2011. Vol.2, No.6, 542-551
Copyright © 2011 SciRes. DOI:10.4236/psych.2011.26084
Childhood Disruptive Behaviour and School Performance across
Comprehensive School: A Prospective Cohort Study
Saija Alatupa1, Laura Pulkki-Råback1,2, Mirka Hintsanen1, Sari Mullola3,
Jari Lipsanen1, Liisa Keltikangas-Järvinen1*
1IBS, Unit of Personality, Work and Health Psychology, University of Helsinki, Helsinki, Finland;
2Finnish Institute of Occupational Health, Topeliuksenkatu 41 A, Helsinki, Finland;
3Centre for Research on Teaching, Department of Teacher Education,
University of Helsinki, Helsinki, Finland.
Email: *
Received June 17th, 2011; revised July 19th, 2011; accepted August 22nd, 2011.
In the present study we examined 1) whether childhood disruptive behaviour, in terms of aggressiveness, hyper-
activity and social adjustment, predicts school performance since toddler age or whether becomes it relevant first
since middle or late childhood, 2) whether gender differences within the associations between school perform-
ance and disruptive behaviour exist, and 3) whether there are trait specific effects in these associations, i.e.
whether hyperactivity is more relevant determinant for later school success than aggression and social adjust-
ment. The subjects were derived from a representative, population based cohort study where 3600 subjects we
followed for 27 years since their childhood. Our sample consisted of 973 participants (516 girls) who were 3, 6
and 9 years of age at baseline and were followed over their whole compulsory education, i.e. 3rd, 6th, and 9th
grades. The most prominent finding was a gender specific association between disruptive behaviour and school
performance: hyperactivity predicted later school performance among girls whereas aggression predicted school
performance among boys. The association between social adjustment and school performance was less clear.
Disruptive behaviour at toddler age (at the age of 3) was not predictable for later school performance but it
started to predict school performance at later age, i.e. when it was assessed at the ages of 6 and 9, and the asso-
ciations were true throughout the whole 9-year comprehensive school. Our findings suggest that early childhood
disruptive behaviour has long-lasting effects. Thus, its intervention before the school entry would be of high
Keywords: Aggression, Disruptive Behaviour, Gender, Hyperactivity, Prospective Study, School Performance,
Social Adjustment, Socioeconomic Position (SEP)
Disruptive behaviour is a composite of co-occuring nega-
tivistic externalizing behaviours that appear in childhood, and
are likely to persist over time (Hinshaw, 1992b). The Diagnostic
and Statistical Manual of Mental Disorders (DSM-IV-TR)
(American Psychiatric Association, 2007) divides disruptive
behaviour into social and cognitive elements. The social com-
ponent of disruptive behaviour incorporates aggressiveness,
opposition, and antisociality, while the cognitive component
comprises hyperactivity, inattention, and impulsivity (Vitaro,
Brendgen, Larose, & Tremblay, 2005). The elements of social
and cognitive disruptiveness are conceptually distinctive even
though certain overlap exists (Hinshaw, 1992a; Hinshaw,
1992b). Further, the cognitive elements of disruptive behaviour,
i.e. hyperactivity, inattention, and impulsivity are close to tem-
peramental characteristics of activity (referring to the vigor and
tempo of motor activity), distractibility (referring to the ease to
get distracted by environmental low-level stimuli), and impul-
sivity (referring to the tendency to act before thinking), respec-
Childhood disruptive behaviour has been associated with poor
educational (Asendorpf, Denissen, & van Aken, 2008; Fergus-
son & Horwood, 1995; Hinshaw, 1992a; Hinshaw, 1992b; Vi-
taro, Larocque, Janosz, & Tremblay, 2001; Vitaro et al., 2005)
and social outcomes (Fergusson & John Horwood, 1995;
Tremblay & Masselink, 1992), although previous studies have
been cross-sectional (Johnson, McGue, & Iacono, 2005), or
follow-ups of rather short endurance (Johnson et al., 2005;
Richman, Stevenson, & Graham, 1982; Rutter, 1974; Trzes-
niewski, Moffitt, Caspi, Taylor, & Maughan, 2006), conducted
in samples from geographically limited area (Asendorpf et al.,
2008; Breslau et al., 2009; Rutter, 1974) and in selected popu-
lations (Frick et al., 1991; Richman et al., 1982). In addition,
most of the studies have focused on the association between
disruptive behaviour and reading ability (Berger, Yule, & Rutter,
1975; Heiervang, Stevenson, Lund, & Hugdahl, 2001; Hinshaw,
1992b; Johnson et al., 2005; Willcutt & Pennington, 2000) or
between disruptive behaviour and achievement in standardized
tests (Fergusson, Horwood, & Lynskey, 1993; Fergusson & John
Horwood, 1995; Frick et al., 1991). It has been suggested,
however, that the link between disruptive behaviour and aca-
demic achievement is stronger when the measurement of
achievement reflects actual performance, such as grades, in the
classroom (Hawkins & Lishner, 1987). School grades may be
considered as a direct feedback of school performance as they
reflect the teacher-student relationship quality, i.e. whether the
disruptive behaviour is a cause of problems between teacher and
pupil (Hinshaw, 1992b). Thus, the GPAs may be seen as an
indicator of actual school performance, as they reflect the grades
in several school subjects. Poor school performance is an im-
portant indicator of educational and social outcomes since it is
known to predict low educational level and low work perform-
ance (Kuncel, Credé, & Thomas, 2005), unemployment (Kokko,
Bergman, & Pulkkinen, 2003), detrimental health behaviour
(Lynch, Kaplan, & Salonen, 1997), and health, such as adult-
hood obesity (Alatupa et al., 2010).
Additional evidence on the risk-proneness of disruptive be-
haviour emerges from studies that have focused on the same
phenomena, but used different terms, such as externalizing
behaviour, undercontrolled behaviour, explosive temper tan-
trums, and lack of emotional control (referring to aggression,
compliance, lability, anxiety, passivity, stability, constructive-
ness and activity) with analogous purpose. Children with ex-
plosive behavioral styles also tend to have erratic life course
patterns characterized by downward occupational mobility,
irregular work lives and poor choices regarding their social lives
as adults (Caspi, Elder, & Bem, 1987). Low self-control is as-
sociated with school social adjustment in adolescence and
long-term unemployment in adulthood, partly via problem
drinking and poor occupational alternatives (Kokko & Pulkki-
nen, 2000; Kokko, Pulkkinen, & Puustinen, 2000).
