Journal of School Psychology 56 (2016) 1325

Contents lists available at ScienceDirect

Journal of School Psychology

journal homepage: www.elsevier.com/locate/jschpsyc

Teacherstudent relationships and adolescent behavioral

engagement and rule-breaking behavior: The moderating role
of dopaminergic genes
Steven De Laet a,, Hilde Colpin a, Karla Van Leeuwen b, Wim Van den Noortgate b, Stephan Claes c,
Annelies Janssens a, Luc Goossens a, Karine Verschueren a
a
b
c

Research Group for School Psychology and Child and Adolescent Psychology (SCAD), University of Leuven, Belgium
Research Group for Parenting and Special Education, University of Leuven, Belgium
Research Group for Psychiatry, Department of Neurosciences, University of Leuven, Belgium

a r t i c l e

i n f o

Article history:
Received 13 May 2015
Received in revised form 18 November 2015
Accepted 17 February 2016
Available online xxxx
Keywords:
Geneenvironment interaction
Teacherstudent relationships
School engagement
Dopamine
DAT1
DRD4

a b s t r a c t
This study examined whether the dopamine transporter DAT1 and the dopamine receptor
DRD4 genes moderate the effect of student-reported teacherstudent relationship afliation
or dissatisfaction on parent-reported adolescent rule-breaking behavior and behavioral engagement. The sample included 1053 adolescents (51% boys, Mage = 13.79) from grades 7 to 9.
Regression analyses were conducted using Mplus while controlling for multiple testing and
nested data. Adolescents who experienced stronger afliation with their teachers were more
engaged in school, whereas greater dissatisfaction predicted more rule-breaking behavior. In
addition, a signicant geneenvironment interaction was found for both genes examined.
The link between low teacherstudent afliation and low engagement was more pronounced
for DAT1-10R homozygotes. The link between high teacherstudent dissatisfaction and more
rule-breaking was stronger for DRD4 non-long carriers. Implications for understanding the
role of teacherstudent relationships in adolescence and suggestions for future research are
outlined.
2016 Society for the Study of School Psychology. Published by Elsevier Ltd. All rights reserved.

1. Introduction
The bio-ecological model of human development (Bronfenbrenner & Morris, 2006) assumes that adolescent development is
shaped by adolescents' social environments, as well as by biological factors such as genes, and the interactions between genes
and environments (i.e., GxE interactions). The GxE interactions take place when the genotype's effect is moderated by the social
environment or, conversely, when the effect of exposure to a socialenvironmental factor on behavior is conditional upon a
person's genotype (Moftt, Caspi, & Rutter, 2006). Regarding social environments, most GxE research focuses on (negative) aspects of parenting such as maltreatment or negative aspects of the social context as a whole such as stressful life events. However,
for several domains of child and adolescent development, including externalizing behaviors such as rule-breaking behaviors and
behavioral school engagement, teacherstudent relationships are important as well (e.g., O'Connor, Dearing, & Collins, 2011).

Funding was provided through grant GOA/12/009 (STRATEGIES project) of the BOF (Bijzonder Onderzoeksfonds), KU LeuvenUniversity of Leuven.
Corresponding author at: School Psychology and Child and Adolescent Development, University of Leuven, Tiensestraat 102box 3717, 3000 Leuven, Belgium.
Tel.: + 32 16 32 62 88.
E-mail address: steven.delaet@ppw.kuleuven.be (S. De Laet).
ACTION EDITOR: Kathy Rudasill.

http://dx.doi.org/10.1016/j.jsp.2016.02.002
0022-4405/ 2016 Society for the Study of School Psychology. Published by Elsevier Ltd. All rights reserved.

14

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

Despite the fact that school is a central context of child and adolescent development (Roeser, Eccles, & Sameroff, 2000), teacher
student relationships have been largely overlooked in GxE research. Not only does the study of the interplay of genes with teacherstudent relationships contribute to the GxE literature, it also adds signicantly to the current knowledge base on the developmental signicance of teacherstudent relationships. Indeed, previous research has suggested that teacherstudent relationships
are particularly inuential for vulnerable students (Sabol & Pianta, 2012). These vulnerabilities, however, have only been assessed
at the behavioral level, raising the question as to whether students with certain genetic markers of vulnerability are also more
susceptible to the quality of teacherstudent relationships.
The present study is the rst GxE study on specic genes (also referred to as a molecular genetics study) that includes teacherstudent relationships as a social environmental factor. The dopamine transporter or DAT1 and the D4 receptor or DRD4 genes
were selected, because they have been associated with aspects of externalizing behavior and behavioral engagement such as concentration problems (e.g., Dmitrieva, Chen, Greenberger, Ogunseitan, & Ding, 2011; Gordon, Stollstorff, Devaney, Bean, & Vaidya,
2012). We sought to investigate whether DAT1 and DRD4 moderate the effect of perceived teacherstudent relationship quality on
adolescents' rule-breaking behavior and behavioral engagement. In other words, we examined whether adolescents with certain
genotypic variants of DAT1 and DRD4 differ in their sensitivity to the effects of teacherstudent relationships.
1.1. Rule-breaking behavior and behavioral school engagement: Importance of teacherstudent relationships
During early adolescence (i.e., ages 10 to 14), there is a general increase in adjustment problems, including an increase in
externalizing behaviors (e.g., Reitz, Dekovic, & Meijer, 2005) and a decrease in behavioral school engagement (e.g., Li & Lerner,
2011). Externalizing behaviors involve conicts with other people or with society as a whole and are clearly noticeable in
overt behaviors such as aggression and rule breaking (Achenbach & Rescorla, 2001). The present study focuses on rulebreaking behavior especially (i.e., not abiding by the rules at home, at school, or anywhere else by means of stealing, lying, hanging out with deviant peers, using drugs, skipping classes) (Achenbach & Rescorla, 2001). As studies including rule-breaking behaviors are rather scarce, relevant studies including other aspects of externalizing problems are discussed as well. Behavioral school
engagement (e.g., Li & Lerner, 2011) involves participation in academics by means of effort, concentration, and doing homework
(Birch & Ladd, 1997). Age-related changes in behavioral school engagement and externalizing behaviors are common (e.g., Reitz
et al., 2005), but they may have detrimental consequences, both concurrently and in the long term. For some adolescents, externalizing behaviors continue into adulthood, expressed by antisocial behavior (Schaeffer, Petras, Ialongo, Poduska, & Kellam, 2003),
addiction, and criminal activity (Reef, Diamantopoulou, van Meurs, Verhulst, & van der Ende, 2011). A lack of behavioral engagement is associated with increased academic failure (e.g., Johnson, McGue, & Iacono, 2006), higher dropout rates (e.g., Alexander,
Entwisle, & Horsey, 1997), and more psychosocial problems (e.g., Li & Lerner, 2011).
A growing body of literature suggests that school is a central context for early adolescents' development (e.g., Roeser et al.,
2000) and that teacherstudent relationships make a unique contribution to early adolescents' academic and socialemotional
development, including externalizing behavior (e.g., O'Connor et al., 2011; Roland & Galloway, 2002) and behavioral school engagement (e.g., O'Farrell, Morrison, & Furlong, 2006). For example, a meta-analysis by Roorda, Koomen, Spilt, and Oort (2011)
(with medium to large effect sizes ranging from r = .25 to .40) revealed that positive aspects of affective teacherstudent relationships, such as teacherstudent support and caring, are related to higher participation in learning and more on-task behavior,
whereas negative aspects of teacherstudent relationships are related to lower school engagement. Drawing from this work, the
present study included adolescent rule-breaking behavior and behavioral engagement as outcomes, and examined the contribution of both positive (i.e., teacherstudent afliation) and negative (i.e., teacherstudent dissatisfaction) aspects of teacherstudent relationships. Distinguishing between positive and negative relationship dimensions is considered important, as it allows
examining differential effects on students' school adjustment (e.g., Birch & Ladd, 1997).
Recent research shows that students with behavioral markers of risk are affected more by the quality of the relationship with
their teachers compared to other students (Sabol & Pianta, 2012). For example, research has indicated that child temperamental
inhibition/disinhibition moderates the effect of early teacherstudent relationships on student mental health in early adolescence
(Essex, Armstrong, Burk, Goldsmith, & Boyce, 2011) and that shy students seem to benet more from close teacherstudent relationships (Arbeau, Coplan, & Weeks, 2010). Based on this research, it seems plausible that students with vulnerability markers
at the genetic level (instead of the behavioral level), may also be more susceptible to the quality of teacherstudent relationships.
To our knowledge, only one study has investigated teachers in a GxE framework (Taylor, Roehrig, Hensler, Connor, &
Schatschneider, 2010). However, this study focused on teaching quality as operationalized by reading skills of the whole class
rather than on (the affective quality of) teacherstudent relationships. Moreover, this study was conducted in a twin-based
study design in which overall genetic and environmental inuences are estimated based on comparisons of siblings with varying
degrees of genetic relatedness. No published study has investigated the joint impact of teacherstudent relationships and genes
on adolescents' academic or socialemotional development, neither in a twin-based study, nor in a molecular genetics design,
in which specic measured gene variants are identied.
1.2. Dopaminergic genes DAT1 and DRD, externalizing behavior and behavioral school engagement
In prior GxE interaction research, externalizing behavior and behavioral engagement have been investigated in relation to dopaminergic genes such as the dopamine transporter gene DAT1 and the dopamine receptor gene DRD4. These genes code for
chemical substances or proteins that regulate the functioning of the neurotransmitter dopamine in the brain. Both genes contain

