Child Personality and Parental Behavior As Moderators of Problem Behavior: Variable-And Person-Centered Approaches

Developmental Psychology Copyright 2004 by the American Psychological Association
2004, Vol. 40, No. 6, 1028 –1046 0012-1649/04/$12.00 DOI: 10.1037/0012-1649.40.6.1028
Child Personality and Parental Behavior as Moderators of Problem

Behavior: Variable- and Person-Centered Approaches
Karla G. Van Leeuwen, Ivan Mervielde, Caroline Braet, and Guy Bosmans
Ghent University
Parenting ⫻ Child Personality interactions in predicting child externalizing and internalizing behavior
were investigated in a variable-centered study and a person-centered study. The variable-centered study
used data from a 3-year longitudinal study of 600 children 7 to 15 years old at Time 1 and 512 children
10 to 18 years old at Time 2. Parents rated child personality (five factor model), negative control, positive
parenting, and child problem behavior, whereas children rated parental behavior. Hierarchical moderated
regression analyses showed significant Parenting ⫻ Child Personality (benevolence and conscientious-
ness) interactions, principally for externalizing behavior. The interactions were largely replicable across
informants and across time. The person-centered study, which classified participants into 3 types, showed
that negative parental control was more related to externalizing behavior for undercontrollers than for
resilients. Negative parental control enhanced internalizing behavior for overcontrollers.
From an ecological or contextual perspective, the child is nested correlations with problem behavior. As O’Connor and Dvorak
in a complex network of interconnected systems (cf. Belsky, 1984; (2001) pointed out, bivariate relationships may be hiding associ-
Bronfenbrenner, 1986; Dishion, French, & Patterson, 1995; ations that are conditional and not universal. For example, negative
Sameroff, 2000), and therefore multiple sources may contribute to parental control is related to externalizing behavior, but this overall
the development of (problem) behavior in children. Besides indi- relation can hide the fact that the relationship holds in particular
vidual risk and protective factors such as intelligence, neuropsy- for children with a difficult temperament but not for resilient
chological deficits, and temperament, variables such as parenting, children. Hence, ignoring this interaction may lead to unwarranted
family climate, the marital relationship, relationships with peers, conclusions about the detrimental effects of parenting behavior. In
and contextual factors (e.g., neighborhood, socioeconomic status) the present study, we aimed to examine the role of interactions
have been taken into account as determinants of problem behavior between parenting and child personality variables as predictors of
(Deater-Deckard, 2001; Deković, 1999; Frosch & Mangelsdorf, child and adolescent externalizing and internalizing behavior. The
2001; Harris, 1998; Jessor, Van Den Bos, Vanderryn, Costa, & study is innovative because it is one of the first to use the five
Turbin, 1995; Stouthamer-Loeber et al., 1993). Deater-Deckard, factor model to assess the child’s personality, adopting a longitu-
Dodge, Bates, and Pettit (1998) demonstrated that child factors dinal as well as a cross-sectional perspective to predict the effect
uniquely explain up to 19% of the variance in externalizing be- of interactions. Furthermore, in order to ensure sufficient power
havior, peer-related factors explain up to 13%, parenting factors for the tests of interaction effects, the study was conducted with a
explain up to 6%, and sociocultural factors explain up to 4%. moderately large sample. In addition, it is the first study to exam-
Although the main effects of child personality and parental ine Child Personality ⫻ Parenting interactions from a person-
behavior on child problem behavior have been extensively docu- centered approach, searching for reliable interaction effects be-
mented in past research, some of which is reviewed briefly below, tween parenting and the resilient, undercontrolled, and
relatively few studies have addressed the combined or interactive overcontrolled personality types.
effects of personality and parental behavior as determinants of
internalizing and externalizing behavior (Barber, 1992; Collins,
Personality and Temperament as Predictors of Child
Maccoby, Steinberg, Hetherington, & Bornstein, 2000; O’Connor
& Dvorak, 2001). However, ignoring personality– environment
Problem Behavior
interactions and considering only main effects can lead to spurious Both temperament and personality refer to individual differ-
ences that evince some stability over time, presumably because
they have a biological basis (Eisenberg, Fabes, Guthrie, & Reiser,
Karla G. Van Leeuwen, Ivan Mervielde, Caroline Braet, and Guy 2000). Temperament refers to “the constitutionally based individ-
Bosmans, Department of Developmental, Personality, and Social Psychol- ual differences in emotional, motor and attentional reactivity and
ogy, Ghent University, Ghent, Belgium. self-regulation” (Rothbart & Bates, 1998, p. 109), whereas the
This study was supported by Grants bel96/32 and PBO99A/48 –50/75 construct of personality refers to “individual differences in the
from the Flemish Community. The study was a part of Karla G. Van
tendency to behave, think, and feel in certain consistent ways”
Leeuwen’s doctoral research, which was conducted under the supervision
of Ivan Mervielde.
(Caspi, 1998, p. 312). Temperamental differences in infancy are
Correspondence concerning this article should be addressed to Karla G. assumed to be the precursors of later personality differences (Caspi
Van Leeuwen, Department of Developmental, Personality, and Social & Silva, 1995; Eisenberg et al., 2000); however, they include only
Psychology, Ghent University, H. Dunantlaan 2, B-9000 Ghent, Belgium. a subset of personality differences in late childhood and adulthood
E-mail: karla.vanleeuwen@ugent.be (Shiner & Caspi, 2003). Several reviews have linked the traditional
1028
MODERATORS OF PROBLEM BEHAVIOR 1029
dimensions of temperament to the five factor model (Caspi, 1998; sions on the one hand and environmental variables (e.g., parenting)
Mervielde & Asendorpf, 2000; Shiner, 1998; Shiner & Caspi, on the other hand.
2003). Recent developments in personality research show a grow- Several researchers studied the replicability of three personality
ing consensus about how individual differences in children’s per- types: resilients, overcontrollers, and undercontrollers (cf. Asen-
sonality can be mapped in a comprehensive taxonomic system, dorpf, Caspi, & Hofstee, 2002). In terms of five factor model
usually referred to as the Big Five personality factors or the five scores, the three types can be described as follows: (a) Resilients
factor model of personality. The five broadband dimensions form- receive average scores on the characteristics of benevolence, ex-
ing the top level of this hierarchical system, Neuroticism, Extra- traversion, openness to experience, and conscientiousness and
version, Openness to Experience, Agreeableness, and Conscien- score below the mean on neuroticism; (b) overcontrollers receive
tiousness, are derived from lexical studies of the personality- high ratings on neuroticism and low ratings on extraversion; and
descriptive language of adults and from the analysis of free (c) undercontrollers receive scores below the mean on agreeable-
parental descriptions of child personality (Kohnstamm, Halverson, ness and conscientiousness. So far, attempts to replicate the three
Mervielde, & Havill, 1998). types across heterogeneous samples, time, informants, methods,
Main-effect-type studies show relations between temperament and variables for clustering have produced mixed results (for an
and problem behavior, with negative emotionality as a general risk overview, see Asendorpf, Borkenau, Ostendorf, & van Aken,
factor (Eisenberg et al., 2000). Externalizing behavior, which 2001, and Asendorpf et al., 2002). This has led to the conclusion
includes hyperactivity, attention problems, antisocial behavior, and that although the three prototypes are frequently recovered, they
conduct disorder, has been related to lack of control (Caspi, Henry, are not necessarily the prototypes that best describe each particular
McGee, Moffitt, & Silva, 1995; Eisenberg et al., 2000). Other sample (Asendorpf, 2003; Van Leeuwen, De Fruyt, & Mervielde,
studies have found evidence for an association between personality 2004).
and delinquency (Krueger et al., 1994; Ruchkin, Eisemann & The use of personality types in developmental psychology has
Hägglöf, 1999). Internalizing behavior, indicating anxiety and several benefits. Instead of considering different features indepen-
depression, has been associated with behavioral inhibition (Eisen- dently, types specify a more parsimonious configuration or com-
berg et al., 2000), flat affect, and passivity (Caspi et al., 1995). bination of characteristics. As such, types facilitate communication
Finch and Graziano (2001) showed that the influence of temper- among researchers and clinicians interested in describing person-
ament on depression in adolescents is entirely mediated by the ality in applied settings and in identifying types at risk for devel-
personality dimensions of agreeableness, extraversion, and oping psychopathology (Asendorpf et al., 2001; Costa, Herbst,
neuroticism. McCrae, Samuels, & Ozer, 2002). Types have utility as predictors
of child and adolescent (mal)adaptive development. Several stud-
Child Personality From Variable- and Person-Centered ies have provided evidence for the link between externalizing
Perspectives behavior and the undercontrolled type, and between internalizing
behavior and the overcontrolled type (Asendorpf et al., 2001; De
Mervielde and Asendorpf (2000) identified and compared two Fruyt, Mervielde, & Van Leeuwen, 2002; Van Leeuwen et al.,
empirical approaches to studying individual differences in chil- 2004).
dren’s dispositions to behave, think, and feel. The variable-
centered approach distinguishes replicable broad categories of Parental Behavior and Child Problem Behavior
variables across individuals, such as the dimensions of the five
factor model of personality. Variable-centered research assesses Although they are often used as interchangeable concepts, it
the correlational structure of the variables across persons within a should be noted that parental behavior is distinct from parenting
particular population. The person-centered approach studies styles as defined by, for example, the typology of Maccoby and
“types,” identifying clusters of individuals with similar personality Martin (1983), in which authoritarian, authoritative, permissive,
patterns. The person-centered approach delineates typical config- and indifferent parenting are based on two dimensions, demand-
urations of variables within the person and hence assesses the ingness and responsiveness. Parenting styles can be regarded as the
common within-person structure of variables. Mervielde and general context, or the climate, in which the more specific parent-
Asendorpf (2000) attempted to clarify the distinction between the ing practices or behaviors are expressed (Darling & Steinberg,
two approaches by referring to a card game as an analogue. The 1993).
variable-centered approach delineates the features (e.g., color, suit, Most studies on risk and protective factors for the development
and values) of the cards that are used in the game, whereas the of problem behavior have focused on externalizing behavior, prob-
person-centered approach studies the typical hands that are re- ably because it is more visible and has more negative social
ceived by the players. From this analogue it is clear that variable- consequences (Deković, 1999). The relationship of parenting to
and person-centered methodologies should be conceived not as child externalizing behavior has been documented both in clinical
competing approaches but rather as complementary methods that and nonreferred samples (Belsky, Hsieh, & Crnic, 1998). Lack of
together provide a more comprehensive picture of how individual parental involvement or poor acceptance or responsiveness, lack of
differences can and should be addressed. The detection of reliable supervision or poor parental monitoring, harsh and inconsistent
and replicable types depends on the identification of adequate punishment, and insufficient rewarding of positive behavior have
dimensions to represent individual differences. Hence, the growing been identified as predictors of externalizing behavior (Deater-
consensus about the utility of the five factor model as a model to Deckard & Dodge, 1997; Forehand, Miller, Dutra, & Chance,
represent individual differences from childhood to adulthood sets 1997; Haapasalo & Tremblay, 1994; Loeber & Dishion, 1983;
the stage not only for the discovery of reliable personality types Loeber & Stouthamer-Loeber, 1986; Patterson & Stouthamer-
but also for the study of the interaction between types and dimen- Loeber, 1984; Rothbaum & Weisz, 1994; Ruchkin et al., 1999;
1030 VAN LEEUWEN, MERVIELDE, BRAET, AND BOSMANS
Stormshak, Bierman, McMahon, Lengua, & Conduct Problems solely on a detrimental form of parenting; (c) investigating both
Prevention Research Group, 2000; Wakschlag & Hans, 1999; internalizing and externalizing behavior as the outcome variables,
Weiss, Dodge, Bates, & Pettit, 1992). Parenting has also been whereas past research has focused predominantly on externalizing
associated with child internalizing behavior, such as anxiety behavior; (d) using a large population sample including both boys
(Gruner, Muris, & Merckelbach, 1999; Siqueland, Kendall, & and girls; and (e) adopting a longitudinal design as well as a
Steinberg, 1996) and depression (Muris, Schmidt, Lambrichs, & cross-sectional design for assessment of the effects of Parenting ⫻
Meesters, 2001; Richter, 1994). Child Personality interactions.
Child Personality and Parental Behavior as Moderators of The Current Research