Altogether, disruptive behaviour seems to have an extensive
influence on several educational and social outcomes. Early
interference is highly important, as the long-term consequences
of disruptive behaviour constitute high-priced major public
health issues. For more efficient intervention, it is important to
identify the age at which the influence of disruptive behaviour
sets in, i.e. whether the preschool years or even the toddler age is
determinant for later school outcomes.
The effect of disruptive behavior may also vary by gender.
Prevalence of externalizing disorders is rather equivalent before
school age (Boylan, Vaillancourt, Boyle, & Szatmari, 2007;
Loeber, Burke, Lahey, Winters, & Zera, 2000), but after the age
of six, the prevalence is reported to be 2 to 4 fold higher among
boys (Boylan et al., 2007; Rucklidge, 2010).
With regard to gender differences, not much research has been
conducted to examine the associations between disruptive beha-
viour and school performance in terms of the GPAs. However,
there is some evidence on the association between antisocial
behaviour and reading ability which has been shown to be
stronger among boys (Reinke, Herman, Petras, & Ialongo, 2008;
Trzesniewski et al., 2006; Trzesniewski et al., 2006; E. G.
Willcutt & Pennington, 2000; Willcutt & Pennington, 2000),
whereas a further study found no difference between girls and
boys (Stevenson, Richman, & Graham, 1985).
To summarize, although the association of disruptive behavior
with school performance has been studied extensively, there are
limitations in previous research such as: 1) lack of studies that
would have measured disruptive behaviour in childhood, that is,
already before school entry, 2) lack of studies examining school
performance systematically at several occasions throughout
comprehensive school, 3) lack of understanding on age- or
gender differences in the associations, and finally 4) lack of
studies conducted in population-based samples that are repre-
sentative of the entire age cohort. These limitations have been
acknowledged in the literature (e.g., Greenfield Spira & Fischel,
2005; Loe & Feldman, 2007).
The purpose of this study was to examine the association of
disruptive behavior with later school performance in a prospec-
tive dataset where disruptive behavior was measured prior to
school entry. The specific goals were: 1) to identify the age at
which disruptive behaviour predicts poor school performance,
and 2) to examine possible gender differences in this association.
We used a prospective dataset of a geographically representative
sample of Finnish pupils whose school grades were assessed on
three occasions covering the entire comprehensive school: at 3rd,
6th, and 9th grades.
On the basis of the literature our hypotheses were: First, dis-
ruptive behaviour predicts school performance over the com-
prehensive school and this predictive association may be seen
already in behaviors before school-age. It has been suggested
that disruptive behaviour is a stronger predictor school per-
formance among boys, and consequently our second hypothesis
was that the association may be more manifest in boys than in
girls. Third, we expected that hyperactivity is a stronger pre-
dictor of school performance in early school years whereas
aggression becomes more relevant at later age.
The Finnish educational system consists of nine years of
compulsory schooling between the age of 7 and 15. Virtually
all (99%) of schools are public schools having parallel curricu-
lums which enables contacting all pupils in a certain age cohort.
From each age group approximately 97% graduate in regular
classes (of which approximately 7% are under special, individ-
ual supervising, approximately 2% in special, “tailored” class,
and less than 1% leaves without this education) from this
state-owned school. After comprehensive school, almost all
students continue either in Senior High Schools (approximately
64%) or Vocational Institutions (approximately 30%) and less
than 5% drop out from this secondary education.
The participants were from a population based prospective
Young Finns study which is a nationally representative, ran-
domly selected sample of 3596 healthy children and adoles-
cents from six age cohorts (3, 6, 9, 12, 15, and 18 years at the
baseline). Based upon the location of the university cities with a
medical school, Finland was divided into five areas (Helsinki,
Kuopio, Oulu, Tampere and Turku). In each area, 360 urban
boys and girls and 360 rural boys and girls were randomly se-
lected on the basis of their personal Social Insurance Institu-
tion’s population register, which covers the whole population of
Finland. Complete details of the sample are given elsewhere
(Raitakari et al., 2003). The study plan was approved by the
local committees of all the participating universities, and the
study protocol of each study phase corresponded to the pro-
posal by the World Health Organization. All subjects gave their
written consent and the study was conducted in accordance
with the Helsinki declaration.
The current study comprises 3-, 6-, and 9-year-old age co-
horts whose disruptive behaviour was measured at the baseline
examination in 1980. Originally, there were 1806 participants
(n = 577, 583, and 646 participants in the respective 3, 6, and
9-year old cohorts). Exclusion of participants based on missing
data resulted in a final sample of 973 participants (n = 225, 347,
and 401 participants in the 3-, 6-, and 9-year-old cohorts, re-
spectively). No data was imputed. The original Young Finns
Sample had an equal proportion of boys and girls, but attrition
analyses showed that girls were slightly over-represented in our
sample (3-year-old 40.4% girls vs. 37.6% boys, 6-year-old
63.7% girls vs. 55.0% boys, and 9-year-old 64.7% girls vs.
59.4% boys). Compared to drop-outs, those who had stayed had
higher GPAs (p-values ranging from < .001 to .002), were
lower in aggression (p-values ranging from .004 and .519), and
in hyperactivity (p-values ranging from .004 and .046) and
higher in social adjustment (p-values ranging from < .001
and .118) and they had mothers with higher education (p-values
ranging from < .001 to .007), although the differences were
rather small (mean differences between GPAs varying from
0.22 to 0.45 and mean differences in maternal years of educa-
tion were 0.68 and 1.15).
Childhood Disruptive Behaviour
The dimensions of disruptive behaviour were assessed in
1980 by the mothers of the participants with a questionnaire
derived from the Health Examination Survey (Wells, 1980).
