Instructional support predicts children's task avoidance in kindergarten

Page 1

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Page 2

Author's personal copy

Early Childhood Research Quarterly 26 (2011) 376–386

Contents lists available at ScienceDirect

Early Childhood Research Quarterly

Instructional support predicts children’s task avoidance in kindergarten�

Eija Pakarinena,∗, Noona Kiurua, Marja-Kristiina Lerkkanenb, Anna-Maija Poikkeusb,Timo Ahonena, Jari-Erik Nurmia

a Department of Psychology, University of Jyväskylä, Finlandb Department of Teacher Education, University of Jyväskylä, Finland

a r t i c l e i n f o

Article history:Received 3 November 2009Received in revised form18 November 2010Accepted 30 November 2010

Keywords:Instructional supportTask-avoidant behaviorMathematicsClassroom observations

a b s t r a c t

This study examined the role of observed classroom quality in children’s task-avoidant behavior andmath skills in kindergarten. To investigate this, 1268 children were tested twice on their math skillsduring their kindergarten year. Kindergarten teachers (N = 137) filled in questionnaires measuring theirprofessional experience and also rated the children on their task-avoidant versus task-focused behaviors.Trained observers used the CLASS instrument (Pianta, La Paro, & Hamre, 2008) to observe 49 kindergartenteachers (out of 137) on their emotional support, classroom organization, and instructional support.The results of multilevel modeling showed that kindergarten classrooms differed in terms of children’stask-avoidant behavior and math skills. Moreover, the more instructional support was evidenced in theclassroom, the less children were rated as showing task-avoidant behavior. Task-avoidant behavior thenpredicted children’s low levels of math skills. The findings of the present study emphasize the importanceof quality of instructional support for children’s adaptive classroom behaviors.

© 2011 Elsevier Inc. All rights reserved.

The kinds of learning-related behavior children show at schoolplay an important role in their academic achievement and perfor-mance (e.g., Lau & Nie, 2008; Onatsu-Arvilommi & Nurmi, 2000;Stephenson, Parrila, Georgiou, & Kirby, 2008). There are also sev-eral studies suggesting that teachers may play an important rolein children’s achievement-related self-beliefs and behaviors (e.g.,Anderman et al., 2001; Turner et al., 2002; Urdan, Midgley, &Anderman, 1998; for a review see, Wigfield, Eccles, Schiefele,Roeser, & Davis-Kean, 2006). This research on the role of teachershas, however, at least two limitations. First, most previous researchhas been conducted among primary school children and adoles-cents, and less is known about the role of teachers in youngerchildren’s achievement behaviors. Second, the majority of previousstudies have examined classroom instruction and classroom goalstructures by using teacher-reports (Aunola, Leskinen, & Nurmi,2006; Urdan et al., 1998) or student-reports (e.g., Lau & Nie,2008; Nolen & Haladyna, 1990; Ryan & Grolnick, 1986), whileless research has been conducted using observational methods(for exceptions, see Rimm-Kaufman, Curby, Grimm, Nathanson,& Brock, 2009; Turner et al., 2002). Consequently, the present

� This study has been carried out in the Centre of Excellence in Learning and Moti-vation Research financed by the Academy of Finland (Nr. 213486 for 2006–2011)and another grant from the same funding agency (Nr. 213353 for 2005–2008).

∗ Corresponding author at: Department of Psychology, University of Jyväskylä,P.O. Box 35, 40014 Jyväskylä, Finland. Tel.: +358 14 260 2797.

E-mail address: [email protected] (E. Pakarinen).

study aimed to add to our understanding of classroom processesby investigating the role of observed classroom quality in chil-dren’s task-avoidant behavior in kindergarten, and whether suchtask-avoidant behavior would contribute to children’s math per-formance.

1. Children’s achievement-related behaviors

Achievement beliefs and behaviors refer to a variety of expec-tations, beliefs, and behaviors that individuals display in variouslearning situations (Nurmi, Onatsu, & Haavisto, 1995; for reviews,see Valentine, DuBois, & Cooper, 2004; Wigfield et al., 2006). Thebasic idea in this framework is that when an individual faces a newchallenging or demanding task, past experiences in learning andachievement situations activate expectations about what will hap-pen (Dweck, 1986). These expectations and related emotions theninfluence the ways in which the individual tries to handle a partic-ular task (Pintrich & De Groot, 1990). Individuals also interpret theoutcomes of their behavior in terms of making causal attributions,such as those related to situation, skills, and effort (Butkowsky &Willows, 1980).

