inhibitory control and teacher–child conflict: reciprocal associations across the...

11
Inhibitory control and teacherchild conict: Reciprocal associations across the elementary-school years Daniel Berry Applied Psychology, Steinhardt School of Culture, Education, and Human Development, New York University 196 Mercer Street, 8th Floor, New York, NY 10012 abstract article info Article history: Received 29 December 2010 Received in revised form 19 October 2011 Accepted 23 October 2011 Available online 5 December 2011 Keywords: Self-regulation Inhibitory control Teacherchild conict Aggression problems Attention problems In the present study, longitudinal data from the NICHD Study of Early Child Care and Youth Development were used to test a theoretical model in which one aspect of children's self-regulation skills their inhibitory- control abilities were hypothesized to show reciprocal relations with their levels of teach of teacherchild across the elementary-school years. The ndings were largely consistent with the hypothesized model. Across multiple points in elementary school, lower levels of inhibitory control were associated with higher subsequent levels of teacherchild conict. In turn, higher levels of teacherchild conict were asso- ciated with lower subsequent levels of inhibitory control. Some evidence suggested that the magnitude of this latter relation was particularly strong for girls in the later elementary-school years. Direct relations be- tween inhibitory control and teacherchild conict were partially mediated by children's inattention and ag- gression problems. Potential implications for theory and practice are discussed. © 2011 Elsevier Inc. All rights reserved. Developing the ability to regulate one's thoughts, emotions, and behavior plays a critical role in children's successful transition into formal schooling (Ladd, Herald, & Kochel, 2006) and has been linked to more positive social and academic outcomes throughout the elementary-school years (Blair & Razza, 2007; Duncan et al., 2007; Ladd, Birch, & Buhs, 1999). Thus, illuminating the processes underly- ing the development of effective self-regulation skills is important to understanding the way children learn to succeed across multiple, developmental domains. An emergent literature suggests that interactions between chil- dren and meaningful adults may support children's development of self-regulation skills (Belsky, Fearon, & Bell, 2007; Kochanska, Murray, & Harlan, 2000). To date, this work has been restricted large- ly to parentchild interactions. However, theory and prior studies considering related outcomes suggest that relationships with other meaningful adults children's teachers may also play a role in the development of self-regulation skills. Like parentchild relationships, emotionally-close teacherchild relationships are theorized to serve as external, self-regulatory supports that help children to organize and manage their thoughts and feelings (Pianta, 1999). Inversely, re- lationships marked by emotionally-negative and conictual ex- changes between children and their teachers are proposed to serve as external stressors that may undermine children's self-regulation skills (Pianta, 1999). When considered developmentally, systems-based perspectives (Ford & Lerner, 1992; Gottlieb, 1991) posit that these processes are likely bidirectional and reciprocal over time (Ladd, Birch, & Buhs, 1999; Rimm-Kaufman & Pianta, 2000). Children bring developmental histories and behavioral patterns to the classroom that affect the quality of their experiences with their teachers. Those evincing self- regulatory struggles tend to have lower quality relationships with their teachers (Ladd & Burgess, 2001; Rudasill & Rimm-Kaufman, 2009). Thus, in a reciprocal and cascading manner, early self- regulation problems may set the stage for classroom experiences that fail to support or actively undermine children's abilities to devel- op more effective skills over time. The goal of the present study is to consider the degree to which one aspect of children's self-regulation skills their inhibitory- control abilities and their levels of teacher-reported teacherchild conict show reciprocal relations across the elementary-school years. A second aim is to test whether the direct associations between inhibitory control and children's subsequent teacherchild conict levels are mediated by their overt inattentive and/or aggressive behavior. Self-regulation and inhibitory control Inhibitory control refers to an individual's ability to withhold a prepotent response in order to enact a sub-dominant response (Miyake et al., 2000). Along with related constructs, such as working memory and attention-shifting, inhibitory control is considered one of a core set of higher order cognitive processes called executive Journal of Applied Developmental Psychology 33 (2012) 6676 Corresponding author. Tel.: + 1 301 437 5548. E-mail address: [email protected]. 0193-3973/$ see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.appdev.2011.10.002 Contents lists available at SciVerse ScienceDirect Journal of Applied Developmental Psychology

Upload: daniel-berry

Post on 04-Sep-2016

248 views

Category:

Documents


7 download

TRANSCRIPT

Page 1: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

Journal of Applied Developmental Psychology 33 (2012) 66–76

Contents lists available at SciVerse ScienceDirect

Journal of Applied Developmental Psychology

Inhibitory control and teacher–child conflict: Reciprocal associations across theelementary-school years

Daniel Berry ⁎Applied Psychology, Steinhardt School of Culture, Education, and Human Development, New York University 196 Mercer Street, 8th Floor, New York, NY 10012

⁎ Corresponding author. Tel.: +1 301 437 5548.E-mail address: [email protected].

0193-3973/$ – see front matter © 2011 Elsevier Inc. Alldoi:10.1016/j.appdev.2011.10.002

a b s t r a c t

a r t i c l e i n f o

Article history:Received 29 December 2010Received in revised form 19 October 2011Accepted 23 October 2011Available online 5 December 2011

Keywords:Self-regulationInhibitory controlTeacher–child conflictAggression problemsAttention problems

In the present study, longitudinal data from the NICHD Study of Early Child Care and Youth Development wereused to test a theoretical model in which one aspect of children's self-regulation skills – their inhibitory-control abilities – were hypothesized to show reciprocal relations with their levels of teach of teacher–child across the elementary-school years. The findings were largely consistent with the hypothesizedmodel. Across multiple points in elementary school, lower levels of inhibitory control were associated withhigher subsequent levels of teacher–child conflict. In turn, higher levels of teacher–child conflict were asso-ciated with lower subsequent levels of inhibitory control. Some evidence suggested that the magnitude ofthis latter relation was particularly strong for girls in the later elementary-school years. Direct relations be-tween inhibitory control and teacher–child conflict were partially mediated by children's inattention and ag-gression problems. Potential implications for theory and practice are discussed.

© 2011 Elsevier Inc. All rights reserved.

Developing the ability to regulate one's thoughts, emotions, andbehavior plays a critical role in children's successful transition intoformal schooling (Ladd, Herald, & Kochel, 2006) and has been linkedto more positive social and academic outcomes throughout theelementary-school years (Blair & Razza, 2007; Duncan et al., 2007;Ladd, Birch, & Buhs, 1999). Thus, illuminating the processes underly-ing the development of effective self-regulation skills is important tounderstanding the way children learn to succeed across multiple,developmental domains.

An emergent literature suggests that interactions between chil-dren and meaningful adults may support children's development ofself-regulation skills (Belsky, Fearon, & Bell, 2007; Kochanska,Murray, & Harlan, 2000). To date, this work has been restricted large-ly to parent–child interactions. However, theory and prior studiesconsidering related outcomes suggest that relationships with othermeaningful adults – children's teachers – may also play a role in thedevelopment of self-regulation skills. Like parent–child relationships,emotionally-close teacher–child relationships are theorized to serveas external, self-regulatory supports that help children to organizeand manage their thoughts and feelings (Pianta, 1999). Inversely, re-lationships marked by emotionally-negative and conflictual ex-changes between children and their teachers are proposed to serveas external stressors that may undermine children's self-regulationskills (Pianta, 1999).

rights reserved.

When considered developmentally, systems-based perspectives(Ford & Lerner, 1992; Gottlieb, 1991) posit that these processes arelikely bidirectional and reciprocal over time (Ladd, Birch, & Buhs,1999; Rimm-Kaufman & Pianta, 2000). Children bring developmentalhistories and behavioral patterns to the classroom that affect thequality of their experiences with their teachers. Those evincing self-regulatory struggles tend to have lower quality relationships withtheir teachers (Ladd & Burgess, 2001; Rudasill & Rimm-Kaufman,2009). Thus, in a reciprocal and cascading manner, early self-regulation problems may set the stage for classroom experiencesthat fail to support or actively undermine children's abilities to devel-op more effective skills over time.

The goal of the present study is to consider the degree to whichone aspect of children's self-regulation skills – their inhibitory-control abilities – and their levels of teacher-reported teacher–childconflict show reciprocal relations across the elementary-schoolyears. A second aim is to test whether the direct associations betweeninhibitory control and children's subsequent teacher–child conflictlevels are mediated by their overt inattentive and/or aggressivebehavior.

Self-regulation and inhibitory control

Inhibitory control refers to an individual's ability to withhold aprepotent response in order to enact a sub-dominant response(Miyake et al., 2000). Along with related constructs, such as workingmemory and attention-shifting, inhibitory control is considered oneof a core set of higher order cognitive processes called executive

Page 2: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

67D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

functions that support cognitive and behavioral self-regulation andfacilitate planning, problem solving, and the initiation and mainte-nance of goal-directed behavior (Miyake et al., 2000; Pennington &Ozonoff, 1996).