Child Problem Behavior Interactions between parenting and child personality as predic-
According to Thomas and Chess’s (1977) goodness-of-fit- tors of child and adolescent externalizing and internalizing behav-
theory, the development of problem behavior has its origins in the ior were investigated in two studies. In the first study, the Parent-
child’s temperament and in its interaction with the socializing ing ⫻ Personality interaction was examined from a variable-
environment. When there is a mismatch between a difficult tem- centered approach. In order to ensure a comprehensive test of the
perament and parenting practices, behavioral disturbances may role of various personality traits as moderators of problem behav-
develop. In theory, temperament does not lead to behavioral prob- ior, a broadband five factor model instrument was used to assess
lems by itself; it only has an effect in conjunction with particular children’s personality. Moreover, to guarantee sufficient power to
environments (Bates, Pettit, Dodge & Ridge, 1998). At present, detect interaction effects, the hypotheses were verified on a large
there is some empirical evidence underscoring the importance of sample and checked for cross-rater and cross-time stability. In line
interactions between child characteristics and parenting in the with the variable-centered methodology, both parenting and per-
prediction of child problem behavior. These studies vary in (a) sonality were entered as (continuous) quantitative variables pre-
design, that is, cross-sectional versus longitudinal (e.g., Bates et dicting child and adolescent problem behavior. As the variable-
al., 1998; Belsky et al., 1998; Rubin, Burgess, Dwyer, & Hastings, centered approach emphasizes interactions between variables and
2003; Stoolmiller, 2001) versus experimental (e.g., Anderson, does not take into account the within-person personality structure,
Lytton, & Romney, 1986); (b) sample composition, that is, non- the first study was supplemented with a second study that adopted
referred versus referred (e.g., Anderson et al., 1986; Colder, Loch- a person-centered approach. In the second study, which used a
man, & Wells, 1997); (c) age, that is, preschool (e.g., Paterson & classic analysis of variance (ANOVA) design, we looked for
Sanson, 1999; Rubin et al., 2003), school-age (e.g., Lengua, interaction effects between the personality types of resilient, un-
Wolchik, Sandler, & West, 2000; Stoolmiller, 2001; Wootton, dercontrolled, and overcontrolled (derived from clustering partic-
Frick, Shelton, & Silverthorn,1997), and adolescent (e.g., Carlo, ipants on the basis of their five factor personality profiles) and
Roesch, & Melby,1998; Olweus, 1980); and (d) gender, that is, discrete categories of parents with high and low levels of positive
some studies include only boys (e.g., Anderson et al., 1986; Belsky parenting and negative parental control.
et al., 1998; Colder et al., 1997; Olweus, 1980; Stoolmiller, 2001).
The studies also differ in how parental behavior and child temper- Study 1
ament are assessed. Most studies linking parenting and child
characteristics have been restricted to the assessment of the child’s Method
temperament, leading to the general conclusion that children with
Participants
a difficult temperament are particularly vulnerable when also ex-
posed to punitive parenting or negative control. Few studies have Study 1 was based on data from a longitudinal study that investigated
examined personality– environment interactions from the perspec- parenting, parental and child personality characteristics, and children’s
tive of the five factor model. O’Connor and Dvorak (2001) showed problem behavior at two assessment periods separated by a 3-year interval.
that parental behaviors matter only for some kinds of children and Participants were parents (both mothers and fathers) and one nonreferred
not for others and hence that specific combinations of personality child per family. At Time 1, 600 families were included (mothers, N ⫽
596; fathers, N ⫽ 533). The target children, 281 boys and 319 girls, had a
characteristics and parenting variables operate as protective or risk
mean age of 10.9 years (SD ⫽ 1.8; range ⫽ 7–15). The age distribution was
factors. Prinzie et al. (2003) also used the five factor model in their as follows: 1.3% of the children were 7 years old, 11.5% were 8 years old,
investigation of interaction effects between personality and par- 12.5% were 9 years old, 15.8% were 10 years old, 11.8% were 11 years
enting on externalizing behavior. They showed (a) that benevo- old, 22.3% were 12 years old, 19.7% were 13 years old, 4.5% were 14
lence functions as a protective factor, buffering the effect of years old, and 0.5% were 15 years old. Of the families, 9.8% were not the
maternal or paternal overreactivity, that is, the tendency of parents original family, and 8.2% included a single parent. The mothers’ mean age
to respond with irritation and/or anger to the problematic behavior was 38.6 years (SD ⫽ 4.7; range ⫽ 20 – 68), and the fathers’ was 40.6 years
of their children, and (b) that coercive parental behavior is more (SD ⫽ 5.0; range ⫽ 26 – 63). Both mothers and fathers had representative
strongly related to externalizing problems for children low on levels of education: The highest level of education was elementary school
conscientiousness. In the present study, we extended previous for 12.9% of the mothers and 7.9% of the fathers; 48.5% of the mothers and
49.5% of the fathers completed secondary education; 29.7% of the mothers
research on child personality and parental behavior as moderators
and 29.4% of the fathers finished higher education; and 8.7% of the
of child problem behavior by (a) adopting the five factor model for mothers and 13.3% of the fathers obtained a university degree. Of the
the assessment of child personality in the prediction of problem mothers, 75.2% were employed, whereas 95.3% of the fathers were
behavior, instead of examining temperament; (b) including both a employed.
more affective and positive component of parenting as well as a At Time 2, 512 families, or about 85% of the families, continued the
control dimension, whereas most previous studies concentrated collaboration (mothers, N ⫽ 501; fathers, N ⫽ 443). The remaining 244
boys and 268 girls had a mean age of 13.9 years (SD ⫽ 1.8; range ⫽ dimensions consistently emerged: Positive Parenting (consisting of the
10 –18). The age distribution was as follows: 0.8% of the children were 10 Positive Parental Behavior, Teaching Rules, and Autonomy scales), and
years old, 10.6% were 11 years old, 13.7% were 12 years old, 16.8% were Negative Control (consisting of the Discipline, Ignoring of Unwanted
13 years old, 13.3% were 14 years old, 22.3% were 15 years old, 18.0% Behavior, and Harsh Punishment scales). These factors are analogous to
were 16 years old, 3.9% were 17 years old, and 0.6% were 18 years old. two dimensions of parenting that are regularly mentioned in the literature
The ratios of boys to girls, ␹2(1, N ⫽ 1,112) ⫽ 0.07, p ⬎ .05, and of (Gallagher, 2002), the first describing the affective nature of the parent–
mothers to fathers, ␹2(1, N ⫽ 2,073) ⫽ 0.03, p ⬎ .05, were equal across the child relationship, as indicated by involvement and providing support
two assessment times. Chi-square analyses showed no significant differ- (Warmth or Responsiveness) and the second referring to efforts of parents
ences between the Time 1 and Time 2 samples for family characteristics, to influence their child’s behavior, such as setting and enforcing standards
␹2(2, N ⫽ 1,112) ⫽ 2.91, p ⬎ .05; social indices for mothers, ␹2(5, N ⫽ of behavior (Control or Demandingness) (Maccoby & Martin, 1983). Table
1,102) ⫽ 2.24, p ⬎ .05, and fathers, ␹2(5, N ⫽ 998) ⫽ 1.42, p ⬎ .05; and 1 shows factor loadings, percentages of explained variance, and Cron-
employment status for mothers, ␹2(2, N ⫽ 1,099) ⫽ 2.36, p ⬎ .05, and bach’s alphas for the two dimensions. The intercorrelations between the
fathers, ␹2(2, N ⫽ 997) ⫽ 0.24, p ⬎ .05. Hence, it can be concluded that two factors were .03 and .13 for Time 1 and Time 2, respectively,
the Time 2 sample characteristics did not differ from the Time 1 sample indicating the independenc of the two parenting dimensions. The correla-
characteristics as a consequence of attrition. tions between the ratings provided by parents and children on the two
parenting dimensions were all positive and significant at p ⬍ .01, ranging
Measures from .19 to .38 for Time 1 and ranging from .21 to .42 for Time 2.
Child personality. The Hierarchical Personality Inventory for Children
Parental behavior. The Ghent Parental Behavior Scale (GPBS; Van
(HiPIC; Mervielde & De Fruyt, 1999) was used to assess the child’s
Leeuwen & Vermulst, in press) is a questionnaire designed to assess
personality. The inventory measures five personality domains based on 18
parenting behavior that is based on concepts from social learning theory
facets hierarchically organized under the five domains: Extraversion (based
(Capaldi & Patterson, 1989; Patterson, Reid, & Dishion, 1992). This
on the facets of Shyness, Optimism, Expressiveness, and Energy), Benev-
theoretical framework organizes parenting into five well-defined constructs
olence (based on Egocentrism, Irritability, Compliance, Dominance, and
derived from observable parental behavior. Validation of a pilot version led
Altruism), Conscientiousness (based on Achievement Motivation, Concen-
to the conclusion that the five Patterson et al. constructs (positive involve-
tration, Perseverance, and Orderliness), Emotional Stability (based on
ment, monitoring, problem solving, structure, and positive reinforcement)
appeared to be too heterogeneous. Refinement of the constructs resulted in Anxiety and Self-Confidence), and Imagination (based on Creativity, Cu-
a new questionnaire with nine scales: Autonomy, Discipline, Positive riosity, and Intellect). Parents rated the 144 items on a 5-point Likert scale.
Parental Behavior, Harsh Punishment, Monitoring, Rules, Ignoring Un- The factor structure has proven to be highly replicable across both child-
wanted Behavior, Material Rewarding, and Inconsistent Discipline. Evi- hood and adolescence (Mervielde & De Fruyt, 2002). In our study, Cron-
dence for the factorial validity and for a moderate to good internal con- bach’s alphas (for maternal and paternal ratings at the two measurement
sistency of the majority of the scales is provided by Van Leeuwen and times) ranged from .93 to .94 for Benevolence (40 items), from .94 to .94
Vermulst (in press). Participants rate the frequency of each behavioral item for Conscientiousness (32 items), from .86 to .88 for Emotional Stability
on a 5-point Likert scale ranging from never to always. The parent version (16 items), from .89 to .91 for Extraversion (32 items) and from .91 to .92
provides self-ratings for parental behavior, and the child version, which has for Imagination (24 items).
the same items as the parent version, allows children and adolescents to Child problem behavior. The Dutch version of the Achenbach Child
rate the parental behavior of their mothers and/or fathers. Behavior Checklist (CBCL; Verhulst, Van der Ende, & Koot, 1996) is used
In order to reduce the number of scales of the GPBS, we conducted as an instrument for screening children with behavioral and emotional
factor analyses with principal axis factoring and oblimin rotation for four problems. Parents rate the frequency of 113 problematic behaviors on a
groups of raters: mothers, fathers, children rating their mothers, and chil- 3-point Likert scale. Two broadband syndromes can be derived: Internal-
dren rating their fathers. A two-factor solution was most appropriate. izing, with items referring to somatic complaints, social withdrawal, and
Because the Monitoring, Material Rewarding, and Inconsistent Discipline anxiety/depression, and Externalizing, including items indexing aggres-
scales did not consistently load on the same dimensions for each of the sion, hyperactivity, and delinquency. Cronbach’s alphas for the Internal-
groups, they were dropped. In each of the four groups of raters, two izing Behavior scale (31 items) ranged from .86 to .88 for maternal and
Table 1
Factor Pattern Matrix for the GPBS Dimensions of Positive Parenting and Negative Control Across a 3-Year Interval
Time 1 Time 2
M F CM CF M F CM CF
GPBS scale Pos Con Pos Con Pos Con Pos Con Pos Con Pos Con Pos Con Pos Con
Positive Parental Behavior .69 ⫺.12 .78 ⫺.11 .73 ⫺.07 .73 ⫺.12 .56 ⫺.18 .65 ⫺.22 .56 .16 .53 ⫺.22
Setting Rules .72 .17 .75 .11 .76 .09 .81 .17 .81 .18 .78 .13 .87 ⫺.26 .88 .11
Autonomy .33 .02 .53 ⫺.01 .63 .00 .61 .01 .28 ⫺.03 .55 .01 .51 .01 .63 .07
Discipline .15 .60 .17 .66 .09 .73 .09 .65 .19 .34 .27 .44 .19 .67 .15 .50
Ignoring Unwanted Behavior ⫺.05 .45 .00 .44 ⫺.12 .76 .02 .49 ⫺.11 .52 ⫺.18 .40 ⫺.05 .61 ⫺.03 .25
Harsh Punishment ⫺.08 .63 ⫺.14 .50 .04 .62 ⫺.11 .73 ⫺.15 .59 ⫺.01 .53 ⫺.11 .31 ⫺.11 .72
% variance 35.47 39.94 50.69 47.07 32.06 35.67 39.82 39.54