Although various types of achievement-related beliefs andbehaviors have been described in the recent literature, there aretwo major kinds of patterns that children display. First, task-focused patterns, such as mastery-orientation (Diener & Dweck,1978; Dweck, 1986) and task-orientation (Nicholls, Cheung, Lauer,& Patashnick, 1989; Salonen, Lepola, & Niemi, 1998), are typi-cally characterized by positive affect, success expectations, high

0885-2006/$ – see front matter © 2011 Elsevier Inc. All rights reserved.doi:10.1016/j.ecresq.2010.11.003

Page 3

Author's personal copy

E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386 377

levels of effort, and persistence. Second, task-avoidant patterns,such as learned helplessness (Diener & Dweck, 1978; Dweck,1986), self-handicapping (Jones & Berglas, 1978), and ego-orientedcoping (Nicholls et al., 1989; Salonen et al., 1998), are typifiedby failure expectations, and low levels of effort and persistencein academic tasks. In the present study, children’s achievementbehaviors were operationalized as teacher-rated task-avoidantversus task-focused behavior (e.g., Onatsu-Arvilommi & Nurmi,2000; Onatsu-Arvilommi, Nurmi, & Aunola, 2002).

Previous research has shown that children’s task-focused versustask-avoidant behavior plays an important role in their aca-demic achievement and performance (Aunola, Nurmi, Niemi,Lerkkanen, & Rasku-Puttonen, 2002; Onatsu-Arvilommi et al.,2002; Stephenson et al., 2008). For example, such behaviors havebeen found to predict the development of math and reading skillsamong children (Aunola, Nurmi, Lerkkanen, & Rasku-Puttonen,2003; Onatsu-Arvilommi & Nurmi, 2000; Onatsu-Arvilommi et al.,2002). Previous studies have also investigated various antecedentsof children’s task-focused versus task-avoidant behaviors. Forexample, good academic performance has been found to contributeto high levels of success expectations and task-focused behavior(Onatsu-Arvilommi & Nurmi, 2000), whereas poor academic per-formance predicts failure expectations and task avoidance (Aunolaet al., 2003; Onatsu-Arvilommi & Nurmi, 2000). Also, parents havebeen found to contribute to children’s achievement behaviors. Forexample, Aunola et al. (2002, 2003) found that high parental expec-tations concerning children’s school performance predicted laterhigh task-focused behavior among children, whereas low parentalexpectations predicted task avoidance. Also some gender differ-ences have been found in task-avoidant behavior. For example,boys deploy more task-avoidant behaviors than girls do (Midgley &Urdan, 1995; Onatsu-Arvilommi & Nurmi, 2000; Urdan et al., 1998).There are also studies suggesting that teachers may have a role inchildren’s achievement behaviors (e.g., Turner et al., 2002; Wigfieldet al., 2006). These studies are discussed later in more detail.

2. Math performance

Learning mathematics is one major academic challenge dur-ing early school years. Math skills begin to develop well beforeformal schooling (Resnick, 1989). For example, Aunola, Leskinen,Lerkkanen, and Nurmi (2004) showed that early math perfor-mance is highly stable in kindergarten and early school years,and increase in mathematical competence is faster among thosechildren who entered kindergarten with an already-high level ofmathematical skills. The skills to discriminate between quantities,to identify numbers, and to master number sequences at the end ofkindergarten have shown to be strong predictors of mathematicsoutcomes at the end of first grade (Chard et al., 2005; Clarke & Shinn,2004; Desoete & Grégoire, 2006). Math performance is also a com-plex skill that is built on several sub-skills, such as knowledge ofnumbers, memory for arithmetical facts, conceptual understand-ing, and procedural knowledge (e.g., Dowker, 1998; Fazio, 1996).Math performance is influenced by many cognitive antecedents,such as working memory (Hecht, Torgesen, Wagner, & Rashotte,2001), counting ability (Geary, 1993), metacognitive knowledge(Desoete, Roeyers, & Buysse, 2001), attention (Ackerman, Anhalt, &Dykman, 1986), language skills (phonological awareness, naming)(Donlan, 1993; Fazio, 1996), and listening comprehension (abilityto understand instructions) (Jordan, Hanich, & Kaplan, 2003).