Effective inhibitory control plays a key role in children's success-ful transition to school, allowing them to override the “inertia”(Kirkham, Cruess, & Diamond, 2003) of their thoughts, emotions,and routines, in order to adopt adaptive classroom learning be-haviors. For example, one must down-regulate the prepotent urgeto shout out an answer or take a desired object from an unsuspectingpeer, in order to enact the learned responses of raising one's handand waiting one's turn. Indeed, empirical evidence suggests that in-hibitory control may play a notable role in children's classroomlearning behaviors. Brock, Rimm-Kaufman, Nathanson, and Grimm(2009) found that children with lower levels of inhibitory controlat the transition to kindergarten (measured using a battery of inhi-bition tasks) spent less time on-task and were less engaged in class-room activities at the end of the school year. Research with childrenin middle childhood has shown similar associations (Valiente,Lemery-Chalfant, Swanson, & Reiser, 2008). In turn, these processesmay lead to lower levels of social and academic success. Increasingevidence indicates that children's executive functioning plays acritical role in supporting the development of adaptive social andacademic skills (Blair & Razza, 2007; Brock et al., 2009; Murray &Kochanska, 2002).

Adult–child interactions and the development ofself-regulation skills

A growing empirical literature suggests that emotionally-close,sensitive, and responsive caregiving may support the developmentof inhibitory control. Across multiple samples and varying measuresof inhibitory control, children who experience higher quality par-ent–child interactions tend to regulate their thoughts and behaviorsmore effectively. These statistical associations have been illustratedacross early and middle childhood (Belsky et al., 2007), and in diversesamples of children (Li-Grining, 2007; NICHD Early Child CareResearch Network [ECCRN], 2003), and are robust in statisticalmodels that control for a wide-array of potential confounds and/orchildren's prior abilities (Belsky et al., 2007; Kochanska et al., 2000;Li-Grining, 2007; NICHD ECCRN, 2003).

Teacher–child relationships and self-regulation skills

Theory suggests that children's relationships with other importantadults, such as their teachers, may play similar roles in their self-regulation development. Like parent–child attachment relationships,children form relationships with their teachers (Howes & Hamilton,1992; Pianta, 1999). Warm, emotionally-close, and supportive ex-changes between children and their teachers are posited to provideself-regulatory supports – or affordance value – for children strugglingto adjust to the social and academic demands of the classroom. In-versely, histories of emotionally-negative and conflictual teacher–child interactions are theorized to serve as stressors that make itmore difficult for children to organize and manage their emotions,thoughts and behaviors (Pianta, 1999).

To date, little empirical work has directly considered relations be-tween teacher–child relationship quality and children's developmentof self-regulation skills. However, a nascent literature indicates thatteacher–child conflict, in particular, may serve as a physiologicalstressor known to have impacts on aspects of higher order cognitionlinked to self-regulatory processes. One recent study of preschool-aged children reported that children with more conflictual teacher–child relationships tended to show increases in their salivary cortisollevels – a biological indicator of Hypothalamic-Pituitary-Adrenal(HPA) Axis stress response – when they engaged in one-on-one

interactions with these teachers (Lisbonee, Mize, Payne, & Granger,2008). Similarly, a recent intervention with Head Start childrenfound that moving children from typical classrooms to smallplaygroups was associated with decreases in children's cortisol levelsover time. Yet, this stress-dampening effect extended only to childrenwith low conflict teacher–child relationships. Children with highlevels of conflict tended to maintain their high cortisol levels, evenafter being placed in small playgroups (Rappolt-Schlichtmann et al.,2009).

In turn, experimental findings from both the human and animalliteratures indicate that, in an inverted-U-shaped manner, eithervery high or very low levels of physiological stress activity can havedetrimental effects on executive cognitive processes (see Lupien,Maheu, Tu, Fiocco, & Schramek, 2007). This is particularly the casewhen learning tasks are novel, difficult, and/or demand cognitive re-sources frommultiple executive processes (Diamond, Campbell, Park,Halonen, & Zoladz, 2007; Schoofs, PreuB, & Wolf, 2008) — typical de-mands of children in the classroom. That is, highly conflictual teacher–child relationships may serve as stressors that – at a cognitive level –make effective self-regulation more challenging.

Indeed, higher levels of teacher–child conflict are predictive ofbroader problematic behavioral patterns thought to partially reflectexecutive functioning and self-regulation. Higher levels of teacher–child conflict in kindergarten are predictive of children's attention(Ladd & Burgess, 2001) and aggression problems (Ladd & Burgess,2001; Pianta, Steinberg, & Rollins, 1995) in the early elementary-school years, even after adjusting for their prior levels of problembehavior. Other work suggests more long-term relations. Moreconflictual teacher–child relationships in kindergarten have been linkedwith more positive attention-problem growth across the elementary-school years (Berry & Willett, 2009) and are predictive of disciplinaryproblems through the end of middle-school (Hamre & Pianta, 2001).Taken together, although the relation between teacher–child conflictand self-regulation development has not been considered directly (tothe author's knowledge), the evidence from related literatures suggeststhat conflictual teacher–child relationships may undermine children'sself-regulation skills.

Teacher–child relationships as dynamic processes: Closingthe loop

Influenced by developmental systems models (Ford & Lerner,1992; Gottlieb, 1991), theoretical work addressing the effects ofteacher–child relationship quality on children's development hasstressed the transactional nature of these relations (Ladd, Birch, &Buhs, 1997; Pianta, 1999; Rimm-Kaufman & Pianta, 2000). Teacher–child relationship quality may have adult-to-child effects on chil-dren's development. Yet, children also bring developmental historiesand behavioral patterns to the classroom. Child behaviors are alsotheorized to influence the quality of the relationships they formwith their teachers. In particular, maladaptive child behavior maylead to more conflictual exchanges between children and theirteachers. Several studies have indicated that children showing higherlevels of externalizing-type problems tend to have teacher–child rela-tionships marked by higher levels of conflict (Hamre, Pianta, Downer,& Mashburn, 2007; Ladd & Burgess, 2001; Murray & Murray, 2004),and this association has been illustrated across multiple points inthe elementary-school years (Stipek & Miles, 2008).

The available evidence suggests similar relations for children dis-playing attention problems. Teachers report that children with atten-tion problems are more stressful to teach than children withoutattention problems (Greene, Beszterczey, Katzenstein, Park, &Goring, 2002). Children with attention-related behavioral strugglestend to receive more criticism (Martin, 1989) and more punitiveand coercive punishment from their teachers (Keogh, 2003; Pullis,1985). Controlling for children's early levels of aggression problems,

Page 3: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

68 D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

higher levels of attention problems in pre-kindergarten are associatedwith more conflictual teacher–child relationships in kindergarten(Berry & Willett, 2009). Children with less effective effortful control –a related construct thought to reflect low inhibitory control – aspre-kindergarteners have been shown to have higher levels of teach-er–child conflict in kindergarten (Rudasill & Rimm-Kaufman, 2009).

These findings regarding broad behavioral patterns cannot be gen-eralized directly to the cognitive processes underlying inhibitory con-trol. However, inhibitory control is theorized to be a key cognitivemechanism underlying individual differences in more overt patternsof child behavior, such as attention and/or aggression problems. Asubstantial literature indicates that executive functioning – in partic-ular, inhibitory-control weaknesses –may be an important neuropsy-chological component underlying broad behavioral measures ofchildren's attention problems (i.e., inattentive and/or impulsive be-havior; Barkley, 1997; Friedman et al., 2007; Willcutt, Doyle, Nigg,Faraone, & Pennington, 2005). Similarly, weaknesses in inhibitorycontrol and related constructs, such as effortful control, have beentheorized to partially explain heightened levels of aggression(Eisenberg, 2000; Séguin & Zelazo, 2005). These findings suggestthat low levels of inhibitory control may be associated with higherlevels of teacher–child conflict. Yet, given the comparatively morecognitive nature of the construct, such relations are likely mediatedbymore overt child behavior. That is, the link between inhibitory con-trol and the quality children's teacher–child relationships may bemediated by more overt behavioral patterns, such as attention and/oraggression problems, that partially reflect lower levels of inhibitorycontrol.

Inhibitory control and teacher–child relationship qualityover time

Taken together, theory and prior empirical work suggest thatlower quality teacher–child relationships – perhaps, particularly,higher levels of conflict – may fail to support or actively underminechildren's development of self-regulation skills. Yet, this work alsosuggests that these processes may be bidirectional, such that chil-dren's self-regulatory skills may also influence the quality of the rela-tionships that they form with their teachers. Because cognitive self-regulatory processes, such as inhibitory control, are thought to reflectmore general cognitive processes, this association is likely exerted bymore overt child behavior.

Further, when considered developmentally, the theorized bidirec-tional effects suggest reciprocal processes between children's self-regulation skills and the quality of their relationships with theirteachers over time. The self-regulation skills that children bring tothe classroom and the patterns of overt behavior that these skills sup-port are theorized to influence the quality of their teacher–child rela-tionships. In turn, these relational experiences are posited to influencechildren's subsequent self-regulation development — each compo-nent of the teacher–child system perpetuates or exacerbates theother over time (Rimm-Kaufman & Pianta, 2000).

Potential moderating roles of sex and ethnicity

Beyond this more general framework, there is some indicationthat these processes may differ systematically for girls and boys.Girls tend to show more effective self-regulation skills (Kochanskaet al., 2000; Li-Grining, 2007) and higher quality teacher–child rela-tionships (Birch & Ladd, 1997; Hamre & Pianta, 2001; Hughes,Cavell, & Wilson, 2001). Some evidence also suggests that relationsbetween gender, teacher–child relationship quality, and children'sdevelopment may be complex. For instance, although mixed, somework suggests that the magnitudes of the respective associations be-tween negatively-valenced aspects of teacher–child relationships(e.g., conflict) and children's academic and behavior problems may

be stronger for boys than girls (Ewing & Taylor, 2009; Hamre &Pianta, 2001). Notably, there is also some indication that positively-valenced aspects, such as teacher–child closeness, may be morestrongly related to these outcomes for girls (Ewing & Taylor, 2009).Thus, although the directions of the potential interaction effects aresomewhat unclear in the literature, some work suggests that the pro-posed reciprocal processes may function differently for girls and boys.