Cronbach’s ␣ .86 .80 .91 .80 .90 .87 .90 .82 .86 .77 .90 .75 .90 .91 .81 .80
No. of items 21 14 21 14 20 14 20 14 21 14 21 14 20 14 20 14
Note. Primary factor loadings are indicated by boldface type. GPBS ⫽ Ghent Parental Behavior Scale; M ⫽ mothers; F ⫽ fathers; CM ⫽ children rating
mothers; CF ⫽ children rating fathers; Pos ⫽ Positive Parenting; Con ⫽ Negative Control.
paternal ratings over the two measurement times and from .90 to .91 for the Statistical Analyses
Externalizing Behavior scale (33 items).
Although interaction effects are frequently reported in experimental
Procedure studies, field researchers often find that moderator effects are extremely
difficult to detect (McClelland & Judd, 1993). This can be attributed in part
Families were recruited via stratified random sampling of elementary to statistical problems with moderated multiple regression (MMR) research
and secondary schools. For elementary schools, the sample was stratified (cf. Aguinis, 1995; Chaplin, 1991; McClelland & Judd, 1993). In general,
by province (East and West Flanders), region (rural or urban), school type tests of hypotheses regarding the effects of moderators often have very low
(public vs. private vs. Catholic schools), and grade (third, fourth, fifth, and statistical power and a high risk of Type II errors. This leads researchers to
sixth years of elementary school). For secondary schools, sampling was prematurely dismiss theoretical models that include moderating effects.
based on province (East and West Flanders), type of curriculum (voca- Several precautions were taken in the present study to reduce statistical
tional, technical, and general education), and grade (first and second years problems with MMR. First, we used reliable measures. Both personality
of secondary school). A letter addressed to the parents informed them about and parenting measures are second-order factors. To reduce measurement
the goal and the procedures of the research project. The response rate was error, we used a cross-rater strategy (Chaplin, 1991) and created aggre-
41% for parents with children in primary schools and 39% for parents with gated scores based on ratings provided by different informants (see the
children in secondary schools. Given this modest response rate, we com- following section on Constructing Composite Scores).
pared the socioeconomic status of the present sample with the character- Second, a nonrestricted large general population sample (N ⫽ 600) was
istics of a representative general community sample (N ⫽ 1,789 families) used. The size of this sample approximates the range of N required to
used to investigate the quality of life in Flemish school-age children (Van produce power of .80 at ␣ ⫽ .05 for small effect sizes and a reliability of
den Bergh, 1997). The family composition was somewhat different, ␹2(2, .80 (cf. Aiken & West, 1991, p. 164). To eliminate multicollinearity, we
N ⫽ 1,840) ⫽ 11.95, p ⬍ .01, mainly because of the higher proportion of centered the predictors, which also provides interpretational advantages
single-parent families in the present study. The educational level differed (Cohen, Cohen, West, & Aiken, 2003, p. 267). Finally, to reduce the
for mothers, ␹2(5, N ⫽ 1,741) ⫽ 12.37, p ⬍ .05, and fathers, ␹2(5, N ⫽ likelihood of capitalizing on chance, we checked for cross-rater stability
1,635) ⫽ 25.46, p ⬍ .05, but not as a consequence of an overrepresentation (i.e., parent vs. child ratings of parental behavior) and for cross-sample
of the higher educational levels in the present study. There was a difference stability (i.e., by replicating the findings across a 3-year interval). We also
in employment status for mothers, ␹2(2, N ⫽ 1,797) ⫽ 9.75, p ⬍ .01, but investigated whether the interaction effects assessed at Time 1 predicted
not for fathers, ␹2(2, N ⫽ 1,735) ⫽ 1.16, p ⬎ .05. Although small Time 2 problem behavior.
differences between the two samples can be observed, it can be concluded Parallel cross-sectional hierarchical multiple regressions were carried
that the present sample represents the broad socioeconomic strata. With out on the Time 1 and Time 2 data in order to detect significant interactions
regard to the presence of problem behavior, we compared the prevalence of between child personality and parenting. In order to check cross-rater
children scoring above the 90th percentile (T score ⬎ 63) on the Total stability, we conducted, at both times, analyses for (a) the parent sample,
Problem Behavior Scale with an epidemiological study in Flanders that with parental self-ratings of Negative Control and Positive Parenting, and
investigated emotional and behavioral problems in children ages 6 to 12 (b) the child sample, with child ratings of Negative Control and Positive
(Hellinckx, De Munter, & Grietens, 1991). In the present study, the Parenting. Child sex (boy was coded as 1, and girl was coded as 2) and age
percentage of children showing serious emotional and/or behavioral prob- were consistently entered in Step 1 as control measures. In Step 2, one of
lems was 13.8, whereas the percentage in the epidemiological study was the five child personality domains (i.e., Benevolence, Conscientiousness,
15.4. Hence it can be concluded that the sample in the present study does Emotional Stability, Extraversion, and Imagination) and one of the two
not consist of children with less problem behavior than might be observed parenting construct variables (i.e., Positive Parenting and Negative Con-
in the target population. Moreover, the present study did not focus on trol) were entered, and in Step 3 the cross-product of the personality
assessment of mean levels or prevalence rates but rather on the relation- variable and the parenting variable was entered. Multiplying the two
ships among variables, and hence the issue of representative sampling centered predictors produced the interaction term. Evidence for a moder-
carries less weight. ator effect is found when there is a significant increase in the multiple R2
At each assessment period, a trained psychology student visited the after entering the interaction term, as indicated by a significant incremental
families at home and instructed the mother, father, and child to indepen- F test. In order to prevent chance capitalization in finding moderator
dently complete a series of questionnaires. Both parents filled out the effects, we applied the Bonferroni correction. In this study, the alpha level
HiPIC, the GPBS, and the CBCL, whereas children were administered the was set at .05, with 40 tests explaining externalizing behavior and 40 tests
GPBS at both measurement times. In addition, teachers filled out the Dutch explaining internalizing behavior. Application of the Bonferroni correction
version of Achenbach’s Teacher Report Form (TRF; Verhulst, Van der indicated that the alpha level for each individual test needed to be lowered
Ende, & Koot, 1997a) at Time 1, and children completed the Dutch version to .001 in order to adjust the overall alpha level to .05.
of the Youth Self Report (YSR; Verhulst, Van der Ende, & Koot, 1997b) For testing and interpreting interactions, we followed the guidelines
and a Dutch shortened version of Goldberg’s (1992) hundred adjectives proposed by Aiken and West (1991) and Cohen et al. (2003). First,
entitled the Questionnaire Big Five (QBF; Gerris et al., 1998) at Time 2, to significant interactions were interpreted by plotting simple regression lines
provide self-ratings of problem behavior and personality, respectively. In for high (1 SD above the mean), mean, and low (1 SD below the mean)
this study, we only used measures that were presented at both measurement values of the moderator variables. Second, the significance of the slopes for
occasions in order to facilitate comparison of the results. We did not opt to these simple regression lines and the difference between the slopes were
query the children about their problem behavior and personality at Time 1 tested with t tests. Finally, interaction patterns were identified, based on the
(a) because the YSR is intended to assess self-reported problem behavior signs of the regression coefficients of the two independent variables (B1
from age 11 onward, but almost half of the children were younger than 11 and B2) and the interaction (B3). Three theoretically meaningful interaction
years at the first assessment period; (b) because at present there is little patterns can be distinguished (Cohen et al., 2003, pp. 285–286): (a)
evidence for the validity, and in particular for the discriminant validity, of synergistic or enhancing interactions, in which all three regression coef-
self-rated personality measures for elementary school children in general; ficients have the same sign, indicating that combining predictors produces
and (c) for practical reasons, that is, we did not want to overload the an incremental effect beyond the additive effects; (b) buffering interac-
children with questionnaires at the first assessment given their age and the tions, in which the two predictors have regression coefficients with oppo-
length of the questionnaires (e.g., the HiPIC consists of 144 items). The site signs, meaning that one predictor weakens the effect of the other
GPBS was specifically designed to obtain ratings from both parents and predictor (for buffering interactions, one predictor represents a risk factor,
from children with reading abilities (ages 7– 8 years and older). whereas the other predictor acts as a protective factor); and (c) interference
or antagonistic interactions, in which both B1 and B2 have the same sign ratings were provided by both mothers and fathers), (b) reduce measure-
and B3 is of the opposite sign, indicating a compensatory or either-or effect ment error, and (c) facilitate comparison of the results. A possible way to
of B1 and B2 on the criterion. Interactions are sometimes described as address the problem of shared method variance is to assess the convergent
ordinal (noncrossing) or disordinal (crossing). This distinction is less useful validity of the data supplied by the informants against external criteria.
because it is mainly determined by the strength of the first-order effects. Correlations between Time 1 TRF and CBCL ratings were all significant at
p ⬍ .001, with values of .44 (father–teacher) and .46 (mother–teacher) for
externalizing behavior and .18 (father–teacher) and .16 (mother–teacher)
Constructing Composite Scores for internalizing behavior. Correlations between Time 2 CBCL and YSR
ratings were all significantly correlated at p ⬍ .001, with values of .41
For the present analyses, aggregated scores for predictor and criterion (father– child) and .41 (mother– child) for Externalizing Behavior and .35
variables were created to reduce the potential number of analyses generated (father– child) and .40 (mother– child) for Internalizing Behavior. Correla-
by the numerous combinations of types of raters (i.e., mothers, fathers, tions between Time 2 parental and adolescent QBF ratings were all
children rating mothers, and children rating fathers), parenting variables significant at p ⬍ .001, with r ⫽ .20 for Benevolence, r ⫽ .46 for
(i.e., Positive Parenting and Negative Control), personality domains (i.e., Conscientiousness, r ⫽ .37 for Neuroticism, r ⫽ .46 for Extraversion, and
Benevolence, Conscientiousness, Emotional Stability, Extraversion, and r ⫽ .34 for Imagination. This type of analysis illustrates that the data
Imagination), dependent variables (Internalizing and Externalizing Behav- provided by the informants exhibit a reasonable degree of convergent
ior), and measurement occasions. This was accomplished by extracting a validity and hence that the results of the present study are not seriously
common factor score from each pair of ratings provided by two different encumbered by problems of shared method variance.
informants by means of principal-components analyses: for mother and
father ratings of their own parental behavior, for mother and father ratings Results
of child personality and problem behavior, and for child ratings of maternal
and paternal behavior. The score of the other rating substituted the missing Cross-Sectional Analyses at Time 1
value in case only one of the two ratings was available (e.g., in single-
parent families). Thus, the common factors reflect (a) self-reported par- Sex and age effects. Results of the hierarchical regression
enting by two different informants (mother and father), (b) other-reports of analyses are shown in Tables 2 and 3. Of the two control measures,
child personality and problem behavior by two different informants sex and age, entered in Step 1, sex accounted for a significant
(mother and father), and (c) other-reports of maternal and paternal behavior proportion of the variance in Externalizing Behavior (R2 ⫽ .039).
by one informant (the child). The use of composite scores is supported by Interaction effects. Three moderator effects were predictive of
the meaningful and significant ( p ⬍ .001) correlations between the paired child externalizing behavior when parents supplied ratings of their
ratings composing the aggregated scores: The correlation between mothers own parental behavior. Significant interactions for Negative Con-
and fathers was .75 for Externalizing Behavior, .59 for Internalizing trol were found with the personality domains of Benevolence,
Behavior, .70 for Benevolence, .81 for Conscientiousness, .69 for Emo-
R2 ⫽ .025 and Fchange(1, 577) ⫽ 30.61, p ⬍ .001, and Conscien-
tional Stability, .69 for Extraversion, .69 for Imagination, .22 for Positive
tiousness, R2 ⫽ .014 and Fchange(1, 577) ⫽ 11.95, p ⬍ .001. A
Parenting, and .44 for Negative Control. The correlation between child
ratings for mothers’ and fathers’ Positive Parenting was .74, and the significant interaction between Positive Parenting and Benevo-
corresponding correlation for Negative Control was .70. The explained lence predicted child externalizing, R2 ⫽ .010 and Fchange(1,
variances across the two time points ranged from 79.50% to 87.63% for the 577) ⫽ 11.17, p ⬍ .001. No interactions significantly predicted
two dependent variables, from 82.46% to 90.76% for the five personality child internalizing behavior.
variables, and from 61.17% to 89.53% for the two parenting variables. Testing and interpreting interaction effects. The tests on the
Parent and child ratings of parental behavior were not aggregated be- significance of the difference between simple slopes (see Table 4)
cause principal-components analyses at both measurement occasions dis- show that the slopes for children rated low (1 SD below the mean)
tinguished two separate factors for Positive Parenting, that is, a factor or around the mean on the personality characteristics of benevo-
including ratings of mothers and fathers and a factor including child ratings lence and conscientiousness are significant. For these children,
of maternal behavior and child ratings of paternal behavior. Not aggregat- parenting has a significant effect on child problem behavior. When
ing scores of parents and children also enables the cross-validation of
rated high (1 SD above the mean) on one of these personality
interaction effects across ratings of parents and children. Moreover, using
domains, the slopes are not significant. This indicates that for
child ratings of parenting in analyses with parent measures of problem
behavior and personality decreases shared method variance.
children characterized by highly adaptive personality characteris-
The use of factor scores is valuable because they reflect the common tics, parenting does not predict externalizing behavior.
core of variables, reduce measurement error, and hence increase the power From the signs of the regression coefficients (see Tables 2 and
of statistical tests for interaction (Chaplin, 1991; Jaccard & Wan, 1995). In 3), two interaction patterns can be identified: (a) buffering inter-
addition, they center the predictor variables, as the mean equals zero and actions and (b) interference or antagonistic interaction patterns.
the standard deviation equals one. For the Negative Control ⫻ Personality interactions (see Figure 1),
the coefficients of the independent predictors have opposite signs,
indicating that one predictor diminishes problem behavior and the
Shared Method Variance other predictor enhances it. Hence, negative control can be con-
sidered as a risk factor for externalizing behavior in particular for
The present study partially controlled for shared method variance by
children rated low or around the mean on benevolence and con-
combining parental ratings of personality and child problem behavior with
child ratings of parenting in Study 1. We acknowledge that it would be
scientiousness. On the other hand, the impact of low scores on
more appropriate to use ratings supplied by different informants for each of benevolence and conscientiousness will be diminished for children
the measures and that extracting the common component from multiple experiencing low parental negative control. Benevolence and con-
ratings ignores the unique information provided by a particular informant. scientiousness can be regarded as protective factors: Children
However, we prefer to use ratings that were made at both measurement scoring high on these personality domains do not show problem
occasions in order to (a) limit the potential number of analyses (given that behavior at all, even when faced with negative parental control.
Table 2
Moderators of Externalizing Behavior
Time 1 Time 2
Parental ratings of parenting Child ratings of parenting Parental ratings of parenting Child ratings of parenting
Variable ⌬F B ⌬F B ⌬F B ⌬F B
Negative Control
Sex, Age 11.63*** ⫺.13**, .02 11.70*** ⫺.16**, .02 8.14*** ⫺.07, .00 8.10*** ⫺.11, .00
BE, CON 271.23*** ⫺.57***, .21*** 238.44*** ⫺.64***, .08** 294.03*** ⫺.66***, .12*** 280.61*** ⫺.69***, .06
BE ⫻ CON 30.61*** ⫺.15*** 8.44** ⫺.08** 51.60*** ⫺.18*** 28.15*** ⫺.15***
Sex, Age 11.63*** ⫺.19**, .00 11.70*** ⫺.20**, .02 8.14*** ⫺.10, .01 8.10*** ⫺.13, .01
CO, CON 105.14*** ⫺.35***, .30*** 76.50*** ⫺.40***, .18*** 101.81*** ⫺.42***, .25*** 85.19*** ⫺.41***, .16***
CO ⫻ CON 11.95*** ⫺.12*** 8.74** ⫺.11** 14.13*** ⫺.14*** 21.84*** ⫺.18***
Sex, Age 11.63*** ⫺.32***, .03 11.70*** ⫺.34***, .05* 8.14*** ⫺.29***, .02 8.10*** ⫺.31***, .03
ES, CON 58.73*** ⫺.15***, .37*** 26.44*** ⫺.18***, .23*** 45.08*** ⫺.17***, .33*** 33.93*** ⫺.19***, .27***
ES ⫻ CON 0.05 .01 0.03 ⫺.01 0.04 ⫺.01 0.13 .01
Sex, Age 11.63*** ⫺.30***, .03 11.70*** ⫺.32***, .04* 8.14*** ⫺.24**, .03 8.10*** ⫺.25**, .03
EX, CON 49.85*** .03, .37*** 15.32*** .04, .21*** 36.02*** .02, .35*** 22.82*** .02, .29***
EX ⫻ CON 0.01 .00 0.04 .01 0.04 .01 0.41 ⫺.03
Sex, Age 11.63*** ⫺.31***, .01 11.70*** ⫺.33***, .02 8.14*** ⫺.26**, .01 8.10*** ⫺.28***, .01
IM, CON 56.59*** ⫺.13***, .37*** 23.41*** ⫺.17***, .21*** 41.84*** ⫺.14***, .33*** 30.86*** ⫺.16***, .27***
IM ⫻ CON 1.71 ⫺.05 2.94 ⫺.07 0.71 ⫺.03 2.26 ⫺.06
Positive Parenting
Sex, Age 11.63*** ⫺.18**, .02 11.85*** ⫺.19**, .02 8.14*** ⫺.11, ⫺.01 8.10*** ⫺.13, ⫺.01
BE, POS 232.02*** ⫺.65***, ⫺.03 233.18*** ⫺.67***, .02 277.42*** ⫺.73***, .04 275.07*** ⫺.73***, .02
BE ⫻ POS 11.17*** .10*** 0.54 ⫺.02 0.82 .02 0.71 .02
Sex, Age 11.63*** ⫺.24***, .01 11.85*** ⫺.25***, .01 8.14*** ⫺.15, .00 8.10*** ⫺.15, .01
CO, POS 64.47*** ⫺.40***, ⫺.09* 60.78*** ⫺.42***, .02 74.82*** ⫺.49***, ⫺.07 72.61*** ⫺.48***, .00
CO ⫻ POS 7.58** .10** 0.33 .02 8.65** .11** 4.59* .09*
Sex, Age 11.63*** ⫺.39***, .03 11.85*** ⫺.39***, .04 8.14*** ⫺.42***, .02 8.10*** ⫺.39***, .03
ES, POS 20.20*** ⫺.16***, ⫺.19*** 9.00*** ⫺.17***, ⫺.03 17.49*** ⫺.19***, ⫺.13** 13.72*** ⫺.21***, ⫺.07
ES ⫻ POS 0.01 .00 0.63 ⫺.03 1.49 ⫺.05 2.02 ⫺.06
Sex, Age 11.63*** ⫺.38***, .04 11.85*** ⫺.39***, .04 8.14*** ⫺.36***, .03 8.10*** ⫺.32***, .03
EX, POS 15.51*** .11**, ⫺.22*** 1.46 .07, ⫺.02 8.17*** .06, ⫺.18*** 2.30 .04, ⫺.09*
EX ⫻ POS 0.09 ⫺.01 3.85* .08* 0.40 ⫺.03 0.03 ⫺.01
Sex, Age 11.63*** ⫺.37***, .02 11.85*** ⫺.39***, .01 8.14*** ⫺.36***, .00 8.10*** ⫺.34***, .01
IM, POS 17.11*** ⫺.13**, ⫺.18*** 8.25*** ⫺.17***, ⫺.01 14.42*** ⫺.17***, ⫺.14** 11.30*** ⫺.19***, ⫺.06
IM ⫻ POS 4.74* .09* 0.59 .03 3.57 .08 0.64 .03
Note. With multiplicative terms, neither traditional unstandardized nor standardized regression coefficients are appropriate to report. However, when the
cross-product is based on z scores, which is known as Friedrich’s procedure, it is appropriate to use the unstandardized solution with interaction terms
(Aiken & West, 1991, pp. 43– 44). In our study, all predictors are standardized, because they are factor scores. BE ⫽ Benevolence; CO ⫽ Conscien-
tiousness; ES ⫽ Emotional Stability, EX ⫽ Extraversion; IM ⫽ Imagination; POS ⫽ Positive Parenting; CON ⫽ Negative Control.
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
An interference or antagonistic interaction is found for the the first-order effects (cf. Cohen et al., 2003, p. 286). In this study,
Positive Parenting ⫻ Benevolence interaction (see Figure 2). Here, the strongest independent effects were found for the personality
the interaction term has the sign opposite to that of both predictors. variables.
This suggests that both positive parenting and child personality are Effect sizes and statistical power of interactions. In order to
negatively related to externalizing behavior but that the importance permit comparison across studies, effect size measures are reported
of positive parenting is lessened by benevolence, or vice versa. in Table 4. Effect sizes for the interactions at Time 1 range from
Figures 1 and 2 show that most interactions are ordinal: The .01 to .05 and can be regarded as small (Aiken & West, 1991, p.
rank order of the outcomes of one predictor is maintained across 158). The statistical power for the interaction terms does not equal,
all levels of the other predictor, within the observed range of the but approximates, the standard of .80. Taking the Negative Con-
second predictor. This appearance is influenced by the strength of trol ⫻ Benevolence interaction as an example, we note reliabilities
Table 3
Moderators of Internalizing Behavior
Time 1 Time 2
Variable ⌬F B ⌬F B ⌬F B ⌬F B
Negative Control
Sex, Age 2.17 .01, .03 2.20 ⫺.03, .03 1.34 .12, .03 1.49 .10, .03
BE, CON 24.38*** ⫺.22***, .09* 23.70*** ⫺.28***, ⫺.08* 36.56*** ⫺.32***, .08 33.91*** ⫺.34***, .01
BE ⫻ CON 9.44** ⫺.12** 1.00 ⫺.04 2.27 ⫺.06 3.15 ⫺.07
Sex, Age 2.17 .00, .02 2.20 ⫺.03, .02 1.34 .11, .03 1.49 .10, .03
CO, CON 23.84*** ⫺.23***, .11** 20.49*** ⫺.26***, ⫺.04 22.14*** ⫺.23***, .13** 17.94*** ⫺.23***, .05
CO ⫻ CON 2.55 ⫺.06 0.56 ⫺.03 0.01 .00 2.41 ⫺.07
Sex, Age 2.17 ⫺.15*, .05** 2.20 ⫺.17*, .05** 1.34 ⫺.17*, .04 1.49 ⫺.17*, .04*
ES, CON 160.73*** ⫺.57***, .12*** 150.31*** ⫺.58***, .01 175.83*** ⫺.63***, .09* 170.95*** ⫺.63***, .06
ES ⫻ CON 0.26 ⫺.02 0.27 .02 0.10 .01 2.72 .05
Sex, Age 2.17 ⫺.03, .01 2.20 ⫺.06, .02 1.34 .05, .01 1.49 .04, .01
EX, CON 54.59*** ⫺.36***, .19*** 40.98*** ⫺.35***, .02 59.58*** ⫺.41***, .18*** 52.76*** ⫺.41***, .13**
EX ⫻ CON 0.25 ⫺.02 0.06 .01 1.52 ⫺.05 0.06 ⫺.01
Sex, Age 2.17 ⫺.09, .00 2.20 ⫺.13, .01 1.34 .01, .01 1.49 ⫺.01, .01
IM, CON 28.67*** ⫺.26***, .14*** 22.93*** ⫺.28***, ⫺.03 24.84*** ⫺.25***, .14*** 21.45*** ⫺.26***, .09*
IM ⫻ CON 1.32 ⫺.04 1.09 ⫺.04 0.04 ⫺.01 2.87 ⫺.07
Positive Parenting
Sex, Age 2.17 ⫺.03, .03 2.23 ⫺.03, .03 1.34 .09, .02 1.49 .10, .02
BE, POS 25.56*** ⫺.24***, ⫺.10* 22.09*** ⫺.27***, ⫺.01 35.45*** ⫺.34***, ⫺.05 34.32*** ⫺.35***, ⫺.04
BE ⫻ POS 1.08 .04 1.09 ⫺.04 0.11 .00 0.20 .02
Sex, Age 2.17 ⫺.03, .02 2.23 ⫺.02, .02 1.34 .07, .02 1.49 .10, .03
CO, POS 23.38*** ⫺.23***, ⫺.10** 19.89*** ⫺.26***, .00 19.59*** ⫺.25***, ⫺.10* 17.75*** ⫺.25***, ⫺.05
CO ⫻ POS 1.31 .04 0.13 .02 1.19 .05 1.51 .05
Sex, Age 2.17 ⫺.17**, .05** 2.23 ⫺.17**, .05** 1.34 ⫺.17*, .04* 1.49 ⫺.18**, .04*
ES, POS 23.38*** ⫺.58***, ⫺.13*** 150.16*** ⫺.58***, ⫺.02 171.47*** ⫺.64***, ⫺.04 169.51*** ⫺.63***, ⫺.04
ES ⫻ POS 1.31 .06 0.02 ⫺.01 5.93* .08* 2.53 .05
Sex, Age 2.17 ⫺.07, .02 2.23 ⫺.07, .02 1.34 ⫺.01, .01 1.49 .01, .01
EX, POS 44.26*** ⫺.33***, ⫺.09* 40.92*** ⫺.35***, .01 48.82*** ⫺.39***, ⫺.06 48.14*** ⫺.39***, ⫺.05
EX ⫻ POS 0.20 .02 2.03 .06 2.61 .07 3.47 .08
Sex, Age 2.17 ⫺.12, .01 2.23 ⫺.12, .01 1.34 ⫺.04, .01 1.49 ⫺.02, .01
IM, POS 26.66*** ⫺.25***, ⫺.11** 22.56*** ⫺.27***, .00 21.31*** ⫺.25***, ⫺.08 20.14*** ⫺.26***, ⫺.05
IM ⫻ POS 0.07 ⫺.01 0.78 .04 1.74 ⫺.05 1.97 .06
Note. BE ⫽ Benevolence; CO ⫽ Conscientiousness; ES ⫽ Emotional Stability, EX ⫽ Extraversion; IM ⫽ Imagination; POS ⫽ Positive Parenting;
CON ⫽ Negative Control.
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
of .80, an interpredictor correlation of ⫺.30, and a squared mul- Interaction effects. The significant interactions with Negative
tiple correlation of .50 for the main effects. According to Aiken Control at Time 1 for externalizing behavior were replicated at
and West (1991, p. 164), these values correspond with a power of Time 2 (see Table 2) with both Benevolence, R2 ⫽ .042 and
.39 –.44 at N ⫽ 392. A power of .80 with these values requires a Fchange(1, 501) ⫽ 51.60, p ⬍ .001, and Conscientiousness, R2 ⫽
sample size between 909 and 1,056. .019 and Fchange(1, 501) ⫽ 14.13, p ⬍ .001. In addition, the
Negative Control ⫻ Benevolence interaction, R2 ⫽ .024 and
Fchange(1, 498) ⫽ 28.15, p ⬍ .001, and the Negative Control ⫻
Replication of the Cross-Sectional Analyses at Time 2
Conscientiousness interaction, R2 ⫽ .030 and Fchange(1, 498) ⫽
Sex and age effects. The control measure sex, entered in Step 21.84, p ⬍ .001, were also replicated when children rated parent-
1, significantly explained about 3% of the variance of externalizing ing behavior. At Time 1, these interactions were significant only
behavior. when the alpha level was set at .01. The Positive Parenting ⫻
Benevolence interaction was not replicable. No interaction terms
Note. Effect size ( f 2) ⫽ (r2Y.MI ⫺ r2Y.M)/(1 ⫺ r2Y.MI) (Aiken & West, 1991, p. 156); r2Y.MI ⫽ the squared multiple correlation from combined predictors by two sets of variables, M (main effects) and
0.554 0.10
0.311 0.03
0.544 0.06
0.278 0.04
f2
significantly predicted child internalizing behavior at Time 2 (see
Effect size
Table 3).
r2Y.M
Testing and interpreting interaction effects. The interaction
patterns found at Time 1 were replicated at Time 2. The Negative
BFM ⫹ 1 SD BFM M BFM ⫺ 1 SD Interaction r2Y.MI
0.596
0.330
0.568
0.308
Control ⫻ Benevolence and the Negative Control ⫻ Conscien-
tiousness interactions were of the buffering type. The effect sizes
⫺7.18***
⫺3.76***
⫺5.31***
⫺4.67***
of the interaction terms were also small, ranging from .01 to .10.1
Time 1 Variables Predicting Time 2 Problem Behavior

