrind, 1978; Locke and Prinz, 2002).
to accurately complete self-report questionnaires. Darling and Steinberg (1993) hypothesized that style
Thus, there is a bias in the literature in which studies moderates the influence of behaviors on the child’s
on younger children utilize parent report or obser- outcome. Based on this model, then, it is important to
vation, while studies on older children more often measure both parenting behaviors and style for all
rely on self-report. Parent reports may be invalidated types of control (Barber, 1996). Parenting style may
by the highly face-valid nature of control questions be best measured by self-report (Darling and Stein-
and the strong social desirability of answers. In berg, 1993), and behaviors by objective observation.
addition, research indicates that parents and adoles- A final issue with respect to control in the par-
cents report differently. Adolescents more often enting literature is that there is currently a lack of
described their parents as permissive or authoritar- studies that focus on cultural variations of control.
ian, while parents more often reported themselves as For example, the only minority group given any
authoritative (Smetana, 1995). Observation largely notable attention in the literature has been African
solves these problems, but social desirability is still an Americans, and research in this area is quite limited.
issue. Naturalistic observations may best reconcile While some research indicates that discipline styles
these shortcomings. and their impact on adolescents are similar in African
The literatures of each type of control often American families (Hill, 1995), some researchers have
overlap, and studies investigate multiple types of argued that Baumrind’s (1967) categories do not
control. While this is not necessarily problematic, the appropriately categorize African American parenting
distinct constructs often lose their individual mean- (Mason et al., 1996). In addition, a curvilinear rela-
ing. This seems to be a measurement problem. Con- tionship was found between psychological and
structs are measured differently across studies and behavioral control and adolescent behavior prob-
often become blended into one another. Discipline, lems. The authors explained that the task of deter-
monitoring, and psychological control need to be mining the appropriate level of control is more
further delineated and continue to be researched as difficult for African Americans parents, who often
separate constructs, as previous findings indicate that live in areas more prone to peer pressure (Mason
they have unique effects on child outcome (Barber, et al., 1996). Clearly, more research is needed in order
1992). One approach would be to utilize measures to determine whether control factors are differentially
with strong psychometric properties consistently related to anxiety disorder development.
across studies. Many have already been developed for
discipline and psychological control (Locke and CURRENT STATE OF THE ANXIETY
Prinz, 2002; Robinson et al., 1995; Schaefer, 1965). LITERATURE
Work needs to continue on the development of an
instrument for monitoring. In addition, utilizing both Overview of the Empirical Research
self-report and observational methods would allow
for validation of these constructs. A multitude of studies have been conducted to
One suggested solution to the measurement assess the differences in family functioning between
problem has been to separate parenting behaviors anxious and nonanxious families, including differ-
and parenting style (Darling and Steinberg, 1993). ences in control. The methods for assessing these
Behaviors are described as more concrete variables, families vary between studies. There are typically
such as spanking, while parenting style is the attitude three different approaches to the methodology. First,
and emotional climate one portrays in childrearing. adults with anxiety disorders are interviewed retro-
Theoretically, parenting style relays to the child how spectively regarding their childhood family environ-
the parent feels about him or her, while parenting ment. This method is plagued by the limitations of
behaviors communicate how the parent feels about self-reported, retrospective data collection, such as
the child’s behavior. Thus, parenting style is a context memory bias. Second, families with an anxious
Control in Anxiety 119

parent are studied concurrently. This methodology rejecting. The construct of control was also measured
represents an ‘‘at risk’’ household because of the in this study. No differences were found between the
child’s increased chance of developing an anxiety anxious and control groups on measures of maternal
disorder (Beidel and Turner, 1997; Turner et al., or paternal overprotection (Laraia et al., 1994).
1987). This method evaluates family functioning
prior to a child developing an anxiety disorder. Thus, Studies of ‘‘at risk’’ Families
predisposing factors for anxiety are at work. Finally,
research is conducted on families with anxious chil- Anxious adults are often assessed with respect to
dren. This method provides a measure of family their current family functioning. The impact of this
functioning following the development of a child- environment on the child is then considered. In one
hood anxiety disorder. Consequently, differences such study, an outpatient sample of anxious adults
found in these families could either be due to factors diagnosed via structured interview and a recruited
that predisposed the child to anxiety or factors that control group both completed self-report question-
changed after the child became anxious, with the naires, and parent–child dyads were observed during
directionality of the effect indeterminable. behavioral interactions in a structured play environ-
Research also differs with respect to how familial ment. Children ranged from 7 to 12-years-old. Anx-
differences are measured. Some studies use self- ious parents perceived their families as being low on
report, while others utilize behavioral observations. cohesion, expressiveness, and moral–religious
Both methods have strengths and weaknesses. Self- emphasis, and high on conflict and achievement ori-
report allows the researcher to tap into cognition and entation. No differences were found on self-reported
other internal processes, but it is also more likely to control between groups. Anxious parents also did not
be distorted by factors such as perception or social report a difference from nonanxious parents in the
desirability. In addition, it must be viewed in light of activities in which they allowed their children to
children’s abilities to accurately provide self-reports participate, but they reported more distress when
on complicated phenomenon. Observational studies their children separated from them and during chil-
provide a more objective representation of family dren’s physical activities. Anxious parents reported
functioning. However, social desirability can still more personal distress during the structured play
impact the family’s behavior in the laboratory setting. task. They were also less likely to play with their child
In addition, valid and reliable coding systems can be during the task than nonanxious parents were
difficult to establish. It is imperative for accurate (Turner et al., 2003).
understanding of the literature to interpret findings Whaley and colleagues (1999) conducted a
with these caveats in mind. behavioral interaction study in which anxious moth-
ers interacted with their child in mild to moderately
Retrospective Studies anxiety-producing tasks. Mothers were recruited and
administered a structured interview about themselves
Retrospective self-report measures of anxious and also a semi-structured interview to obtain the
adults were among the earliest type of data collection child’s diagnostic information. Children, who ranged
in familial anxiety studies. The participants for these in age from 7 to 14, were also administered the semi-
studies were often drawn from a clinical population. structured diagnostic interview to determine diag-
In one study, adults previously diagnosed by a mental nostic status. The groups were created based on
health worker completed retrospective control mea- presence or absence of an anxiety diagnosis in the
sures for their mother and father. A sample of 60 mother, and 36 families met criteria. Anxious moth-
adults with panic disorder recalled their fathers as ers were found to be less warm, positive, and granting
being high on overprotection and low on care (Leon of autonomy. In addition, they were more critical and
and Leon, 1990). In another study, 80 female patients catastrophizing. Maternal behavior predicted child
with panic disorder with agoraphobia, diagnosed anxiety, contributing more to its development than
using a structured interview, described their child- maternal psychopathology or ongoing stress. Fur-
hood environments as conflictual, lacking in parental ther, anxious mothers whose children were also
warmth and support, and involving parental chronic anxious were significantly less likely to display
physical illness and substance abuse in comparison to behaviors consistent with autonomy-granting than
100 women with no self-reported clinical diagnosis. were either anxious mothers with nonanxious chil-
Mothers of the anxious group were described as more dren or the nonanxious mother–child dyads, lending
120 Ballash et al.