Besides cognitive factors, math performance is also sensi-tive to various motivational and affective factors (Aunola et al.,2003; De Corte, 1995; Pajares & Miller, 1994). For example, mathperformance has been shown to be reduced by children’s anxi-ety (Wigfield & Meece, 1988). Moreover, children’s maladaptiveachievement behavior has been found to hamper their subsequent

progress in mathematical skills (Onatsu-Arvilommi et al., 2002).Children’s math skills are also sensitive to teachers’ instructionalgoals (Aunola et al., 2006; Lau & Nie, 2008; Turner et al., 2002). Inthe present study, we examine the extent to which classroom qual-ity, on the one hand, and children’s task-avoidant behavior, on theother, would predict children’s math performance in kindergarten.

3. Observed classroom quality

Recent research has differentiated three domains in classroomquality, that is, emotional support, classroom organization, andinstructional support (Pianta et al., 2008). Emotionally supportiveteachers are sensitive to children’s needs and interests, and theyare responsive and warm toward children (Pianta et al., 2008). Inclassrooms with a high level of classroom organization, teachers useproactive rather than reactive approaches to discipline, establishclear and stable routines, monitor students carefully to keep theminvolved in academic tasks, and provide activities that are inter-esting to children (Emmer & Stough, 2001; Pianta et al., 2008). Inclassrooms with high instructional support, teachers provide scaf-folding and support (Yates & Yates, 1990), create opportunities forconceptual development, and offer appropriate questioning andfeedback for their students (La Paro, Pianta, & Stuhlman, 2004;Pianta et al., 2008).

Previous research has shown that these three domains areimportant for the development of children’s academic skills andadaptive classroom behaviors. For example, emotional support pro-vided by the teacher promotes children’s engagement (McWilliam,Scarborough, & Kim, 2003) and academic achievement (Hamre& Pianta, 2005; Perry, Donohue, & Weinstein, 2007). Moreover,classroom organization is associated with children’s achieve-ment (Wharton-McDonald, Pressley, & Hampton, 1998), behavioralengagement, and on-task behavior (Rimm-Kaufman et al., 2009).Finally, instructional quality contributes to students’ academic out-comes (Hamre & Pianta, 2005; Howes et al., 2008; Mashburn et al.,2008) and engagement in academic activities (Bogner, Raphael, &Pressley, 2002; Dolezal, Welsh, Pressley, & Vincent, 2003). Despitethe great amount of studies showing that classroom quality is animportant contributor to child outcomes and adaptive behavior,no research to our knowledge has been conducted on how dif-ferent domains of classroom quality are connected with children’sacademic achievement and achievement behaviors outside the US.

4. Classroom quality and children’s task-avoidant behavior

It might be assumed that classroom quality provides an impor-tant context for the development of children’s task-avoidant andtask-focused behavior. For example, teachers may either encour-age children’s engagement and enthusiasm in learning activitiesor, alternatively, discourage their efforts and inadvertently aug-ment their anxiety and task-avoidant behavior. Previous researchhas shown that less task avoidance occurs in classrooms in whichteachers provide instructional and motivational support for learn-ing compared to classrooms in which teachers do not provide suchsupport (Turner et al., 2002; Turner, Meyer, Midgley, & Patrick,2003; Urdan et al., 1998). In classrooms with high instructionaland motivational support, teachers help students to build under-standing, emphasize learning and enjoyment, support effort, andprovide personal attention and encouragement (Turner et al.,2002). Moreover, in classrooms with a higher observed class-room quality, off-task behavior among children is less commonthan in classrooms with lower quality (Pianta et al., 2002; Ponitz,Rimm-Kaufman, Grimm, & Curby, 2009; Rimm-Kaufman et al.,2009; Rimm-Kaufman, La Paro, Downer, & Pianta, 2005). Effec-tive classroom management and high instructional quality arealso associated with children’s greater engagement in school work

Page 4

Author's personal copy

378 E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386

(Downer, Rimm-Kaufman, & Pianta, 2007; Wharton-McDonaldet al., 1998). Anderman et al. (2001) also showed that studentsexperienced decrements in achievement values in classroomswhere teachers used performance-oriented instructional practices,that is, they tended to point out ability differences and to use com-petitive instructional methods.