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

15

a brief sequence of genetic information that it is repeated a different number of times in different people (i.e., a variable number
of tandem repeats or VNTR) (Bakermans-Kranenburg, van IJzendoorn, Pijlman, Mesman, & Juffer, 2008). For DAT1, the number of
repeats ranges from 3 to 13, with the 10-repeat (10R) and 9-repeat (9R) polymorphisms being the two most prevalent alleles in
most human populations (Mitchell et al., 2000). For DRD4, the number of repeats ranges from 2 to 11 repeats, with the 2-repeat
(2R), 4-repeat (4R), and 7-repeat (7R) alleles being the most prevalent in most human populations (Chang, Kidd, Livak, Pakstis, &
Kidd, 1996).
Although the number of repeats is thought to affect gene functioning, the specic biological mechanisms linking individual
variability on these genes to gene functioning and neural structures remain largely unclear. Nevertheless, researchers have compared participants, grouped into carriers of the short and long variants (i.e., with comparatively smaller and higher numbers of
repeats), for both DAT1 and DRD4 on a host of behaviors assumed to be associated with the dopamine system. As children inherit
two copies of each gene, one from their mother and one from their father, multiple combinations of alleles or genotypes can be
distinguished. For the DAT1 gene, for instance, one can distinguish carriers of the short/long (i.e., 9R/10R) genotype (called heterozygous), and the short/short (i.e., 9R/9R) and long/long (i.e., 10R/10R) genotypes (which are both called homozygous). As
there is no prior molecular GxE interaction research including teacherstudent relationships, we draw upon genetic association
studies that examine direct effects (or main effects) of those genes on behavior, and GxE interaction studies on parentchild rather than teacherstudent relationships. The latter studies have often included preschoolers and elementary school children, as GxE
interaction research including adolescents is scarce.

1.2.1. DAT1 in genetic association studies and GxE interaction research

The large majority of genetic association studies investigating the direct effects of DAT1 on externalizing behavior show a positive association between problem behavior and the DAT1-10R (long) allele (e.g., Beaver, Wright, & DeLisi, 2008; Blum et al.,
2011; Burt & Mikolaiewski, 2008; Chen et al., 2007; Guo, Roettger, & Shih, 2007; Lee et al., 2007). For example, carriers of the
DAT1-10R allele (i.e., 9R/10R and/or 10R/10R) have an increased risk for antisocial behavior, ADHD (e.g., Cornish et al., 2005;
Guo, Cai, Guo, Wang, & Harris, 2010), risk taking (Mata, Hau, Papassotiropoulos, & Hertwig, 2012), and concentration and attention problems in a memory task (which can be considered aspects of low behavioral engagement; Gordon et al., 2012). Studies in
neurobiology indicate that the DAT1-10R (long) allele is associated with increased gene expression and therefore decreased
dopamine in the striatum (i.e., Brookes et al., 2007; Heinz et al., 2000; Mill, Asherson, Browes, D'Souza, & Craig, 2002). Hence,
DAT1-10R allele homozygotes would have the highest gene expression and the lowest striatal dopamine. That is why prior research on polygenetic risk scores has coded the DAT1-10R allele homozygotes as having a low dopamine genotype while
DAT1-9R allele carriers as having a high dopamine genotype (Nikolova, Ferrell, Manuck, & Hariri, 2011).
The results from GxE interaction research have been inconsistent about whether carrying (various combinations of) the DAT19R allele or the DAT1-10R allele makes people more susceptible to adverse environments. Some studies revealed a stronger association between aspects of parenting (e.g., parental expressed emotion and negative maternal parenting) and conduct disorder in
DAT1-9R (short) carriers (Lahey et al., 2011; Sonuga-Barke et al., 2009). Other studies found that the effects of both supportive
and adverse family factors on various outcomes (e.g., delinquency and ADHD symptoms) were stronger for DAT1-10R (long)
carriers (Boardman et al., 2014; Li & Lee, 2013) and, more specically, for DAT1-10R homozygotes (Laught et al., 2007).
Drawing from this work, we formed a group of students with a so-called double genetic risk (i.e., DAT1-10R homozygotes)
versus the rest (i.e., genotypes homozygous for the DAT1-9R and heterozygous genotypes possessing both 9- and 10-repeats). We
expected that DAT1-10R homozygotes would display more externalizing behavior and less behavioral engagement, as logic would
dictate that they have the highest level of gene expression and, therefore, the lowest level of available dopamine. Given the inconsistent results from prior GxE interaction research, we did not have a clear hypothesis about whether DAT1-10R homozygotes
would be more or less susceptible to the effects of teacherstudent relationships.

1.2.2. DRD4 in genetic association studies and GxE interaction research

The vast majority of genetic association studies have consistently shown a positive association between externalizing behavior
and carrying the DRD4-7R (long) allele (e.g., Dmitrieva et al., 2011; Hohmann et al., 2009; Schmidt, Fox, Rubin, Hu, & Hamer,
2002) and aspects of behavioral engagement (e.g., concentration; Fossella et al., 2002; Swanson, Flodman et al., 2000). In addition,
most GxE interaction research including DRD4 found that carrying DRD4-long genotypes (i.e., 68 repeats) makes children more
susceptible to certain aspects of their environment. A meta-analysis by Bakermans-Kranenburg and van IJzendoorn (2011),
including 12 GxE studies on a total of 1232 Caucasian children, found that children with the DRD4-7R allele are more susceptible
to both positive and negative parental inuences. However, some studies found that the absence and not the presence of the
DRD4-7R allele makes children more susceptible to family adversity (e.g., Propper, Willoughby, Halpern, Carbone, & Cox, 2007).
The 7-R allele, for example, was found to protect early adults against the adverse effect of childhood trauma on emotional resilience (Das, Cherbuin, Tan, Anstey, & Easteal, 2011).
Drawing from this work, the present study sample was divided into DRD4-long carriers (i.e., genotypes possessing at least one
7R or longer) versus DRD4-non-long carriers (i.e., genotypes possessing two alleles shorter than 7R). Based on genetic association
research, we expected that DRD4-long carriers display more rule-breaking behavior and less behavioral engagement compared to
DRD4-non-long carriers. In line with most prior GxE research, we additionally expected that DRD4-long carriers are more susceptible to the effects of teacherstudent relationships.