support to the notion that control behaviors exhib- The majority of studies involving anxious chil-
ited by anxious mothers may be reciprocally influ- dren are conducted using behavioral data. Often both
enced by child anxiety. behavioral and self-report data are combined. One
A follow-up study to the original Whaley et al. study demonstrated that children are accurate
(1999) research found a main effect for child anxiety reporters of familial interactions involving themselves
status on maternal granting of autonomy, indicating (Siqueland et al., 1996). Children between the ages of
that it is anxiety in the child that contributed most to 9 and 12 were designated to either anxious or control
mothers’ control behaviors. Maternal displays of groups. Children in the anxious group were diag-
warmth toward their children were most associated nosed by structured interviews conducted with parent
with child anxiety status, not maternal anxiety and child. Controls were recruited participants who
(Moore et al., 2004). did not receive an elevated anxiety score on a parent-
In a similar study, 51 parents interacted with report measure. Diagnostic status of parents was not
their children during moderately anxiety-provoking obtained. Objective observers rated parents with
tasks. Parents were designated as either anxious or clinically anxious children as less granting of auton-
nonanxious following administration of a clinical omy (i.e., more controlling). Anxious children rated
interview. Children, who ranged in age from 6 to 12, their parents as low in acceptance. Mothers’ and
were also administered a structured diagnostic inter- fathers’ self-report ratings coincided with one
view to determine anxiety status. Woodruff-Borden another, but not with their children’s. Interestingly,
and colleagues (2002) observed that anxious parents the children’s reports, but not the parents’, correlated
were less productively engaged, more withdrawn, and with observer ratings.
more disengaged than control parents. A mean dif- Another study, however, failed to find a rela-
ference in control was not found between groups. tionship between children’s reports and observed
However, following a child’s display of negative behavior (Greco and Morris, 2002) in a group of high
affect, anxious parents were more likely than non- versus low socially anxious children. Groups were
anxious parents to attempt to control the situation. defined by the highest and lowest quartile scores on a
child social anxiety measure. The 48 children were
Studies with Anxious Offspring either in elementary or middle school, but age infor-
mation was not provided. Children also completed
Finally, researchers have approached the study of self-report measures of perceived parenting style.
characteristics in anxious families from the perspective Children and their fathers completed a brief, chal-
of families with anxious children. In one study, 51 lenging task together, which was coded for behavioral
children between the ages of 9 and 14 were separated indicators of parental overprotection and rejection.
into diagnostic groups on the basis of a semi-structured There was also no difference in socially anxious and
interview with the child. Diagnostic status of parents socially nonanxious children’s perceptions of their
was not obtained, however, mothers of the children fathers. However, observers rated fathers of socially
completed a self-report measure of family functioning. anxious children as more controlling than fathers of
Anxious and depressed children perceived their fami- nonanxious children.
lies as less supportive, more conflictual, more disen- A sample of children in treatment for an anxiety
gaged from outside activities, more enmeshed, less disorder and a sample of children who had never
promoting of independence, and less democratic than sought treatment were compared on interactions with
control children did (Stark et al., 1990). a parent during cognitively challenging tasks. The
Messer and Beidel (1994) compared 100 children clinical group was diagnosed via structured inter-
on their reports of family functioning. The children views with parent and child. Diagnostic status of the
were separated into groups on the basis of a composite parents was not obtained. The total sample included
diagnosis from structured interviews with the parent 95 children between the ages of 7 and 15. Parents of
and child. The children had a mean age of 10, with no anxious children were rated by objective observers as
age range reported. Anxious children reported that being more negative and more intrusively involved in
their families promoted less independence. Results of the task (Hudson and Rapee, 2001).
a child self-report measure did not reveal perceived In a similar behavioral task, only behavior was
differences in terms of control. However, correlational coded as opposed to verbal interactions (Krohne and
analyses revealed that child anxiety was positively Hock, 1991). Children ranged between 10 and
related to reported parental control. 13-years-old, and they completed a difficult
Control in Anxiety 121

problem-solving task with their mothers. Maternal groups. Anxious children were more likely to be
anxiety and child anxiety were assessed via a self- noncompliant and coercive, underscoring the impor-
report measure of trait and state anxiety. Mothers of tance of examining the bidirectional nature of an
highly anxious girls were more likely than mothers of interaction between parent and child.
girls with low anxiety to intrude upon the child’s
problem-solving behaviors in a competitive manner. Conclusions and Limitations
In addition, the mothers had a tendency to maintain
that control. This relationship was not found among Overall, these findings indicate that there is a
boys. However, it is important to note that in this difference in parental behavior between anxious and
study maternal anxiety was significantly associated nonanxious families. The retrospective reports and
with anxiety in the female children, but not the male at-risk family studies provide partial support for
children. control being more prevalent in anxious households.
In a behavioral observation study examining the However, studies involving anxious children provide
effect of parental discussions on cognition in anxious overwhelming support for this notion. Both self-re-
children, Barrett and colleagues (1996) presented a port and behavioral observation studies revealed an
series of ambiguous situations, which could be con- increased level of control in anxious families.
strued as either threatening or nonthreatening, to Another important finding to be extracted from
children with anxiety, oppositional defiant disorder, these data is the importance of parental behavior.
or no diagnosis, and their parents. Results revealed Parental behavior was found to contribute more to
that anxious children interpreted the situations as child anxiety than maternal psychopathology or
significantly more threatening, after discussion with stress (Whaley et al., 1999). This finding demonstrates
their parents, compared to the nonanxious and the importance of conducting behavioral studies and
aggressive children. The anxious group also reported further exploring the impact of controlling behavior
more avoidant responses than the other groups. on children. Woodruff-Borden and colleagues (2002)
Parents of the anxious children concurred, and were illustrated a situation in which controlling behavior is
significantly more likely to report that their children most likely to occur for anxious parents. Anxious
would use more avoidant strategies to manage the parents are more likely to become controlling when
situations. In a follow-up paper examining the family their child expresses negative affect. This may be the
processes during discussion of these ambiguous situ- parent’s strategy for trying to alleviate perceived
ations, parents of nonanxious children were signifi- anxiety in the child. Dumas and colleagues (1995)
cantly more likely to discuss the positive results of elaborated on the bidirectional nature of a control
engaging in specific behaviors, and parents of the exchange between parents and children. Data taking
anxious children were more likely to respond to their into consideration both the parent and child’s con-
children’s avoidance with avoidance behaviors or tribution are vital. They demonstrate the effect that
suggestions (Dadds et al., 1996). one is having on another and what behaviors lead
Finally, interactions between parents and their them to behave in certain ways, such as becoming
young children were explored (Dumas et al., 1995). A more controlling.
total of 126 children between the ages of 2 and 6 were Unfortunately, in all of these studies examining
divided into three groups on the basis of teacher’s control in anxious families, only two assessed the
reports: socially competent, aggressive, and anxious. impact of child age or developmental level. In those
Maternal anxiety status was not obtained. Mothers studies, child age was found to influence both parent
of anxious children were found to initiate more and child behavior. Hudson and Rapee (2001) found
control exchanges and demonstrate aversive affect that child age was related to maternal involvement,
more often than other mothers. They were more but not negativity. Older children received less help
coercive and unresponsive to their children. More- from their mothers. Dumas and colleagues (1995)
over, mothers of anxious children complied with their found an impact of child age on child behavior.
children’s control exchange behaviors significantly Younger children were more likely to express non-
less often than did mothers of aggressive or compe- compliance behaviors than older children. Thus, it
tent children. There were no significant differences appears that child age could have significant conse-
among the three maternal groups in response to quences on behavioral interactions between parent
aversive child behavior versus positive child behavior. and child. How these differences impact the expres-
Children’s behaviors also differed among the three sion of control and the development of anxiety is still
122 Ballash et al.

unknown, as studies to date either have not analyzed specifically related to anxiety than to depression. He
age as a variable, or have not included a sample of surmised that anxiety is related to the children’s
children with an age range large enough to investigate perception of parents as being controlling and not
differences. Moreover, other factors directly related accepting. He also noted that controlling behaviors
to age such as children’s emotional and social devel- and anxiety could have a cyclical relationship be-
opment can impact the development of anxiety and tween parent and child, in that one increases the
the expression of control. likelihood of the other.
Despite findings to date, the literature on control Rapee (1997) also noted many of the aforemen-
in anxious and nonanxious families has several tioned methodological limitations, including vague
methodological shortcomings. First, control has not and numerous labels of control, which made the
been systematically studied. Although many positive combination of data difficult, a lack of rationale, the
results regarding control were found, control was not use of retrospective data, and a lack of clinical
necessarily identified as the main variable in these comparison groups. Rapee (1997) stated that the
studies. Often, no specific results were hypothesized. relationship found between control and anxiety was
Instead, researchers explored what differences might not necessarily related to control by one parent only.
exist between anxious and nonanxious families. In The other parent, typically the father, also needs to
addition, there is a lack of consistency in the opera- be assessed. In addition, causality cannot be deter-
tionalization of control. For example, it has been mined by this type of research, and while it is an
defined as overprotection (Leon and Leon, 1990) or important first step to determine that a relationship
lack of promoting independence (Messer and Beidel, exists, it is now time for the literature to move beyond
1994; Stark et al., 1990) in self-report studies and that. Studies also often have small sample sizes,
operationalized as behavioral devices, such as com- which yield inadequate power to detect an effect.
mands (Woodruff-Borden et al., 2002) or taking over Finally, he noted that research has tended to focus on
the task (Krohne and Hock, 1991), in observational perceptions of child-rearing as opposed to actual
studies. Third, control is rarely measured by both observation, and he advocated for more studies using
self-report and objective observation. Such a method behavioral observation.
would allow for comparison of perception of control A more recent qualitative review of the anxiety
to actual behavioral differences. Fourth, the issue of literature has been conducted, building on Rapee’s
diagnosis is paramount. Many studies rely upon self- (1997) earlier review (Wood et al., 2003). Wood and
report measures of anxiety to separate participants his colleagues (2003) focused on literature from 1990
into groups, which precludes examination of differ- to 2002 that assessed parenting and anxiety. Overall,
ences between clinical and nonclinical groups. The they concluded that a higher level of observed
literature has shown an increased risk for develop- parental control was related to an increased risk for a
ment of an anxiety disorder in children with parents child meeting criteria for an anxiety disorder. They
with an anxiety diagnosis, not anxiety in general summarized this effect as medium to large and of
(Beidel and Turner, 1997; Turner et al., 1987). clinical significance.
Finally, it is vital to take both the parent and child’s Wood and colleagues (2003) also summarized
anxiety into consideration. If both are not known, the limitations in the literature and made suggestions for
inferences that can be made are limited. For inter- future research. First, research is generally conducted
pretation, it is important to consider whether both on nonrepresentative samples, often excluding
the parent and child are anxious or if only one is. The minority ethnic groups. Second, they cited an over-
bidirectional effects of parents and children on one reliance on self-report measures. Third, a reliance on
another should be considered. cross-sectional designs limits the ability to make
inferences regarding causality and development.
Reviews Finally, they criticized the use of global parenting
measures. They categorized parenting into three cat-
Rapee (1997) conducted a qualitative review of egories: acceptance, control, and modeling of anxious
empirical studies focusing on the impact of childre- behaviors. They also differentiated between general
aring on the development of anxiety and depression. parenting styles, as often assessed by global parenting
He summarized the research as investigating two of family measures, and discrete parenting behaviors.
main parental factors: acceptance and control. Based They suggested that it is important for empirical
on his review, he concluded that control is more research to do this as well. They described the study
Control in Anxiety 123