Previous research on the role of teacher instruction and class-room practices has, however, at least four limitations. First, themajority of studies have been carried out among older studentsin middle and secondary schools (for exceptions, see Perry et al.,2007; Rimm-Kaufman et al., 2005, 2009). Second, most researchin the field has used either student-reports of classroom practices(e.g., Lau & Nie, 2008; Nolen & Haladyna, 1990; Ryan & Grolnick,1986) or teacher-reports (e.g., Aunola et al., 2006; Urdan et al.,1998). Third, few previous studies have investigated and controlledfor the entry level of student variables when investigating theimpacts of teacher’s instruction on children’ achievement beliefsand behaviors (as an exception, see Anderman et al., 2001). Fourth,few previous studies have investigated teachers’ role using meth-ods that are appropriate for analyzing hierarchical data, which istypical of classrooms (for exceptions, see Anderman et al., 2001;Rimm-Kaufman et al., 2009). Consequently, the present study madean effort to increase our understanding of the role that observedteaching practices play in children’s task-avoidant behavior, aftercontrolling for their previous skill level.

5. Kindergarten education in Finland

In Finland, compulsory education begins in the year in whichthe child turns 7 years of age and ends when he/she is 16. Thegreat majority of 6-year-olds are entitled to kindergarten educa-tion for one year before starting nine-year comprehensive school.At present, about 98% of six-year-olds attend kindergarten edu-cation provided either at a day-care center (78%) or in a school(22%) (Rautanen, 2007). Irrespective of the setting, the curricula forkindergarten education are the same.

Kindergarten education in Finland aims at developmentallyappropriate instruction and child-centered activities, and cen-ters on learning through imagination and play (National Boardof Education, 2000). Children can learn at their own pace, andtheir interests are taken into account in activities. The curriculumcovers seven subject areas including mathematics, but instructionand activities are not divided into subject area lessons (NationalBoard of Education, 2000). Instead, teacher-led activities are inte-grated into thematic learning taking place throughout the day.For example, instead of formal teaching of mathematics, childrenare encouraged to play with numbers and mathematical concepts.Kindergarten education (minimum 700 h per year), meals, healthcare, and travel are provided free of charge. During kindergarten,children also have the right to day care if it is needed.

6. Aims and hypotheses

The present study examined the following research questions:

(1) Do kindergarten classrooms differ in the extent to which chil-dren show task-avoidant behavior? On the basis of previousresearch (Urdan et al., 1998), we expected (Hypothesis 1) thatchildren in different kindergarten classrooms would differ inrespect to their task-avoidant behavior.

(2) Do kindergarten classrooms differ from each other in terms ofchildren’s math skills? We expected (Hypothesis 2) that chil-dren in different kindergarten classrooms would differ fromeach other with respect to their performance in mathematics(Aunola et al., 2004; see also Perry et al., 2007).

(3) Does observed classroom quality predict classroom differencesin children’s task-avoidant behavior, after controlling for class-room differences in teachers’ professional experience, mother’seducation, gender, children’s age, and previous skills in math?Since observed instructional practices have been found to beassociated with students’ achievement behaviors (e.g., Piantaet al., 2002; Ponitz et al., 2009; Rimm-Kaufman et al., 2009;Turner et al., 2002), we assumed that low levels of emotionalsupport, classroom organization, and instructional supportwould predict a high level of task avoidance among children(Hypothesis 3).

(4) Do observed classroom quality and classroom differencesin task-avoidant behavior predict classroom differences inchildren’s math skills? We expected that besides individualdifferences in children’s task-avoidant behavior, classroomdifferences in such behavior would also predict classroom dif-ferences in children’s math skills (Hypothesis 4a). We alsoassumed that classroom quality would predict classroom dif-ferences in children’s math skills (Hypothesis 4b, Aunola et al.,2006; Caprara, Barbanelli, Steca, & Malone, 2006; Perry et al.,2007; Turner et al., 2002).

7. Method

7.1. Participants and procedure

In the present study, 1268 (613 girls, 655 boys) kindergarten-ers (M = 73.58 months old, SD = 3.40 months) and 137 kindergartenteachers (130 female, 7 male) were investigated. The study is partof the ongoing First Steps Study (Lerkkanen et al., 2006). The chil-dren participating in this study represent a whole age-cohort fromthree municipalities, two of them located in Central Finland andone in Eastern Finland.