Time 2
Table 5 reports the results of a hierarchical multiple regression

7.87***
7.48***
5.19***
6.48*** (HMR) analysis with Time 1 variables predicting Time 2 exter-
nalizing and internalizing behavior. When parents provided self-
t values
ratings of parental behavior, the Negative Control ⫻ Benevolence

Simple slopes
interaction, R2 ⫽ .024 and Fchange(1, 465) ⫽ 18.84, p ⬍ .001, and

3.88***
6.53***
4.17***
the Negative Control ⫻ Conscientiousness interaction, R2 ⫽ .022

1.76
and Fchange(1, 465) ⫽ 13.95, p ⬍ .001, predicted Time 2 child

externalizing behavior. The results also show that Time 1 child
personality—in particular, benevolence and conscientiousness—is
1.98*
⫺2.08*
⫺1.62
⫺0.26
a predictor for Time 2 externalizing behavior. Time 2 internalizing

behavior is predicted by each of the Big Five personality dimen-
I (interaction effect); r2Y.M ⫽ the squared multiple correlation resulting from prediction by set M. BFM ⫽ Big Five measure.
sions, with emotional stability and extraversion as the strongest

predictors. Examination of the parenting variables shows that Time
0.504 0.05
0.295 0.02
0.474 0.01
0.240 0.02
0.467 0.02
f2
1 Negative Control is a predictor of both Time 2 externalizing

Effect size
behavior and Time 2 internalizing behavior for parental self-

r2Y.M
Tests of Significance of Difference Between Simple Slopes (T Values) and Effect Sizes of Interactions
ratings and also a predictor of externalizing behavior when child

ratings are used in the analyses. Time 1 Positive Parenting is not
BFM ⫺ 1 SD Interaction r2Y.MI
0.529
0.309
0.481
0.252
0.477
a strong predictor of Time 2 child problem behavior.