Similarly, the proposed processes may vary across ethnicity.Racial-minority children are more likely to experience risk factors(e.g., poverty) that are predictive of less effective self-regulation skills(Raver, 2004). Empirical work has also shown that teachers tend toperceive their relationships with African American children as beinglower quality (Hughes, Gleason, & Zhang, 2005; Murray & Murray,2004). Like gender, it is possible that links between inhibitory controland teacher–child relationship quality may reflect effects due to eth-nicity that are unassociated with the key question variables (i.e., eth-nicity as a confound). Ethnicity may also moderate the structuralrelations of interest. Although no prior work has considered such in-teractive effects in the context of associations between teacher–childrelationship quality and children's self-regulation development, priorfindings suggest that the relations between teacher–child relation-ship quality and children's future academic achievement may bestronger for African American children compared to non-AfricanAmerican children (Burchinal, Peisner-Feinberg, Pianta, & Howes,2002). Together, sex and ethnicity serve as potentially importantconfounds, as well as potential moderators of the relations of interest.

The present study

The goal of this study was to test a model that is consistent withthese theorized reciprocal processes. Specifically, it was hypothesizedthat lower levels of inhibitory control prior to school-entry wouldbe predictive of more conflictual teacher–child relationships in kin-dergarten, and in turn cascading, bidirectional relations betweenteacher–child conflict and children's inhibitory-control abilities overtime. Secondly, it was hypothesized that the direct associationsbetween children's inhibitory-control levels and the quality of theirsubsequent teacher–child relationships would be mediated by theirovert inattentive and aggressive behavior. A related aim was to con-sider the extent to which these theorized relations differed acrosssex and ethnicity. Notably, given the dearth of prior work consideringsuch interactions in the context of children's self-regulation develop-ment, these moderated relations were considered exploratory.

Control covariates

In addition to sex and ethnicity, several other control covariateswere added to the key statistical models. Family income and maternaleducation were controlled, as they have been linked with both chil-dren's self-regulation skills (Raver, 2004) and the quality of theirteacher–child relationships (Ladd et al., 1999). Early maternal sensi-tivity was also controlled, given that prior work has suggested thathigher quality parent–child interactions are also associated with bet-ter inhibitory control (see above) as well as higher quality-teacher–child relationships (O'Connor & McCartney, 2007). Finally, ninedummy variables representing research site (see below) were includ-ed to adjust for site differences and potentially correlated residualswithin-site.

Research design

Sample

The present data were obtained from a sample of children,mothers, and teachers, participating in Phases I, II and III of theNICHD Study of Early Child Care and Youth Development (SECCYD), a

Page 4: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

69D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

longitudinal study of 1364 children (50.1% male) and their families in10 sites across the U.S. Potential participants were originally recruitedfrom among 8986 mothers who gave birth within a 24-hour samplingperiod in 31 hospitals across the sites (see NICHD ECCRN, 1997 for re-view). Although the sample is not nationally representative, it doesshow some socio-demographic diversity. Based on the 1153 fami-lies/teachers who provided some inhibitory control, teacher–child re-lationship, or behavior problem data, 81% of children were Caucasian,12% were African American, and 7% were of other ethnicities. Motherswere fairly well-educated. Approximately 56% had some college or afour-year degree and 16% held advanced degrees. Twenty-percent ofthe mothers reported high school as their highest level of educationand 9% did not have a high school diploma. However, there was stilla notable representation of families below the federal poverty thresh-old, when the child was 4.5-years old, approximately 11%.

As is common in longitudinal studies, there were some missingdata. There were 211 cases (16% of original sample) in which therewere no data available for any of the key question variables. In an ef-fort to reduce potential biases introduced by missing data, all modelswere fitted using Full Information Maximum Likelihood (FIML) esti-mation. FIML estimation helps to adjust for biases due to missingdata, under the assumption that missing data is “missing at random” —

that is, missingness is conditional on observed variables that are includ-ed in themodel and, after adjusting for missingness, are not conditionalon unobserved values of the variables with missing data (Allison, 2003;Collins, Schafer, & Kam, 2001).

Procedure

Trained research assistants measured children's inhibitory-controlabilities using a standardized laboratory procedure, when the chil-dren were pre-kindergarten-aged (approximately, 4.5 years old)and then again in first and fourth grade. Mothers provided demo-graphic and questionnaire data at multiple points between birthand pre-kindergarten. Observations of maternal sensitivity duringsemi-structured play tasks were made when children were 6, 15,24, and 36 months old. Teachers rated children's inattention and ag-gression problems when the children were in pre-kindergarten andagain in first grade. In kindergarten and second grade, teachersrated the quality of their relationship with the child, using a commonquestionnaire. The teacher–child relationship assessment periodswere selected to occur temporally between the three inhibitory-control assessments.

Measures

Inhibitory controlTrained research assistants measured children's inhibitory-control

abilities by assessing their performance on the Continuous Perfor-mance Test (CPT; Mirsky, Anthony, Duncan, Ahearn, & Kellam,1991). The CPT is a computer-delivered task in which children viewtarget and non-target pictorial stimuli on a computer screen and arerequired to follow two critical rules: (1) the child is to press a re-sponse key when the target stimulus is presented on the screen,and (2) the child is to withhold from pressing the response keywhen non-target stimuli are presented on the screen. Although theoverall task was quite similar at each assessment point, the exact na-ture of the target and non-target stimuli changed slightly over time.At the pre-kindergarten assessment, children viewed dot-matrixillustrations of non-target familiar objects (e.g., butterfly, fish, flower)and were asked to respond only when the target item (a chair) waspresented. At the first-grade assessment, children viewed non-target items that were represented by objects similar to those at thepre-kindergarten assessment, but the target stimulus was repre-sented by an ‘X’. At the fourth-grade assessment, children viewed se-quences of letters (X, A, B, C, E, F, L, M, N, P) shown one-by-one on a

computer screen. The target stimulus was when the letter ‘A’ was fol-lowed by the letter ‘X.’ All other cases represented non-target stimuli.At all assessments, each stimulus appeared on the screen for 500 ms,with a 1500 ms-span between stimuli. At the pre-kindergarten as-sessment, children were presented ten stimuli per block for 22 blocks.In first grade, they were presented with ten stimuli per block for 30blocks. In fourth grade, children were presented 12 stimuli perblock for 45 blocks. For all assessments, the target stimulus was pre-sented randomly, twice per block. Inhibitory control was operationa-lized as the number of commission errors (i.e. response to non-targetitems) made during the entirety of the CPT, scaled on the proportionof delivered trials. Natural-logarithm transformations were adoptedat each assessment period to adjust for skewness. The CPT shows rea-sonable test-retest reliability in non-clinical samples (rs = .65 to .74;Halperin, Sharma, Greenblatt, & Schwartz, 1991).

Teacher–child conflictTeacher–child conflict was measured using items from Student

Teacher Relationship Scale (STRS; Pianta, 1992). Specifically, at eachtime point, teacher–child conflict was modeled as a unitary latent fac-tor theorized to influence children's ratings on seven observed indica-tor items tapping emotionally-negative and conflictual relationalaspects between the study child and his/her teacher. The items tap as-pects of the teacher–child dyad, child behaviors thought to be reflec-tive of the conflictual relationships, as well as teacher self-reflectionsindicative of conflictual relationships. Teachers rate the degree towhich each item is representative of his/her relationship with thestudy child (1 = Definitely does not apply to 5 = Definitely applies),with higher scores reflective of more conflictual relationships.Natural-logarithm transformations were used for each indicator vari-able to reduce skewness. Although prior work suggests that the STRSconflict scale shows reasonable internal consistency and test-retestreliability (Pianta, 1992), as well as concurrent and predictive validity(Hamre & Pianta, 2001), a latent variable approach affords at leasttwo advantages. First, latent representations improve precision byallowing one to model measurement error (Brown, 2006). Second,they allow one to test and adjust for measurement invariance (seebelow). A latent representation was, thus, adopted to capitalize onthese advantages.

Aggression problemsChildren's aggression problems were measured using raw scores

from the Aggression Problems scale of the Teacher Report Form(TRF; Achenbach, 1991). Non-parental caregivers (pre-kindergarten)and teachers (first grade) rated the degree to which a series ofaggressive behaviors (e.g., cruel, mean, or bullies others; destroysproperty belonging to others) were reflective of the child's behavior,on a three-point Likert-type scale (0 = not true, 1 = somewhat orsometimes true, 2 = very true or often true). Aggression raw scoreswere calculated as the sum across items; higher scores indicatemore aggressive behavior. The items administered at the pre-kindergarten assessment differed slightly from those administeredin first grade. Only the 18 common items were used presently.Natural-logarithm transformations were used to reduce skewness.Internal-consistency reliability was acceptable at each assessmentperiod (α = .93 and .91, respectively).