16

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

1.2.3. Geneenvironment interaction: Three alternative hypotheses

The present study is the rst molecular GxE study to include teacherstudent relationships as a socialenvironmental factor. In
contrast to most prior GxE interaction research that focused exclusively on negative environmental factors (e.g., family stress) and
negative behavioral outcomes (e.g., aggression), the present study includes both negative and positive aspects of teacherstudent
relationships (i.e., teacherstudent dissatisfaction and afliation) and adolescent behavior (i.e., adolescent rule-breaking behavior
and behavioral engagement).
This differentiated approach enables us to adequately test three alternative GxE interaction hypotheses. First, according to the
dual-risk or diathesisstress model of development, the detrimental effects of a negative social environment on problem behavior
hold especially for adolescents carrying specic variants of certain genes (i.e., vulnerability genes or risk alleles; Pluess & Belsky,
2013). Second, the vantage sensitivity hypothesis assumes that individuals carrying certain genetic variants (i.e., vantagesensitivity factors) are more sensitive to positive inuences from their social environments, resulting in more positive behavior,
while these genetic variants may not make them more susceptible to detrimental effects of negative environments (Pluess &
Belsky, 2013). Third, according to the differential susceptibility hypothesis, adolescents carrying specic genetic variants not
only display more problem behavior when confronted with a negative social environment but also display more positive behavior
when confronted with a positive social environment (e.g., Belsky & Pluess, 2009).
1.3. Current study goals and hypotheses
Three types of expectations were formulated for the present study. First, regarding main effects of teacherstudent relationships, we expected that adolescents who experience more afliation and less dissatisfaction with their teachers would display
less rule-breaking behavior and more behavioral engagement, whereas the opposite was expected for adolescents experiencing
greater dissatisfaction and less afliation with their teachers. Second, regarding main effects of DAT1, we expected DAT1-10R homozygotes to display more rule-breaking behavior and less behavioral engagement compared to DAT1-9R carriers. As regard main
effects of DRD4, we expected that DRD4-long carriers would display more rule breaking behavior and less behavioral engagement
compared to DRD4-non-long carriers. Third, regarding GxE interactions including DAT1, the absence of prior GxE research including teacherstudent relationships and the inconsistent ndings from prior GxE studies including parentchild relationships does
not allow us to put forward specic hypotheses. Based on GxE research including DRD4 and parentchild relationships, we tested
whether DRD4-long carriers would be more susceptible to the effects of teacherstudent relationships as well. In case a signicant
GxE interaction effect was found, three alternative GxE interaction hypotheses were tested (i.e., differential susceptibility, diathesisstress, and vantage-sensitivity) using a procedure proposed by Roisman et al. (2012). Specically, we analyzed for which
values of teacherstudent relationships there were signicant differences in student outcomes between two genetic groups
(also called regions of signicance or RoS).
All hypotheses were tested in a multi-informant study in which perceived teacherstudent relationship was measured using
student reports and student outcome variables were measured using parent reports. In line with most other research on teacherstudent relationships in secondary education (cf. review by Roorda et al., 2011), students were asked about their relationship
with teachers in general. This way we aimed to capture students' overall relational experiences in school, which have been found
to impact students' school-related behaviors and attitudes.
2. Method
2.1. Participants and procedure
The present study was conducted on the rst data-wave of an ongoing longitudinal multiple cohort study on the development
of adolescents called STRATEGIES (i.e., Studying Transactions in Adolescence: Testing Genes in Interaction with Environments).
The STRATEGIES study is carried out in Belgium. Approval for the procedure was obtained from the Institutional Review Board
of the Faculty of Medicine at the University of Leuven. The STRATEGIES sample was drawn with a randomized multistage sampling approach. First, secondary schools from different provinces were asked to participate in the project. The schools were stratied by educational track in order to include students from general, technical, and vocational tracks. Second, within the nine
participating schools, we randomly selected a total of 121 classes from grades 7 to 9. Third, all students within these classes
were asked to participate. 49% of the students (N = 1111) participated in the rst wave of the STRATEGIES project (51% boys,
Mage = 13.79, SDage = .93). Nearly all participating adolescents had the Belgian nationality (95%), 13% had a home language
other than Dutch (e.g., French, Turkish, Arabic, and Berber). Seventy-four students in our study had a non-European background,
meaning that at least one grandparent was born outside Europe. All analyses were conducted with and without these students.
The results were similar which seem to suggest that effects of population stratication (i.e., people from different ancestry groups
have a different genotypic background) are limited in our specic research design. Hence, we chose to report the analyses and
results conducted upon the most complete sample of students.
Parents and adolescents signed active informed consent forms. The researchers visited the schools to supervise the adolescents
when lling out the questionnaires and to assist them when donating a saliva sample. These samples were collected with the
Oragene DNA collection kits (DNA Genotek; Ontario, Canada). A total of 1103 adolescents was present during the saliva collection
and could be genotyped. In addition, most adolescents' parents lled out questionnaires (Nmothers = 838, Nfathers = 720). A small
majority of the parents completed higher education (58% of mothers, 52% of fathers). The remaining parents nished (some years

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

17

of) high school (30% of mothers, 34% of fathers) or completed primary school (2% of mothers, 2% of fathers). Our study sample
contained 50 pairs of siblings. Due to genetic similarity of siblings, one adolescent from each pair of siblings (N = 50) was removed randomly. Hence, our nal study sample consisted of 1053 adolescents.
2.2. Measures
2.2.1. Teacherstudent afliation and dissatisfaction
Teacherstudent relationship quality was assessed by means of a slightly adapted version of the People in My Life Questionnaire (PIML; Cook, Greenberg, & Kusche, 1995). Adolescents were asked to respond to the items on a rating scale ranging from 1
(almost never true) to 5 (almost always true). We used the Afliation Scale, which consists of 10 items (e.g., My teachers accept
me the way that I am; = .90), and the Dissatisfaction Scale, which consists of 4 items (e.g., I nd it difcult to talk with my
teachers; = .85). These scales have been found to be both reliable (i.e., good internal consistency) and valid (i.e., signicant
medium expected correlations with students' social and emotional adjustment; Murray & Greenberg, 2000).
2.2.2. Rule-breaking behavior
Parents completed the 17 item Rule-Breaking Behavior Scale of the Child Behavioral Checklist (CBCL; Achenbach & Rescorla,
2001) on a 3-point rating scale: 0 (not true), 1 (somewhat or sometimes true), and 2 (very true or often true) (e.g., My son/daughter does not abide to rules at home, at school, or anywhere else; = .71). This scale has been found to be highly reliable
(i.e., excellent internal consistency) and valid (i.e., both criterion-related and construct validity) across many studies
(Achenbach & Rescorla, 2001).
2.2.3. Behavioral school engagement
We assessed behavioral school engagement by means of the Parent-Report of School Liking and Avoidance Questionnaire (PSLAQ; Ladd, Buhs, & Seid, 2000). Prior research conrmed both the reliability and the validity of this scale (Jillian & Ladd, 2011;
Ladd & Dinella, 2009). Parents were asked to respond to the items on a rating scale ranging from 1 (almost never) to 5 (almost
always). The Behavioral Engagement Scale was used which consists of 5 items (e.g., My son/daughter always does his/her homework; mothers = .84, fathers = .85; rmothers&fathers = .81, p b .001). When both parents completed the questionnaire, scores of
mothers and fathers were averaged to obtain a general mean measure of rule-breaking behavior (63%). In all other cases the separate scores of mothers or fathers were used.
2.2.4. DAT1 and DRD4 genotyping
Genotyping of the saliva samples was performed by the Center for Human GeneticsLeuven University. We used the following
primers. For the 40-bp VNTR in the 3 UTR of the DAT1 gene: forward primer: 5-VICTGCGGTGTAGGGAACGGCC TGAG-3; reverse
primer: 5-CTTCCTGGAGGTCACGGCTCAAGG-3. For the 48-bp VNTR in exon 3 of the DRD4 gene: forward primer: 5-NEDGCGACTACGTGGTCTACTCG-3; reverse primer: 5-AGGACCCTCATGGCCTTG-3. The amplication mixture for PCR of the DAT1VNTR and the DRD4-VNTR included 50 ng genomic DNA, 12.5 l Master Mix (Promega, USA), .5 mol/l of each forward and reverse primer, 1 M betaine solution (Sigma-Aldrich, USA), and 1.5 l water. The cycling conditions for the PCR started with
5 min at 95 C, followed by 35 cycles of 30 s at 95 C, 30 s at 60 C, and 90 s at 72 C, afterward followed by 7 min at 72 C.
After nishing PCR, the DNA mixture was cooled down to 4 C and fragment analysis was performed. After a nal denaturation
step at 95 C for 3 min, analysis followed on an ABI 3730xl Genetic Analyzer (Applied Biosystems, USA). Results were printed with
the GeneMarker software Version 1.91 (SoftGenetics, 2010).
Based on the classication used in previous research (e.g., Mill et al., 2002) and the genotype frequencies found in the present
study sample, we dichotomized the DAT1 genotypes into DAT1-10R homozygotes (n = 575) versus DAT1-9R carriers
(i.e., genotypes homozygous for the 9-repeat and heterozygous genotypes possessing both 9- and 10-repeats) (n = 450). A
total of 35 participants with other genotypes were left out of the analyses and for 28 participants the DAT1 genotype information
could not be computed by the Center for Human Genetics. Again based on the classication used in previous research
(e.g., Bakermans-Kranenburg et al., 2008) and the genotype frequencies found in the present study sample, we dichotomized
the DRD4 genotypes into long carriers (i.e., genotypes possessing at least one 7R or longer) (N = 360) versus non-long carriers
(i.e., shorter than 7R) (N = 664). For 29 participants the DRD4 genotype information could not be computed by the Center for
Human Genetics.
2.2.5. Control variables
Three variables were included as control variables: adolescent gender, adolescent age, and maternal education as indicators of
socioeconomic status. Maternal education was coded on an ordered metric representing the highest degree. As GxE interactions
are difcult to interpret in the presence of geneenvironment correlations (e.g., Rathouz, Van Hulle, Rodgers, Waldman, &
Lahey, 2008), we controlled for direct relationships between genes and environment in our analyses.
3. Statistical analyses
Statistical analyses were conducted in two steps. In a rst step, eight GxE interaction effects were investigated. More specically, four separate regression models were tested per outcome variable (i.e., rule-breaking behavior and behavioral engagement)