of discrete behaviors as contributing more to the by the parent could contribute to the development of
understanding of parent–child interactions and as anxiety in two ways. First, it relays to the child that
being more clinically useful. They suggested that there is a constant threat to them, leading to hyper-
researchers study more specific hypotheses in this vigilance and fear. Second, it limits the child’s
area. In addition, they suggested that researchers pay opportunity to learn how to cope successfully with
more attention to the expected time interval between situations on their own, thus preventing the devel-
a parental behavior and development of anxiety in a opment of realistic expectations for themselves and
child as well as the cumulative impact of this the world.
behavior. Wood and colleagues (2003) suggest a similar
mechanism. Specifically, they postulate that parents’
Summary controlling behaviors prevent children from devel-
oping age-appropriate behaviors, such as dressing
Overall, the parenting literature lacks a uniform one’s self, and age-appropriate boundaries between
theory with respect to the understanding of the role parent and child, such as decreased clinging behav-
of behavioral and psychological control on the iors. This impairment in appropriate development
expression of anxiety. The studies often examine the then leads to a lack of a sense of control, mastery,
relationship between control and multiple outcome and autonomy for the child. The child is dependent
variables. This is not only a statistical problem due to on his parent, leading to possible separation anxiety.
inflated alpha level, but it also impedes conceptual They argue that the specificity of this model allows
progress. More theoretical progress could be made by for hypothesis-testing. For example, children below
moving beyond simple relationships and examining age-appropriate norms could be compared to chil-
the paths among variables. dren above those norms in terms of anxiety symp-
Despite these limitations, studies of control in toms. However, no research has yet been conducted
the parenting literature improve upon studies in the to test this model.
anxiety literature by more often including fathers, As a result of the emphasis on parental control,
conducting naturalistic behavioral observations, and the construct of perceived control has also received
utilizing longitudinal methodology. Overall, this lit- significant attention in the anxiety literature as well as
erature has consistently demonstrated that high levels in the developmental literature. For example, Skinner
of control, especially in the absence of warmth, lead (1996) suggests that perceived control can be under-
to negative outcomes for children and adolescents. stood in terms of a developmental process that is
The division of control into three distinct types and influenced by contextual factors such as the child’s
the conceptualization of parenting as both behavior interaction with his or her parents, motivation, and
and style allows for investigation of the unique cognition. Thus, changes along these dimensions will
aspects of control that may contribute to childhood result in changes in perceived control. Related to this
problems. However, this finer level of analysis has yet is Barlow’s (2001) model, which proposes that the
to be applied to psychopathology in children and perception that events are outside of one’s control is a
would enhance models of the development of anxiety. key factor in the psychopathology of anxiety. Thus,
researchers have investigated the role it plays in the
CURRENT CONCEPTUAL MODELS RELATED development of anxiety. The rationale behind these
TO ANXIETY DEVELOPMENT studies is that overcontrolling parental behavior leads
children to believe that events are outside of their
The construct of control is well established and control (Davis and Phares, 1969). In order to prop-
accepted as a contributor to the development of erly present theoretical models in this area, it is first
anxiety. More conceptual models of how this trans- necessary to establish that there is a difference in the
mission occurs and what specific processes operate family environments of children with different per-
are needed to further develop models of the devel- ceptions of control.
opment of anxiety.
Both Rapee (1997) and Wood et al. (2003) pro- Empirical Studies
pose mechanisms by which control may lead to the
development of anxiety. Rapee (1997) suggests that Like parental control, the construct of perceived
controlling behavior might teach children that events control has been defined in numerous ways. One of
are outside of their control. Overcontrolling behavior the first and most often used constructs is that of locus
124 Ballash et al.

of control. There are two dimensions of locus of ing than children with an external locus of control
control. First, an internal locus of control indicates perceived their fathers to be. Finally, objective
that individuals view events as predictable and within observers characterized the family environment with
their control. Second, individuals with an external children with an internal locus of control as more
locus of control view events as out of their control and articulate, fun, relaxed, encouraging, and helpful.
as being unpredictable. This type of locus of control They also noted that parents of children with an
would be hypothesized to be related to anxiety. internal locus of control were more verbal, partici-
An early study evaluated this construct among pated more often, made more helpful suggestions,
undergraduate students (Davis and Phares, 1969). made less negative statements, and praised the child
Comparisons involved 120 students, half of whom more often than parents of children with an external
scored on the extreme internal end of a self-report locus of control did.
measure of locus of control, and half of whom scored
on the extreme external end. Students with an inter- Conclusions and Limitations
nal locus of control reported their parents to be more
positive and lower in rejection, control, inconsistent It is noteworthy that many of the same charac-
discipline, and love withdrawal as compared to those teristics of families with anxious children were also
with an external locus of control. Davis and Phares found in families with children who reported an
(1969) suggested that children might develop an external locus of control. The most relevant finding to
external locus of control in order to avoid the anxiety this review is the construct of parental control, which
associated with the increased rejection and hostility. differed between families with children whose locus of
Instead of accepting the failure or deprecation this control was internal versus external. Family envi-
may cause, they attribute aversive events as outside of ronments of children with an external locus of control
their control. were perceived or observed as more controlling and
Other studies have measured locus of control in hostile than family environments of children with an
young children and found similar results. Sixty 2nd internal locus of control.
grade students were divided into groups based on a The limitations of these studies are similar to
median split of scores on a locus of control ques- those reviewed for parental control. The samples are
tionnaire. The researchers observed the mother and small, and there is an overreliance on self-report data.
child while they worked on three puzzles of varying Additionally, perceived control is defined in several
difficulty. They found mothers of children with an different ways. However, findings with behavioral
internal locus of control to be warmer, more nur- observation were similar to those based on self-re-
turing, more granting of autonomy, and less critical. ports. The study that utilized undergraduates (Davis
Mothers of children with an external locus of control, and Phares, 1969) did not specify if the measure of
however, interfered, criticized, and took over the parental characteristics was retrospective or current.
child’s task more often than the mothers of internals. This would make a significant difference in interpre-
They also disengaged from the task more often and tation. Further, it is unknown how much contact the
engaged in their own play activities (Gordon et al., college students still had with their parents. It likely
1981). varies widely, along with the impact that parents still
A similar methodology was utilized in a study on have on their young adult children. Overall, the
5th graders (Chandler et al., 1980). Fifteen children direction of these effects cannot be determined. It is
scoring at the extreme internal end of a locus of possible that children with low perceived control elicit
control measure and 11 children scoring at the more assistance from their parents due to a percep-
extreme external end were compared. They partici- tion that they are not in control of the task. Finally,
pated in challenging behavioral tasks with both par- there is no consideration of the impact of a child’s age
ents. The tasks were designed so that it was necessary in this literature. It seems of utmost importance to
for the parents to provide assistance. Maternal self- consider this in regards to this cognitive variable, as
ratings of locus of control were related to her child’s, cognitive abilities change rapidly in children.
while paternal self-ratings were not. Fathers of chil-
dren with an internal locus of control perceived Review
themselves to be more accepting and encouraging of
independence and less restrictive. Their children also Schneewind (1995) reviewed literature on the
perceived them to be more nurturing and less reject- relationship between family processes and perceptions
Control in Anxiety 125