Inattention problemsChildren's inattention problems were measured using items from

the Attention Problems scale of the TRF (Achenbach, 1991). Basedon face-validity, three items tapping inattention were used (fails tofinish things he/she starts; can't concentrate/pay attention for long;inattentive, easily distracted). Teachers rated the degree to which in-attentive behaviors were representative of the child's behavior, on anidentical three-point Likert-type scale as that for Aggression. Theitems were summed, with higher scores indicating higher levels of

Page 5: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

70 D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

inattention. Natural-logarithm transformations were used to reduceskewness. Internal-consistency reliability was acceptable (α = .83and .86, respectively).

Maternal sensitivityObservational measures of maternal sensitivity were conducted

when children were approximately 6-, 15-, 24- and 36-months old(see NICHD ECCRN, 1999). At all time-points mother–child interac-tions were videotaped during a 15-minute semi-structured play pro-cedure. Blind raters coded the videotaped interactions for maternalsensitivity. At the 6-, 15-, and 24-month assessments, sensitivitywas measured as a composite across maternal sensitivity to non-distress, positive regard and intrusiveness (reversed). Each wasrated on a 1- to 4-point scale and subsequently summed in the com-posite, with higher ratings indicating higher levels of maternal sensi-tivity. The 36-month composite was adjusted for developmentalappropriateness and included ratings of maternal supportive pres-ence, respect for autonomy, and hostility (reversed). Each was ratedon a scale from 1 to 7 and summed to create the composite. Approx-imately, 20% of the videotapes were coded by two raters, at eachassessment period. Inter-rater reliability estimates based uponrepeated-measures ANOVA (Winer, 1971) ranged from .83 to .87.Internal-consistency reliability was acceptable at each time point(α = .70–.81). A composite representing average maternal sensi-tivity across infancy and early childhood was created by standard-izing each scale and averaging the scores across the longitudinalassessments.

Income-to-needs ratioA continuous income-to-needs ratio variable was computed by di-

viding reported family income in pre-kindergarten by the federalpoverty threshold for family size. A natural-logarithm transformationwas adopted to reduce skewness.

Maternal educationMothers reported the number of years that they attended school

when the child was one-month old.

EthnicityA dichotomous representation of ethnicity was adopted, such that

African American children were coded as one and all other ethnicitieswere coded as zero.

Sex of childSex was dummy-coded, such that males were coded as one.

SiteNine dummy-variables were included to adjust for research-site

differences.

Data-analytic strategy

Using the robust maximum likelihood estimator available1 in theMplus 5.21 statistical platform (Muthén & Muthén, 2007), Confirma-tory Factor Analyses (CFAs) were first conducted to test the validity ofthe teacher–child conflict construct in kindergarten and secondgrade, respectively. The within-time CFAs were then integrated intosingle longitudinal CFA. Residual covariances between identical

1 Satorra-Bentler (S-B; Satorra & Bentler, 2001) adjusted likelihood ratio tests wereused in all but one set of comparisons of nested models. Unadjusted log-likelihood es-timates were used in the multi-group analyses considering whether the four cross-lagged relations varied across ethnicity, as robust estimation led to convergence prob-lems. Tests of latent and observed variable interactions (Klein & Moosbrugger, 2000)using robust χ2 likelihood ratio tests confirmed the same (null) interactions as thosefound using the multi-group analyses fitted with unadjusted maximum likelihoodestimation.

items over time were freely estimated, as were the residual covari-ances between items sharing similar phrasing (e.g., items using theword “anger”). The longitudinal CFAs were then tested for measure-ment invariance across time, ethnicity, and sex (see Vandenberg &Lance, 2000).

A taxonomy of structural equation models were then fitted, toaddress the questions regarding reciprocal relations between chil-dren's inhibitory-control and teacher–child conflict levels over time.These first models are represented by a simplified version of the com-paratively more complex path diagram illustrated in Fig. 1. Unlike thediagram displayed in Fig. 1, however, this first set of models testedonly the auto-regressive and the cross-lagged relations representedby the solid pathways in Fig. 1. No mediating pathways (i.e., dashedpaths) were included in the model, nor were the proposed mediatingvariables (i.e., inattention, aggression) included in the observed (co)variance matrix to which the model was fitted. Multi-group analyseswere used to test whether the four cross-lagged regression parame-ters of substantive interest varied across sex and ethnicity.Bonferonni corrections were made for families of related tests of in-variance. For example, an alpha level of .006 (.05/8) was adopted inthe interaction models, in order to account for the eight interactionsof substantive interest.

To test whether the respective direct associations between inhib-itory control and children's subsequent teacher–child conflict levelswere mediated by their overt aggressive and inattentive behavior, in-direct pathways were added to the model to test these mediated ef-fects (i.e., dashed paths; Fig. 1). Confidence intervals (95%) for themediated relations (i.e., product of the regression parameters com-prising the mediational path) were estimated using a 3000-iterationbias-corrected bootstrap (see MacKinnon, Fairchild, & Fritz, 2007;Preacher & Hayes, 2008).

Results

Measurement models

Tests of longitudinal invariance indicated that the latent teacher–child conflict factors showed metric and scalar invariance over time;the longitudinally-invariant CFA model fit the data reasonably well(χ2 = 148.25, df = 76, p b .001, CFI = .98, RMSEA = .03; Hu &Bentler, 1999). Multi-group analyses also revealed metric invarianceacross sex and ethnicity; however, there was some evidence of partialscalar variability. Teachers rated boys as being more of a struggle(STRG), as well as more of a drain on his/her energy (ENER), on aver-age. On average, teachers also rated African American children higheron an item tapping the degree to which the child “stays angry” withhim/her (STYAN). Partial measurement invariance was adopted inthe substantive analyses (Byrne, Shavelson, & Muthén, 1989) byregressing the observed STYAN indicator on the ethnicity dummyvariable and by allowing the STRG and ENER indicator intercepts tovary across sex in the final multi-group model (Brown, 2006).

Preliminary analyses

As shown in Table 1, the (maximum-likelihood) zero-order corre-lations indicated that there was moderate rank-order stability in chil-dren's latent teacher–child conflict levels between kindergarten andsecond grade. There was also moderate rank-order stability in chil-dren's levels of inhibitory control, inattention, and aggression levelsover time. On average, children with lower levels of inhibitory control(i.e., more commission errors) tended to show higher levels of teach-er–child conflict. Higher levels of inattention and aggression wereeach associated with lower levels of inhibitory control and higherlevels of teacher–child conflict.

Page 6: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

Pre-KInhibitory Control

1st GradeInhibitory Control

4th GradeInhibitory Control

Kindergarten Teacher-Child

Conflict

Pre-KInhibitory Control

Pre-KInattention

Pre-KAggression

First GradeAggression

2nd GradeTeacher-Child

Conflict

First GradeInattention

Pre-KInhibitory Control

Fig. 1. Path diagram representing cross-lagged relations between inhibitory control and teacher–child conflict across multiple points in middle-childhood. Direct paths linking chil-dren's inhibitory control abilities to their subsequent levels of teacher–child conflict are represented by bold, non-dashed paths. Mediated paths are represented by dashed paths.Non-bolded, non-dashed paths represent select controls within the model. Pathways for the remaining control covariates were are not included in the diagram for visual clarity.

71D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

Reciprocal associations between inhibitory control and teacher–childconflict over time

Across different specifications, the theorized models fit the datareasonably well, and the fitted structural parameters were consistentwith the underlying hypotheses that children's inhibitory-control andteacher–child conflict levels would show cross-lagged reciprocal as-sociations across multiple points in elementary school. Notably, themulti-group analyses indicated that the relation between second-grade teacher–child conflict and fourth-grade inhibitory control wasmoderated by sex (ΔS-Bχ2= 9.35, df = 1, p = .002). No other statis-tically significant interactions emerged. Therefore, the estimatedstructural parameters and fit statistics described below are basedon a multi-group model in which this regression parameter, theintermediate/distal outcome intercepts, and residual (co)variances(measurement and disturbances) were estimated freely across girls

Table 1Zero-order correlations between children's inhibitory-control abilities, inattention andaggression problems, and teacher–child conflict levels across multiple points in (or justprior to) the elementary-school years (n = 1,153).

1 2 3 4 5 6 7 8 Mean(SD)

1 KindergartenT-C conflict

1.00 0.00(0.19)

2 Grade 2T-C conflict

.47 1.00 0.02(0.19)

3 Pre-Kinhibitory control

.26 .29 1.00 1.02(0.45)

4 Grade 1inhibitory control

.17 .24 .36 1.00 0.77(0.45)

5 Grade 4inhibitory control

.18 .26 .30 .35 1.00 0.94(0.41)

6 Pre-Kinattention

.32 .25 .29 .17 .23 1.00 0.53(0.63)

7 Grade 1inattention

.23 .39 .31 .24 .25 .38 1.00 0.48(0.70)

8 Pre-K aggression .39 .42 .25 .18 .2 .53 .27 1.00 0.99(1.00)

9 Grade 1aggression

.48 .55 .28 .23 .21 .33 .52 .44 0.64(0.91)

Note. All correlations are statistically significant at the p b .001 level.

and boys (Table 2). All other estimates were constrained to be equalacross sex.