⫺5.53***
⫺3.46***
3.34***
⫺2.90**
⫺2.96**
Discussion
In this study, we investigated Parenting ⫻ Child Personality
interactions for the prediction of child and adolescent problem
Time 1
behavior. Study 1 examined interactions that have been detected in

8.80***
8.66***
4.21***
5.76***
⫺3.13**
previous studies, for example, between negative parental control

and low agreeableness, as predictors of externalizing behavior
t values
Simple slopes
6.95***
8.57***
4.95***
BFM M
2.58**
1
⫺0.99
We also checked for nonlinear relationships between independent and

dependent variables, although (a) the inclusion of curvilinear relationships
in this study is not driven by theoretical assumptions and (b) there is an
BFM ⫹ 1 SD
ongoing debate over whether one should include curvilinear effects if the
3.72***
central interest is in the interaction term. Curvilinear effects show lower

1.30
⫺0.02
1.32
1.51
reliability than the cross-product terms, and the inclusion of multiple higher
order terms introduces multicollinearity and instability of the regression
equation (Cohen et al., 2003, pp. 299 –300). Testing quadratic relationships
was limited to models with significant Parenting ⫻ Personality interac-
Negative Control ⫻ Conscientiousness
Negative Control ⫻ Conscientiousness
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
tions. For the Time 1 data, significant quadratic relationships were present
Positive Parenting ⫻ Benevolence
Negative Control ⫻ Benevolence
Negative Control ⫻ Benevolence
for benevolence and conscientiousness in the prediction of child external-

Positive Parenting: Parental ratings
Negative Control: Parental ratings
izing behavior and for benevolence predicting child internalizing behavior.

Externalizing behavior
Negative Control: Child ratings
These quadratic effects did not influence the Parenting ⫻ Personality

interactions for the majority of the models, indicating that the interactions
were not the result of the curvilinear nature of the regression model. Only
two interactions could be attributed to the curvilinear effects: the Benev-
olence ⫻ Positive Parenting (parental ratings) interaction and the Benev-
olence ⫻ Negative Control (child ratings) interaction in the prediction of
externalizing behavior. For the Time 2 data, the Conscientiousness ⫻
Table 4
Positive Parenting (child ratings) interaction was not a significant predictor