This questionnaire was originally designed to screen children
with potential behavioural problems, and can be completed by
non-professionals (by persons without a background in psycho-
logy). Table 1 presents the items of the three dimensions of
disruptive behaviour, i.e. aggression, hyperactivity and social
adjustment. These scales have been tested for construct validity
(Katainen & Raikkonen, 1999; Räikkönen, Katainen, Keski-
vaara, & Kelikangas-Järvinen, 2000) and predictive validity
(Pesonen, Räikkönen, Keskivaara, & Keltikangas-Järvinen,
2003; Pulkki-Råback, Elovainio, Kivimäki, Raitakari, & Kelti-
kangas-Järvinen, 2005) in relation to similar constructs.
School Performance
School performance was assessed by grade point averages
(GPA) which is a standard measure of school performance in
Finland. Grade point averages (GPAs) were based on school
reports in the 3rd, 6th and 9th grades with the respective ages of
the participants being 9, 12, and 15 years. The GPAs are the
means of marks in all school subjects, and are assessed on a
scale from 4 to 10 (4 = fail, 5 - 6 = poor, 7 - 8 = good and 9 - 10
= excellent). GPAs were assessed twice a year and all pupils
are evaluated on the same subjects (e.g. math, biology, history)
using similar criteria in each school. The GPAs were reported
by the participants’ mothers at the age of 9 (GPA at 3rd school
grade), and self-reported by the participants at the ages of 12
(6th grade) and 15 (9th grade).
Maternal Educational Level
Based on a recent meta-analysis, the impact of parental SEP
on students’ school performance ranges between r = .28 and .30
(Sirin, 2005). In Finland, maternal educational level explains
38% of the variance of a student’s performance while the re-
spective figure in 32% for paternal education (Kuusela, 2003).
To examine possible socioeconomic confounding, maternal
education was used as a covariate in the analyses.
Study Design
Table 2 shows the study design and the number of partici-
pants at each study phase. With an exception of the oldest co-
hort (9-year-olds), disruptive behaviour was assessed before
school beginning, i.e. at the ages of three and six. With an ex-
ception of the youngest cohort (three years), GPAs were meas-
ured at three occasions (3rd , 6th, and 9th grades).
Statistical Methods
As one study focus was on gender differences, all analyses
were performed separately for girls and boys. To examine the
association of childhood disruptive behaviour (aggression, hy-
peractivity, and social adjustment) with GPAs in the 3rd, 6th,
and 9th grades, we computed linear regressions with the GPAs
as the continuous dependent variables. These analyses were
conducted separately in each age cohort to examine age-related
differences in the association between disruptive behaviour and
school performance.
We additionally used the repeated measures ANOVA proce-
Table 1.
Items of the childhood disrupti v e b e haviour of aggressi on , hyperactivity and social adjustment.
Disruptive Behaviour Range of scale / Item value
Aggression, 6 items (Range 1-5a)
1. Child shows physical aggression towards other children 1-5
2. Other children’s parents often complain about the child's behavior 1-5
3. The child often fights 1-5
4. The child often swears 1-5
5. Other children often tell tales about him/her 1-5
6. The child often comes home to tell he/she has hurt himself/herself 1-5
Hyperactivity, 1 item (Range 1-4)
Child is always controlled 1
Child is overactive or restless only occasionally, for instance when tired 2
Child is continuously more active than the average child or youth 3
Child is always extremely active and energetic, even restless 4
Social adjustment, 1 item (Range 1-3)
Child is always very co-operative and responsive to others 1
Child has sometimes problems with peers, but is mostly co-operative 2
Child shows continuous problems with peers 3
a1 = the statement doesn't fit the child, 5 = the statement totally fits the child.
Table 2.
Study design. Participant s age (years) at the measurements of disrupt ive behaviour and grade point averages (GPAs).
Grade point average measurements
Baseline Follow-up
Participants’ age (years) at the
measurement of disruptive behaviour in 1980
n 1980 1983 1986 1989
3 225 9 (3rd) 12 (6th)
6 347 9 (3rd) 12 (6th) 15 (9th)
9 401 9 (3rd) 12 (6th) 15 (9th)
# = The analyses were conducted separately for the 3, 6, and 9 years-olds.
dure to examine association of disruptive behaviour with GPAs
over the three measurements, i.e. 3rd, 6th, and 9th. For this, the
GPA measurements were employed as continuous dependent
variable and each disruptive behaviour trait was entered sepa-
rately as independent binary variable. For this, each of the traits
was divided into low and high through median split. The GPA
means were then plotted over the three measurements by
childhood aggression, hyperactivity, and social adjustment.
All of the aforementioned analyses were conducted in two
steps: without-adjustments and with an adjustment for years of
maternal education. We used the Bonferroni correction in order
to control for Type I error rate (Abdi, 2007). The critical α level
of .050 was divided by 3, which was the number of analyses
performed in examining the three measurements of disruptive
behaviour. We then used the adjusted α level of .016 (.050/3
= .016) as the critical significance value. All analyses were
performed using the SPSS software (version 15.0)
The characteristics of the participants are presented in Table
Boys scored lower on social adjustment in each of the co-
horts (p values varying from .003 to .017). Girls had consis-
tently higher GPAs than boys (p values in each school grade
< .001).
The results of disruptive behaviour predicting the GPAs are
shown separately for girls (Table 4) and for boys (Table 5).
While disruptive behaviour at the age of three had no associa-
tion with GPAs, consistent associations were found between
disruptive behaviour at ages 6 and 9 and subsequent GPAs.
Among girls, high hyperactivity at the age of six years pre-
dicted poorer GPAs at 3rd, 6th, and 9th grade. Additionally, high
aggression at the age of nine years was associated with poorer
GPAs in 6th grade whereas social adjustment, assessed at the
age of nine, predicted poor GPAs at 9th grade.
Table 5 shows that among boys, high aggression at the age of
nine years predicted poorer GPAs in the 3rd and 6th grade. In
addition, we found that social adjustment at the age of nine
predicted poor performance at 3rd and 9th grades. In girls and
boys, the associations were robust against adjustment for ma-
ternal education (see Table 5).