The multi-group model fit the data reasonably well (χ2 = 787.77,df = 609, p b .001, CFI = .97, RMSEA = .02). Controlling for themother, child, and family covariates, children exhibiting lower levelsof inhibitory control (i.e., more commission errors) prior to school-entry tended to show higher levels of teacher–child conflict, on aver-age. This association corresponded to a standardized effect of .24.

In turn, children's teacher–child conflict levels in kindergartenwere positively associated with their subsequent inhibitory-controlabilities in first grade, adjusting for their prior inhibitory-control abil-ities and the control covariates. On average, those with more conflic-tual teacher–child relationships tended to show lower levels ofinhibitory control in first grade. This corresponded to a somewhatmodest standardized effect of approximately .07.

Table 2Fitted unstandardized regression coefficients from a multi-group structural equationmodel (SEM) testing reciprocal relations between inhibitory control and teacher–child conflict across the elementary-school years.

GK T-Cconflict

G1 inhibitorycontrol

G2T-Cconflict

G4 inhibitorycontrol

B B B B

Intercept (α) 0.00/0.05⁎⁎⁎ 0.66⁎⁎⁎/0.87⁎⁎⁎ 0.00/0.03⁎ 0.69/0.89⁎⁎⁎

Pre-K inhibitorycontrol

0.10⁎⁎⁎ 0.25⁎⁎⁎ – 0.10⁎⁎

GK T-C conflict – 0.15⁎ 0.41⁎⁎⁎ –

G1 inhibitorycontrol

– – 0.03⁎ 0.19⁎⁎⁎

G2T-C conflict – – – 0.38⁎⁎⁎/ 0.10Controlcovariates

Family income −0.01 −0.02 −0.03⁎ −0.03†

Maternaleducation

0.00 −0.01 0.00 −0.01

Maternalsensitivity

−0.01 −0.01 −0.04⁎⁎ 0.01

AfricanAmerican 0.07⁎⁎⁎ 0.07 0.06† 0.00Goodness of fit: χ2 = 787.44, df = 609, CFI = .97, RMSEA = .02

Note. Parameters that vary across gender in the multi-group analyses are presented asgirl/boy. The fitted parameters for the site dummies are not included, for visual clarity.They are available from the author, upon request.† pb .10. ⁎ pb .05. ⁎⁎ pb .01. ⁎⁎⁎ pb .001.

Page 7: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

72 D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

Similar to the effect noted for children's pre-kindergarteninhibitory-control abilities, first-grade inhibitory control was posi-tively associated with their subsequent levels of teacher–child con-flict in second grade. Holding prior conflict level and the controlsconstant, children with less effective inhibitory control in first gradetended to show higher levels of teacher–child conflict in secondgrade. This corresponded to a modest standardized effect of approxi-mately .07.

Finally, higher levels of second-grade teacher–child conflict were,in turn, predictive of lower levels of inhibitory control in fourth grade,adjusting for children's first-grade and pre-kindergarten inhibitory-control levels and the control covariates. However, as noted above,multi-group analyses indicated that this relation differed betweengirls and boys. As displayed in Fig. 2, on average, girls tended toshow higher levels of inhibitory control (i.e., fewer commission er-rors) than did boys, irrespective of level of teacher–child conflict.Yet, the conditional relation between teacher–child conflict and in-hibitory control was comparatively stronger for girls than for boys.The simple slope for girls was statistically significant (BGirls = .38,p b .001) and corresponded to a standardized regression coefficientof .18. In contrast, the simple slope did not reach statistical signifi-cance for boys (BBoys = .10, p = .16).

Inhibitory control→ behavior problems→ teacher–child conflict

As indicated in Table 3, the meditational model fit the data reason-ably well (χ2 = 1020.41, df = 799, CFI = .97, RMSEA = .02), andthe inclusion of the indirect pathways contributed significantly tomodel fit (ΔS-Bχ2 = 225.57, df = 8, p b .001). Specifically, the re-sults indicated that the association between pre-kindergarten inhibi-tory control and kindergarten teacher–child conflict was partiallymediated by children's overt inattention problems as pre-kindergarteners (product of the parameter estimates = .01 [.002,.02]). Children with lower levels of inhibitory control as pre-kindergartens (i.e., more commission errors) tended to have higherlevels of teacher-rated inattention problems as pre-kindergarteners.In turn, children with higher levels of inattention as pre-kindergarteners tended to have more conflictual teacher–child rela-tionships in kindergarten. Each of these effects was somewhat mod-est in size, corresponding to standardized regression coefficients of.22 and .11, respectively.

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

-3 -2 -1 0 1 2 3

4th

Gra

de L

og C

PT

Com

mis

ion

Err

ors

2nd Grade Teacher-Child Conflict

Boy

Girl

Fig. 2. Fitted relations between children's levels of second-grade teacher–child conflictand their subsequent levels of fourth-grade inhibitory control, conditional on gender.The teacher–child conflict axis is mean-centered and presented in SD-units. The scaleof the log-commission-errors axis represents a range from the mean – 1SD to themean +1SD.

Children's pre-kindergarten aggression levels also partially medi-ated the association between pre-kindergarten inhibitory controland teacher–child conflict in kindergarten (product of the parameterestimates = .03 [.02, .05]). Less effective pre-kindergarten inhibitorycontrol was associated with higher concurrent levels of aggression. Inturn, higher levels of aggression were predictive of more conflictualteacher–child relationships in kindergarten. These relations werealso somewhat modest to moderate in size, corresponding to stan-dardized regression coefficients of .23 and .31, respectively. The re-spective associations between teacher–child conflict and pre-kindergarten inattention and aggression appeared to be additive;the association between inattention and teacher–child conflict wasnot moderated by aggression.

Similar mediated relations were evident later in the elementary-school years. Accounting for the control covariates, as well as chil-dren's prior levels of inattention and teacher–child conflict, lowerlevels of first-grade inhibitory control were associated with higherlevels of first-grade inattention. In turn, higher levels of first-grade in-attention were predictive of more conflictual teacher–child relation-ships in second grade (product of the parameters = .01 [.002, .01]).Notably, although higher levels of aggression in first-grade were pre-dictive of more conflictual teacher–child relationships in second-grade, the relation between children's residualized first-gradeinhibitory-control levels and first-grade aggression only reachedmar-ginal levels of statistical significance in the full control covariatemodel. This was explained, in part, by the conservative nature of theestimate — it represents the association between inhibitory controlthat is unique to first grade (i.e., residualized on prior inhibitory con-trol) and children's aggression levels that are unique to first grade(i.e., residualized on prior aggression). Post-hoc probes indicatedthat this statistically marginal association was largely accounted forby the rank-order stability of aggression over time; the relation be-tween inhibitory control and aggression was stronger and statisticallysignificant, upon constraining the auto-regressive parameter for ag-gression to zero. As somewhat expected, the indirect relation be-tween first-grade inhibitory control and children's second-gradeteacher–child conflict levels via their first-grade aggression problemsalso failed to reach traditional levels of statistic significance in the fullmodel (product of the parameters = .01 [0.00 02]).

Discussion

Growing theoretical and empirical literatures suggest that the de-velopment of self-regulation in childhood may be shaped by recipro-cal processes between children and their interpersonal experienceswith meaningful adults — including their teachers. The first goal ofthe present study was to consider whether there was evidence con-sistent with these theorized reciprocal processes, when children'sself-regulation skills were measured using a rather objective measureof inhibitory control. The second goal was to consider whether the re-lations between inhibitory control and children's subsequent levels ofteacher–child conflict were mediated by their overt inattentive andaggressive behaviors.

Teacher–child conflict → inhibitory control

Across multiple points in the elementary-school years, teacher–child relationships marked by higher levels of conflict were associat-ed with lower subsequent levels of inhibitory control. Specifically, rel-ative to children with lower conflict ratings, those with teachers whorated their relationships with the child as involving higher levels ofconflict tended to make more commission errors (i.e., have lowerlevels of inhibitory control) on an objective neuropsychologicalassessment of inhibitory control. The relations were robust, control-ling for children's prior inhibitory-control abilities and extend agrowing parenting literature indicating that lower quality adult–

Page 8: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

Table 3Fitted unstandardized regression coefficients from amulti-group SEM testing the role of children's inattention and aggression problems as mediators linking their inhibitory-controlabilities with their subsequent levels of teacher–child conflict.

Pre-Kinattention

Pre-Kaggression

GK T-Cconflict

G1 inhibitory control(IC)

G1inattention

G1aggression

G2 T-Cconflict

G4 inhibitory control(IC)

B B B B B B B B

Intercept (α) 0.44⁎⁎⁎/0.64⁎⁎⁎ 0.91⁎⁎⁎/1.10⁎⁎⁎ 0.00/0.05⁎⁎⁎ 0.64⁎⁎⁎/0.86⁎⁎⁎ 0.25⁎⁎⁎/0.42⁎⁎ 0.14⁎⁎/0.35⁎ 0.00/0.05⁎⁎⁎ 0.66⁎⁎⁎/0.89⁎⁎⁎

Pre-K IC 0.29⁎⁎⁎ 0.48⁎⁎⁎ 0.06⁎ 0.25⁎⁎⁎ – – – 0.09⁎⁎

Pre-K Inatt – – 0.04⁎ – 0.28⁎⁎⁎ – – –

Pre-Agg. – – 0.06⁎⁎⁎ – – 0.21⁎⁎⁎ – –

Kindergarten T-C conflict – – – 0.15⁎ 0.29⁎ 1.40⁎⁎⁎ 0.25⁎ –

Grade 1 IC – – – – 0.19⁎⁎⁎ 0.10~ 0.02† 0.19⁎⁎⁎

Grade 2T-C conflict – – – – – – – 0.38⁎⁎⁎/0.11Grade1 inattention – – – – – – 0.03⁎⁎ –

Grade 1 aggression. – – – – – – 0.07⁎⁎⁎ –

Control covariatesIncome −0.02 0.02 −0.01 −0.02 −0.09⁎⁎ −0.01 −0.02† −0.03Maternal education −0.02 −0.02 0.00 −0.01 −0.02⁎ −0.01 0.00 −0.01Maternal sensitivity −0.12⁎⁎ −0.10 −0.01 −0.01 −0.07⁎ −0.08† −0.03⁎ 0.01African American −0.09 0.12 0.06⁎ −0.07 0.08 0.33⁎⁎ 0.03 0.00Goodness of fit: χ2 = 1020.41, df = 799, CFI = .97, RMSEA = .02.