of externalizing behavior when the quadratic effect of conscientiousness
was included in the regression model.
Figure 1. Interaction between personality and negative control predicting externalizing behavior at Time 1.
BE ⫽ benevolence; CO ⫽ conscientiousness.
(e.g., O’Connor & Dvorak, 2001; Prinzie et al., 2003). However, children rated high on these personality domains are not likely to
integrating all Big Five personality domains in this study (i.e., develop externalizing behavior even when exposed to restrictive
Agreeableness, Conscientiousness, Emotional Stability, Extraver- parental control. Significant interactions were also found between
sion, and Imagination) enabled us to test for interactions between positive parenting and benevolence.
parenting and each of the five broad child personality domains as From these results it can be concluded that children rated low on
predictors of both externalizing and internalizing behavior. benevolence are likely to show externalizing behavior in particular
HMR analyses showed prominent interactions between negative when parents are not supportive. When children show high levels
parental control and the personality domains of benevolence and of benevolence they are not affected by low levels of positive
conscientiousness, predicting externalizing behavior, and hence parenting. However, this Positive Parenting ⫻ Benevolence inter-
corroborating the findings of previous studies. These interactions action is not stable across judges and across time.
were found for parental self-ratings of parenting behavior and were With internalizing behavior as the outcome variable, there were
partly replicable across raters and a 3-year interval. Moreover, the no significant interactions. Our findings suggest mainly indepen-
Time 1 interactions predicted Time 2 externalizing behavior. The dent or additive contributions of parenting and child characteristics
results indicate that negative control is a risk factor for external- to internalizing behavior. The significant main effects suggest that,
izing behavior, in particular for children rated low or around the in particular, child personality is an important independent predic-
mean on benevolence and conscientiousness. On the other hand, tor of internalizing behavior.
Figure 2. Interaction between personality and positive parenting predicting externalizing behavior at Time 1.
BE ⫽ benevolence.
In Study 1, we adopted a variable-centered approach to child (2001). In a first step, we conducted hierarchical cluster analyses with
personality, considering the effects of single Big Five domains and Ward’s method, using raw HiPIC domain scores. The resulting three-
parenting variables as well as the effects of personality–parenting cluster solutions of these analyses were then used as initial cluster centers
interactions. The variable-centered approach identifies the impor- in a nonhierarchical K-means clustering procedure.
tant personality and parenting dimensions, but it does not allow us
to target particular groups of children at risk. Moreover, it does not Delineation of Parenting Categories
take into account the prevalence and relevance of particular com- The continuous variables positive parenting and negative control, rated
binations of risk factors, such as a low score on both benevolence by the parents, were recoded into three categories: scores of 1 SD ⱕ the
and conscientiousness, and therefore it is not clear to what extent mean, scores around the mean, and scores of 1 SD ⱖ the mean. In the
such configurations constitute an additional risk factor for devel- analyses, only participants scoring low or high on parenting variables were
oping externalizing or internalizing problem behavior. In Study 2, included. The number of children exposed to high positive parenting was
we adopted a person-centered approach, grouping children into 105, whereas the number exposed to low positive parenting was 100. The
types based on their personality profiles. We assigned the same category of high negative control included 97 participants, and the category
of low negative control grouped 91 participants.
sample of children to three replicable personality types: resilient,
overcontrolled, and undercontrolled. Hence the next study focused
on the interactions between these personality types and the par- Results
enting dimensions. Personality Types
Study 2 The final three-cluster solution obtained after applying the two-
step procedure (Ward’s method and K-means clustering) clearly
Method resembled the resilient, undercontrolled, and overcontrolled pro-
totypes. The first cluster grouped 185 children with above-average
Participants cluster centers for the personality domains benevolence, extraver-
The sample in Study 2 is part of the Time 1 sample of the longitudinal sion, conscientiousness, emotional stability and imagination. This
data set described in Study 1 and consisted of 539 families with 251 boys group was identified as the resilient type. The second cluster
and 288 girls. Participants with missing variables were dropped, which included 162 children scoring low on emotional stability and
accounts for the reduced sample size in comparison with the Time 1 sample extraversion and resembled the overcontrolled type. The third
of Study 1. cluster, with 192 children characterized by low scores on consci-
entiousness and benevolence and scores around the mean for
Measures and Procedure emotional stability, extraversion and imagination, was designated
The results of this study are based on the same questionnaires admin- as the undercontrolled type.
istered in Study 1: the GPBS (Van Leeuwen & Vermulst, in press), the
HiPIC (Mervielde & De Fruyt, 1999), and the Dutch version of the CBCL Parenting ⫻ Personality Type Interactions
(Verhulst et al., 1996). The data collection procedure was the same as that
described in Study 1. General linear model (GLM) univariate analyses of variance
were conducted to detect significant Parenting ⫻ Personality Type
Assessment of Personality Types2
2
In order to derive the overcontrolled, undercontrolled, and resilient A full description of the derivation of the three types for this sample is
personality types, we followed the method described by Asendorpf et al. given in De Fruyt et al. (2002).
Table 5
Time 1 Variables Predicting Time 2 Problem Behavior
Externalizing Time 2 Internalizing Time 2
Time 1 variables ⌬F B ⌬F B ⌬F B ⌬F B
Negative Control
Sex, Age 5.76** ⫺.05, .03 5.78** ⫺.10, .03 0.80 .13, .03 0.80 .09, .03
BE, CON 115.47*** ⫺.45***, .17*** 101.69*** ⫺.51***, .03 15.84*** ⫺.17***, .11* 12.74*** ⫺.21***, ⫺.02
BE ⫻ CON 17.84*** ⫺.15*** 2.53 ⫺.05 1.23 ⫺.05 0.07 ⫺.01
Sex, Age 5.76** ⫺.11, .01 5.78** ⫺.13, .01 0.80 .10, .02 0.80 .08, .02
CO, CON 63.36*** ⫺.31***, .23*** 46.24*** ⫺.35***, .10** 13.65*** ⫺.15***, .13** 9.12*** ⫺.18***, .00
CO ⫻ CON 13.95*** ⫺.14*** 8.18** ⫺.11** 1.72 .05 0.00 ⫺.00
Sex, Age 5.76** ⫺.20*, .03 5.78** ⫺.24**, .04 0.80 .01, .04 0.80 ⫺.02, .04
ES, CON 32.64*** ⫺.11**, .31*** 10.18*** ⫺.14***, .15*** 71.61*** ⫺.42***, .15*** 62.30*** ⫺.43***, .04
ES ⫻ CON 0.86 .04 1.88 ⫺.06 0.17 .02 0.96 ⫺.04
Sex, Age 5.76** ⫺.19*, .03 5.78** ⫺.23**, .04 0.80 .09, .01 0.80 .07, .02
EX, CON 28.56*** .04, .31*** 5.74** .05, .14** 25.48*** ⫺.24***, .17*** 16.49*** ⫺.24***, .05
EX ⫻ CON 0.05 ⫺.01 0.27 ⫺.02 1.83 ⫺.05 0.43 ⫺.03
Sex, Age 5.76** ⫺.21*, .02 5.78** ⫺.25**, .02 0.80 .04, .01 0.80 .01, .01
IM, CON 30.29*** ⫺.09*, .30*** 9.19*** ⫺.14**, .14*** 13.31*** ⫺.15***, .14*** 7.67*** ⫺.18***, .02
IM ⫻ CON 0.36 ⫺.03 6.11* ⫺.10* 0.12 ⫺.02 4.70* ⫺.09*
Positive Parenting
Sex, Age 5.76** ⫺.11, .03 5.91** ⫺.12, .03 0.80 .09, .02 0.80 .10, .03
BE, POS 101.12*** ⫺.51***, ⫺.01 102.33*** ⫺.52***, .00 15.62*** ⫺.19***, ⫺.10* 14.08*** ⫺.21***, ⫺.06
BE ⫻ POS 0.79 .03 0.00 .00 0.16 .02 0.04 .01
Sex, Age 5.76** ⫺.17*, .01 5.91** ⫺.18*, .01 0.80 .08, .02 0.80 .08, .02
CO, POS 42.60*** ⫺.36***, .04 41.95*** ⫺.36***, .00 12.21*** ⫺.16***, ⫺.11* 10.13*** ⫺.17***, ⫺.06
CO ⫻ POS 2.59 .06 0.98 .04 1.32 ⫺.05 0.50 .03
Sex, Age 5.76** ⫺.28***, .03 5.91** ⫺.28***, .04 0.80 ⫺.03, .04 0.80 ⫺.02, .04
ES, POS 7.80*** ⫺.12**, ⫺.11 4.61** ⫺.12**, ⫺.03 67.48*** ⫺.42***, ⫺.11** 63.59*** ⫺.42***, ⫺.06
ES ⫻ POS 2.50 ⫺.07 0.37 ⫺.03 3.28 .07 1.30 .04
Sex, Age 5.76** ⫺.27**, .03 5.91** ⫺.27**, .04 0.80 .06, .01 0.80 .06, .02
EX, POS 5.95** .09, ⫺.14** 1.33 .07, ⫺.02 18.66*** ⫺.21***, ⫺.09* 17.20*** ⫺.23***, ⫺.06
EX ⫻ POS 0.21 ⫺.02 6.24* .11* 0.02 .01 0.31 .02
Sex, Age 5.76** ⫺.28***, .02 5.91** ⫺.28***, .02 0.80 .00, .01 0.80 .01, .01
IM, POS 7.06*** ⫺.11*, ⫺.10* 4.41* ⫺.13**, ⫺.02 11.43*** ⫺.15***, ⫺.11* 8.70*** ⫺.16***, ⫺.06
IM ⫻ POS 0.02 .01 0.16 .02 1.68 ⫺.06 0.05 .01
Note. BE ⫽ Benevolence; CO ⫽ Conscientiousness; ES ⫽ Emotional Stability, EX ⫽ Extraversion; IM ⫽ Imagination; POS ⫽ Positive Parenting;
CON ⫽ Negative Control.
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
interactions. In the analyses with externalizing behavior as the Table 6 shows the category means and the results of the
dependent variable, the undercontrollers versus resilients dichot- analyses of variance. Significant Parenting ⫻ Personality Type
omy was entered as one of the independent variables, whereas in interactions were found only when negative control was in-
the analyses with internalizing behavior as the dependent variable, cluded as the parenting variable, for both externalizing and
the overcontrollers versus resilients dichotomy was entered as one internalizing behavior. The results indicate that undercontrolled
of the independent variables. This choice was based on previous children were rated significantly higher on externalizing behav-
research that showed a meaningful relationship between the un- ior than were resilient children, with the greatest difference
dercontrolled personality type and externalizing behavior and a occurring for children exposed to high levels of parental neg-
relationship between the overcontrolled personality type and in- ative control. This suggests that undercontrolled children sub-
ternalizing behavior (Asendorpf et al., 2001; De Fruyt et al., 2002; jected to negative parental control are especially at risk for
Van Leeuwen et al., 2004). externalizing behavior. The difference in problem behavior
1040
Table 6
Category Means and Results of General Linear Model Analysis With Parenting Categories and Personality Types
Resilients Overcontrollers Undercontrollers Parenting Personality type Parenting ⫻ Personality Type
Dependent variable Lowa Higha Low High Low High F ␩2 Power F ␩2 Power F ␩2 Power
Negative Control
Externalizing T1 ⫺0.66 0.05 ⫺0.25 1.32 38.38*** .247 1.00 20.97*** .152 0.99 5.53* .045 0.65
(38) (16) (28) (39)
Externalizing T2 ⫺0.56 ⫺0.31 ⫺0.30 1.07 17.84*** .154 .987 18.27*** .157 .988 8.49** .080 .823
(35) (11) (23) (33)
Internalizing T1 ⫺0.45 ⫺0.49 0.18 0.99 4.23* .038 0.53 30.85*** .225 1.00 4.92* .044 0.59
(38) (16) (20) (36)
Internalizing T2 ⫺0.54 ⫺0.39 ⫺0.02 0.77 4.96* .055 .594 17.84*** .173 .987 2.16 .025 .306
(35) (11) (18) (25)
Positive Parenting
Externalizing T1 ⫺0.01 ⫺0.55 0.71 0.02 10.67*** .080 0.90 11.61*** .086 0.92 0.15 .001 0.07
(19) (44) (42) (22)
Externalizing T2 ⫺0.22 ⫺0.54 ⫺0.41 0.16 6.87** .066 .738 9.08** .086 .847 0.57 .006 .117
(15) (32) (37) (17)
Internalizing T1 ⫺0.25 ⫺0.51 0.74 0.28 4.26* .038 0.53 25.99*** .193 0.99 0.33 .003 0.09
(19) (44) (28) (22)
Internalizing T2 ⫺0.08 ⫺0.50 0.30 0.24 1.63 .019 .243 9.19** .099 .850 0.94 .011 .160
VAN LEEUWEN, MERVIELDE, BRAET, AND BOSMANS
(15) (32) (23) (18)
Note. Ns appear in parentheses below the means. T1 ⫽ Time 1; T2 ⫽ Time 2. For Negative Control, dfs for Fs are (3, 117), (3, 101), (3, 109), and (3, 88) for Externalizing T1, Externalizing T2,
Internalizing T1, and Internalizing T2, respectively. The corresponding dfs for Positive Parenting are (3, 126), (3, 100), (3, 112), and (3, 87).
a
Low and high refer to categories of parenting.
* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.
between undercontrollers and resilients faded when the level of problem behavior at all. In addition, the results showed that the
received negative control was low. strongest Time 1 interaction effect still predicted Time 2 external-
The overcontrolled children were rated significantly higher on izing behavior. The Time 1 main effects of personality types and
internalizing behavior than were the resilient children, with the parenting categories largely held up over time.
overcontrolled children in negative control families showing the The results of Study 2 are partially in line with the results of
highest levels of internalizing problem behavior. The results also Study 1. Children scoring below the mean on the benevolence and
indicate that high or low negative parental control did not make a conscientiousness dimensions in Study 1 correspond to the under-
difference for the resilient children: The mean scores on internal- controlled children in Study 2. These children are at risk for
izing behavior were almost equal.3 externalizing behavior when exposed to high levels of negative
The value of eta-squared (␩2; see Table 6), which is the regres- parental control. Children with above-average scores on benevo-
sion coefficient (R2) for a nonlinear regression curve, can be used lence and conscientiousness in Study 1 resemble the resilient
as an estimate of effect size. For the interaction effects in this children in Study 2. These children are not affected by restrictive
study, we found values of .04, which resemble the effect sizes of control. However, in contrast to the results of Study 1, positive
the interaction effects in Study 1 (see Table 4). Values of the parenting, benevolence, and conscientiousness were not identified
observed statistical power of the interaction effects were .59 and as moderators in the prediction of externalizing behavior, because
.65 for internalizing and externalizing, respectively, and are also there was no significant interaction between positive parenting and
comparable to the power estimations of Study 1. the undercontrolled/resilient dichotomy.
No significant interactions were found between the personality This study demonstrates the utility of personality types as mod-
types and positive parenting. However, there were significant main erators of the relationship between environmental variables such as
effects for both positive parenting and personality type. Undercon- parenting and problem behavior. Groups of children with similar
trollers were rated significantly higher on externalizing behavior configurations of personality characteristics were a priori identi-
than were resilients, and overcontrollers were rated significantly fied. Note that to group children using a variable-centered ap-
higher on internalizing behavior than were resilients. Children who proach, researchers or clinicians have to define arbitrary ad hoc
were supported by their parents showed significantly lower levels cut-off scores on the dimensions or divide the children into groups
of problem behavior in comparison with children who received on the basis of a median split. The present study illustrates that a
low levels of positive parental behavior. priori defined types have predictive validity in explaining mal-
adaptive child behavior and that they should therefore be consid-
Time 1 Categorization Predicts Time 2 Problem Behavior ered as a viable alternative to the variable-centered approach in
developmental and clinical research. However, it has to be recog-
Table 6 shows how Time 1 categorization predicts Time 2 nized that identifying the relevant personality dimensions (adopt-
problem behavior. The Time 1 Negative Control ⫻ Undercon- ing the variable-centered approach) remains important because the
trollers/Resilients interaction predicted Time 2 externalizing be- scores on the dimensions are indispensable for grouping children
havior. The results further illustrate that the Time 1 personality into the types and for describing the typical personality profiles of
types significantly predict Time 2 externalizing and internalizing the resilient, undercontrolled, and overcontrolled children. The
behavior. The Negative Control categories also predicted Time 2 knowledge that undercontrolled children are at risk for developing
externalizing and, to a lesser extent, internalizing behavior, externalizing behavior, in particular when they have experienced a
whereas the Positive Parenting categories mainly predicted exter- highly negative parental style, may have practical utility when one
nalizing behavior.
has to decide whether they are at risk. However, in order to explain
why they are at risk, it may be useful to refer to the dimensions
Discussion defining the particular personality profile of each personality type.
Therefore, types and dimensions should be considered as useful
In Study 2, we adopted a person-centered approach and exam-
classes of information generated by two complementary research
ined the interactive effects of parenting and the resilient, under-
controlled, and overcontrolled personality types. The types were
assigned according to the standard procedure outlined by Asen-
dorpf et al. (2001). Children were assigned to groups experiencing 3
Parenting ⫻ Personality Type interactions were also examined with
differences in parenting on the basis of their scores on the parent- HMR analysis, with personality types as dummy-coded variables (resilients
ing variables: children experiencing below- or above-average pos- vs. undercontrollers and resilients vs. overcontrollers) and parenting di-
itive parenting and children exposed to below- or above-average mensions as continuous variables. This revealed one significant interaction
negative control. effect, Undercontrolled/Resilient Types ⫻ Negative Control, in predicting
GLM analyses showed two significant interaction effects. Un- child externalizing behavior, ⌬R2 ⫽ .027; Fchange(1, 369) ⫽ 15.77, p ⬍
dercontrolled children (characterized by below-average levels on .001. This interaction was of the synergistic or enhancing type (Cohen et
both conscientiousness and benevolence) showed significantly al., 2003, p. 285), with both predictors affecting externalizing behavior in
more externalizing behavior than resilient children, especially the same direction and their combination producing an additional effect.
The effect size of this interaction was .04, which equals the effect size of
when exposed to highly negative parental control. Overcontrolled
the GLM interaction effect (see Table 6), although the number of included
children (scoring low on emotional stability and extraversion) participants was larger in the HMR analysis (N ⫽ 374) than in the GLM
exhibited higher levels of internalizing behavior than resilient analyses (N ⫽ 121). The Overcontrolled/Resilient Types ⫻ Negative
children, with the highest levels of internalizing behavior shown Control interaction effect on internalizing behavior, which was significant
by children experiencing high negative parental control. Resilient in the GLM analysis (N ⫽ 110), showed a statistical trend ( p ⫽ .085) in
children were not affected by negative control and showed no the HMR analysis (N ⫽ 345).
strategies, that is, the variable- and the person-centered Finch and Graziano (2001) which showed that the benevolence,
approaches. extraversion, and neuroticism personality dimensions are strongly
related to depression.
General Discussion The weaker associations between the parenting variables and
internalizing behavior, in comparison with externalizing behavior,
A Comprehensive Test of Personality as a Moderator can be explained by the theoretical and empirical underpinnings of
the parenting questionnaire, that is, the heavy emphasis in social
The present research extends in several ways previous empirical learning theory on antisocial behavior and parenting. For example,
studies that documented the role of personality– environment in- the majority of the behavior control items in the GPBS predomi-
teractions in the development of child problem behavior. First, nantly describe how a parent reacts when a child misbehaves, and
other studies that linked parenting and child characteristics as- hence there is a limited emphasis on parental behavior that may
sessed only the child’s temperament or various specific personality reduce internalizing behavior. A study by Barber (1996) showed
characteristics. Our first study included a broadband measure of that behavioral control, referring to attempts to control or manage
child personality based on the five factor model, which can be children’s behavior, was uniquely associated with externalizing,
regarded as a robust reference model (De Raad & Perugini, 2002) and not with internalizing, adolescent problem behavior. However,
applicable to both children and adolescents (Shiner, 1998). In our there was no evidence for a particular association between inter-
second study, we considered child personality from a person- nalizing behavior and psychological control aimed at the psycho-
centered approach, studying the role of personality types in inter- logical and emotional development of the child.
action with parenting. Second, a review of the literature indicates
that most studies that have investigated Parenting ⫻ Temperament
Types of Interaction
or Personality interactions were limited to negative child parenting
practices such as coerciveness or restrictive control. The present A buffering interaction pattern was identified for the Negative
research included both a positive and a negative dimension of Control ⫻ Personality interactions. Negative parental control
parenting behavior. Finally, both internalizing and externalizing tended to be a risk factor for externalizing behavior, especially for
behavior were studied as outcome variables, whereas past research children rated low or around the mean on benevolence and con-
has focused mostly on externalizing behavior. scientiousness. Hence, these personality domains function as pro-
tective factors: Children rated high on these domains do not show
Reliable Moderator Effects problem behavior at all, even in the presence of parental restrictive
control. Interference, or antagonistic, interactions were present for
Past moderator research has been hampered by various meth- positive parenting and benevolence. Again, the results show that
odological and statistical shortcomings. To reduce these problems, children rated high on benevolence do not show problem behavior
we used (a) a nonrestricted large population sample, (b) reliable at all, even when they are deprived of positive parenting. The
measures, (c) more than one informant, and (d) a follow-up mea- combination of above-mean-level positive parenting and below-
surement within a 3-year interval. The HMR analyses provided mean-level benevolence diminishes externalizing behavior.
evidence for several significant Parenting ⫻ Child Personality
interactions. Six major conclusions can be drawn: (a) Several
The Replicability of the Interactions
significant Parenting ⫻ Personality interactions were identified,
but they were more prominent for externalizing than for internal- Some of the findings in this study were replicable across judges
izing behavior; (b) two types of interactions were identified, buff- and across a 3-year interval. Our study included two measurement
ering interactions and interference interactions; (c) the interactions occasions: the first when most of the participants were school-age
were largely replicable across informants (i.e., parents and chil- children, the second when 88% of the sample were adolescents.
dren) and across a 3-year interval (i.e., for children and adoles- The Benevolence ⫻ Negative Control and the Conscientious-
cents); (d) the most prominent personality domains that figured in ness ⫻ Negative Control interactions predicting externalizing
the interactions were benevolence and conscientiousness; (e) the problem behavior for the school-age children remained 3 years
person-centered study also showed meaningful Personality ⫻ Par- later. These interactions measured at Time 1 were also predictive
enting interactions; and (f) interactions measured at Time 1 pre- of Time 2 externalizing behavior. The first-order effects of per-
dicted problem behavior across a 3-year interval. sonality showed great stability over time, whereas the effects of
parenting behavior, especially on internalizing behavior, dimin-
Differences in the Prediction of Externalizing and ished slightly over time.
Internalizing Behavior
The Role of Benevolence and Conscientiousness
The present research identified moderating effects predomi-
nantly for externalizing behavior, whereas for internalizing behav- The most prominent interactions included the benevolence and
ior, parenting and child personality turned out to be primarily conscientiousness personality domains. This finding is consistent
additive effects. Strong negative first-order effects of extraversion with the results of a study by Prinzie et al. (2003) and is related to
and emotional stability on internalizing behavior were clearly the results of studies that have investigated Parenting ⫻ Temper-
present: Children scoring low on these adaptive traits were more ament interactions. The interactions of parenting with benevolence
prone to internalizing behavior. Independent effects of extraver- and conscientiousness may reflect “evocative person– environment
sion and emotional stability on externalizing behavior were much transactions” (Caspi, 1998, p. 357): On the basis of their unique
smaller. This result corroborates previous research findings by personality characteristics, individuals act, and the environment
reacts, resulting in mutually interlocking evocative transactions. want to screen children, and for parents, who will benefit from
Patterson (1982) also described this process as part of the coercive treatment such as enrollment in a parenting training program.
family process. The present analyses showed no significant inter-
actions for imagination, extraversion, and emotional stability, and
hence the present results are consistent with the findings of Prinzie Variable- Versus Person-Centered Approaches
et al. (2003).
Including personality types in interactions with parenting, in-
stead of personality variables, enriched the findings of the first
The Importance of Interaction Effects study. Neither the number of significant effects nor the effect sizes
The emphasis in the present study on the importance of Parent- of Study 2 are remarkable in comparison with the findings of
ing ⫻ Child Personality interaction effects is not meant to mini- Study 1. After all, it was not our goal to enhance the likelihood of
mize or to cloud the main effects of the child’s personality or the finding interaction effects by using categorical variables, a strategy
effects of parenting. The importance of the main effects is clearly that McClelland and Judd (1993) have tried to dissuade researchers
documented in Tables 2, 3, 5, and 6 and has also been recognized from using. The main advantage of the personality type approach
and stressed in several studies referred to in the introduction. is that, instead of considering only one variable at a time, it
Although it is evident that main effects explain a greater propor- combines children with a given profile of scores on multiple
tion of the variance in externalizing and internalizing behavior, personality variables into one category. From Study 1 we con-