18

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

using Mplus version 6 (Muthn & Muthn, 19982010). Each regression model consisted of two steps. In step 1, one gene variable
and one teacherstudent variable were added while controlling for adolescents' age, gender, maternal education, and geneenvironment correlations. In step 2, the GxE interaction variable was added. The teacherstudent relationship variables were standardized before computing interaction terms (Aiken, West, & Reno, 1991). As eight separate models were tested, we controlled
for multiple testing using the conservative Bonferroni correction. To take into account the clustering of adolescents in classes,
we used the complex analysis feature in all models, which adjusts the standard errors of the estimated path coefcients for clustering (Williams, 2000). To deal with missing data, we analyzed the models using the full information maximum likelihood
(FIML) algorithm (Jelicic, Phelps, & Lerner, 2009), which borrows information from the observed portion of the data instead of
imputing data. For student-reported teacherstudent dissatisfaction and afliation 1.5% of the data was missing, for parentreported rule-breaking behavior and behavioral engagement 25% of the data was missing. The MCAR test (missing completely
at random test) shows the data was missing at random ( (10) = 15.330, p = .120). To take into account the non-normality
of our data, we used robust standard error estimates (MLR).
In a second step, we investigated whether the signicant interaction effects found in step 1 were in line with differential susceptibility, diathesisstress, or vantage sensitivity. To this end, we conducted three critical tests (Roisman et al., 2012). First, the
regions of signicance (RoS) were identied on the environmental predictor (i.e., teacherstudent relationships) (Dearing &
Hamilton, 2006; Kochanska, Kim, Barry, & Philiber, 2011). These regions examine the full range of the environmental predictor
for which there is signicant spread in the outcome variable (i.e., behavioral engagement or rule-breaking behavior) associated
with the two genotypes. Hence, RoS indicate all the values of the environmental predictor for which the outcome signicantly
differs between the two genetic groups (Roisman et al., 2012). Moreover, when probing the interaction, we bracketed the
range of interest by 2 SD from the mean of the environmental predictor (Aiken et al., 1991). Second, in addition to RoS testing
on the environmental predictor, we calculated the proportion of the sample above the crossover point (i.e., the value of that predictor at which the regression lines cross over). The crossover point of the interaction, proportion above the crossover point, and
RoS were calculated using the web-based program at http://www.yourpersonality.net/interaction. Third, to check whether any
signicant interaction effect was an artifact of imposing a linear model on a nonlinear phenomenon, we estimated an additional
regression model that included quadratic terms (Roisman et al., 2012).
4. Results
4.1. Correlations
Correlations among the core variables are reported in Table 1. Genotype classications for the two genes were independent
from one another, and the correlations between the two aspects of teacherstudent relationships and between the two parentreported outcomes were negative, signicant, and moderate in size (i.e., .30 b r b .50; Cohen, 1988). Teacherstudent afliation
was negatively associated with rule-breaking behavior and positively with behavioral engagement, whereas teacherstudent dissatisfaction showed the opposite pattern of correlations. DAT1-9R carriers displayed more behavioral engagement. Geneenvironment correlations (i.e., associations between the genotypes and teacherstudent relationships) were non-signicant.
Signicant effects were found for all three control variables. Regarding gender, boys had signicantly higher mean levels of
rule-breaking behavior (t(834) = 4.00, p b .001) compared to girls and lower mean levels of behavioral engagement
(t(841) = 8.06, p b .001). Regarding age, we found that older adolescents reported less teacherstudent afliation (r =
.16, p b .001), more teacherstudent dissatisfaction (r = .12, p b .001), and more rule-breaking behavior (r = .07, p b .05) compared to younger adolescents. Regarding maternal education, adolescents displayed more rule-breaking behavior when their
mothers had a lower level of education (r = .08, p b .05).
4.2. Step 1: Regression analyses
The results from the regression analyses are reported in Table 2. The left-hand side of the table depicts the four separate regression models including rule-breaking behavior (models 14), whereas the four models including behavioral engagement are

Table 1
Bivariate correlations, means, standard deviations, and minimum and maximum values of the main study variables.
Variable

1. DAT1
2. DRD4
3. TS afliation
4. TS dissatisfaction
5. Rule-breaking behavior
6. Behavioral engagement

1
.00
.03
.03
.02
.07

1
.03
.02
.06
.05

1
.47
.21
.32

.13
.22

1
.40

M
.45
.35
2.60
1.83
.18
3.96

SD

Min
.50
.48
.63
.58
.12
.78

.00
.00
1.00
1.00
.00
1.00

Max
1.00
1.00
4.00
4.00
1.29
5.00

Note: DAT1 = dopamine transporter gene, dummy coded (0 = 10R homozygotes, 1 = 9R carriers); DRD4 = dopamine D4 receptor gene, dummy coded
(0 = non-long carriers, 1 = long carriers); TS = teacherstudent.
p b .05.
p b .001.

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

19

Table 2
Regression of rule-breaking behavior and behavioral engagement on main and interaction effects of genes and environments; p b .006 (Bonferroni corrected .05).

Predictor
Step 1 Afliation
DAT1
Step 2 DAT1 afliation

Rule-breaking behavior
Regression estimates

Behavioral engagement
Regression estimates

Model 1

Model 5

1 = .28
2 = .03
3 = .13

1 = .34
2 = .08
3 = .26

Step 1 Dissatisfaction
DAT1
Step 2 DAT1 dissatisfaction

Model 2
1 = .19
2 = .03
3 = .12

Model 6
1 = .34
2 = .07

Step 1 Afliation
DRD4
Step 2 DRD4 afliation

Model 3
1 = .20
2 = .05
3 = .02

Model 7
1 = .18
2 = .04
3 = .07

Step 1 Dissatisfaction
DRD4
Step 2 DRD4 dissatisfaction

Model 4
1 = .31
2 = .06
3 = .25

Model 8
1 = .36
2 = .04
3 = .23

3 = .14

Note: N = 1058. In all analyses, we controlled for gender, age, maternal education, geneenvironment correlations and nesting of students within classrooms. The
dopamine transporter gene DAT1 was dummy coded (0 = 10R homozygotes, 1 = 9R carriers). The dopamine D4 receptor gene DRD4 was dummy coded
(0 = non-long carriers, 1 = long carriers).

p b .10.
p b .05.
p b .01.
p b .001

depicted on the right-hand side (models 58). In each model, step 1 provides the regression estimates that indicate whether a
signicant main effect was found. Step 2 provides the regression estimate that indicates whether a signicant interaction effect
was found. To check whether gender moderated the GxE interactions, 8 multi-group analyses were conducted in which a
constrained model (with the GxE effect set to be equal across gender) was compared with an unconstrained model (in which
the GxE effect was allowed to vary across gender). Fit comparisons were made by means of the S-B2 (Satorra & Bentler,
2001) with signicant S-B2 indicating moderation by gender. None of these tests appeared to be signicant.

Fig. 1. Teacherstudent afliation predicting adolescent behavioral engagement for DAT1-10R homozygotes versus DAT1-9R carriers.

20

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

Five out of eight regression models did not yield a signicant interaction term (i.e., models 1, 2, 3, 6, and 7). Hence, the main
effects of these models may be interpreted. Results show that more teacherstudent afliation predicts more behavioral engagement and less rule-breaking behavior whereas more teacherstudent dissatisfaction predicts less behavioral engagement. In addition, DAT1-10R homozygotes display less behavioral engagement compared to DAT1-9R carriers.
Three out of eight regression models yielded a signicant interaction term. Step 2 of model 5 shows that behavioral engagement is predicted not only by teacherstudent afliation and DAT1 (RStep1 = .12), but also by their interaction (R = .01,
p b .05). Step 2 of model 4 shows that rule-breaking behavior is predicted not only by teacherstudent dissatisfaction and
DRD4 (RStep1 = .03), but also by their interaction (R = .01, p b .05). Finally, model 8 shows that DRD4 also moderates the
main effect of teacherstudent dissatisfaction on behavioral engagement (R = .01, p b .05). After controlling for multiple testing
using the Bonferroni correction, the interactions found in models 4 and 5 remained signicant. Hence, we only visualized these
two interactions (see Figs. 1 and 2 respectively). The interactions were probed from 2 to +2 SD from the mean of the environmental predictor (Aiken et al., 1991). Both gures show a crossover interaction with a stronger association between teacherstudent afliation and behavioral engagement for DAT1-10R homozygotes (Fig. 1) and a stronger association between teacher
student dissatisfaction and rule-breaking behavior for DRD4-non-long carriers (Fig. 2). Additional critical tests were conducted
to nd out whether these signicant interaction effects are in line with the differential susceptibility, diathesisstress, or vantage
sensitivity hypothesis.