To sum up, different aspects of disruptive behaviour seemed
to predict school grades in boys and in girls. Hyperactivity was
the strongest predictor of poor outcomes for girls, while high
aggression was a more important determinant of boys’ per-
For illustrative purposes, we plotted the GPA means over the
three measurements by disruptive behaviour as a binary out-
come variable separately for girls (Figure 1) and boys (Figure
2). Figure 1 demonstrates that boys with high aggression had
lower GPAs over the whole comprehensive school, i.e. at 3rd,
6th, and 9th grade (adjusted for age and maternal education). The
pairwise comparisons showed that the difference between boys
with high and low aggression was significant in 6th grade (p
= .014) and almost significant in 9th grade (p =.018).
Figure 2 shows that girls with high hyperactivity had lower
GPAs throughout the three measurements. The pairwise com-
parisons demonstrate that the GPA differences among girls with
high and low hyperactivity were significant in 6th and 9th grade
(p values in 6th and 9th grade were .001 and .011, respectively).
In regard to social adjustment, the pairwise comparisons
showed that the GPA differences were significant among girls
in 3rd and 9th grade (p values .010 and .003, respectively), and
among boys in 6th grade (p = .014).
The current study examined in a population-based sample
whether the early, middle and late childhood disruptive beha-
viour in terms of aggression, hyperactivity, or social adjustment,
predicts the school performance over the whole comprehensive
school education.
Probably the most important result was that no trait specific
effect but a gender specific effect was found. That means that
against our hypothesis hyperactivity and aggression were as
important factors in predicting school performance but they
played different roles in different gender groups. High hyperac-
tivity predicted poor GPAs among girls whereas high aggres-
sion predicted poor GPAs among boys. The association be-
tween disruptive behaviour and school performance was evi-
dent throughout the whole school career, even though it was
likely to decrease along with age maintaining, however, statis-
tical significance. The impact of social adjustment on school
performance, however, was less consistent.
We found that disruptive behaviour in middle and late child-
hood predicted later school performance, whereas no associa-
tion was found when children’s behaviour was measured at
toddler age. It is known that in age of three years disruptive
behaviour is at least to certain degree age-appropriate, and not
relevant predictor for later academic success as shown in the
current study. In accordance with previous research (Caspi &
Henry, 1995) we found that disruptive behaviour becomes
relevant to school performance when it is measured more
proximal to the school beginning. In regard to children’s age by
the measurement of disruptive behaviour, we further support
earlier research (Hinshaw, 1992b) by showing that hyperactiv-
ty associates stronger with school performance in elementary i
Table 3.
Participant characteristics. The descriptive analyses were conducted separately for girls and boys separately within the 3-, 6-, and 9 year-old co-
Girls Boys
n M ± SD n M ± SD
3-Year-Old Cohort (N = 225)
Aggression (range 1 - 5) 114 1.06 0.10 111 1.09 0.14 0.203
Hyperactivity (range 1 - 4) 114 2.11 0.59 111 2.17 0.75 0.527
Social adjustment (range 1 - 3) 114 1.36 0.63 111 1.58 0.72 0.017
School performance
GPA in 3rd grade 114 8.07 0.53 111 7.76 0.57 <0.001
GPA in 6th grade 114 8.18 0.64 111 7.81 0.65 <0.001
Maternal education (range 6-22 years) 114 11.34 3.15 111 11.34 3.33 0.999
6-Year-Old Cohort (N = 347)
Aggression (range 1 - 5) 193 1.06 0.13 154 1.06 0.11 0.988
Hyperactivity (range 1 - 4) 193 2.05 0.61 154 2.12 0.58 0.232
Social adjustment (range 1 - 3) 193 1.47 0.68 154 1.71 0.84 0.003
School performance
GPA in 3rd grade 193 8.03 0.50 154 7.75 0.59 <0.001
GPA in 6th grade 193 8.20 0.64 154 7.84 0.73 <0.001
GPA in 9th grade 193 8.39 0.83 154 7.81 0.98 <0.001
Maternal education (range 4 - 21 years) 193 11.0 3.24 154 11.58 3.39 0.103
9-Year-Old Cohort (N = 401)
Aggression (range 1 - 5) 209 1.04 0.11 192 1.07 0.14 0.060
Hyperactivity (range 1 - 4) 209 2.07 0.64 192 2.04 0.46 0.584
Social adjustment (range 1 - 3) 209 1.54 0.72 192 1.57 0.74 0.711
School performance
GPA in 3rd grade 209 7.95 0.58 192 7.62 0.55 <0.001
GPA in 6th grade 209 8.10 0.70 192 7.69 0.72 <0.001
GPA in 9th grade 209 8.19 0.82 192 7.68 0.89 <0.001
Maternal education (range 6 - 22 years) 209 10.44 2.91 192 10.62 3.43 0.563
a = School performance is measured as grade point average (GPA) ranging from 4 to 10; The ages of participants in 3rd, 6th, and 9th grade are 9, 12, and 15, respectively.
grades whereas aggression associates with school performance
first by adolescence.
The association between aggression and school performance
was stronger among boys than girls although aggression was
not more common among boys. This finding is in contradiction
with previous research showing that boys in general are more
likely to show higher levels of aggressive behaviour (Archer,
2004; Rhee & Waldman, 2002). Previous research suggests a
gender related difference in regard to the acceptance of aggres-
sive behaviour. For instance, toddler and preschool-aged girls’
aggressive behaviour is more likely to be ignored by the teach-
ers and the peers whereas it is reinforced among boys (Fagot &
Hagan, 1985; Serbin, O'leary, Kent, & Toniek, 1973). Conse-
quently, even though girls and boys show similar levels of ag-
gression, it may be ignored among girls but not among boys.
Previous literature has shown that the association between dis-
ruptive behaviour and academic performance is stronger among
boys (Williams & McGee, 1994), among girls (Maughan,
Table 4.
Standardized beta coefficients of disruptive behaviour in 3 different age groups in predicting grade point averages (GPAs) in the 3rd, 6th, and 9th
grades for girls. The results are shown separately for 3-, 6-, and 9-year old cohorts.