of control. He commented on the paucity of research suggested that early childhood experiences with a
in this area. He stated that specific parenting practices lack of control, perhaps due to overprotective par-
have been associated with the development of chil- enting, lead to an increased likelihood to perceive
dren’s control beliefs. For instance, internal control subsequent events as outside of one’s control. This
orientation has been associated with granting of perception consequently acts as a psychological vul-
autonomy, less intrusiveness, less hostility, more nerability to the development of anxiety. Thus, they
inductive discipline, warmth, and supportiveness. placed greater weight on childhood experiences with
Parents of children with external orientations have a lack of control, although they still acknowledged
been found to be less stimulating, less responsive, and that later experiences can reshape that perception.
more authoritarian, intrusive, overprotecting, and Chorpita and Barlow (1998) formulated two
rejecting. He also cited longitudinal evidence that sequential models. First, in early childhood, they
early perceptions of a stimulating family environment argued that family characteristics are associated
are associated with control beliefs as a young adult. with the development of a lack of a sense of control
Finally, Schneewind examined the correlation in children. If these same family characteristics are
between control beliefs in parent–child dyads. He associated with the development of anxiety, then the
stated that a relationship had been found between perceptions of control serve as a mediator in the
parents and older children and adolescents, but not development of anxiety. The research reviewed in
with preschool children. This suggests that children the current paper indicates that parental overcon-
may be incapable of modeling at a young age, but that trol is common to both. Further, Chorpita and
it acts as an important mechanism of transmission in Barlow (1998) hypothesized that certain familial
later development. characteristics can put a child at risk for developing
In sum, this literature suggests that a controlling both control-related cognitions and anxiety, sug-
environment may be a predisposing factor for gesting that perceptions of control mediated the
developing control-related cognitions. However, it is relationship between family characteristics and
not without its limitations. Schneewind (1995) dis- anxiety. In addition, they suggested that in later
cussed many shortcomings of the literature. First, development these cognitions begin to interact with
measuring the cognitive phenomenon of perceptions negative experiences, and it is this interaction that
of control in children is difficult, and the methods leads to anxiety. This corresponds with a more
that researchers use vary widely. Second, the majority traditional diathesis-stress, or moderating model of
of the research is conducted on mother–child rela- psychopathology (Chorpita, Brown, and Barlow,
tionships, precluding consideration of the more 1998).
complex family system. Third, longitudinal data is To date, empirical support for this model is
almost nonexistent. Finally, analyses investigating limited. Support has largely been borrowed from the
specific behaviors and their impact on control beliefs depression literature. Depression researchers have
are still needed. long recognized and empirically supported the notion
that children’s increasing cognitive abilities will
Mediating/Moderating Models impact the attributional process inherent in these
models (Finchman and Cain, 1986; Hammen et al.,
The next step, then, in conceptualizing the rela- 1988; Nolen-Hoeksema, et al., 1992; Rholes et al.,
tionship between parental control, perceptions of 1980). Nolen-Hoeksema and colleagues (1992) found
control, and anxiety is to develop a model of the that children’s cognitions stabilize with age and begin
sequential relationship among them. Chorpita and to contribute more to the development of depression.
Barlow (1998) have proposed a theoretical model to These developmental changes lead to different
accomplish this. It is currently the only model in the expectations for mediating versus moderating effects
literature to connect all of these constructs in a linear of negative cognitions on depression. Cole and
fashion. They reviewed extensive literature from Turner (1993) found support for both mediating and
depression, locus of control, explanatory style, ani- moderating models in 356 4th, 6th, and 8th grade
mal learning, biological, and attachment literature, children. However, the fit of the model differed
which supported the notion that a lack of a sense of depending on the type of cognition. However, using
control influences the development of anxiety. They 409 children from the same age groups, Turner and
defined control as an individual’s capacity to influ- Cole (1994) found that a moderating model only fit
ence situations and outcomes in the environment, and for the oldest children.
126 Ballash et al.

Chorpita and Barlow (1998) hypothesized that obviously vital to a better understanding and vali-
progressive stabilization of cognition represents the dation of this model. It is the first model to so con-
transition from a mediating to a moderating model. cisely draw these variables together, with initially
Empirical support is needed in the anxiety literature promising results. Most importantly, however, the
to support Chorpita and Barlow’s (1998) model, age at which the transition occurs from control as a
especially measuring perceived control as a specific mediator to a moderator is unclear. There is limited
type of cognition. Chorpita and colleagues (1998) support for a change in cognitive effects on depres-
have initiated this process. They studied 93 families sion between 6th and 8th grades (Turner and Cole,
with children ranging from 6 to 15-years-old, with a 1994). Research needs to strive toward the elucida-
mean of 11.15 years. Of these children 62 were drawn tion of this transition.
from a clinical setting and were diagnosed with an The extant empirical data and theoretical con-
anxiety disorder using structured interviews with the ceptualizations of the relationship between parental
parent and child. The remaining 31 families were control and anxiety have been presented and cri-
recruited from the community. Their children did not tiqued. To summarize, there is a difference in the level
have a primary diagnosis of anxiety, as assessed by a of parental control between anxious and nonanxious
structured interview with the parent and child. Based families, a model linking control and anxiety via
on self-report and parent-report measures, they perceptions of control has been proposed, and there
found support for a mediational model among are limited data to support this model. For the
school-age children. Locus of control mediated the further conceptualization of the construct of control,
relationship between control in the family environ- a combination of theoretical models would be useful
ment and negative affect. Negative affect was then (Chorpita and Barlow, 1998). Chorpita and Barlow
related to clinical symptoms. A controlling family (1998) drew from varied literatures to compose
environment appears to contribute to a cognitive their model. Unfortunately, there has been a lack of
vulnerability of a lack of perceived sense of control, integrative research across fields, resulting in the
which then leads to anxiety. A large portion of locus development of diverse constructs (Masia and
of control was accounted for by control in the family Morris, 1998) and subsequently impeding progress as
(31%), and 27% of negative affect was explained by research is done concurrently, but separately, on
locus of control. Thus, these relationships are not similar constructs.
only significant, but likely meaningful and important
for understanding the development of anxiety. In a A DEVELOPMENTAL APPROACH
study examining the role of control in the association TO ANXIETY
between family characteristics and anxiety in sample
of young adults, Ballash et al. (2006) found that One commonality between both the anxiety and
control served to mediate the link between perceived parenting literatures is the lack of attention given to
family environment and self-reported anxiety; how- developmental changes in control. As previously
ever, no moderating effects of control were found. reviewed, few studies have considered it. The par-
Chorpita and colleagues (1998) recognized a enting literature has tended to analyze age differences
need for further replication because of the complexity more often than the anxiety literature has. However,
of the model. They particularly suggested replication neither literature has systematically studied the effect
using more sophisticated measures of control, longi- of children’s age and other developmental factors on
tudinal methodology, and more diverse samples. parental control or how control impacts outcomes at
Chorpita and Barlow (1998) advocated for future different ages with respect to anxiety.
research to investigate the precise time in develop- Several researchers have recognized the impor-
ment that the switch occurs between these two tance of a developmental analysis in understanding
models. In addition, they suggested combining the etiology of anxiety disorders in children
existing theoretical models in order to expand our (Manassis and Bradley, 1994; Vasey and Dadds,
understanding of the development of anxiety. 2001). Westenberg et al. (2004) and Warren and
The developmental combination of two models Sroufe (2004) have proposed models suggesting that
presents a complicated picture, and research thus far manifestation of anxiety symptoms changes
is extremely limited and inconsistent within the throughout the course of development. For example,
depression literature. Variance may be due to the way at preschool age, children are fearful of being alone
control is measured and defined. Replication is and of darkness. Worrying about death and danger
Control in Anxiety 127