4.3. Step 2: Critical tests proposed by Roisman et al. (2012)

Model 4 indicates that that rule-breaking behavior is predicted by teacherstudent dissatisfaction, DRD4, and the interaction
between them. Hence, the area in Fig. 2 above the crossover point at .39 can be labeled w (i.e., for worse) and the area
below the crossover point can be labeled b (i.e., for better). The lower-bound RoS on the environmental predictor lie below
2 SD from the mean ( 2.908). None of the adolescents had such a low score on that predictor. The upper-bound RoS on
the predictor lie within the bounds of 2 to +2 SD (.092). The RoS are depicted in gray in Fig. 2. Hence, the difference between
DRD4 long allele carriers and DRD4 non-long allele carriers is only signicant when adolescents experience average or high levels
of teacherstudent dissatisfaction which seems to be consistent with diathesisstress. More specically, lower levels of teacher
student dissatisfaction predict less rule-breaking behavior for both DRD4 long and non-long carriers whereas average to high
levels of teacherstudent dissatisfaction predict more rule-breaking behavior especially for DRD4 non-long carriers. The proportion
of the study sample above the crossover point is 76%. Thus, the vast majority of cases in our study sample are situated above the
crossover point which seems to be consistent with diathesisstress. Non-linear predictor terms were not signicant, implying that
the interaction effect was not an artifact of imposing a linear model on a plausibly nonlinear phenomenon (Roisman et al., 2012)

Fig. 2. Teacherstudent dissatisfaction predicting adolescent rule-breaking behavior for DRD4-long carriers versus DRD4-non-long carriers.

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

21

Model 5 indicates that behavioral engagement is predicted by teacherstudent afliation, DAT1, and the interaction between
them. Hence, the area in Fig. 1 above the crossover point at .56 can be labeled b (i.e., for better) and the area below the crossover
point can be labeled w (i.e., for worse). The upper-bound RoS on the environmental predictor lie above 2 SD from the mean
(2.393). None of the adolescents in our sample had such a high score on that predictor. The lower-bound RoS on the predictor
lie within the bounds of 2 to +2 SD (.089). The RoS are depicted in gray in Fig. 1. Hence, the difference between DAT1-10R
homozygotes and DAT1-9R carriers is only signicant when adolescents experience average or low levels of teacherstudent afliation, which seems to be consistent with diathesisstress. More specically, higher levels of teacherstudent afliation predict
more behavioral engagement for both DAT1-10R homozygotes and DAT1-9R carriers whereas low to average levels of teacherstudent afliation predict less behavioral engagement especially for DAT1-10R homozygotes. The proportion of the study sample
above the crossover point is 20%. Hence, the vast majority of cases in our study sample are situated below the crossover point
which also indicates that diathesisstress is at work. Non-linear predictor terms were not signicant. In sum, whereas the gures
of the two cross-over interactions seem to suggest that differential susceptibility is at work, the Roisman analyses show our data
can be interpreted as support for the diathesisstress model over the differential susceptibility and vantage sensitivity models.
5. Discussion
The main aim of this study was to contribute to research on the developmental signicance of teacherstudent relationships
and on the moderating role of child factors. More specically, we examined whether students with particular genetic variants of
the dopaminergic genes DAT1 and DRD4 are more susceptible to the effects of teacherstudent relationships on rule-breaking behavior and behavioral engagement. After controlling for multiple testing, two regression models yielded a signicant interaction
effect. Both models provided support for the diathesisstress hypothesis.
First, adolescents were more behaviorally engaged in school when experiencing more highly supportive, afliative teacher
student relationships, but this association was stronger for DAT1-10R homozygotes. More specically, average or low levels of
teacherstudent afliation predicted less behavioral engagement especially for adolescents homozygous for DAT1-10R. Hence, it
seems the DAT1-10R homozygotes were more responsive or affected by the negative effect of lower teacherstudent afliation
on behavioral engagement. These results seem to be in line with GxE research by Laucht et al. (2007) who found that the effect
of adverse family factors on inattention and hyperactivityimpulsivity was stronger for DAT1-10R homozygotes.
Second, adolescents displayed more rule-breaking behavior when experiencing higher levels of dissatisfaction with their
teachers, regardless of their DRD4 genotype. However, this association was stronger for DRD4 non-long carriers. More specically,
average to high levels of teacherstudent dissatisfaction predicted more rule-breaking behavior especially for DRD4 non-long carriers. Hence, it seems that DRD4 non-long carriers were more affected by the adverse effect of teacherstudent dissatisfaction on
rule-breaking behavior. Thus, in relation to rule-breaking behavior, the DRD4-long allele (i.e., at least one 7R allele or longer) acts
as a protective rather than a risk allele. These results are not in line with our hypothesis and contradict most prior GxE research
including parenting as an environmental factor (Bakermans-Kranenburg & van IJzendoorn, 2011). Nevertheless there are other
studies that found similar interactive effects. Specically, the DRD4-long allele (and especially the 7R allele) has been found to
make children less susceptible to family adversity (e.g., Propper et al., 2007) and to protect young adults against the adverse effect
of childhood trauma on emotional resilience (Das et al., 2011).
In the ve regression models without a signicant GxE interaction, several signicant main effects of genes and/or environments were found. First, adolescents who experienced more afliation with their teachers were more behaviorally engaged in
school and displayed less rule-breaking behavior. Contrarily, adolescents who experienced more dissatisfaction with their teachers
displayed more rule-breaking behavior and less behavioral engagement. These results are line with previous research by O'Connor
et al. (2011) and O'Farrell et al. (2006). Second, adolescents homozygous for DAT1-10R displayed less behavioral engagement
compared to DAT1-9R carriers. This nding can be linked to previous research that found DAT1-10R homozygotes to have
more concentration and attention problems, which can be considered aspects of behavioral engagement (e.g., Gordon et al., 2012).
Interestingly, the genotypes that were associated with either more rule-breaking behavior (i.e., DRD4 non-long allele carriers)
or less behavioral engagement (i.e., DAT1-10R homozygotes) are the same genotypes that were previously associated with less
dopamine availability (e.g., Asghari et al., 1995; Heinz et al., 2000). In that way, our study provides additional support for the notion that low dopamine availability is related to an unpleasant emotional state which makes children and adolescents prone to
more thrill- and sensation seeking behavior, egocentrically driven behavior (Matthys et al., 2013), and risk-taking (Robinson &
Berridge, 2008). However, these scarce neurobiological research ndings have just scratched the surface with regard to how/
which genetic markers have actual effects on cellular function. Therefore, caution is needed when trying to explain genebehavior
associations from a neurobiological point of view.
5.1. Strengths and limitations
Despite the need for further research unraveling the neurobiological processes involved, the current study contributes to extant literature in several ways. With regard to the teacherstudent relationship literature, this study showed that adolescents
with particular genetic variants are more vulnerable to the effects of negative or lack of positive teacherstudent relationships.
Although GxE interactions accounted for only a limited amount of the variance in outcomes, these ndings contribute to prior research on teacherstudent relationships that has found that students with markers of vulnerability at the behavioral level are
more susceptible to the quality of the relationship with their teachers (Sabol & Pianta, 2012). The nding that this also applies