Note. Parameters that vary across gender in the multi-group analyses are presented as girl/boy. The fitted parameters for the site dummies are not included, for visual clarity. Theyare available from the author, upon request.† pb.10. ⁎ pb .05. ⁎⁎ pb.01. ⁎⁎⁎ pb.001.

73D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

child interactions may undermine the development of effective self-regulation skills (Belsky et al., 2007; Kochanska, et al., 2000; NICHDECCRN, 2003) . Although the mechanisms underlying these associa-tions remain largely unclear, recent studies of children's stress phys-iology (Lisbonee et al., 2008; Rappolt-Schlichtmann et al., 2009)suggest the possibility that conflictual teacher–child exchanges mayserve as physiological stressors known to undermine higher ordercognitive processes (Lupien et al., 2007).

Interestingly, in the later elementary-school years, the relation be-tween teacher–child conflict and children's subsequent levels of in-hibitory control extended only to girls. In some ways, this finding isconsistent with prior work. For instance, some have reported thatthe positive associations between emotionally-close teacher–child re-lationships – measured as global quality (i.e., high closeness and lowconflict) or teacher–child closeness – and more effective academicand social outcomes are particularly strong for girls (Baker, 2006;Ewing & Taylor, 2009). However, other studies consideringnegatively-valenced aspects of children's teacher–child relationships(e.g., conflict) in the context of their development of behavior prob-lems (e.g., aggression, maladaptive classroom behavior) have shownmore pronounced associations for boys than for girls (Ewing &Taylor, 2009; Hamre & Pianta, 2001). Given that behavior problemsare thought to partially reflect less effective inhibitory control, theseprior findings may contradict the interaction with sex reportedpresently.

Due to both the inconsistent nature of the findings across studies,as well as the fact that the present study is the first to consider suchinteractive effects in the development of inhibitory control, the inter-pretation of the sex difference is speculative. As some have proposed,it may be the case that girls are somewhat more attuned to social re-lationships – and, in turn, influenced by these relationships – than areboys (Baker, 2006; Ewing & Taylor, 2009; Maccoby, 2000). In turn,due to these differing relational approaches, the effects of social rela-tionships on children's self-regulatory processes may be more pro-nounced for girls. This may be particularly the case toward the endof middle-childhood and early adolescence. For example, some evi-dence raises the possibility that associations between social-relationship quality and children's physiological stress-responsesmay be stronger for girls. In girls – but not boys – lower self-reported feelings of parental acceptance have been linked with low

or “blunted” levels of trait-like cortisol production (Booth, Granger,& Shirtcliff, 2008). In a similar fashion, it is possible that teacher–child conflict may be linked with physiological processes that chal-lenge the development of effective inhibitory control, and that thismay be particularly the case for girls. Further, in the study citedabove, Booth and colleagues (2009) reported that the sex differencesemerged in early adolescence but not earlier in middle childhood,suggesting a possible developmental change across this period. Ulti-mately, however, the mechanisms underlying the interaction in thepresent study remain unclear and questions to be addressed in futurework.

Inhibitory control → teacher–child relationships

Based on the theorized model, children with lower levels of inhib-itory control were hypothesized to show more conflictual relation-ships with their teachers, on average. Further, these relations wereproposed to be (at least, partially) mediated by their overt inattentionand aggression problems. The findings largely supported the first hy-pothesis. Temporally-lagged associations between children'sinhibitory-control abilities and their subsequent levels of teacher–child conflict were evident across multiple points in theelementary-school years. Children showing lower levels of inhibitorycontrol prior to school-entry tended to have higher teacher–childconflict ratings in kindergarten. Similarly, lower levels of first-gradeinhibitory control were predictive of more conflictual teacher–childrelationships in second grade. This association remained statisticallysignificant, after adjusting for children's prior levels of inhibitory con-trol and teacher–child conflict.

Subsequent analyses suggested that the relations between inhib-itory control and children's later teacher–child conflict levels werepartially mediated by their overt inattentive and aggressive be-haviors. On average, children with lower levels of pre-kindergarteninhibitory control tended to show more inattentive and aggressivebehavior as pre-kindergarteners. In turn, higher levels of pre-kindergarten inattention and aggression, respectively, were associ-ated with more conflictual teacher–child relationships in kindergar-ten. These latter relations appeared to be additive, rather thaninteractive. Similar indirect effects were evident later in theelementary-school years, yet extended only children's inattention

Page 9: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

74 D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

problems. The association between first-grade inhibitory control andchildren's concurrent levels of aggressive behavior only reachedmarginal levels of statistical significance. Notably, the difference inthis relation over time was explained largely by differences in thelevels of statistical control that were built into the respectiveestimates. That is, the ‘residualized’ nature of both the aggressionproblem and teacher–child conflict variables in the model makesthe present estimates quite conservative, representing the associa-tion between inhibitory control that is unique to first grade andaggression that is unique to first grade. When these, perhaps overlystringent, controls were relaxed in post-hoc models, the relationbetween first-grade inhibitory control and aggression was positiveand statistically significant.

Overall, these findings are consistent with prior work suggestingaggression and attention problems may be partially explained byexecutive functioning problems (Barkley, 1997; Séguin & Zelazo,2005; Willcutt et al., 2005). They also align with an extant literatureindicating that children with attentional, self-regulatory (Keogh,2003; Ladd & Burgess, 2001; Pullis, 1985, and/or aggression prob-lems (Hamre et al., 2007; Ladd & Burgess, 2001; Murray & Murray,2004) tend to experience lower quality relationships with theirteachers.

Reciprocal associations between inhibitory control and teacher–childconflict

Developmental systems perspectives posit that children's devel-opmental trajectories are shaped by reciprocal, self-organizing pro-cesses between children and their experiences over time (Ford &Lerner, 1992; Gottlieb, 1991). These dynamic processes are theorizedto be at play across contexts, including children's experiences inschool (Pianta, 1999; Rimm-Kaufman & Pianta, 2000). Taken togeth-er, the results were largely consistent with these theoretical models.On average, less effective self-regulation skills prior to school-entry– partially mediated by children's overt inattentive and aggressive be-haviors –may set the stage for reciprocal relations between children'sexperiences of teacher–child conflict and their self-regulation devel-opment across the elementary-school years.

Limitations and future directions

Given the descriptive nature of these data, it is impossible to makecausal inferences about any of the theorized relations. Indeed, the re-ciprocal processes underlying most developmental theories present aparticularly vexing methodological problem for researchers— there isno clear way to manipulate these theorized processes repeatedly andexogenously over time. Nonetheless, the present findingswere at leastconsistent with what might be expected, if reciprocal processes wereat play. In addition, many of the structural parameters of substantiveinterest were fitted in models in which the intermediate/distaloutcome was ‘residualized’ on children's prior level of the same con-struct. This may have reduced unobserved variables bias. However,there are possibly unobserved or unobservable variables notincluded in the model that may have biased the reported estimates.

Although the overall cross-lagged trends shown presently are con-sistent with the theorized model, inconsistencies in the lengths of thetemporal lags between the longitudinal variables make it difficult tomeaningfully interpret differences between substantively similar re-lations at different points in time. This was unfortunately unavoid-able, given the original design of the study. The timing of the cross-lagged associations did, however, allow one limit shared rater bias.Given the temporal lag, it was rarely the case that the same teachersrated both children's behavior problems and the quality of theirteacher–child relationships.

There were also advantages and disadvantages to using teacherreports of inattention and aggression. One advantage is that teacher

reports tap children's behaviors in the context in which they are en-gaging with teachers. This analytic choice did, however, reduce thepre-kindergarten inattention data to children attending ten or morehours of non-parental care at the age of 4.5. The majority of childrenat 54-months of age (~88%) were in non-parental care. Further, cov-ariates associated with selection into non-parental care wereaccounted for in the model. However, it is possible that unobservedvariables associated with selection into care may have biased thereported estimates.

Potential implications for practice

If the reciprocal processes underlying the theoretical model are infact causal, changes in children's self-regulation skills and/or theemotional quality of their relationships with their teachers mayslow the growth of children's self-regulation struggles, problem be-haviors, and maladaptive teacher–child relationship trajectoriesover time. On average, children who enter school with better self-regulation skills often experience the types of relationships withtheir teachers that tend to perpetuate or enhance these early skillsover time. Building these skills early may help to prevent maladaptiveteacher–child patterns before they start.