may become apparent at ages 10–13, and worries control occur over time for younger children versus
related to social situations at age 14–17 (Weems and adolescents. This indicates that a child’s age may play
Costa, 2005; Weems and Stickle, 2005). The content an important role in parents’ adjustment of control.
of worries and fears is influenced by various devel- However, method of reporting confounds these re-
opmental processes. For example, it may change sults. Parents reported their own control in the study
from concrete to more abstract threats (Vasey, 1993) with younger children, while the adolescents reported
as the child develops the capacity for abstract on their parents’ control. In addition, parental or
thinking. Worries may become more focused on child anxiety was not considered in these analyses.
evaluation as the adolescent begins developing social Only one study including anxious families
understanding. examined the changes in parental control throughout
Empirical evidence supports the developmental the child’s development. In an observational study,
approach. Studies consistently demonstrate that mothers of anxious older children provided less help
younger children are at great risk for developing than mothers of anxious younger children (Hudson
separation anxiety disorder, while adolescents are and Rapee, 2001). Changes in parental control across
more likely to develop social anxiety disorder (Weems child age have not been directly compared between
et al., 1998; Weems and Costa, 2005). Panic attacks anxious and nonanxious parents.
do not typically begin until adolescence (Weiss and The developmental changes in both perceived
Last, 2001). Among normal children, younger chil- control and parental control may have important
dren are more likely than adolescents to be afraid of implications for child anxiety. Overall, it appears
the dark, harm coming to them, and insects, as well as that, among nonanxious parents, control tends to
endorsing greater social withdrawal (Bell-Dolan increase until a child is approximately 9 years of age,
et al., 1990). Age differences are also evident in chil- and then tends to decrease. These findings can guide
dren who refuse to go to school. There appear to be future work in this area and aid in the development of
two primary anxiety diagnoses related to this conceptual models. There is almost no research on
behavior: phobias and separation anxiety. Phobic this trend in anxious parents.
school refusers tended to have a later age of onset and One study examined this relationship in 5th and
a more severe and pervasive course of school refusal 8th grade students. Psychological control was more
(Last and Strauss, 1990). strongly related to delinquency in younger children,
Several studies have examined developmental while depression and psychological control were
issues related to parental control and the child’s more strongly related in older children (Barber,
perceived control and they suggest a relationship 1996). No studies have examined this issue in regard
between the child’s age and child’s perceived control. to anxiety as an outcome. The developmental changes
Extensive research in this area has been conducted by in both perceived control and parental control may
Skinner (1990, 1991, 1996), indicating that develop- have important implications for child anxiety.
mental changes in perceived control are a reflection of
the change in beliefs about causes. For example, at A PROPOSED MODEL FOR A DEVELOPMEN-
age 7–8 children’s beliefs about causes are categorized TAL APPROACH TO CONTROL IN ANXIETY
as ‘‘known’’ and ‘‘unknown’’ (Connell, 1985). At
9–10 years of age, causes are categorized as internal Many gaps remain in the literatures on control
or external (e.g., Nowicki and Strickland, 1973). and anxiety. In order to progress conceptually with
The child’s development may also affect parental these constructs, we must begin to integrate the
control. In children studied longitudinally at information from various fields and develop
33 months and 9 years of age, parents indicated that hypotheses regarding the missing information.
their level of control increased over time (Hagekull Schneewind (1995) suggested a mechanism by which
et al., 2001). High school students reported that their this literature becomes cohesive. He stated that it is one
parents’ levels of monitoring decreased over the of the most subtle and difficult challenges of parenting
course of 2 years (Freeman and Newland, 2002; to adjust parental behavior to an age-appropriate level
Smetana, 2000). In addition, older children are trea- for the child. However, extensive research indicates
ted more permissively (Steinberg, 1987) and describe that anxious parents or parents of anxious children are
their families as more participatory and allowing more controlling than nonanxious parents at various
more autonomy than younger children (Yee and child ages (2–14-years-old; Dumas et al., 1995; Greco
Flanagan, 1985). Thus, different patterns of change in and Morris, 2002; Hudson and Rapee, 2001; Leon and
128 Ballash et al.

Leon, 1990; Messer and Beidel, 1994; Siqueland et al., control. Child age is hypothesized to moderate this
1996; Stark et al., 1990; Whaley et al., 1999). It could be relationship. Anxious parents’ levels of control will
hypothesized that anxious parents are unable to adjust remain high and constant regardless of child’s age,
their controlling behavior appropriately to their child’s whereas nonanxious parents’ control will likely
developmental level. increase until middle childhood and then decrease
Woodruff-Borden and colleagues (2002) sug- with child’s age. This difference in parental modifi-
gested that anxious parents tend to withdraw from cation of behavior across time results in an ordinal
their children until they behave negatively. Then, the interaction. However, child age is hypothesized to
parent acts in an overcontrolling manner. They only affect parenting behaviors, while parenting style
hypothesized that the initial withdrawal is due to the should remain more stable regardless of child age
parent’s preoccupation with his or her own worries or (Locke and Prinz, 2002).
fears. Thus, it could be speculated that anxious par- Prevalence studies of anxiety disorders have
ents are not as attentive to the subtle developmental demonstrated a higher prevalence in females than
needs of their children and act in an overcontrolling males (e.g., Kessler et al., 1994). This suggests that
manner, failing to appropriately consider a child’s parent gender will influence parental anxiety. Most
age. This failure to properly modulate their behavior studies to date have only examined mother–child
could prevent children from learning age-appropriate dyads, or have included fathers along with mothers
tasks and foster the belief that they are unable to deal without consideration of the differential effects of
with events on their own, leading to a sense of inse- parent gender. However, some studies have found a
curity, dependence, and anxiety (Rapee, 1997; Wood difference in discipline styles between mothers and
et al., 2003). fathers (Conrad and Ho, 2001; Russell et al., 1998).
The extant literature is summarized in a theo- In addition, mothers were found to be more influ-
retical model to support the hypothesis that anxious ential in their child’s development of perceptions of
parents differ from nonanxious parents in how they control (Chandler et al., 1980), and adolescents’ self-
adjust their levels of control across child age. The esteem and shyness were more closely related to
implications of this difference and other potential mothers’ psychological control than to fathers’
impacts of child age/developmental level are also (Eastburg and Johnson, 1990; Litovsky and Dusek,
outlined in Fig. 1. 1985).
It has been established in the literature (e.g., In the current model, parental control is repre-
Whaley et al., 1999) that the presence or absence of sented as both a style and behavior. This dichotomy
an anxiety diagnosis will lead to a certain level of has been suggested in both the anxiety and parenting

Parental Control
Style
Child
Perceived
Control

Child Anxiety

Parent Parent
Gender Anxiety Parental Control
Behaviors

Child Age/Developmental
Level

Fig. 1. Hypothesized path model with child developmental level moderating the relationship between parent anxiety, parental control,
child’s perceived control, and child’s anxiety.
Control in Anxiety 129

literature (Barber, 1996; Darling and Steinberg, 1993; RECOMMENDATIONS FOR FUTURE
Wood et al., 2003). The parenting control styles RESEARCH
include self-report measures of discipline, monitor-
ing, and psychological control (Barber, 1996; Darling The resulting model is complex, but provides
and Steinberg, 1993). The controlling behaviors are cohesion to a vast and previously disconnected body
assessed via observed interactions with the parent and of literature. Empirical testing of this model will
child. The parenting style acts as a moderator on the require several studies and replication. However,
relationship between controlling behaviors and child studies on individual parts of the model can be con-
outcome (Darling and Steinberg, 1993). ducted and revisions made as necessary. In doing this,
Child outcome is first conceptualized by the validation of other models proposed in the literature
cognitive construct of perceived control. The anxiety is possible. For instance, Chorpita and Barlow’s
literature indicates that parental control leads to a (1998) model involving a switch from a mediating to
child’s perceptions of control (Schneewind, 1995). a moderating model of development can be clarified.
The current model allows for the examination of The proposed model will allow for explication of the
influence of each construct of control as well as style age and developmental level at which cognitions
versus behavior on these perceptions. Again, child change and become more stable.
age impacts this relationship. It could moderate the Several suggestions for improvement in meth-
relationship between parental control and children’s odology have been proposed throughout this review.
perceptions of control. Research has indicated that To summarize, an anxiety diagnosis for both parent
child age has an impact on types of negative cogni- and child should be determined via clinical interviews
tions associated with anxiety (Weems et al., 2001). as opposed to a self-report measure of anxiety. A
Child age can also moderate the impact of parent young child’s ability to complete questionnaires must
anxiety on parental control behaviors. Studies indi- be determined (Alfano Hagekull et al., 2001; 2002). In
cate that parental level of control changes with child addition, if children from a large range of ages are
age (e.g., Freeman and Newland, 2002; Hagekull being assessed, common reporting procedures must
et al., 2001; Smetana, 2000). Finally, the depression be followed by all ages in order to avoid a potential
literature indicates that cognitions stabilize and con- confound. Attention to developmental level is also
tribute more to the development of depression as a critical. Reporting of control should consist of both
child ages (Nolen-Hoeksema et al., 1992). Thus, child self-report and observations. Naturalistic observa-
age and developmental level could serve as a mod- tions are suggested. Control should be measured in
erator in the final path in which a lack of perceived terms of discipline, monitoring, and psychological
control will lead to the development of anxiety control behaviors and styles in order to determine the
(Chorpita and Barlow, 1998; Chorpita et al., 1998). specific effects of each type. Ideally, behavioral cod-
This model incorporates the bidirectional influ- ing systems should be developed in order to corre-
ence of parent–child interactions. As previously sug- spond with the self-reported styles for each type of
gested (Parker, 1993), overcontrolling parental control. Children ranging from early childhood to
behavior may be elicited by children with anxious late adolescence should be studied. A longitudinal
temperaments, or anxious children might report their methodology would provide the best test of Chorpita
parents more negatively. For instance, a longitudinal and Barlow’s model; however, cross-sectional
study found that child manageability during their first research to test specific components of the model or a
year and externalizing behaviors at 2½-years-old more elaborate structural equation model would
predicted later parental control (Hagekull et al., provide a useful starting point in beginning to test the
2001). In addition, anxious children between 2 and proposed model.
6 years of age were observed to be noncompliant and Finally, there are two largely understudied pop-
coercive (Dumas et al., 1995). Children evidently play ulations in the control and anxiety literatures. First,
a role in control exchanges with their parents, and most studies to date have only examined mother–
this role likely changes across development. As indi- child dyads, or have included fathers along with
viduals change, so do the nature of their relationships mothers without consideration of the differential
(Collins and Russell, 1991). For instance, between the effects of parent gender. However, some studies have
ages of 13 and 15, African American adolescents found a difference in discipline styles between moth-
decreased in their beliefs that personal issues were ers and fathers (Conrad and Ho, 2001; Russell et al.,
under parental authority (Smetana, 2000). 1998). While mothers’ influence appears to be greater
130 Ballash et al.