GPA in 3rd grade¤ GPA in 6th grade¤ GPA in 9th grade¤
n β Adjusted R2 of
the model
change n β Adjusted R2
of the model
change n β Adjusted R2 of
the model
3-Year-Old Cohort
Aggression 114 .121 .006 .015 114.004 .009 .000
+ SEPa 114 .075 .097 .006 114.073 .207 .005
Hyperactivity 114 .079 .003 .006 114.040 .007 .002
+ SEPa 114 .091 .103 .008 114.023.202 .001
Social adjustment 114 .151 .014 .023 114.074 .003 .006
+ SEPa 114 .136 .110 .019 114.053.204 .003
6-Year-Old Cohort
Aggression 193 .111 .007 .012 193.156.019 .024 193.127 .011 .016
+ SEPa 193 .105 .022 .011 193.144.092 .021 193.117 .065 .014
Hyperactivity 193 .185* .029 .034 193.189*.031 .036 193.185* .029 .034
+ SEPa 193 .175* .042 .031 193.170*.100 .029 193.168* .079 .028
Social adjustment 193 .203* .036 .041 193.134.013 .018 193.134 .013 .018
+ SEPa 193 .185* .045 .034 193.095.080 .009 193.100 .061 .010
9-Year-Old Cohort
Aggression 209 .085 .002 .007 209.179*.027 .032 209.122 .010 .015
+ SEPa 209 .069 .026 .005 209.158.077 .025 209.098 .071 .010
Hyperactivity 209 .024 .004 .001 209.132.013 .017 209.109 .007 .012
+ SEPa 209 .022 .022 .000 209.130.069 .017 209.107 .073 .011
Social adjustment 209 .085 .002 .007 209.104.006 .011 209.197** .034 .039
+ SEPa 209 .096 .031 .009 209.119.066 .014 209.213** .107 .045
Note: The time of the measurement of disruptive behaviour within the cohorts is the same as the cohort age; * p < .017,**p < .001; a = Childhood socioeconomic position
in terms of maternal years of education; ¤ = The ages of participants in the 3rd, 6th, and 9th grades are 9, 12, and 15, respectively.
Pickles, Hagell, Rutter, & Yule, 1996) and to be similar among
both gender groups (E. G. Willcutt, Pennington, & DeFries,
2000). Here we found that the association between aggression
and school performance was true only among boys suggesting a
gender specific role of aggression in academic achievement.
A gender specific association between hyperactivity and
school performance was found, too. Hyperactivity was associ-
ated with poor school performance only among girls even
though there were no gender differences in the mean levels of
hyperactivity at any of the cohorts. Consequently, hyperactivity
seems to play a different role for girls and boys. This is in ac-
cordance with a previous study carried out in the present sam-
ple showing that childhood hyperactivity predicted adulthood
atherosclerosis (indicated by ultrasound measurements of ca-
rotid intima-media thickness) over 21 years of follow-up in
women but not in men (Keltikangas-Jarvinen, Pulkki-Raback,
Puttonen, Viikari, & Raitakari, 2006). There is, however, also
evidence of stronger association among boys (McGee, Prior,
Williams, Smart, & Sanson, 2002).
Social adjustment is known to correlate with evoking dislik-
ing from the teachers, and with poor social status among peers
(Dougherty, 2006; Newcomb, Bukowski, & Pattee, 1993). In
girls and boys, there were moderate evidence for the association
between social adjustment and school performance. Previous
research has suggested that a role of social adjustment in child’s
social status and social popularity is far from clear. Actually, a
direct association exists in kindergarten, only (Johnson, Iron-
smith, Snow, & Poteat, 2000). Our finding suggests that the
association between social adjustment and school performance
is not consistent, either.
The most prominent result of the present study was the gender
related difference between the ssociations of disruptive be- a
Table 5.
Standardized be ta coe fficie nts of disru ptiv e Behav iour in 3 diffe rent age gr ou ps in pre dicti ng gr ade po int av era ges (GPAs) in the 3rd, 6th, an d 9th grades for boys.
The results are shown separately for 3-, 6-, and 9-year old cohorts.
GPA in 3rd grade¤ GPA in 6th grade¤ GPA in 9th grade¤
n β Adjusted R2
of the model
change n β Adjusted R2
of the model
change n β Adjusted R2
of the model
3-Year-Old Cohort
Aggression 111 .045 .007 .002 111.029 .008 .000
+ SEPa 111 .013 .100 .000 111.056.066 .003
Hyperactivity 111 .122 .006 .006 111.126 .007 .016
+ SEPa 111 .056 .103 .003 111.072 .068 .005
Social adjustment 111 .046 .007 .002 111.116.005 .014
+ SEPa 111 .004 .100 .000 111.083.070 .007
6-Year-Old Cohort
Aggression 154 .037 .005 .001 154.115.007 .013 154 .168 .022 .028
+ SEPa 154 .099 .074 .009 154.038.131 .001 154 .092 .145 .008
Hyperactivity 154 .015 .006 .000 154.105.005 .011 154 .184 .028 .034
+ SEPa 154 .041 .066 .002 154.071.135 .005 154 .149 .159 .022
Social adjustment 154 .000 .007 .000 154.116.007 .014 154 .115 .007 .013
+ SEPa 154 .062 .068 .004 154.037 .131 .001 154 .033 .138 .001
9-Year-Old Cohort
Aggression 192 .292**.080 .085 192.200*.035 .040 192 .180* .027 .032
+ SEPa 192 .255**.158 .064 192.164.111 .026 192 .140 .119 .019
Hyperactivity 192 .076 .001 .006 192.100.005 .010 192 .118 .009 .014
+ SEPa 192 .035 .094 .001 192.061.088 .004 192 .076 .106 .006
Social adjustment 192 .214*.041 .046 192.155.019 .024 192 .180* .027 .032
+ SEPa 192 .199*.133 .040 192.141.105 .020 192 .165* .127 .027
Note: The time of the measurement of disruptive behaviour within the cohorts is the same as the cohort age; * p < .017,**p < .001; a = Childhood socioeconomic position in
terms of maternal years of education; ¤ = The ages of participants in the 3rd, 6th, and 9th grades are 9, 12, and 15, respectively.
haviour and school performance. Even though the mean levels of
aggression and hyperactivity were not different between the
gender groups, their impact on school performance was different.