documenting reliable interaction effects essentially qualifies the cluded that children rated low on conscientiousness or low on
main effects. benevolence were at risk for externalizing behavior when exposed
The study of interaction effects is important from both theoret- to parental negative control. Including personality types in the
ical and practical perspectives. Reliable interaction effects qualify interactions led to the conclusion that children rated below the
theories that are based on straightforward and unqualified effects mean on both conscientiousness and benevolence, that is, the
of parenting and child personality. Moreover, because most theo- undercontrolled children, are at risk for externalizing behavior
ries are tested by adopting a variable-centered approach, the im- when exposed to parental negative control, in comparison with
plications at the individual or person level are usually underesti- resilient children. This study further showed that children scoring
mated. The practical implications at the individual level can be low on emotional stability and low on extraversion, that is, the
judged more appropriately by adopting a person-centered approach overcontrolled children, are at risk for internalizing behavior when
because it specifies the type of participants for whom the general
exposed to parental negative control. Resilient children are pro-
rule does or does not apply (e.g., Negative Control has detrimental
tected against possible negative effects of parental negative con-
effects on Time 1 internalizing; see Table 6). For instance, high
trol. With the variable-centered approach in Study 1, we found
versus low negative parental control did make a difference for the
evidence only for main effects of personality and parenting in the
degree of internalizing shown by the 162 overcontrollers, but the
prediction of internalizing behavior. However, studying overcon-
same differences in parenting did not affect the internalizing scores
of the 185 resilient children. Hence, although this interaction effect trolled children with a profile indicating low scores on the two
explains only 4% of the variance, it nevertheless shows that relevant dimensions (emotional stability and extraversion) pro-
parenting was an additional risk factor for less than half of the vides extra information on moderators of internalizing behavior.
sample. Moreover, in this case, the main effect of parenting ex- These results emphasize the relevance of using personality types
plains less of the internalizing variance than does the interaction in developmental research and clinical practice. Types clearly have
effect. predictive validity in the study of maladaptive child behavior and
A similar procedure can be applied to the data reported in the should be further integrated in research about personality–
variable-centered study. Although this procedure requires setting environment interactions. Types also have practical value because
arbitrary ad hoc cut-off points to calculate the number of partici- they combine information on several personality domains in a
pants to whom the general rule does or does not apply, the logic is single case. This facilitates, for example, the diagnostic process as
essentially the same. In Table 4 we report the interaction between well as clinical decision making.
personality and parenting on externalizing as well as the simple To our knowledge, there is only one other study that has
slopes for the relationship between parenting and externalizing for investigated interactions between parenting and personality types
three (ad hoc) groups: those scoring high, moderate, and low on in predicting problem behavior. Dubas, Gerris, Janssens, and Ver-
benevolence. Figure 1A shows that the parenting– externalizing mulst (2002) found that undercontrolled adolescents exposed to
relationship was significant for the group scoring below and high levels of restrictive control scored higher on internalizing
around the mean but not for the group scoring above the mean. behavior than did resilients and overcontrollers in less restrictive
Moreover, calculation of the R2s for each of these groups shows families. This finding is in contrast to findings of previous studies
that the effect of parenting explains 11% of the variance of that evaluated the external validity of the types. Dubas et al.
externalizing for the low-benevolence group and 10% of the vari- attributed this to the co-morbidity of externalizing and internaliz-
ance for the medium-benevolence group, whereas it explains only ing problems.4 Our study corroborates the externalizing tendencies
5% for the high-benevolence group. This comparison illustrates for undercontrollers and the internalizing tendencies for
that the overall main effect of parenting underestimates the im- overcontrollers.
portance of parenting for children scoring low or around the mean
on benevolence but overestimates it for the other ad hoc person-
ality group. Finally, this sort of data not only has theoretical 4
In an additional GLM analysis of variance, we checked whether this
significance but is also important for clinical researchers, who interaction could also be found in our data, but this was not the case.
Limitations of This Study and Future Research research on externalizing behavior that showed evidence for a
developmental typology of delinquency, that is, that the childhood-
A study with biologically related parents and children does not onset and the adolescent-onset categories are characterized by
rule out that associations between parental behavior and child different etiologies (Moffitt, Caspi, Harrington, & Milne, 2002).
measures are due to common factors instead of environmental Three-way interactions in HRM analyses could verify whether
influences. Environmental influences, such as parenting, partly Parenting ⫻ Personality interactions differ for boys and girls and
reflect genetic influences, that is, genotype– environment covari- according to age.
ance (Lahey, Waldman, & McBurnett, 1999). An example of such The finding that stable personality characteristics play a major
a passive type of genotype– environment interaction is a child role in child problem behavior should not be equated with the
diagnosed as having a conduct disorder who was reared by parents pessimistic or deterministic view that changing problematic be-
showing aggressive behavior. In such a case, parental and child havior is impossible; rather, it should be taken as evidence for a
behavior are correlated (Rowe, 2003). more realistic perspective (Lytton, 1990). Parents can have an
The present study found evidence for a moderator effects model impact on their children’s behavior, but they will not be able to
that explains child outcomes by the simple interaction of consti- change major personality traits, such as extraversion and neuroti-
tutional vulnerability (child personality) and environmental factors cism (Rowe, 1990). The present studies support the view that
(parenting). However, such a model does not imply that conclu- parents may receive too much blame for the behavioral difficulties
sions can be drawn about reciprocal and recurrent interactions over of their children because the moderating role of the child’s per-
time between the organism and the environment, as postulated by sonality is often ignored. On the other hand, environmental factors
the transactional effects model (Lytton, 1990). combined with certain personality types may increase the likeli-
The reliance in the present study on a questionnaire measure to hood of the development of problem behavior.
assess parenting is a potential limitation. Studies investigating
effects of parenting on social development that are based on direct References
behavioral observations usually show larger effect sizes than do
studies centered on parental reports (Collins et al., 2000). How- Aguinis, H. (1995). Statistical power problems with moderated multiple
regression in management research. Journal of Management, 21, 1141–
ever, the need for a large sample in order to maximize statistical
1158.
power forced us to use self-rating questionnaires. The GPBS was Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
developed with great care and has good psychometric properties. interpreting interactions. Newbury Park, CA: Sage.
Besides, the questionnaire was administered to two informants, Anderson, K. E., Lytton, H., & Romney, D. M. (1986). Mothers’ interac-
parents and children. The combination of child ratings of parenting tions with normal and conduct-disordered boys: Who affects whom?
with parental ratings of personality and problem behavior reduces Developmental Psychology, 22, 604 – 609.
shared method variance and allows replication of the results found Asendorpf, J. B. (2003). Head-to-head comparison of the predictive valid-
for parental ratings. We acknowledge that the use of aggregated ity of personality types and dimensions. European Journal of Person-
scores for parental behavior might mask differential effects of ality, 17, 327–346.
maternal and paternal parental behavior, in particular for positive Asendorpf, J. B., Borkenau, P., Ostendorf, F., & van Aken, M. (2001).
Carving personality description at its joints: Confirmation of three
parenting because the correlation between ratings supplied by
replicable personality prototypes for both children and adults. European
mothers and fathers was below .3. We also recognize the possi- Journal of Personality, 15, 169 –198.
bility that parental disagreement about child rearing also might Asendorpf, J. B., Caspi, A., & Hofstee, W. K. B. (Eds.). (2002). The puzzle
predict problem behavior. of personality types. [Special issue]. European Journal of Personality,
Although we used measures at two assessment times in order to 16.
replicate the basic findings, these follow-up measures are not Barber, B. (1992). Family, personality, and adolescent problem behaviors.
independent, and hence replication in independent samples is Journal of Marriage and the Family, 54, 69 –79.
warranted. Nevertheless, the five broadband measures of person- Barber, B. K. (1996). Parental psychological control: Revisiting a ne-
ality and the two parenting dimensions used in this study can be glected construct. Child Development, 67, 3296 –3319.
considered as useful measures for detecting Personality ⫻ Parent- Bates, J. E., Pettit, G. S., Dodge, K. A., & Ridge, B. (1998). Interaction of
temperamental resistance to control and restrictive parenting in the
ing interactions.
development of externalizing behavior. Developmental Psychology, 34,
Main effects of parental behavior on the problem behavior of the 982–995.
child often conceal interactions with the personality of the child. Belsky, J. (1984). The determinants of parenting: A process model. Child
This often leads to the erroneous conclusion that some parenting Development, 55, 83–96.
practices (e.g., restrictive parenting) are detrimental or bad and Belsky, J., Hsieh, K., & Crnic, K. (1998). Mothering, fathering, and infant
should be avoided at all costs. The present study documents that negativity as antecedents of boys’ externalizing problems and inhibition
the effects of parenting behavior should be qualified and hence that at age 3 years: Differential susceptibility to rearing experience? Devel-
it may be premature to blame all parents adopting a common opment and Psychopathology, 10, 301–319.
parenting style for the problem behavior of children without taking Bronfenbrenner, U. (1986). Ecology of the family as a context for human
into account the moderating effects of the child’s personality. development. Developmental Psychology, 22, 521–530.
Capaldi, D., & Patterson, G. R. (1989). Psychometric properties of fourteen
Although in our analyses we consistently controlled for effects
latent constructs from the Oregon Youth Study. New York: Springer-
of child sex and age, we did not look for differential effects of age Verlag.
and sex. Further research could examine whether Parenting ⫻ Carlo, G., Roesch, S. C., & Melby, J. (1998). The multiplicative relations
Personality interactions are different for girls and boys or depend of parenting and temperament to prosocial and antisocial behaviors in
on age groups, for example, school-age children versus adoles- adolescence. Journal of Early Adolescence, 18, 266 –290.
cents. The importance of this differentiation can be illustrated by Caspi, A. (1998). Personality development across the life course. In W.
Damon (Series Ed.) & N. Eisenberg (Vol. Ed.), Handbook of child Frosch, C. A., & Mangelsdorf, S. C. (2001). Marital behavior, parenting
psychology: Vol. 3. Social, emotional, and personality development (pp. behavior, and multiple reports of preschoolers’ behavior problems: Me-
311–388). New York: Wiley. diation or moderation? Developmental Psychology, 37, 502–519.
Caspi, A., Henry, B., McGee, R. O., Moffitt, T. E., & Silva, P. A. (1995). Gallagher, K. C. (2002). Does child temperament moderate the influence of
Temperamental origins of child and adolescent behavior problems: From parenting on adjustment? Developmental Review, 22, 623– 643.
age three to age fifteen. Child Development, 66, 55– 68. Gerris, J. R. M., Houtmans, M. J. M., Kwaaitaal-Roosen, E. M. G., de
Caspi, A., & Silva, P. A. (1995). Temperamental qualities at age 3 predict Schipper, J. C., Vermulst, A. A., & Janssens, J. M. A. M. (1998).
personality traits in young adulthood: Longitudinal evidence from a birth Parents, adolescents and young adults in Dutch families: A longitudinal
cohort. Child Development, 66, 486 – 498. study. Nijmegen, Netherlands: Institute of Family Studies.
Chaplin, W. F. (1991). The next generation of moderator research in Goldberg, L. R. (1992). The development of markers for the Big Five
personality psychology. Journal of Personality, 59, 143–178. factor structure. Psychological Assessment, 4, 26 – 42.
Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple Gruner, K., Muris, P., & Merckelbach, H. (1999). The relationship between
regression/correlation analysis for the behavioral sciences (3th ed.). anxious rearing behaviors and anxiety disorders symptomatology in
Mahwah, NJ: Erlbaum. normal children. Journal of Behavior Therapy and Experimental Psy-
Colder, C. R., Lochman, J. E., & Wells, K. C. (1997). The moderating chiatry, 30, 27–35.
effects of children’s fear and activity level on relations between parent- Haapasalo, J., & Tremblay, R. E. (1994). Physically aggressive boys from
ing practices and childhood symptomatology. Journal of Abnormal ages 6 to 12: Family background, parenting behavior, and prediction of
Child Psychology, 25, 251–263. delinquency. Journal of Consulting and Clinical Psychology, 62, 1044 –
Collins, W. A., Maccoby, E. E., Steinberg, L., Hetherington, E. M. & 1052.
Bornstein, M. H. (2000). Contemporary research on parenting. The case Harris, J. R. (1998). The nurture assumption: Why children turn out the
for nature and nurture. American Psychologist, 55, 218 –232. way they do. New York: Free Press.
Costa, P. T., Herbst, J. H., McCrae, R. R., Samuels, J., & Ozer, D. J. Hellinckx, W., De Munter, A., & Grietens H. (Eds.). (1991). Gedrags- en
(2002). The replicability and utility of three personality types. European emotionele problemen bij kinderen. Een epidemiologisch onderzoek bij
Journal of Personality, 16, S73–S87. 6- tot 12-jarigen in Vlaanderen. Deel 1: Beschrijving en screening van
Darling, N., & Steinberg, L. (1993). Parenting style as context: An inte- probleemgedrag [Behavioral and emotional problems in children. An
grative model. Psychological Bulletin, 113, 487– 496. epidemiological study of 6- to 12 year olds. Part 1: Description and
Deater-Deckard, K. (2001). Annotation: Recent research examining the screening of problem behavior]. Leuven-Apeldoorn, Belgium: Garant.
role of peer relationships in the development of psychopathology. Jour- Jaccard, J., & Wan, C. K. (1995). Measurement error in the analysis of
nal of Child Psychology and Psychiatry and Allied Disciplines, 42, interaction effects between continuous predictors using multiple regres-
565–579. sion: Multiple indicator and structural equation approaches. Psycholog-
Deater-Deckard, K., & Dodge, K. A. (1997). Externalizing behavior prob- ical Bulletin, 117, 348 –357.
lems and discipline revisited: Nonlinear effects and variation by culture, Jessor, R., Van Den Bos, J., Vanderryn, J., Costa, F. M., & Turbin, M. S.
context, and gender. Psychological Inquiry: An International Journal of (1995). Protective factors in adolescent problem behavior: Moderator
Peer Commentary and Review, 8, 161–175. effects and developmental change. Developmental Psychology, 31, 923–
Deater-Deckard, K., Dodge, K. A., Bates, J. E., & Pettit, G. S. (1998). 933.
Multiple risk factors in the development of externalizing behavior prob- Kohnstamm, G. A., Halverson, C. F., Mervielde, I., & Havill, V. L. (1998).
lems: Group and individual differences. Development and Psychopa- Parental descriptions of child personality: Developmental antecedents
thology, 10, 469 – 493. of the Big Five? Mahwah, NJ: Erlbaum.
De Fruyt, F., Mervielde, I., & Van Leeuwen, K. (2002). The consistency of Krueger, R. F., Schmutte, P. S., Caspi, A., Moffitt, T. E., Campbell, K., &
personality type classification across samples and five-factor measures. Silva, P. A. (1994). Personality traits are linked to crime among men and
European Journal of Personality, 16, S57–S72. women: Evidence from a birth cohort. Journal of Abnormal Psychology,
Deković, M. (1999). Risk and protective factors in the development of 103, 328 –338.
problem behavior during adolescence. Journal of Youth and Adoles- Lahey, B. B., Waldman, I. D., & McBurnett, K. (1999). Annotation: The
cence, 28, 667– 685. development of antisocial behavior: An integrative causal model. Jour-
De Raad, B., & Perugini, M. (2002). Big Five factor assessment: Intro- nal of Child Psychology and Psychiatry and Allied Disciplines, 40,
duction. In B. de Raad & M. Perugini (Eds.), Big Five assessment (pp. 669 – 682.
1–26). Göttingen, Germany: Hogrefe & Huber. Lengua, L. J., Wolchik, S. A., Sandler, I. N., & West, S. G. (2000). The
Dishion, T. J., French, D. C., & Patterson, G. R. (1995). The development additive and interactive effects of parenting and temperament in predict-
and ecology of antisocial behavior. In D. Cicchetti & D. J. Cohen (Eds.), ing problems of children of divorce. Journal of Clinical Child Psychol-
Developmental psychopathology: Vol. 1. Theory and methods (pp. 421– ogy, 29, 232–244.
471). New York: Wiley. Loeber, R., & Dishion, T. (1983). Early predictors of male delinquency: A
Dubas, J. S., Gerris, J. R. M., Janssens, J. R. M., & Vermulst, A. A. (2002). review. Psychological Bulletin, 94, 68 –99.
Personality types of adolescents: Concurrent correlates, antecedents, and Loeber, R., & Stouthamer-Loeber, M. (1986). Family factors as correlates
type ⫻ parenting interactions. Journal of Adolescence, 25, 79 –92. and predictors of juvenile conduct problems and delinquency. In M.
Eisenberg, N., Fabes, R. A., Guthrie, I. K., & Reiser, M. (2000). Disposi- Tonry & N. Morris (Eds.), Crime and justice: A review of research (Vol.
tional emotionality and regulation: Their role in predicting quality of 7, pp. 29 –149). Chicago: University of Chicago Press.
social functioning. Journal of Personality and Social Psychology, 78, Lytton, H. (1990). Child and parent effects in boys’ conduct disorder: A
136 –157. reinterpretation. Developmental Psychology, 26, 683– 697.
Finch, J. F., & Graziano, W. G. (2001). Predicting depression from tem- Maccoby, E. E., & Martin, J. A. (1983). Socialization in the context of the
perament, personality, and patterns of social relations. Journal of Per- family: Parent– child interaction. In P. H. Mussen (Ed.), Handbook of
sonality, 69, 27– 49. child psychology: Vol. 4. Socialization, personality and social develop-
Forehand, R., Miller, K. S., Dutra, R., & Chance, M. W. (1997). Role of ment (pp. 1–101). New York: Wiley.
parenting in adolescent deviant behavior: Replication across and within McClelland, G. H., & Judd, C. M. (1993). Statistical difficulties of detect-
two ethnic groups. Journal of Consulting and Clinical Psychology, 65, ing interactions and moderator effects. Psychological Bulletin, 114,
1036 –1041. 376 –390.
Mervielde, I., & Asendorpf, J. B. (2000). Variable-centered and person- Sameroff, A. J. (2000). Developmental systems and psychopathology.
centered approaches to childhood personality. In S. E. Hampson (Ed.), Development and Psychopathology, 12, 297–312.
Advances in personality psychology (Vol. 1, pp. 37–76). Philadelphia: Shiner, R. L. (1998). How shall we speak of children’s personalities in
Psychology Press. middle childhood? A preliminary taxonomy. Psychological Bulletin,
Mervielde, I., & De Fruyt, F. (1999). Construction of the Hierarchical 124, 308 –332.
Personality Inventory for Children (HiPIC). In I. Mervielde, I. Deary, F. Shiner, R., & Caspi, A. (2003). Personality differences in childhood and
De Fruyt, & F. Ostendorf (Eds.), Personality psychology in Europe (pp. adolescence: Measurement, development, and consequences. Journal of
107–127). Tilburg, Netherlands: Tilburg University Press. Child Psychology and Psychiatry and Allied Disciplines, 44, 2–32.
Mervielde, I., & De Fruyt, F. (2002). Assessing children’s traits with the Siqueland, L., Kendall, P. C., & Steinberg, L. (1996). Anxiety in children:
Hierarchical Personality Inventory for Children. In B. de Raad & M. Perceived family environments and observed family interaction. Journal
Perugini (Eds.), Big Five assessment (pp. 129 –146). Göttingen, Ger- of Clinical Child Psychology, 25, 225–237.
many: Hogrefe & Huber. Stoolmiller, M. (2001). Synergistic interaction of child manageability
Moffitt, T. E., Caspi, A., Harrington, H., & Milne, B. J. (2002). Males on problems and parent-discipline tactics in predicting future growth in
the life-course-persistent and adolescence-limited antisocial pathways: externalizing behavior for boys. Developmental Psychology, 37, 814 –
Follow-up at age 26 years. Development and Psychopathology, 15, 825.
179 –207. Stormshak, E. A., Bierman, K. L., McMahon, R., Lengua, L. J., & Conduct
Muris, P., Schmidt, H., Lambrichs, R., & Meesters, C. (2001). Protective Problems Prevention Research Group (2000). Parenting practices and
and vulnerability factors of depression in normal adolescents. Behaviour child disruptive behavior problems in early elementary school. Journal
Research and Therapy, 39, 555–565. of Clinical Child Psychology, 29, 17–29.
O’Connor, B. P., & Dvorak, T. (2001). Conditional associations between Stouthamer-Loeber, M., Loeber, R., Farrington, D. P., Zhang, Q., van
parental behavior and adolescent problems: A search for personality– Kammen, W., & Maguin, E. (1993). The double edge of protective and
environment interactions. Journal of Research in Personality, 35, 1–26. risk factors for delinquency: Interrelations and developmental patterns.
Olweus, D. (1980). Familial and temperamental determinants of aggressive Development and Psychopathology, 5, 683–701.
behavior in adolescent boys: A causal analysis. Developmental Psychol- Thomas, A., & Chess, S. (1977). Temperament and development. New
ogy, 16, 644 – 660. York: Brunner/Mazel.
Paterson, G., & Sanson, A. (1999). The association of behavioural adjust- Van den Bergh, B. (1997). Kindertijd. Kinderen en ouders over de leefsi-
ment to temperament, parenting and family characteristics among 5-year
tuatie van kinderen op lagere schoolleeftijd in Vlaanderen [Childhood.
old children. Social Development, 8, 293–309.
Children and parents about the quality of life of elementary school-aged
Patterson, G. R. (1982). Coercive family process. Eugene, OR: Castalia.
children]. Leuven/Apeldoorn, Belgium: Garant.
Patterson, G. R., Reid, J., & Dishion, T. (1992). Antisocial boys: Vol. 4. A
Van Leeuwen, K., De Fruyt, F., & Mervielde, I. (2004). A longitudinal
social interactional approach. Eugene, OR: Castalia.
study of the utility of the resilient, overcontrolled and undercontrolled
Patterson, G. R., & Stouthamer-Loeber, M. (1984). The correlation of
personality types as predictors of children’s and adolescents’ problem
family management practices and delinquency. Child Development, 55,
behavior. International Journal of Behavioral Development, 28, 210 –
1299 –1307.
220.
Prinzie, P., Onghena, P., Hellinckx, W., Grietens, H., Ghesquière, P., &
Van Leeuwen, K. G., & Vermulst, A. A. (in press). The Ghent Parental
Colpin, H. (2003). The additive and interactive effects of parenting and
Behavior Scale: Some psychometric properties. European Journal of
children’s personality on externalising behaviour. European Journal of
Psychological Assessment.
Personality, 17, 95–117.
Verhulst, F. C., Van der Ende, J., & Koot, H. M. (1996). Handleiding voor
Richter, J. (1994). Parental rearing and aspects of psychopathology with
de CBCL/4 –18 [Manual of the CBCL/4 –18]. Erasmus Universiteit Rot-
special reference to depression. In C. Perris, W. A. Arrindell, & M.
Eisemann (Eds.), Parenting and psychopathology (pp. 235–251). Chi- terdam, Afdeling Kinder- en Jeugdpsychiatrie.
chester, England: Wiley. Verhulst, F. C., Van der Ende, J., & Koot, H. M. (1997a). Handleiding voor
Rothbart, M. K., & Bates, J. E. (1998). Temperament. In W. Damon (Series de Teacher’s Report Form [Manual of the TRF]. Erasmus Universiteit
Ed.) & N. Eisenberg (Vol. Ed.), Handbook of child psychology: Vol. 3. Rotterdam, Afdeling Kinder- en Jeugdpsychiatrie.
Social, emotional and personality development (pp. 105–176). New Verhulst, F. C., Van der Ende, J., & Koot, H. M. (1997b). Handleiding
York: Wiley. voor de Youth Self Report [Manual of the YSR]. Erasmus Universiteit
Rothbaum, F., & Weisz, J. R. (1994). Parental caregiving and child Rotterdam, Afdeling Kinder- en Jeugdpsychiatrie.
externalizing behavior in nonclinical samples: A meta-analysis. Psycho- Wakschlag, L. S., & Hans, S. L. (1999). Relation of maternal responsive-
logical Bulletin, 116, 55–74. ness during infancy to the development of behavior problems in high-
Rowe, D. C. (1990). As the twig is bent? The myth of child-rearing risk youths. Developmental Psychology, 35, 569 –579.
influences on personality development. Journal of Counseling and De- Weiss, B., Dodge, K. A., Bates, J. E., & Pettit, G. S. (1992). Some
velopment, 68, 606 – 611. consequences of early harsh discipline: Child aggression and a maladap-
Rowe, D. C. (2003). Assessing genotype– environment interactions. In R. tive social information processing style. Child Development, 63, 1321–
Plomin, J. C. Defries, I. W. Craig, & P. McGuffin, (Eds.), Behavioral 1335.
genetics in the postgenomic era (pp. 71– 86). Washington DC: American Wootton, J. M., Frick, P. J., Shelton, K. K., & Silverthorn, P. (1997).
Psychological Association. Ineffective parenting and childhood conduct problems: The moderating
Rubin, K. H., Burgess, K. B., Dwyer, K. M., & Hastings, P. D. (2003). role of callous-unemotional traits. Journal of Consulting and Clinical
Predicting preschoolers’ externalizing behaviors from toddler tempera- Psychology, 65, 301–308.
ment, conflict and maternal negativity. Developmental Psychology, 39,
164 –176.
Ruchkin, V. V., Eisemann, M., & Hägglöf, B. (1999). Coping styles in Received August 27, 2003
delinquent adolescents and controls: The role of personality and parental Revision received May 6, 2004
rearing. Journal of Youth and Adolescence, 28, 705–717. Accepted May 20, 2004 䡲