Indeed, early childhood curricula and intervention work are be-ginning to pay particular attention children' early self-regulationskills. For instance, the Tools of the Mind curriculum (Bodrova &Leong, 2006) adopts Vygotskian principles to facilitate children's de-velopment of attentional and cognitive self-regulatory skills. Initialevaluations of this curriculum suggest that providing children withtools, such as “private speech” and external supports that to helpthem remember and maintain goal-driven behavior, may have posi-tive impacts on their self-regulation skills and social behavior(Barnett et al., 2008; Diamond, Barnett, Thomas, & Munro, 2007).Similarly, interventions such as the Head Start REDI program thatimbed self-regulation training (i.e., PATHS curriculum) within abroader academic intervention framework have shown benefits forboth children's executive functioning, as well as their academic andsocial skills in early childhood (Bierman, Nix, Greenberg, Blair, &Domitrovich, 2008). It is unclear if these positive changes in child be-havior also lead to improvements in the quality of their relationshipswith their teachers. Yet, given the present results, it seems plausiblethat children's teacher–child relationships may also improve.

Enhancing the quality of children's teacher–child relationshipsmay be an additional route to improve children's self-regulation skillsand change the course of maladaptive teacher–child processes overtime. Although formal evaluations of programs aiming to improveteacher–child relationships are scarce, there is some evidence thatteacher–child relationships can be improved when teachers setaside opportunities to facilitate sensitive one-on-one interactionswith their students. For example, as part of the Banking Time inter-vention (Pianta & Hamre, 2001) teachers meet regularly to engagein one-on-one activities with a child with whom he/she is struggling.A recent exploratory evaluation of Banking Time (Driscoll & Pianta,2010) found that teachers who were randomly assigned to receivethe intervention showed significant self-rated gains in both their per-ceived relationship quality and their abilities to sensitivity engagewith the student. Although the self-rated nature of these measuresshould be considered a caveat, these initial findings suggest that trou-bled teacher–child relationships can be repaired. The impact of suchinterventions on children's developmental outcomes has yet to beconsidered, but may become clearer in the future.

Finally, child-based and teacher-based intervention strategies arenot mutually exclusive. They likely function optimally when theyare implemented together. This new wave of intervention work re-flects this realization. For instance, the same child self-regulationtraining strategy used in the Head Start REDI intervention describedabove is imbedded within the My Teaching Partner intervention, a

Page 10: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

75D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

large scale intervention to improve teacher–child relationships andoverall teacher quality (Kinzie et al., 2006). Simply, even interven-tions designed to enhance children's self-regulation skills are likelymost effective when they are delivered in the context of high-quality teacher–child relationships.

Ultimately, the long-term effects of either child- or teacher–childrelationship-based interventions remain unclear. However, theory,prior empirical work, and the present findings suggest that theymay be promising mechanisms for breaking maladaptive cycles be-tween children's self-regulation problems and their experienceswith their teachers over time.

Acknowledgement

Chronologically, different stages of this research were supportedby the Julius B. Richmond Fellowship granted by Harvard UniversityCenter on the Developing Child, a dissertation fellowship granted bythe Spencer Foundation, and post-doctoral research support providedby Dr. Clancy Blair (NIH R01 HD051502 and New York University, In-stitute of Human Development and Social Change). This researchwould not have been possible without the work of the NICHD EarlyChild Care Research Network and research staff, who designed andimplemented the overall study or the dedicated children, families,and teachers who participated. I would like to thank Peter Blake,Clancy Blair, Joanna Christodoulou, Kathleen McCartney, StephanieJones, Scott Seider, and John Willett for their thoughtful reviews ofprior versions of this manuscript.

References

Achenbach, T. M. (1991). Integrative guide for the 1991 CBCL/4-18, YSR, and TRF profiles.Burlington: University of Vermont, Department of Psychiatry.

Allison, P. D. (2003). Missing data techniques for structural equation modeling. Journalof Abnormal Psychology, 112, 545–557.

Baker, J. A. (2006). Contributions of teacher–child relationships to positive school ad-justment during elementary school. Journal of School Psychology, 44, 211–229.

Barkley, R. A. (1997). Behavioral inhibition, sustained attention, and executive func-tions: Constructing a unifying theory of ADHD. Psychological Bulletin, 121, 65–94.

Barnett, S. W., Jung, K., Yarosz, D. J., Thomas, J., Hornbeck, A., Stechuk, R., & Burns, S.(2008). Educational effects of the Tools of the Mind curriculum: A randomizedtrial. Early Childhood Research Quarterly, 23, 299–313.

Belsky, J., Fearon, R. M. P., & Bell, B. (2007). Parenting, attention and externalizing prob-lems: Testing mediation longitudinally, repeatedly and reciprocally. Journal of ChildPsychology and Psychiatry, and Allied Disciplines, 48, 1233–1242.

Berry, D., & Willett, J. B. (2009, April). Pre-kindergarten attention problems, teacher–childconflict, and achievement in late elementary school: A child-by-environment media-tional growth model. Poster session presented at the Biennial Conference of theSociety for Research in Child Development Denver, CO.

Bierman, K. L., Nix, R. L., Greenberg, M. T., Blair, C., & Domitrovich, C. E. (2008). Executivefunctions and school readiness intervention: Impact, moderation, and mediation inthe Head Start REDI program. Development and Psychopathology, 20, 821–843.

Birch, S. H., & Ladd, G. W. (1997). The teacher–child relationship and children's earlyschool adjustment. Journal of School Psychology, 35, 61–79.

Blair, C., & Razza, R. P. (2007). Relating effortful control, executive function, and false-belief understanding to emerging math and literacy ability in kindergarten. ChildDevelopment, 78, 647–663.

Bodrova, E., & Leong, D. J. (2006). Self-regulation as a key to school readiness. In M.Zaslow & I. Martinez-Beck (Eds.), Critical issues in early childhood professionaldevelopment (pp. 223–270). Baltimore, MD: Paul H. Brookes.

Booth, A., Granger, D. A., & Shirtcliff, E. A. (2008). Gender- and age-related differencesin the association between social relationship quality and trait levels of salivarycortisol. Journal of Research on Adolescence, 18, 239–260.

Brock, L. L., Rimm-Kaufman, S. E., Nathanson, L., & Grimm, N. A. (2009). The contribu-tions of ‘hot’ and ‘cool’ executive function to children's academic achievement,learning-related behaviors, and engagement in kindergarten. Early ResearchQuarterly, 24, 337–349.

Brown, T. A. (2006). Confirmatory Factor Analysis for Applied Research. New York, NY:Guilford.

Burchinal, M. R., Peisner-Feinberg, E., Pianta, R., & Howes, C. (2002). Development ofacademic skills from preschool through second grade: Family and classroom pre-dictors of developmental trajectories. Journal of School Psychology, 40, 415–436.

Byrne, B. M., Shavelson, R. J., & Muthén, B. O. (1989). Testing for the equivalence of fac-tor covariance and mean structures: The issue of partial measurement invariance.Psychological Bulletin, 105, 456–466.

Collins, L. M., Schafer, J. L., & Kam, C. M. (2001). A comparison of inclusive and restrictivestrategies in modern missing data procedures. Psychological Methods, 6, 330–351.

Diamond, A., Barnett, W. S., Thomas, J., & Munro, S. (2007). Preschool program im-proves cognitive control. Science, 318, 1387–1388.

Diamond, D. M., Campbell, A. M., Park, C. R., Halonen, J., & Zoladz, P. R. (2007). The tem-poral dynamics model of emotional memory processing: A synthesis on the neuro-biological basis of stress induced amnesia, flashbulb and traumatic memories, andthe Yerkes–Dodson law. Neural Plasticity, 1–33.

Driscoll, K. C., & Pianta, R. C. (2010). Banking time in Head Start: Early efficacy of anintervention designed to promote supportive teacher–child relationships. EarlyEducation & Development, 21, 38–64.

Duncan, G. J., Dowsett, C. J., Claessens, A., Magnuson, K., Huston, A. C., Klebanov, P., et al.(2007). School readiness and later achievement. Developmental Psychology, 43,1428–1446.

Eisenberg, N. (2000). Emotion, regulation, and moral development. Annual Review ofPsychology, 51, 665–697.

Ewing, A. R., & Taylor, A. R. (2009). The role of child gender and ethnicity in teacher–child relationship quality and children's behavioral adjustment in preschool.Early Childhood Research Quarterly, 24, 92–105.

Ford, D. H., & Lerner, R. M. (1992). Developmental systems theory: An integrativeapproach. Newbury Park, CA: Sage.

Friedman, N. P., Haberstick, B. C., Willcutt, E. G., Miyake, A., Young, S. E., Corley, R. P.,et al. (2007). Greater attention problems during childhood predict poorer execu-tive functioning in late adolescence. Psychological Science, 18, 893–900.

Gottlieb, G. (1991). Experiential canalization of behavioral development: Theory.Developmental Psychology, 27, 4–13.

Greene, R. W., Beszterczey, S. K., Katzenstein, T., Park, K., & Goring, J. (2002). Arestudents with ADHD more stressful to teach? Journal of Emotional and BehavioralDisorders, 10, 79–89.