on child outcomes, fathers apparently still have an the extant literature into a ‘‘roadmap’’ for the
important impact. The impact of and difference in relationship between control and anxiety across
paternal control needs to be further explored. development.
Unfortunately, the anxiety literature has largely ig-
nored this issue. Greco and Morris (2002) examined
control in a sample of father–child dyads, but these
data were not compared with mother–child dyads. REFERENCES
Another understudied population, as stated
Alfano, C. A., Beidel, D. C., and Turner, S. M. (2002).
previously, is that of minority ethnic groups. It has Cognition in childhood anxiety: Conceptual, methodological,
been argued that disparities between White and Black and developmental issues. Clinical Psychology Review 22:
samples are due to socioeconomic differences. For 1209–1238.
Anderson, J. C., Williams, S., McGee, R., and Silva, P. A. (1987).
example, one study compared socioeconomically DSM-III disorders in preadolescent children: Prevalence in a
equivalent samples of African American and Euro- large sample from the general population. Archives of General
pean American kindergarten children and their par- Psychiatry 44: 69–76.
Ballash, N. G., Pemble, M. K., Usui, W. M., Buckley, A. F., and
ents, and hostile control was related more strongly Woodruff-Borden, J. (2006). Family functioning, perceived
with conduct problems for European Americans than control, and anxiety: A mediational model. Journal of Anxiety
for African Americans. Overall, however, parents Disorders 20: 486–497.
Barber, B. K. (1992). Family, personality, and adolescent
were similar in their parenting practices (Hill and problem behaviors. Journal of Marriage and the Family 54:
Bush, 2001). Research in this area is inconsistent and 69–79.
very preliminary. More research needs to be con- Barber, B. K. (1996). Parental psychological control: Revisiting a
neglected construct. Child Development 67: 3296–3319.
ducted to determine if control differs in minority Barber, B. K., Olsen, J. E., and Shagle, S. C. (1994). Associations
families, and if it does, how that difference impacts between parental psychological and behavioral control and
the development of anxiety. Specifically, the current youth internalized and externalized behaviors. Child Develop-
ment 65: 1120–1136.
model needs to be tested on these understudied Barlow, D. H. (2001). Anxiety and its disorders: The nature and
populations, including fathers and African Ameri- treatment of anxiety and panic (2nd ed.). Guilford Press, New
cans. This ensures generalizability of the model and York.
Barrett, P. M., Rapee, R. M., Dadds, M. M., and Ryan, S. M.
allows for further investigation of differences across (1996). Family enhancement of cognitive study in anxious and
groups. aggressive children. Journal of Abnormal Child Psychology 24:
Finally, it is important to consider the bidirec- 187–203.
Baumrind, D. (1967). Child care practices anteceding three patterns
tional influence of parent–child interactions. As pre- of preschool behaviour. Genetic Psychology Monongraphs 75:
viously suggested (Parker, 1993), overcontrolling 43–88.
parental behavior may be elicited by child anxious Baumrind, D. (1971). Current patterns of parental authority.
Developmental Psychology Monograph 4: 1–102.
temperament, or anxious children might report their Baumrind, D. (1978). Parental disciplinary patterns and social
parents more negatively. This transactional relation- competence in children. Youth and Society 9: 239–276.
ship and the role of the child in the elicitation of Baumrind, D. (1991). The influence of parenting style on adoles-
cent competence and substance use. Journal of Early Adoles-
parental control need to be explored. cence 11: 56–95.
In conclusion, there are many areas of parental Baumrind, D., and Black, A. E. (1967). Socialization practices
control in the development of anxiety to be further associated with dimensions of competence in preschool boys
and girls. Child Development 38: 291–327.
studied, and they are worthy of significant attention Beidel, D. C., and Turner, S. M. (1997). At risk for anxiety:
in the literature. Control and anxiety are extremely I. Psychopathology in the offspring of anxious parents. Journal
common, they have significant consequences for of American Academy of Child and Adolescent Psychiatry 36:
918–924.
children, and their relationship is well established. Bell-Dolan, D. J., Last, C. G., and Strauss, C. C. (1990). Symptoms
Further, the interactions between anxious parents of anxiety disorders in normal children. Journal of the
and their children have serious ramifications for American Academy of Child and Adolescent Psychiatry 29:
759–765.
families. These behavioral strategies are likely pas- Bernstein, G. A., and Brochardt, C. M. (1991). Anxiety disorders
sed along with anxiety itself, causing an intergen- of childhood and adolescence: A critical review. Journal of
erational concern. A clearer understanding of the the American Academy of Child and Adolescent Psychiatry 30:
519–532.
role of control in the development of anxiety would Boyes, M. C., and Allen, S. G. (1993). Styles of parent–child
enable psychological treatment to target dysfunc- interaction and moral reasoning in adolescence. Merrill-
tional family patterns. The literature has only Palmer Quarterly 39: 551–570.
Brown, B. B., Mounts, N., Lamborn, S. D., and Steinberg, L.
recently begun to approach its study from a more (1993). Parenting practices and peer group affiliation in
conceptual standpoint. This review brings together adolescence. Child Development 64: 467–482.
Control in Anxiety 131