Recent findings in Finland have shown that the level of aggres-
sion in girls and boys has narrowed (Tilastokeskus, 2010), but it
seems that gender related behavioural expectations still remain
the same. This difference may result from cultural expectations,
i.e. depending on the cultural gender role norms, some behav-
ioural tendencies may be viewed as more or less appropriate or
desirable in girls and boys (see meta-analysis and review by
(Else-Quest, Hyde, Goldsmith, & Hulle, 2006; McIntyre &
Edwards, 2009). For example, higher motor activity and ag-
gressive tendencies may be better condoned among boys
whereas silent and persistent working may be more expected
from girls. However, these often tacit expectations regarding
gender specific behaviours may have an important role on
teacher’s ratings. Indeed, teachers’ ratings have been shown to
be more gender typed than are parents’ ratings, and it has been
suggested that this may result from teachers’ frequent observa-
tions of students’ interactions among same gendered students,
which in turn, have shown to magnify gender role differences
(Else-Quest et al., 2006; Maccoby, 1990).
The strengths of the present study are as follows. We were
able to use 1) a nationally representative, although somewhat
selected, 2) prospective study over the whole comprehensive
school (i.e. for over nine years). 3) We had the opportunity to
examine disruptive behaviour before entering the school, i.e. at
the time, when the school has not yet contributed on child’s
behaviour. Finally, 4) as the teachers receive the same educa-
tion and all schools follow the same curriculum the Finnish
school system can be seen as rather homogenous.
Certain limitations have to be taken into account when inter-
preting the present results. First limitation is linked with attri-
tion. The persons lost to follow-up in each of the cohort were
Figure 1.
The fully adjusted grade point averages (GPAs) over the three meas-
urements (3rd , 6th, and 9th grade) among girls with low and high ag-
gression, hyperactivity, and social adjustment, respectively.
more likely to be boys than girls, had lower GPAs throughout
the measurements, and were from less educated families. This
may at least partly be explained so that the most aggressive
participants were excluded from the study.
Second limitation is related with the self reports of the school
performance. Even though the self-reported GPAs may be bi-
ased through recall problems or social desirability, a recent
meta-analysis suggest, however, that they reflect the actual
performance reasonably well (Kuncel et al., 2005). In Finland,
GPAs are often the most important source of student’s aca-
demic achievement when applying for the secondary schooling
and a certain level of achievement is mostly required to get a
place to study. Previous research has also shown that GPAs at
the end of comprehensive school predicts later academic, social
and health outcomes (Alatupa et al., 2010; K. Kokko et al.,
2003; Kuncel et al., 2005). Moreover, it has been shown that in
Finland, school grades given by teachers are a more reliable
Figure 2.
The fully adjusted grade point averages (GPAs) over the three meas-
urements (3rd , 6th, and 9th grade) among boys with low and high ag-
gression, hyperactivity an d s oc ial adjustment, respectively.
predictor of later academic success than standardized tests (e.g.
baccalaureate) (Rantanen, 2004).
Third limitation associates with the non-standardized meas-
urement of disruptive behaviour. By the start of our study in
1980, there were not many appropriate measurements of dis-
ruptive behaviour. Although it is possible that the full domain of
disruptive behaviour has not been reached, the reliability and
validity of the measurement has been shown to be reasonable in
several previous studies (Katainen & Raikkonen, 1999; Räik-
könen et al., 2000).
Finally, we may not exclude the possibility that mothers’ re-
port of child’s behaviour may be reflected by the quality of
mother-child relationship and even by mother’s own personality.
It has for example been shown that maternal depression has an
impact on their perceptions of child’s behaviour (Pesonen et al.,
2003; Whiffen, 1990).
In the present study we showed a gender specific association
of the different components of disruptive behaviour with school
performance. Disruptive behaviour started to predict school
performance before school age, and predicted school perform-
ance throughout the compulsory education, even after control-
ling for maternal education. Our results emphasize the impor-
tance of early identifying of children with disruptive behaviour.
This study was supported by the Academy of Finland (grant
124399 for L.K.-J. and grant 123621 for L.P.-R.), The Research
Funds of the University of Helsinki (L.P.-R.), Alfred Korde-
lin’s Foundation (S.A.), Finnish Cultural Foundation, Päi-
jät-Häme Regional fund (S.A.), Oskar Öflund’s Foundation
(S.A.), Emil Aaltonen Foundation (M.H.) and Ella and Georg
Ehrnrooth Foundation (M.H).
Abdi, H. (2007). Bonferroni and sidak corrections for multiple com-
parisons. In N. J. Salkind (Ed.), Encyclopedia of measurement and
statistics (pp. 103-107). Thousand Oaks (CA): Sage.
Alatupa, S., Pulkki-Råback, L., Hintsanen, M., Ravaja, N., Raitakari, O.
T., Telama, R., et al. (2010). School performance as a predictor of
adulthood obesity: A 21-year follow-up study. European Journal of
Epidemiology, 25, 267-74. doi:10.1007/s10654-010-9428-6
American Psychiatric Association. (2007). Attention-deficit and disrup-
tive behaviour disorders. DSM-IV-TR. Fourth ed. Arlington: VA:
American Psychiatric Association, 85-102.
Archer, J. (2004). Sex differences in aggression in real-world settings:
A meta-analytic review. Review of General Psychology, 8, 291-322.
Asendorpf, J. B., Denissen, J. J., & van Aken, M. A. (2008). Inhibited
and aggressive preschool children at 23 years of age: Personality and
social transitions into adulthood. Developmental Psychology, 44,
997-1011. doi:10.1037/0012-1649.44.4.997
Berger, M., Yule, W., & Rutter, M. (1975). Attainment and adjustment
in two geographical areas: IIthe prevalence of specific reading re-
tardation. The British Jo ur na l of Psychiatry, 126, 510-519.
Boylan, K., Vaillancourt, T., Boyle, M., & Szatmari, P. (2007). Comor-
bidity of internalizing disorders in children with oppositional defiant
disorder. European Child & Adolescent Psychiatry, 16, 484-494.