22

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

at the level of genes extends this research signicantly and makes a stronger case for the presence of bio-ecological processes in
schools. Likewise, this study adds to prior geneenvironment interaction research by extending it to schools. Indeed, in contrast to
family-related factors, the interplay of genes with school-related factors has hardly been investigated (Brendgen, 2012).
Despite the study's general contributions and methodological strengths, such as the large and heterogeneous sample of students and the use of different informants for environmental factors (i.e., student perceptions) and outcomes (i.e., parent reports),
it is important to consider some limitations and suggestions for future research as well. First, as our study is the rst to include
teacherstudent relationships in molecular GxE interaction research, hypotheses regarding GxE interactions were drawn from research including parenting. However, as both relationships clearly differ (in terms of emotional investment, genetic similarity, ),
the interplay between teacherchild relationships and genes is not necessarily similar to the interplay between parentchild relationships and genes. Thus, there is need for more GxE research including teacherstudent relationships and we suggest future
research to investigate the differential role of parents and teachers within a GxE framework. Second, the inconsistent ndings
in GxE interaction research on parentchild relationships made it difcult to formulate hypotheses and warrant caution. A possible explanation for the contradicting results in GxE interaction research may be that the candidate gene approach and the focus
on a single location within a gene are too simplistic to grasp the true complexity of human genetics. We strongly encourage future
research to include multiple co-acting genes, to look at multiple genetic markers on each gene, and to create an index of genetic
vulnerability across a network of genes (e.g., Civelek & Lusis, 2014). Third, in line with most research on affective teacherstudent
relationships among secondary school students, students were asked to report on the relationship they had with their teachers in
general. An alternative option would be to let students report on their relationship with one specic teacher, for example their
mentor. Although this gives an indication of dyadic teacherchild relationship quality instead of the more general relational climate at school, this option has drawbacks too. Especially in school systems where students have a lot of teachers and typically
do not have frequent contact with their mentors (such as the school system in XXX), this assessment may not capture the
most signicant relationship experiences for the student. Fourth, parents reported on their children's behavioral school engagement and rule-breaking behavior. Although parents may not have full access to all school-related behaviors of students, the
fact that we found the expected associations with student-reported teacherstudent relationships adds to the validity of these
parent reports. Moreover, these results add tso the generalizability of teacherstudent relationship correlates beyond behaviors
perceived by actors in the school context (i.e., same or other teachers). Fifth, due to the cross-sectional nature of the present
study, the associations that were found may not reect causal relations. It could be that the identied genetic risks are associated
with an underlying behavior disposition that places students at risk for behavioral problems and poor relationships with teachers.
Therefore, longitudinal GxE studies are needed.
5.2. Practical implications
In general, the present study shows that students with particular genetic variants are more vulnerable to the effects of negative (or lack of positive) teacherstudent relationships, conrming that student outcomes depend on the interplay of teacherstudent relationships and child biological factors. These ndings bring more insight into the complex underlying mechanisms of
student psychosocial (mal) adjustment and, thus, contribute to the knowledge base in school psychology. Although the immediate
signicance for school practice appears to be limited, we argue that these insights may nonetheless yield implications for teachers
and schools, in the short or in the long run.
First, all people have implicit ideas about the relevance of genes for behavior. Teachers also encounter arguments that genetic
predispositions explain student behavioral problems. This genetic attribution may yield biased thoughts regarding this behavior,
such as that it is unchangeable, which may reduce teachers' willingness to invest in interventions with these students (DarNimrod & Heine, 2011). The present study (among others) elucidates the complexity of the role of genes and environments by
showing that it is not either genes or environment, but the interplay of both that shapes behavior. Moreover, our study reveals
that providing a supportive school environment is important for all students (even though effects may be less pronounced in
some individuals compared to others). Research that advocates the importance of the environment while taking into account
the possible role of genes (such as the present study), is even more convincing in that respect than research that only includes
environments. In order to translate the insights of our study to practical guidelines for teachers and schools, school psychologists
seem to be well qualied. Through their role as school-based consultant (Gutkin & Curtis, 2009), school psychologists can increase
teachers' awareness and knowledge of how both genetic predispositions and environmental factors explain student behavior. For
example, school psychologists could highlight the fact that bad student behavior is not fully predetermined by a genetic risk, and
therefore unchangeable, but that aspects of teacherstudent relationships may buffer detrimental effects of genetic risks. This
psycho-education may decrease teachers' genetic attribution bias and empower them to invest their time and resources in students with psychosocial adjustment problems. In addition, school psychologists can guide schools in implementing interventions
to create a supportive class and school environment (Gutkin & Curtis, 2009). As our study shows that the perceived affective quality of teacherstudent interactions (e.g., feeling accepted by your teachers) is important for students' academic and psychosocial
adjustment, interventions should promote positive, supportive teacherstudent relationships. My Teaching Partner (Allen, Pianta,
Gregory, Mikami, & Lun, 2011) is an intervention program that focuses on improving classroom-based teacher behavior and has
been implemented in secondary school. Banking Time (Pianta & Hamre, 2001) and Playing-2-gether (Vancraeyveldt et al., 2015)
are alternative interventions aimed at improving teacherstudent relationships but they are aimed at preschoolers. In order to implement their principles in secondary schools, we recommend future research to create a version that is tailored to the specic
social world and characteristics of older students.

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

23

Second, some scholars (e.g., Mufano, Zammit, & Flint, 2014) have argued that it would be theoretically possible for GxE
research to inform psychological practice through the identication of high-risk groups. More specically, school psychological
interventions could potentially benet from our research as it allows targeting specic groups of students with a heightened
genetic risk for adjustment problems (i.e., targeted interventions). However, although some scholars have argued that
genetically-based targeted interventions are possible, it is obviously too early to carry them out in practice, given the inconsistent
results from prior GxE studies, the understudied biological pathways that link genes with behavior, and, evidently, the
fundamental ethical and practical complexities (Fisher & McCarthy, 2013). Nevertheless, as the eld of GxE studies continues to
evolve, GxE studies may eventually facilitate psychological interventions in the school context.
6. Conclusion
Prior research on GxE interactions has largely overlooked teacherstudent relationships as a social environmental factor. The
present study examined whether and how dopaminergic genes (i.e., the dopamine transporter or DAT1 and the dopamine D4 receptor or DRD4 genes) moderate the effect of student-reported teacherstudent relationships (i.e., afliation and dissatisfaction)
on parent-reported rule-breaking behavior and behavioral school engagement. As expected, adolescents who experienced stronger afliation with their teachers were more behaviorally engaged in school, whereas greater dissatisfaction predicted more rulebreaking behavior. In addition, a signicant geneenvironment interaction was found for each of the genes examined. The link
between low teacherstudent afliation and low behavioral engagement was more pronounced for adolescents who had two copies of the long variant of the DAT1 gene. The link between high teacherstudent dissatisfaction and more rule-breaking behavior
was more pronounced in adolescents who did not have the long variant of the DRD4 gene.
In sum, students with particular genetic variants were found to be more vulnerable to the effects of negative (or lack of positive) teacherstudent relationships, conrming that student outcomes depend on the interplay of teacherstudent relationships
and child biological factors. These results conrm and extend teacherstudent relationship research that used child risk factors at
the behavioral level (Sabol & Pianta, 2012). In general, they conrm the basic assumption of Bronfenbrenner and Morris' (2006)
bio-ecological model, that a complex interplay of genetic and environmental inuences underlies child behavior and development.
References
Achenbach, T. M., & Rescorla, L. A. (2001). Manual for ASEBA school-age forms & profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth, &
Families.
Aiken, L. S., West, S. G., & Reno, R. R. (1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.
Alexander, K. L., Entwisle, D. R., & Horsey, C. S. (1997). From first grade forward: Early foundations of high school dropout. Sociology of Education, 70, 87107. http://dx.
doi.org/10.2307/2673158.
Allen, J. P., Pianta, R. C., Gregory, A., Mikami, A. Y., & Lun, J. (2011). An interaction-based approach to enhancing secondary school instruction and student achievement.
Science, 333, 10341037. http://dx.doi.org/10.1126/science.1207998.
Arbeau, K. A., Coplan, R. J., & Weeks, M. (2010). Shyness, teacher-child relationships, and socio-economic adjustment in grade 1. International Journal of Behavioral
Development, 34(3), 259269. http://dx.doi.org/10.1177/0165025409350959.
Asghari, V., Sanyal, S., Buchwaldt, S., Paterson, A., Jovanovic, V., & Van Tol, H. H. (1995). Modulation of intercellular cyclic AMP levels by different human dopamine D4
receptor variants. Journal of Neurochemistry, 65, 11571165. http://dx.doi.org/10.1046/j.1471-4159.1995.65031157.x.
Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2011). Differential susceptibility to rearing environment depending on dopamine-related genes: New evidence
and a meta-analysis. Development and Psychopathology, 23, 3952. http://dx.doi.org/10.1017/S0954579410000635.
Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., Pijlman, F. T., Mesman, J., & Juffer, F. (2008). Experimental evidence for differential susceptibility: Dopamine D4
receptor polymorphism (DRD4 VNTR) moderates intervention effects on toddlers' externalizing behavior in a randomized controlled trial. Developmental
Psychology, 44, 293300. http://dx.doi.org/10.1037/0012-1649.44.1.293.
Beaver, K. M., Wright, J. P., & DeLisi, M. (2008). Delinquent peer group formation: Evidence of a gene X environment correlation. Journal of Genetic Psychology, 169,
227244. http://dx.doi.org/10.3200/Gntp.169.3.227-244.
Belsky, J., & Pluess, M. (2009). Beyond diathesis-stress: Differential susceptibility to environmental influences. Psychological Bulletin, 135, 885890.
Birch, S., & Ladd, G. (1997). The teacherchild relationship and children's early school adjustment. Journal of School Psychology, 35, 6179. http://dx.doi.org/10.1016/
S0022-4405(96)00029-5.
Blum, K., Chen, A. L. C., Oscar-Berman, M., Chen, T. J. H., Lubar, J., White, N., ... Bailey, J. A. (2011). Generational association studies of dopaminergic genes in reward
deficiency syndrome (RDS) subjects: Selecting appropriate phenotypes for reward dependence behaviors. International Journal of Environmental Research and
Public Health, 8, 44254459. http://dx.doi.org/10.3390/ijerph8124425.
Boardman, J. D., Menard, S., Roettger, M. E., Knight, K. E., Boutwell, B. B., & Smolen, A. (2014). Genes in the dopaminergic system and delinquent behaviors across the
life course: The role of social controls and risks. Criminal Justice and Behavior.. http://dx.doi.org/10.1177/0093854813514227.
Brendgen, M. (2012). Genetics and peer relations: A review. Journal of Research on Adolescence, 22, 419437. http://dx.doi.org/10.1111/j.1532-7795.2012.00798.x.
Bronfenbrenner, U., & Morris, P. A. (2006). The bioecological model of human development. In R. M. Lerner, & W. Damon (Eds.), Handbook of child psychology (6th ed.)1.
(pp. 793828). Hoboken, NJ: Wiley.
Brookes, K. J., Neale, B. M., Sugden, K., Khan, N., Asherson, P., & D'Souza, U. M. (2007). Relationship between VNTR polymorphisms of the human dopamine transporter
gene and expression in post-mortem midbrain tissue. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 144, 10701078. http://dx.doi.org/10.
1002/ajmg.b.30572.
Burt, S. A., & Mikolaiewski, A. J. (2008). Preliminary evidence that specific candidate genes are associated with adolescent-onset antisocial behavior. Aggressive
Behavior, 34, 437445. http://dx.doi.org/10.1002/Ab.20251.
Chang, F. M., Kidd, J. R., Livak, K. J., Pakstis, A. J., & Kidd, K. K. (1996). The world-wide distribution of allele frequencies at the human dopamine D4 receptor locus.
Human Genetics, 98, 91101. http://dx.doi.org/10.1007/s004390050166.
Chen, T. J. H., Blum, K., Mathews, D., Fisher, L., Schnautz, N., Braverman, E. R., ... Comings, D. E. (2007). Preliminary association of both the dopamine D2 receptor (DRD2)
[Taq1 A1 Allele] and the dopamine transporter (DAT1) [480 bp Allele] genes with pathological aggressive behavior, a clinical subtype of reward deficiency syndrome (RDS) in adolescents. Gene Therapy and Molecular Biology, 11, 93101.
Civelek, M., & Lusis, A. J. (2014). Systems genetics approaches to understand complex traits. Nature Reviews Genetics, 15, 3448. http://dx.doi.org/10.1038/nrg3575.
Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Erlbaum.
Cook, E., Greenberg, M., & Kusche, C. (1995). People in My Life: Attachment relationships in middle childhood. Paper presented at the biennial meeting of the Society for
Research in Child Development. Indiana: Indianapolis.