Cantwell, D. P., and Baker, L. (1989). Stability and natural history Dumas, J. E., LaFreniere, P. J., and Serketich, W. J. (1995).
of DSM-III childhood diagnoses. Journal of the American ‘‘Balance of power’’: A transactional analysis of control in
Academy of Child and Adolescent Psychiatry 28: 691–700. mother–child dyads involving socially competent, aggressive,
Chandler, T. A., Wolf, F. M., Cook, B., and Dugovics, A. (1980). and anxious children. Journal of Abnormal Psychology 104:
Parental correlates of locus of control in fifth graders: An 104–113.
attempt at experimentation in the home. Merrill-Palmer Eastburg, M., and Johnson, W. B. (1990). Shyness and perceptions
Quarterly 26: 183–195. of parental behavior. Psychological Reports 66: 915–921.
Chapell, M. S., and Overton, W. F. (1998). Development of logical Fauber, R., Forehand, R., Thomas, A. M., and Wierson, J. (1990).
reasoning in the context of parental style and test anxiety. A mediational model of the impact of marital conflict on
Merrill-Palmer Quarterly 44: 141–156. adolescent adjustment in intact and divorced families: The role
Chorpita, B. F., and Barlow, D. H. (1998). The development of of disrupted parenting. Child Development 61: 1112–1123.
anxiety: The role of control in the early environment. Fincham, F. D., and Cain, K. M. (1986). Learned helplessness in
Psychological Bulletin 124: 3–21. humans: A developmental analysis. Developmental Review 6:
Chorpita, B. F., Brown, T. A., and Barlow, D. H. (1998). Perceived 301–333.
control as a mediator of family environment in etiological Freeman, H. S., and Newland, L. A (2002). Family transitions
models of childhood anxiety. Behavior Therapy 29: 457–476. during the adolescent transition: Implications for parenting.
Cobham, V. E., Dadds, M. R., and Spence, S. H. (1999). Anxious Adolescence 37: 457–475.
children and their parents: What do they expect?. Journal of Gordon, D., Nowicki, S. Jr., and Wichern, F. (1981). Observed
Clinical Child Psychology 28: 220–231. maternal and child behaviors in a dependency producing task
Colder, C. R., Lochman, J. E., and Wells, K. C. (1997). The as a function of children’s locus of control orientation. Merrill-
moderating effects of children’s fear and activity level on Palmer Quarterly 27: 43–51.
relations between parenting practices and childhood symp- Greco, L. A., and Morris, T. L. (2002). Paternal child-rearing style
tomatology. Journal of Abnormal Child Psychology 25: 251– and child social anxiety: Investigation of child perceptions and
263. actual father behavior. Journal of Psychopathology and
Cole, D. A., and Turner, J. E. Jr. (1993). Models of cognitive Behavioral Assessment 24: 259–267.
mediation and moderation in child depression. Journal of Grolnick, W. S., and Slowiaczek, M. L. (1994). Parents’ involve-
Abnormal Psychology 102: 271–281. ment in children’s schooling: A multidimensional conceptu-
Collins, W. A., and Russell, G. (1991). Mother–child and father– alization and motivational model. Child Development 65:
child relationships in middle childhood and adolescence: A 237–252.
developmental analysis. Developmental Review 11: 99–136. Grotevant, H. D., and Cooper, C. R. (1986). Individuation in
Connell, J. P. (1985). A new multidimensional measure of family relationships: A perspective on individual differences in
children’s perceptions of control. Child Development 56: the development of identity and role-taking skill in adoles-
1018–1041. cence. Human Development 29: 82–100.
Conrad, G., and Ho, R. (2001). Differential parenting styles for Hagekull, B., Bohlin, G., and Hammarberg, A. (2001). The role of
fathers and mothers: Differential treatment for sons and parental perceived control in child development: A longitudi-
daughters. Australian Journal of Psychology 53: 29–35. nal study. International Journal of Behavioral Development 25:
Cummings, E. M., Davies, P. T., and Campbell, S. B. (2000). 429–437.
Developmental psychopathology and family process: Theory, Hammen, C., Adrian, C., and Hiroto, D. (1988). A longitudinal
research, and clinical implications, Guilford Press, New York. test of the attributional vulnerability model in children at
Dadds, M. R., Barrett, P. M., Rapee, R. M., and Ryan, S. (1996). risk for depression. British Journal of Clinical Psychology 27:
Family process and child anxiety and aggression: An obser- 37–46.
vational analysis. Journal of Abnormal Child Psychology 24: Hart, C. H., DeWolf, M., Wozniak, P., and Burts, D. C. (1992).
715–734. Maternal and paternal disciplinary styles: Relations with
Dadds, M. R., Holland, D. E., Laurens, K. P., Mullins, M., preschoolers’ playground behavioral orientations and peer
Barrett, P. M., and Spence, S. H. (1999). Early intervention status. Child Development 63: 879–892.
and prevention of anxiety disorders in children: Results at 2- Hart, C. H., Nelson, D. A., Robinson, C. C., Olsen, S. F., and
year follow-up. Journal of Consulting and Clinical Psychology McNeilly-Choque, M. K. (1998). Overt and relational
67: 145–150. aggression in Russian nursery-school-age children: Parenting
Darling, N., and Steinberg, L. (1993). Parenting style as context: style and marital linkages. Developmental Psychology 34:
An integrative model. Psychological Bulletin 113: 487–496. 687–697.
Davis, W. L., and Phares, J. (1969). Parental antecedents of Hess, R. D., and McDevitt, T. M. (1984). Some cognitive
internal–external control of reinforcement. Psychological consequences of maternal intervention techniques: A longitu-
Reports 24: 427–436. dinal study. Child Development 55: 2017–2030.
Dishion, T. J., and McMahon, R. J. (1998). Parental monitoring Hettema, J. M., Neale, M. C., and Kendler, K. S. (2001). A review
and the prevention of child and adolescent problem behavior: and meta-analysis of the genetic epidemiology of anxiety
A conceptual analysis and empirical formulation. Clinical disorders. American Journal of Psychiatry 158: 1568–1578.
Child and Family Psychology Review 1: 61–75. Hill, N. E. (1995). The relationship between family environment
Dishion, T. J., Patterson, G. R., Stoolmiller, M., and Skinner, M. and parenting style: A preliminary study of African American
L. (1991). Family, school, and behavioral antecedents to early families. Journal of Black Psychology 21: 408–423.
adolescent involvement with antisocial peers. Developmental Hill, N. E., and Bush, K. R. (2001). Relationships between
Psychology 27: 172–180. parenting environment and children’s mental health among
Dornsbusch, S. M., Carlsmith, J. M., Bushwall, S. J., Ritter, P. L., African American and European American mothers and
Leiderman, H., Hastorf, A. H., and Gross, R. T. (1985). Single children. Journal of Marriage and Family 63: 954–966.
parents, extended households, and the control of adolescents. Hudson, J. L., and Rapee, R. M. (2001). Parent–child interactions
Child Development 56: 326–341. and anxiety disorders: An observational study. Behavior
Dornsbusch, S. M., Ritter, P. L., Leiderman, P. H., Roberts, D. F., Research and Therapy 39: 1411–1427.
and Fraleigh, M. J. (1987). The relation of parenting style to Kashani, J. H., and Orvaschel, H. (1990). A community study of
adolescent school performance. Child Development 58: 1244– anxiety in children and adolescents. American Journal of
1257. Psychiatry 147: 313–318.
132 Ballash et al.