Breslau, J., Miller, E., Breslau, N., Bohnert, K., Lucia, V., &
Schweitzer, J. (2009). The impact of early behavior disturbances on
academic achievement in high school. Pediatri cs, 123, 1472-1476.
Caspi, A., Elder, G. H., & Bem, D. J. (1987). Moving against the world:
Life-course patterns of explosive children. Developmental Psycho-
logy, 23, 308-313. doi:10.1037/0012-1649.23.2.308
Caspi, A., & Henry, B. (1995). Temperamental origins of child and
adolescent behavior problems: From age three to age fifteen. Child
Development, 66, 55-68. doi:10.2307/1131190
Dougherty, L. R. (2006). Children’s emotionality and social status: A
meta-analytic review. Social Development, 15, 394-417.
Else-Quest, N. M., Hyde, J. S., Goldsmith, H. H., & Hulle, C. A. V.
(2006). Gender differences in temperament: A meta-analysis. Psy-
chological Bulletin, 1 3 2 , 33-72. doi:10.1037/0033-2909.132.1.33
Fagot, B. I., & Hagan, R. (1985). Aggression in toddlers: Responses to
the assertive acts of boys and girls. Sex Roles, 1 2 , 341-351.
Fergusson, D. M., Horwood, L. J., & Lynskey, M. T. (1993). The ef-
fects of conduct disorder and attention deficit in middle childhood on
offending and scholastic ability at the age 13. Journal of Child Psy-
chology and Psychiatry, 34, 899-916.
Fergusson, D. M., & John Horwood, J. (1995). Early disruptive behav-
ior, IQ, and later school achievement and delinquent behavior. Jour-
nal of Abnormal Child Psychology, 23, 183-99.
Frick, P. J., Kamphaus, R. W., Lahey, B. B., Loeber, R., Christ, M. A.,
Hart, E. L., et al. (1991). Academic underachievement and the dis-
ruptive behaviour disorders. Journal of Consulting & Clinical Psy-
chology, 59, 289-94. doi:10.1037/0022-006X.59.2.289
Spira, E. G., & Fischel, J. E. (2005). The impact of preschool inatten-
tion, hyperactivity, and impulsivity on social and academic devel-
opment: A review. Journal of Child Psychology and Psychiatry, 46,
755-773. doi:10.1111/j.1469-7610.2005.01466.x
Hawkins, J. D., & Lishner, D. M. (1987). Schooling and delinquency.
In E. H. Johnson (Ed.), (pp. 179-221). New York: Guilford Press.
Heiervang, E., Stevenson, J., Lund, A., & Hugdahl, K. (2001). Behav-
iour problems in children with dyslexia. Nordic Journal of Psychia-
try, 55(4), 251-256. doi:10.1080/080394801681019101
Hinshaw, S. (1992a). Academic underachievement, attention deficits,
and aggression: Comorbidity and implications for intervention.
Journal of Consulting and Clinical Psychology, 60, 893-903.
Hinshaw, S. (1992b). Externalizing behavior problems and academic
underachievement in childhood and adolescence: Causal relation-
ships and underlying mechanisms. Psychological Bulletin, 111, 108-
126. doi:10.1037/0033-2909.111.1.127
Johnson, C., Ironsmith, M., Snow, C. W., & Poteat, G. M. (2000). Peer
acceptance and social adjustment in preschool and kindergarten.
Early Childhood Education Journal, 27, 207-212.
Johnson, W., McGue, M., & Iacono, W. G. (2005). Disruptive behavior
and school grades: Genetic and environmental relations in
11-year-olds. Journal of Edu catio nal Psychology, 97, 391-405.
Katainen, S., & Räikkönen, K. (1999). Maternal child-rearing attitudes
and role satisfaction and children’s temperament as antecedents of
adolescents’ depressive tendencies: Follow-up study of6- to 15-
years-olds. Journal of Youth & Adolescence, 28, 139-16.
Keltikangas-Järvinen, L., Pulkki-Råback, L., Puttonen, S., Viikari, J., &
Raitakari, O. T. (2006). Childhood hyperactivity as a predictor of
carotid artery intima media thickness over a period of 21 years: The
cardiovascular risk in young finns study. Psychosomatic Medicine,
68, 509-516.
Kokko, K., Bergman, L. R., & Pulkkinen, L. (2003). Child personality
characteristics and selection into long-term unemployment in finnish
and swedish longitudinal samples. International Journal of Behav-
ioral Development, 27, 134-144. doi:10.1080/01650250244000137
Kokko, K., & Pulkkinen, L. (2000). Aggression in childhood and
long-term unemployment in adulthood: A cycle of maladaptation and
some protective factors. Prevention & Treatment, 3, 1-17.
Kokko, K., Pulkkinen, L., & Puustinen, M. (2000). Selection into
long-term unemployment and its psychological consequences. Inter-
national Journal of B ehavioral Development, 24, 310-320.
Kuncel, N. R., Credé, M., & Thomas, L. L. (2005). The validity of
self-reported grade point averages, class ranks, and test scores: A
meta-analysis and review of the literature. Review of Educational
Research, 75, 63-82. doi:10.3102/00346543075001063
Kuusela, J. (2003). Lukioiden tuloksiin vaikuttavista tekijöistä (from the
factors affecting the results in senior high schools) (Copy No. 13).
Helsinki: Edita Prima.
Loe, I. M., & Feldman, H. M. (2007). Academic and educational out-
comes of children with ADHD. Journal of Pediatric Psychology, 32,
643-654. doi:10.1093/jpepsy/jsl054
Loeber, R., Burke, J., D., Lahey, B. B., Winters, A. B. A., & Zera, M. B.
A. (2000). Oppositional defiant and conduct disorder: A review of
the past 10 years, part I. Journal of the American Academy of Child
& Adolescent Psychiatry., 39, 1468-1484.
Lynch, J. W., Kaplan, G. A., & Salonen, J. T. (1997). Why do poor
people behave poorly? Variation in adult health behaviours and psy-
chosocial characteristics by stages of the socioeconomic lifecourse.
Social Science & Medicine, 44, 809-819.
Maccoby, E. E. (1990). Gender and relationships: A developmental
account. American Psychologist, 45, 513-520.