24

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

Cornish, K. M., Manly, T., Savage, R., Swanson, J., Morisano, D., Butler, N., ... Hollis, C. P. (2005). Association of the dopamine transporter (DAT1) 10/10-repeat genotype
with ADHD symptoms and response inhibition in a general population sample. Molecular Psychiatry, 10, 686698. http://dx.doi.org/10.1038/sj.mp.4001641.
Dar-Nimrod, I., & Heine, S. (2011). Genetic essentialism: On the deceptive determinism of DNA. Psychological Bulletin, 137(5), 800818.
Das, D., Cherbuin, N., Tan, X., Anstey, K. J., & Easteal, S. (2011). DRD4-exonIII-VNTR moderates the effect of childhood adversities on emotional resilience in youngadults. PLoS One, 6, e20177. http://dx.doi.org/10.1371/journal.pone.0020177.
Dearing, E., & Hamilton, L. C. (2006). Contemporary approaches and classic advice for analyzing mediating and moderating variables. Monographs of the Society for
Research in Child Development, 71, 88104.
Dmitrieva, J., Chen, C., Greenberger, E., Ogunseitan, O., & Ding, Y. -C. (2011). Gender-specific expression of the DRD4 gene on adolescent delinquency, anger, and thrill
seeking. Social Cognitive and Affective Neuroscience, 6, 8289.
Essex, M. J., Armstrong, J. M., Burk, L. R., Goldsmith, H. H., & Boyce, W. T. (2011). Biological sensitivity to context moderates the effects of early teacherchild relationships on the development of mental health by adolescence. Developmental Psychology, 23, 149161. http://dx.doi.org/10.1017/S0954579410000702.
Fisher, C.,. B., & McCarthy, E.,. L.,. H. (2013). Ethic in prevention science involving genetic testing. Prevention Science, 14, 310318. http://dx.doi.org/10.1007/s11121012-0318-x.
Fossella, J., Sommer, T., Fan, J., Wu, Y., Swanson, J. M., Pfaff, D. W., et al. (2002). Assessing the molecular genetics of attention networks. BioMedical Central Neuroscience,
3, 14. http://dx.doi.org/10.1186/1471-2202-3-14.
Gordon, E. M., Stollstorff, M., Devaney, J. M., Bean, S., & Vaidya, C. J. (2012). Effect of dopamine transporter genotype on intrinsic functional connectivity depends on
cognitive state. Cerebral Cortex, 22, 21822196. http://dx.doi.org/10.1093/cercor/bhr305.
Guo, G., Cai, T., Guo, R., Wang, H., & Harris, K. M. (2010). The dopamine transporter gene, a spectrum of most common risky behaviors, and the legal status of the behaviors. PLoS ONE, 5, e9352.
Guo, G., Roettger, M. E., & Shih, J. C. (2007). Contributions of the DAT1 and DRD2 genes to serious and violent delinquency among adolescents and young adults. Human
Genetics, 121, 125136. http://dx.doi.org/10.1007/s00439-006-0244-8.
Gutkin, T. B., & Curtis, M. (2009). School-based consultation: The science and practice of indirect service delivery. In T. B. Gutkin, & C. R. Reynolds (Eds.), The handbook
of school psychology (pp. 463496) (4th ed.). New York, NY: Wiley.
Heinz, A., Goldman, D., Jones, D. W., Palmour, R., Hommer, D., & Gorey, J. G. (2000). Genotype influences in vivo dopamine transporter availability in human striatum.
Neuropsychopharmacology, 22, 133139. http://dx.doi.org/10.1016/S0893-133X(99)00099-8.
Hohmann, S., Becker, K., Fellinger, J., Banaschewski, T., Schmidt, M. H., Esser, G., & Laucht, M. (2009). Evidence for epistasis between the 5-HTTLPR and the dopamine
D4 receptor polymorphisms in externalizing behavior among 15-year-olds. Journal of Neural Transmission, 116, 16211629. http://dx.doi.org/10.1007/s00702-0090290-1.
Jelicic, H., Phelps, E., & Lerner, R. (2009). Use of missing data methods in longitudinal studies: The persistence of bad practices in developmental psychology.
Developmental Psychology, 45, 11951199. http://dx.doi.org/10.1037/a0015665.
Jillian, S., & Ladd, G. W. (2011). Measuring school engagement: A longitudinal evaluation of the school liking and avoidance questionnaire from kindergarten through
sixth grade. (ASU Electronic Dissertation. Retrieved from) http://hdl.handle.net/2286/R.A.56844
Johnson, W., McGue, M., & Iacono, W. G. (2006). Genetic and environmental influences on academic achievement trajectories during adolescence. Developmental
Psychology, 42, 514532. http://dx.doi.org/10.1037/0012-1649.42.3.514.
Kochanska, G., Kim, S., Barry, R. A., & Philibert, R. A. (2011). Children's genotypes interact with maternal responsive care in predicting children's competence:
Diathesisstress or differential susceptibility? Development and Psychopathology, 23, 605616. http://dx.doi.org/10.1017/S0954579411000071.
Ladd, G. W., & Dinella, L. M. (2009). Continuity and change in early school engagement: Predictive of children's achievement trajectories from first to eighth grade?
Journal of Educational Psychology, 101, 190206. http://dx.doi.org/10.1037/a0013153.
Ladd, G. W., Buhs, E. S., & Seid, M. (2000). Children's initial sentiments about kindergarten: Is school liking an antecedent of early classroom participation and achievement? Merrill-Palmer Quarterly, 46, 255279.
Lahey, B. B., Rathouz, P. J., Lee, S. S., Chronis-Tuscano, A., Pelham, W. E., Waldman, I. D., & Cook, E. H. (2011). Interactions between early parenting and a polymorphism
of the child's dopamine transporter gene in predicting future child conduct disorder symptoms. Journal of Abnormal Psychology, 120, 3345. http://dx.doi.org/10.
1037/a0021133.
Laucht, M., Skowronek, M. H., Becker, K., Schmidt, M. H., Esser, G., Schulze, T. G., & Rietschel, M. (2007). Interacting effects of the dopamine transporter gene and psychosocial adversity on attention-deficit/hyperactivity disorder symptoms among 15-year-olds from a high-risk community sample. Archives of General Psychiatry,
654, 585590. http://dx.doi.org/10.1001/archpsyc.64.5.585.
Lee, S. S., Lahey, B. B., Waldman, I., Van Hulle, C. A., Rathouz, P., Pelham, W. E., ... Cook, E. H. (2007). Association of dopamine transporter genotype with disruptive
behavior disorders in an eight-year longitudinal study of children and adolescents. American Journal of Medical Genetics Part B: Neuropsychiatric Genetics, 144B,
310317. http://dx.doi.org/10.1002/Ajmg.B.30447.
Li, J. J., & Lee, S. S. (2013). Interaction of dopamine transporter gene and observed parenting behaviors on attention-deficit/hyperactivity disorder: A structural equation
modeling approach. Journal of Clinical Child and Adolescent Psychology, 42, 174186. http://dx.doi.org/10.1080/15374416.2012.736355.
Li, Y., & Lerner, R. M. (2011). Trajectories of school engagement during adolescence: Implications for grades, depression, delinquency, and substance use.
Developmental Psychology, 47, 233247. http://dx.doi.org/10.1037/a0021307.
Mata, R., Hau, R., Papassotiropoulos, A., & Hertwig, R. (2012). DAT1 polymorphism is associated with risk taking in the balloon analogue risk task (BART). PLoS ONE, 7,
e39135. http://dx.doi.org/10.1371/journal.pone.0039135.
Matthys, W., Vanderschuren, L., & Schutter, D. (2013). The neurobiology of oppositional defiant disorder and conduct disorder: Altered functioning in three mental
domains. Development and Psychopathology, 25, 193207. http://dx.doi.org/10.1017/S0954579412000272.
Mill, J., Asherson, P., Browes, C., D'Souza, U., & Craig, I. (2002). Expression of the dopamine transporter gene is regulated by the 3 UTR VNTR: Evidence from brain and
lymphocytes using quantitative RT PCR. American Journal of Medical Genetics, 114, 975979. http://dx.doi.org/10.1002/ajmg.b.10948.
Mitchell, R. J., Howlett, S., Earl, L., White, N. G., McComb, J., Schanfield, M. S., ... Crawford, M. H. (2000). Distribution of the 3' VNTR polymorphism in the human dopamine transporter gene in world populations. Human Biology, 72, 295304.
Moffitt, T. E., Caspi, A., & Rutter, M. (2006). Measured geneenvironment interactions in psychopathology: Concepts, research strategies, and implications for research,
intervention, and public understanding of genetics. Perspectives on Psychological Science, 1, 527. http://dx.doi.org/10.1111/j.1745-6916.2006.00002.x.
Mufano, M. R., Zammit, S., & Flint, J. (2014). Practitioner review: A critical perspective on gene-environment interaction modelswhat impact should they have on
clinical perceptions and practice? Journal of Child Psychology and Psychiatry, 55, 10921101. http://dx.doi.org/10.1111/jcpp.12261.
Murray, C., & Greenberg, M. T. (2000). Children's relationship with teachers and bonds with school: An investigation of patterns and correlates in middle childhood.
Journal of School Psychology, 38, 423445. http://dx.doi.org/10.1016/S0022-4405(00)00034-0.
Muthn, L. K., & Muthn, B. O. (1998-2010). Mplus user's guide. The brain: Understanding neurobiology through the study of addiction (6th ed.). Los Angeles, CA: Muthn
& Muthn. National Institutes of Health. (2000) (Retrieved from http://science.education.nih.gov/supplements/nih2/addiction/guide/pdfs/Entire.pd).
Nikolova, Y. S., Ferrell, R. E., Manuck, S. B., & Hariri, A. R. (2011). Multilocus genetic profile for dopamine signaling predicts ventral striatum reactivity.
Neuropsychopharmacology, 36, 19401947. http://dx.doi.org/10.1038/npp.2011.82.
O'Connor, E., Dearing, E., & Collins, B. (2011). Teacherstudent relationship trajectories: Predictors of behavior problem trajectories and mediators of child and family
factors. American Educational Research Journal, 48, 120162.
O'Farrell, S. L., Morrison, G. M., & Furlong, M. J. (2006). School engagement. In G. G. Bear, & K. M. Minke (Eds.), Children's needsIII: Development, prevention, and intervention (pp. 4558). Bethesda, MD: National Association of School Psychologists.
Pianta, R. C., & Hamre, B. (2001). Students, teachers, and relationship support (STARS). Lutz, FL: Psychological Assessment Resources.
Pluess, M., & Belsky, J. (2013). Vantage sensitivity: Individual differences in response to positive experiences. Psychological Bulletin, 139, 901916. http://dx.doi.org/10.
1037/a0030196.
Propper, C., Willoughby, M., Halpern, C. T., Cox, M., & Carbone, M. A. (2007). Parenting quality, DRD4, and the prediction of externalizing and internalizing behaviors in
early childhood. Developmental Psychobiology, 49, 619632. http://dx.doi.org/10.1002/dev.20249.