Halperin, J. M., Sharma, V., Greenblatt, E., & Schwartz, S. T. (1991). Assessment ofthe Continuous Performance Test: Reliability and validity in a nonreferred sample.Psychological Assessment, 3, 603–608.

Hamre, B. K., & Pianta, R. C. (2001). Early teacher–child relationships and the trajectoryof children's school outcomes through eighth-grade. Child Development, 72,625–638.

Hamre, B. K., Pianta, R. C., Downer, J. T., & Mashburn, A. J. (2007). Teachers' perceptionof conflict with young students: Looking beyond problem behaviors. Social Devel-opment, 17, 115–136.

Howes, C., & Hamilton, C. E. (1992). Children's relationships with child care teachersstability and concordance with parental attachments. Child Development, 63,867–878.

Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structureanalysis: Conventional criteria versus new alternatives. Structural Equation Modeling,6, 1–55.

Hughes, J. N., Cavell, T. A., & Wilson, V. (2001). Further support for the developmentalsignificance of the quality of the teacher-student relationship. Journal of SchoolPsychology, 39, 289–301.

Hughes, J. N., Gleason, K. A., & Zhang, D. (2005). Relationship influences on teachers'perceptions of academic competence in academically at-risk minority and majorityfirst grade students. Journal of School Psychology, 43, 303–320.

Keogh, B. K. (2003). Temperament in the classroom: Understanding individual differences.Baltimore, MD: Brookes.

Kinzie, M. B., Whitaker, S. B., Neesen, K., Kelley, M., Matera, M., & Pianta, R. C. (2006).Innovative web-based professional development for teachers of at-risk preschoolchildren. Educational Technology & Society, 9, 194–204.

Kirkham, N. Z., Cruess, L. M., & Diamond, A. (2003). Helping children apply their knowl-edge to their behavior on a dimension-switching task. Developmental Science, 6,449–467.

Klein, A., & Moosbrugger, H. (2000). Maximum likelihood estimation of latent interac-tion effects with the LMS method. Psychometrika, 65, 457–474.

Kochanska, G., Murray, K. T., & Harlan, E. T. (2000). Effortful control in early childhood:Continuity and change, antecedents, and implications for social development.Developmental Psychology, 36, 220–232.

Ladd, G. W., Birch, S. H., & Buhs, E. S. (1999). Children's social and scholastic lives inkindergarten: Related spheres of influence? Child Development, 70, 1373–1401.

Ladd, G. W., & Burgess, K. B. (2001). Do relational risks and protective factors moderatethe linkages between childhood aggression and early psychological and schooladjustment? Child Development, 72, 1579–1602.

Ladd, G. W., Herald, S. L., & Kochel, K. P. (2006). School readiness: Are there social pre-requisites? Early Education and Development, 17, 115–150.

Li-Grining, C. P. (2007). Effortful control among low-income preschoolers in three cities:Stability, change, and individual differences. Developmental Psychology, 43, 208–221.

Lisbonee, J. A., Mize, J., Payne, A. L., & Granger, D. A. (2008). Children's cortisol andthe quality of teacher– child relationships in child care. Child Development, 79,1818–1832.

Lupien, S. J., Maheu, F., Tu, M., Fiocco, A., & Schramek, T. E. (2007). The effects of stressand stress hormones on human cognition: Implications for the field of brain andcognition. Brain and Cognition, 65, 209–237.

Maccoby, E. E. (2000). Perspectives on gender development. International Journal ofBehavioral Development, 24, 398–406.

MacKinnon, D. P., Fairchild, A. J., & Fritz, M. S. (2007). Mediation analysis. AnnualReview of Psychology, 58, 593–614.

Martin, R. P. (1989). Activity level, distractibility, and persistence: Critical characteris-tics in early schooling. In G. A. Kohnstamn, J. E. Bates, & M. K. Rothbart (Eds.),Temperament in childhood (pp. 451–462). Chichester, England: Wiley.

Mirsky, A. F., Anthony, B. J., Duncan, C. C., Ahearn, M. B., & Kellam, S. G. (1991). Analysisof the elements of attention: A neuropsychological approach. NeuropsychologyReview, 2, 109–145.

Page 11: Inhibitory control and teacher–child conflict: Reciprocal associations across the elementary-school years

76 D. Berry / Journal of Applied Developmental Psychology 33 (2012) 66–76

Miyake, A., Friedman, N. P., Emerson, M. J., Witzki, A. H., Howerter, A., & Wager, T. D.(2000). The unity and diversity of executive functions and their contributions tocomplex “frontal lobe” tasks: A latent variable analysis. Cognitive Psychology, 41,49–100.

Murray, K. T., & Kochanska, G. (2002). Effortful control: Factor structure and relation toexternalizing and internalizing behaviors. Journal of Abnormal Child Psychology, 30,503–514.

Murray, C., & Murray, K. M. (2004). Child level correlates of teacher–student relation-ships: An examination of demographic characteristics, academic orientations andbehavioral orientations. Psychology in the Schools, 41, 751–762.

Muthén, L. K., & Muthén, B. O. (2007). Mplus. Statistical Analysis with Latent Variables:User's Guide. Los Angeles, CA: Muthén & Muthén.

NICHD Early Child Care Research Network (1997). The effects of infant child care oninfant–mother attachment security: Results of the NICHD Study of Early ChildCare. Child Development, 68, 860–879.

NICHD Early Child Care Research Network (1999). Child care and mother-child interac-tion in the first three years of life. Developmental Psychology, 35, 1399–1413.

NICHD Early Child Care Research Network (2003). Do children's attention processesmediate the link between family predictors and school readiness? DevelopmentalPsychology, 39, 581–593.

O'Connor, E., & McCartney, K. (2007). Examining teacher–child relationships andachievement as part of an ecological model of development. American EducationalResearch Journal, 44, 340–369.

Pennington, B. F., & Ozonoff, S. (1996). Executive functions and developmental psycho-pathology. Journal of Child Psychology and Psychiatry, 37, 51–87.

Pianta, R. C. (1992). Conceptual and methodological issues in research on relationshipsbetween children and nonparental adults. In R. C. Pianta (Ed.), Beyond the parent:The role of other adults in children's Lives. New directions for child development. SanFrancisco, CA: Jossey-Bass.

Pianta, R. C. (1999). Enhancing relationships between children and teachers. Washington,DC: American Psychological Association.

Pianta, R. C., & Hamre, B. (2001). Students, teachers, and relationship support (STARS).Lutz, FL: Psychological Assessment Resources.

Pianta, R. C., Steinberg, M. S., & Rollins, L. B. (1995). The first two years of school: Teacher–child relationships and deflections in children's classroom adjustment. Developmentand Psychopathology, 7, 295–312.

Preacher, C. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessingand comparing indirect effects in multiple mediator models. Behavior ResearchMethods, 40, 879–891.

Pullis, M. (1985). Students' temperament characteristics and their impact on decisionsby resource and mainstream teachers. Learning Disability Quarterly, 8, 109–122.

Rappolt-Schlichtmann, G., Willett, J. B., Ayoub, C., Lindsley, R., Hulette, A. C., & Fischer,K. W. (2009). Poverty, relationship conflict, and the regulation of cortisol in smalland large group contexts at child care. Mind Brain & Education, 3, 131–142.

Raver, C. C. (2004). Placing emotional self-regulation in sociocultural and socioeco-nomic contexts. Child Development, 75, 346–353.

Rimm-Kaufman, S. E., & Pianta, R. C. (2000). An ecological perspective on the transitionto kindergarten: A theoretical framework to guide empirical research. Journal ofApplied Developmental Psychology, 21, 491–511.

Rudasill, K. M., & Rimm-Kaufman, S. E. (2009). Teacher–child relationship quality: Theroles of child temperament and teacher–child interactions. Early ChildhoodResearch Quarterly, 23, 27–50.

Satorra, A., & Bentler, P. M. (2001). A scaled difference chi-square test statistic for mo-ment structure analysis. Psychometrika, 66, 507–514.

Schoofs, D., PreuB, D., & Wolf, O. T. (2008). Psychosocial stress induces workingmemory impairments in an n-back paradigm. Psychoneuroendocrinology, 33, 643–653.

Séguin, J. R., & Zelazo, P. D. (2005). Executive function in early physical aggression. InR. E. Tremblay, W. W. Hartup, & J. Archer (Eds.), Developmental origins of aggression(pp. 307–329). New York, NY: Guilford.

Stipek, D., & Miles, S. (2008). Effects of aggression on achievement: Does conflict withthe teacher make it worse? Child Development, 79, 1721–1735.

Valiente, C., Lemery-Chalfant, K., Swanson, J., & Reiser, M. (2008). Prediction ofchildren's academic competence from their effortful control, relationships, andclassroom participation. Journal of Educational Psychology, 100, 67–77.

Vandenberg, R. J., & Lance, C. E. (2000). A review and synthesis of the measurementinvariance literature: Suggestions, practices, and recommendations for organiza-tional research. Organizational Research Methods, 3, 4–69.

Willcutt, E. G., Doyle, A. E., Nigg, J. T., Faraone, S. V., & Pennington, B. F. (2005). Validityof the executive function theory of Attention Deficit/Hyperactivity Disorder: Ameta-analytic review. Biological Psychiatry, 57, 1336–1346.

Winer, B. J. (1971). Statistical principles in experimental design (Second edition). NewYork, NY: McGraw-Hill.