Maughan, B., Pickles, A., Hagell, A., Rutter, M., & Yule, W. (1996).
Reading problems and antisocial behaviour: Developmental trends in
comorbidity. Journal of Child Psychology & Psychiatry & Allied
Disciplines, 37, 405-418. doi:10.1111/j.1469-7610.1996.tb01421.x
McGee, R., Prior, M., Williams, S., Smart, D., & Sanson, A. (2002).
The long-term significance of teacher-rated hyperactivity and reading
ability in childhood: Findings from two longitudinal studies. Journal
of Child Psychology & Psychiatry & Allied Disciplines, 43, 1004-
1017. doi:10.1111/1469-7610.00228
McIntyre, M. H., & Edwards, C. P. (2009). The early development of
gender differences. Annual Review of Anthropology, 38, 83-97.
Newcomb, A. F., Bukowski, W. M., & Pattee, L. (1993). Children's
peer relations: A meta-analytic review of popular, rejected, neglected,
controversial, and average sociometric status. Psychological Bulletin,
113, 99-128.
Pesonen, A., Räikkönen, K., Keskivaara, P., & Keltikangas-Järvinen, L.
(2003). Difficult temperament in childhood and adulthood: Continu-
ity from maternal perceptions to self-ratings over 17 years. Persona-
lity & Individual Differences, 34, 19-31.
Pulkki-Råback, L., Elovainio, M., Kivimäki, M., Raitakari, O. T., &
Keltikangas-Järvinen, L. (2005). Temperament in childhood predicts
body mass in adulthood: The cardiovascular risk in young Finns
study. Health Psychology, 24, 307-15.
Räikkönen, K., Katainen, S., Keskivaara, P., & Keltikangas-Järvinen, L.
(2000). Temperament, mothering, and hostile attitudes: A 12-year
longitudinal study. Personality and Social Psychology Bulletin, 26,
Raitakari, O. T., Juonala, M., Kähönen, M., Taittonen, L., Laitinen, T.,
Mäki-Torkko, N., et al. (2003). Cardiovascular risk factors in child-
hood and carotid artery intima-media thickness in adulthood: The
cardiovascular risk in young finns study. JAMA, 290, 2277-2283.
Rantanen, P. (2004). Valinnasta työelämään: Ammatillisen koulutuksen
ja ammattikorkeakoulujen opiskelijavalinnan tarkastelua (No. 19).
Helsinki: Opetusministeriö, Koulutus- ja Tiedepolitiikan Osasto.
Reinke, W. M., Herman, K. C., Petras, H., & Ialongo, N. S. (2008).
Empirically derived subtypes of child academic and behavior prob-
lems: Co-occurrence and distal outcomes. Journal of Abnormal Chi ld
Psychology, 36, 759-770. doi:10.1007/s10802-007-9208-2
Rhee, S., & Waldman, I. (2002). Genetic and environmental influences
on antisocial behaviour: A meta-analysis of twin and adoption stud-
ies. Psychological Bulletin, 128, 490-529.
Richman, N., Stevenson, J., & Graham, P. (1982). Preschool to school:
A behavioural study. San Diego, CA: Academic Press.
Rucklidge, J. J. (2010). Gender differences in attention-Deficit/Hype-
ractivity disorder. Psychiatric Clinics of North America, 33, 357-373.
Rutter, M. (1974). Emotional disorder and educational underachieve-
men t. Archives of Disease in Childhood, 49, 249-256.
Serbin, L. A., O’leary, K. D., Kent, R. N., & Toniek, I. J. (1973). A
comparison of teacher response to the preacademic and problem be-
havior of boys and girls. Child Devel opmen t, 44, 796-804.
Sirin, S. R. (2005). Socioeconomic status and academic achievement: A
meta-analytic review of research. Review of Educational Research,
75, 417-53. doi:10.3102/00346543075003417
Stevenson, J., Richman, N., & Graham, P. (1985). Behaviour problems
and language abilities at three years and behavioural deviance at
eight years. Journal of Child Psychology & Psychiatry & Allied Dis-
ciplines, 26, 215-230. doi:10.1111/j.1469-7610.1985.tb02261.x
Tilastokeskus. (2010). Rikoksista selvitetttiin 59%. URL (last checked
17 June 2011)
Tremblay, R. E., & Masselink, B. (1992). Early disruptive behavior,
poor school achievement, delinquent behavior, and delinquent.
Journal of Consulting & Clinical Psychology, 60, 64-72.
Trzesniewski, K. H., Moffitt, T. E., Caspi, A., Taylor, A., & Maughan,
B. (2006). Revisiting the association between reading achievement
and antisocial behavior: New evidence of an environmental explana-
tion from a twin study. Child Development, 77, 72-88.
Vitaro, F., Larocque, D., Janosz, M., & Tremblay, R. E. (2001). Nega-
tive social experiences and dropping out of school. Educational
Psychology, 21, 401-415. doi:10.1080/01443410120090795
Vitaro, F., Brendgen, M., Larose, S., & Trembaly, R. E. (2005). Kin-
dergarten disruptive behaviors, protective factors, and educational
achievement by early adulthood. Journal of Educational Psychology,
97, 617-629. doi:10.1037/0022-0663.97.4.617
Wells, E. (1980). Behavioral pattern of children in school. Vitality
Health Statistics, No. 113.
Whiffen, V. E. (1990). Maternal depressed mood and perceptions of
child temperament. Journal of Genetic Psychology, 151, 329-339.
Willcutt, E. G., Pennington, B. F., & DeFries, J. C. (2000). Twin study
of the etiology of comorbidity between reading disability and atten-
tion-deficit/hyperactivity disorder. American Journal of Medical
Genetics, 96, 293-301.
Willcutt, E. G., & Pennington, B. F. (2000). Psychiatric comorbidity in
children and adolescents with reading disability. Journal of Child
Psychology & Psychiatry & Allied Disciplines, 41, 1039-1048.
Williams, S., & McGee, R. (1994). Reading attainment and juvenile
delinquency. Journal of Child Psychology & Psychiatry & Allied
Disciplines, 35, 441-459. doi:10.1111/j.1469-7610.1994.tb01733.x