S. De Laet et al. / Journal of School Psychology 56 (2016) 1325

25

Rathouz, P. J., Van Hulle, C. A., Rodgers, J. L., Waldman, I., & Lahey, B. B. (2008). Specification, testing, and interpretation of gene-by-measured-environment interaction
models in the presence of geneenvironment correlation. Behavior Genetics, 38, 301315. http://dx.doi.org/10.1007/s10519-008-9193-4.
Reef, J., Diamantopoulou, S., van Meurs, I., Verhulst, F. C., & van der Ende, J. (2011). Developmental trajectories of child to adolescent externalizing behavior and adult
DSM-IV disorder: Results of a 24-year longitudinal study. Social Psychiatry and Psychiatric Epidemiology, 46, 12331241. http://dx.doi.org/10.1007/s00127-0100297-9.
Reitz, E., Dekovi, M., & Meijer, A. M. (2005). The structure and stability of externalizing and internalizing problem behavior during early adolescence. Journal of Youth
and Adolescence, 34, 577588. http://dx.doi.org/10.1007/s10964-005-8947-z.
Robinson, T. E., & Berridge, K. C. (2008). The incentive sensitization theory of addiction: Some current issues. Philosophical Transactions of the Royal Society, B: Biological
Sciences, 363, 31373146. http://dx.doi.org/10.1098/rstb.2008.0093.
Roeser, R. W., Eccles, J. S., & Sameroff, A. J. (2000). School as a context of early adolescents' academic and socialemotional development: A summary of research findings. The Elementary School Journal, 100, 443471. http://dx.doi.org/10.1086/499650.
Roisman, G. I., Newman, D. A., Fraley, R. C., Haltigan, J. D., Groh, A. M., & Haydon, K. C. (2012). Distinguishing differential susceptibility from diathesisstress: Recommendations for evaluating interaction effects. Development and Psychopathology, 24, 389409. http://dx.doi.org/10.1017/S0954579412000065.
Roland, E., & Galloway, D. (2002). Classroom influences on bullying. Educational Research, 44, 299312. http://dx.doi.org/10.1080/0013188022000031597.
Roorda, D. L., Koomen, H. M. Y., Spilt, J. L., & Oort, F. J. (2011). The influence of affective teacherstudent relationships on children's school engagement and achievement: A meta-analytic approach. Review of Educational Research, 81, 493529. http://dx.doi.org/10.3102/0034654311421793.
Sabol, T. J., & Pianta, R. C. (2012). Recent trends in research on teacherchild relationships. Attachment & Human Development, 14, 213231. http://dx.doi.org/10.1080/
14616734.2012.672262.
Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square test statistic for moment structure analysis. Psychometrika, 66, 507514. http://dx.doi.org/10.1007/
BF02296192.
Schaeffer, C. M., Petras, H., Ialongo, N., Poduska, J., & Kellam, S. (2003). Modeling growth in boys' aggressive behavior across elementary school: Links to later criminal
involvement, conduct disorder, and antisocial personality disorder. Developmental Psychology, 39, 10201035. http://dx.doi.org/10.1037/0012-1649.39.6.1020.
Schmidt, L. A., Fox, N. A., Rubin, K. H., Hu, S., & Hamer, D. H. (2002). Molecular genetics of shyness and aggression in preschoolers. Personality and Individual Differences,
33, 227238.
SoftGenetics (2010). GeneMarker (Version 1.91) [Computer software]. State College, PA:SoftGenetics.
Sonuga-Barke, E. J., Oades, R. D., Psychogiou, L., Chen, W., Franke, B., Buitelaar, J., ... Faraone, S. V. (2009). Dopamine and serotonin transporter genotypes moderate
sensitivity to maternal expressedemotion: The case of conduct and emotional problems in attention deficit/hyperactivity disorder. Journal of Child Psychology
and Psychiatry, 50, 10521063. http://dx.doi.org/10.1111/j.14697610.2009.02095.x.
Swanson, J., Flodman, P., Kennedy, J., Spence, M. A., Moyzis, R., Schuck, S., et al. (2000). Dopamine genes and ADHD. Neuroscience and Biobehavioral Reviews, 24, 2125.
http://dx.doi.org/10.1016/S0149-7634(99)00062-7.
Taylor, J., Roehrig, A. D., Hensler, B. S., Connor, C. M., & Schatschneider, C. (2010). Teacher quality moderates the genetic effects on early reading. Science, 328, 512514.
http://dx.doi.org/10.1126/science.1186149.
Vancraeyveldt, C., Verschueren, K., Wouters, S., Van Craeyevelt, S., Van den Noortgate, W., & Colpin, H. (2015). Improving teacherchild relationship quality and
teacher-rated behavioral adjustment amongst externalizing preschoolers: Effects of a two-component intervention. Journal of Abnormal Child Psychology, 43,
243257. http://dx.doi.org/10.1007/s10802-014-9892-7.
Williams, R. L. (2000). A note on robust variance estimation for cluster-correlated data. Biometrics, 56, 645646. http://dx.doi.org/10.1111/j.0006-341X.2000.00645.x.