Child Personality and Parental Behavior As Moderators of Problem Behavior: Variable-And Person-Centered Approaches

Uploaded by

Document Informationclick to expand document informationJurnal

Document Informationclick to expand document information

Copyright:

Available Formats

Child Personality and Parental Behavior As Moderators of Problem Behavior: Variable-And Person-Centered Approaches

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Child Personality and Parental Behavior As Moderators of Problem Behavior: Variable-And Person-Centered Approaches

Uploaded by

Copyright:

Available Formats

Developmental Psychology Copyright 2004 by the American Psychological Association

2004, Vol. 40, No. 6, 1028 –1046 0012-1649/04/$12.00 DOI: 10.1037/0012-1649.40.6.1028

Child Personality and Parental Behavior as Moderators of Problem

Child Personality and Parental Behavior as Moderators of The Current Research

% variance 35.47 39.94 50.69 47.07 32.06 35.67 39.82 39.54

Benevolence interaction was not replicable. No interaction terms

BFM ⫹ 1 SD BFM M BFM ⫺ 1 SD Interaction r2Y.MI

Time 1 Variables Predicting Time 2 Problem Behavior

Table 5 reports the results of a hierarchical multiple regression

ratings of parental behavior, the Negative Control ⫻ Benevolence

interaction, R2 ⫽ .024 and Fchange(1, 465) ⫽ 18.84, p ⬍ .001, and

the Negative Control ⫻ Conscientiousness interaction, R2 ⫽ .022

and Fchange(1, 465) ⫽ 13.95, p ⬍ .001, predicted Time 2 child

a predictor for Time 2 externalizing behavior. Time 2 internalizing

sions, with emotional stability and extraversion as the strongest

1 Negative Control is a predictor of both Time 2 externalizing

behavior and Time 2 internalizing behavior for parental self-

ratings and also a predictor of externalizing behavior when child

a strong predictor of Time 2 child problem behavior.

behavior. Study 1 examined interactions that have been detected in

previous studies, for example, between negative parental control

We also checked for nonlinear relationships between independent and

central interest is in the interaction term. Curvilinear effects show lower

Negative Control ⫻ Conscientiousness

* p ⬍ .05. ** p ⬍ .01. *** p ⬍ .001.

Negative Control ⫻ Benevolence

for benevolence and conscientiousness in the prediction of child external-

izing behavior and for benevolence predicting child internalizing behavior.

Negative Control: Child ratings

These quadratic effects did not influence the Parenting ⫻ Personality

Positive Parenting (child ratings) interaction was not a significant predictor

Externalizing Time 2 Internalizing Time 2

Resilients Overcontrollers Undercontrollers Parenting Personality type Parenting ⫻ Personality Type

(15) (32) (23) (18)

You might also like

* p ⬍ .05. p ⬍ .01. * p ⬍ .001.