Keller, M. B., Lavori, P. W., Wunder, J., Beardslee, W. R., Parker, G. (1993). Parental overprotection: A risk factor in
Schwartz, P., and Roth, T. (1992). Chronic course of psychosocial development, Grune and Stratton, New York.
anxiety disorders in children and adolescents. Journal of the Pettit, G. S., Bates, J. E., and Dodge, K. A. (1993). Family
American Academy of Child and Adolescent Psychiatry 31: interaction patterns and children’s conduct problems at home
595–599. and school: A longitudinal perspective. School Psychology
Kendler, K. S., Neale, M. C., Kessler, R. C., Heath, A. C., and Review 22: 403–420.
Eaves, L. J. (1992). The genetic epidemiology of phobias in Pettit, G. S., Laird, R. D., Dodge, K. A., Bates, J. E., and Criss, M.
women. Archives of General Psychiatry 49: 273–281. M. (2001). Antecedents and behavior-problem outcomes of
Kessler, R. C., McGonagle, K. A., Zhao, S., Nelson, C. B., parental monitoring and psychological control in early ado-
Hughes, M., and Eshleman, S. et al., (1994). Lifetime and 12- lescence. Child Development 72: 583–598.
month prevalence of DSM-III-R psychiatric disorders in the Rapee, R. M. (1997). Potential role of childrearing practices in the
United States: Results from the National Comorbidity Survey. development of anxiety and depression. Clinical Psychology
Archives of General Psychiatry 51: 8–19. Review 17: 47–67.
Krohne, H. W., and Hock, M. (1991). Relationships between Rholes, W. S., Blackwell, J., Jordan, C., and Walters, C. (1980). A
restrictive mother–child interactions and anxiety of the child. developmental study of learned helplessness. Developmental
Anxiety Research 4: 109–124. Psychology 16: 616–624.
Laraia, M. T., Stuart, G. W., Frye, L. H., Lydiard, R. B., and Robins, L. N., Helzer, J. E., Weissman, M. M., Orvaschel, H.,
Ballenger, J. C. (1994). Childhood environment of women Gruenberg, E., and Burke, J. et al., (1984). Lifetime
having panic disorder with agoraphobia. Journal of Anxiety prevalence of specific psychiatric disorders in three sites.
Disorders 8: 1–17. Archives of General Psychiatry 41: 949.
Last, C. G., Hersen, M., Kazdin, A. E., Orvaschel, H., and Robinson, C. C., Mandleco, B., Olsen, S. F., and Hart, C. H.
Perrin, S. (1991). Anxiety disorders in children and their (1995). Authoritative, authoritarian, and permissive parenting
families. Archives of General Psychiatry 48: 928–934. practices: Development of a new measure. Psychological
Last, C. G., and Strauss, C. C. (1990). School refusal in anxiety- Reports 77: 819–830.
disordered children and adolescents. Journal of the American Russell, A., Aloa, V., Feder, T., Glover, A., Miller, H., and
Academy of Child and Adolescent Psychiatry 2: 31–35. Palmer, G. (1998). Sex-based differences in parenting styles in
Lenane, M. C., Swedo, S. E., Leonard, H., and Pauls, D. L. et al., a sample with preschool children. Australian Journal of
(1990). Psychiatric disorders in first degree relatives of children Psychology 50: 89–99.
and adolescents with obsessive compulsive disorder. Journal of Schaefer, E. S. (1965). A configurational analysis of children’s
the American Academy of Child and Adolescent Psychiatry 29: reports of parent behavior. Journal of Consulting Psychology
407–412. 29: 552–557.
Leon, C. A., and Leon, A. (1990). Panic disorder and parental Schneewind, K. A. (1995). Impact of family processes on control
bonding. Psychiatric Annals 20: 503–508. beliefs. In Bandura, A. (Ed.), Self-efficacy in changing societies
Litovsky, V. G., and Dusek, J. B. (1985). Perceptions of child (pp. 114–148). New York: Cambridge University Press.
rearing and self-concept development during the early adoles- Siqueland, L., Kendall, P. C., and Steinberg, L. (1996). Anxiety in
cent years. Journal of Youth and Adolescence 14: 373–388. children: Perceived family environments and observed family
Locke, L. M., and Prinz, R. J. (2002). Measurement of parental interaction. Journal of Clinical Child Psychology 25: 225–237.
discipline and nurturance. Clinical Psychology Review 22: 895– Skinner, E. A. (1990). Age differences in the dimensions of
929. perceived control during middle childhood: Implications for
Manassis, K., and Bradley, S. J. (1994). The development of developmental conceptualizations and research. Child Devel-
childhood anxiety disorders: Toward an integrated model. opment 61: 1882–1890.
Journal of Applied Developmental Psychology 15: 345–366. Skinner, E. A. (1991). Development and perceived control: A
Masia, C. L., and Morris, T. L. (1998). Parental factors associated dynamic model of action in context. In Gunnar, M., and
with social anxiety: Methodological limitations and sugges- Sroufe, L. A. (Eds.), Minnesota symposium on child develop-
tions for integrated behavioral research. Clinical Psychology: ment (Vol. XXIII). Erlbaum, Hillsdale, NJ, pp. 167–216.
Science and Practice 5: 211–228. Skinner, E. A. (1996). A guide to constructs of control. Journal of
Mason, C. A., Cauce, A. M., Gonzales, N., and Hiraga, Y. (1996). Personality and Social Psychology 71: 549–570.
Neither too sweet nor too sour: Problems peers, maternal Spence, S. H. (2001). Prevention Strategies. In Vasey, M. W., and
control, and problem behavior in African American adoles- Dadds, M. R. (Eds.), The developmental psychopathology of
cents. Child Development 67: 2115–2130. anxiety New York: Oxford University Press.
Messer, S. C., and Beidel, D. C. (1994). Psychosocial correlates of Smetana, J. G. (1995). Parenting styles and conceptions of parental
childhood anxiety disorders. Journal of the American Academy authority during adolescence. Child Development 66: 299–316.
of Child and Adolescent Psychiatry 33: 975–983. Smetana, J. G. (2000). Middle-class african american adolescents’
Moore, P. S., Whaley, S. E., and Sigman, M. (2004). Interactions and parents’ conceptions of parental authority and parenting
between mothers and children: Impacts of maternal and child practices: A longitudinal investigation. Child Development 71:
anxiety. Journal of Abnormal Psychology 113: 471–476. 1672–1686.
Nolen-Hoeksema, S., Girgus, J. S., and Seligman, M. E. P. (1992). Smetana, J. G., and Daddis, C. (2002). Domain-specific anteced-
Predictors and consequences of childhood depressive symp- ents of parental psychological control and monitoring: The
toms: A 5-year longitudinal study. Journal of Abnormal role of parenting beliefs and practices. Child Development 73:
Psychology 101: 405–422. 563–580.
Nowicki, S Jr., and Strickland, B. R. (1973). A locus of control Stark, K. D., Humphrey, L. L., Crook, K., and Lewis, K. (1990).
scale for children. Journal of Consulting and Clinical Psychol- Perceived family environments of depressed and anxious
ogy 40: 148–154. children: Child’s and maternal figure’s perspectives. Journal
Paquette, D., Bolte, C., Turcotte, G., Dubeau, D., and of Abnormal Child Psychology 18: 527–547.
Bouchard, C. (2000). A new typology of fathering: Defining Steinberg, L. (1987). Single parents, stepparents, and the suscep-
and associated variables. Infant and Child Development 9: tibility of adolescents to antisocial peer pressure. Child
213–230. Development 58: 269–275.
Control in Anxiety 133

Steinberg, L., Elmen, J. D., and Mounts, N. S. (1989). Journal of the American Academy of Child and Adolescent
Authoritative parenting, psychosocial maturity, and aca- Psychiatry 44: 656–663.
demic success among adolescents. Child Development 60: Weems, C. F., Hammond-Lawrence, K., Silverman, W. K., and
1424–1436. Ginsburg, G. S. (1998). Testing the utility of the anxiety
Stemberger, R. T., Turner, S. M., Beidel, D. C., and Calhoun, K. S. sensitivity construct in children and adolescents referred for
(1995). Social phobia: An analysis of possible developmental anxiety disorders. Journal of Clinical Child Psychology 27:
factors. Journal of Abnormal Psychology 104: 526–531. 69–77.
Tambs, K., and Moum, T. (1993). Low genetic effect and age- Weems, C. F., and Stickle, T. R. (2005). Anxiety disorders in
specific family effect for symptoms of anxiety and depression childhood: Casting a nomological net. Clinical Child and
in nuclear families, halfsibs, and twins. Journal of Affective Family Psychology Review 8: 107–134.
Disorders 27: 183–195. Weiss, D. D., and Last, C. G. (2001). Developmental variations in
Torgenson, S. (1983). Genetic factors in anxiety disorders. Archives the prevalence and manifestation of anxiety disorders. In
of General Psychiatry 40: 1085–1089. Vasey, M. W., and Dadds, M. R. (Eds.), The developmental
Turner, J. E. Jr., and Cole, D. A. (1994). Developmental differences psychopathology of anxiety (pp. 27–42). New York: Oxford
in cognitive diatheses for child depression. Journal of Abnor- University Press, Inc.
mal Child Psychology 22: 15–32. Weissman, M. M., Leckman, J. F., Merikangas, K. R., Gammon,
Turner, S. M., Beidel, D. C., and Costello, A. (1987). Psychopa- G. D., and Prusoff, B. A (1984). Depression and anxiety
thology in the offspring of anxiety disorder patients. Journal of disorders in parents and children: Results for the Yale family
Consulting and Clinical Psychology 55: 229–235. study. Archives of General Psychiatry 41: 845–852.
Turner, S. M., Beidel, D. C., Roberson-Nay, R., and Tervo, K. Westenberg, P. M., Drewes, M. J., Siebelink, B. M., and Treffers,
(2003). Parenting behaviors in parents with anxiety disorders. P. D. A. (2004). A developmental analysis of self-reported
Behavior Research and Therapy 41: 541–554. fears in late childhood through mid-adolescence: Social-
Vasey, M. W. (1993). Development and cognition in childhood evaluative fears on the rise?. Journal of Child Psychoogy and
anxiety: The example of worry. In Ollendick, T. H., and Prinz, Psychiatry and Allied Disciplines 45: 481–496.
R. J. (Eds.), Advances in clinical child psychology (Vol. 15). Whaley, S. E., Pinto, A., and Sigman, M. (1999). Characterizing
Plenum, New York, pp. 1–39. interactions between anxious mothers and their children.
Vasey, M. W., and Dadds, M. R. (2001). An introduction to the Journal of Consulting and Clinical Psychology 67: 826–836.
developmental psychopathology of anxiety. In Vasey, M. W., Wood, J. J., McLeod, B. D., Sigman, M., Hwang, W. C., and Chu,
and Dadds, M. R. (Eds.), The developmental psychopathology of B. C. (2003). Parenting and childhood anxiety: Theory,
anxiety (pp. 3–26)). New York: Oxford University Press, Inc. empirical findings, and future directions. Journal of Child
Warren, S., and Sroufe, L. A. (2004). Developmental Issues. In Psychology and Psychiatry 44: 134–151.
Ollendick, T., and Marsh, J. (Eds.), Phobic and anxiety Woodruff-Borden, J., Morrow, C., Bourland, S., and Cambron, S.
disorders in children and adolescents: A clinician’s guide to (2002). The behavior of anxious parents: Examining mecha-
effective psychosocial and pharmacological intervention (pp. 92– nisms of transmission of anxiety from parent to child. Journal
115). New York: Oxford University Press. of Clinical Child and Adolescent Psychology 31: 364–374.
Weems, C. F., Berman, S. L., Silverman, W. K., and Saavedra, L. Yee, D. K., and Flanagan, C. (1985). Family environments and
M. (2001). Cognitive errors in youth with anxiety disorders: self-consciousness in early adolescence. Journal of Early
The linkages between negative cognitive errors and anxious Adolescence 5: 59–68.
symptoms. Cognitive Therapy and Research 25: 559–575.
Weems, C. F., and Costa, N. M. (2005). Developmental differences
in the expression of childhood anxiety symptoms and fears.