Before starting the study, children’s (N = 1580) parents werecontacted and 80.3% of them gave written consent for their chil-dren’s and their own participation. Subsequently, 1268 childrenwere examined twice during their kindergarten year. Their pre-math skills (n = 1268) were measured in October (Time 1, T1)and their math skills (n = 1239) in April (Time 3, T3). The testswere carried out by trained investigators. Children’s (n = 1231)task-avoidant versus task-focused behavior was rated by theirkindergarten teachers in March (Time 2, T2). Mothers were alsoasked to fill in a background questionnaire in March (Time 2, T2)and the response rate was 81.6%.

Children came from 137 kindergarten classes and from 121kindergarten sites. Eighty-seven of kindergarten sites were situatedin day-care centers and 34 in elementary schools. The kindergartenclass sizes ranged from 3 to 24 (M = 13.70; SD = 5.51). All the groupswere Finnish-speaking. Although most of the groups were com-posed of kindergarten-age children (6-year-olds) exclusively, theage composition was wider in some of the groups: some groupsthat were taught in day-care centers also enrolled younger chil-dren (most often 5-year-olds), and some groups that were taughtin elementary schools also enrolled 1st and/or 2nd graders. How-ever, only children at kindergarten age were included in the study.Some kindergartens were in semi-rural and some in urban areas.

Kindergarten teachers were first asked in March for their writ-ten consent to participate the study. Ninety-one percent of teachersagreed (137 teachers). Teachers were then asked to complete ques-tionnaires on their education and teaching experience (Time 2, T2).Questionnaires were sent to the kindergarten teachers and theywere returned by mail. The kindergarten teachers’ working expe-rience in day care ranged from less than a year to more than 15years (Mode = more than 15 years). Their experience in an elemen-tary school setting ranged from not at all to more than 15 years

Page 5

Author's personal copy

E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386 379

(Mode = not at all). All teachers had at least a Bachelor degree orequivalent. The teachers participating in the classroom observa-tions (n = 49) were selected on a voluntary basis from the total of137 kindergarten teachers.

7.2. Testing missing-at-random assumption

As only a subsample of 49 out of 137 teachers participated inclassroom observations, the present study contained a problem ofmissingness by design. To use all the data in the case of missingby design, the missingness should be random (Muthén & Muthén,1998–2009). Consequently, the assumption of missingness-at-random was tested in two ways. First, we conducted a set ofindependent samples t-tests to compare the 49 kindergartenteachers who participated in observations to those who did not par-ticipate. The results showed no statistically significant differencesbetween the compared groups in their educational background,professional experience, number of children in the classrooms,mean age of the children, number of personnel available, teacherstress, and teachers’ efficacy beliefs. Thus, the observed teach-ers did not differ from the rest of the sample. As a second step,we tested the missing-at-random assumption in regards to allkinds of missingness in the variables of this study. To accomplishthis, we conducted Little’s tests of Missing-Completely-At-Random(MCAR tests; Little, 1988) separately for the class-level variablesand individual-level variables (H0: Missingness is independent ofobserved values or missing values in the data set; H1: Missing-ness depends on the observed or missing values in the data set).The results of MCAR tests showed that the missingness was ran-dom both at the individual-level: �2(76) = 94.05, p = 0.08 and at theclass-level: �2(24) = 31.45, p = 0.14.

Since the missing data proved to be missing-at-random, statisti-cal analyses were carried out using the full-information maximumlikelihood estimation (FIML), which allows all the informationavailable to be used without imputing data (Muthén & Muthén,1998–2009). Consequently, we were able to use the data for allkindergarten teachers (N = 137) and children (N = 1268) in furtheranalyses.

7.3. Measures

7.3.1. Children’s measures7.3.1.1. Task-avoidant behavior. The kindergarten teachers wereasked to evaluate the behavior of each child in their class usingthe Behavior Strategy Rating Scale (BSR; Onatsu & Nurmi, 1995;Onatsu-Arvilommi & Nurmi, 2000). They were asked to considerhow a certain child typically behaved in classroom situations andthen to rate his or her behavior using five statements (e.g., Does thechild have a tendency to find something else to do instead of focusingon the task at hand?; Does the child show persistence even in the moredifficult tasks? (reversed)) assessed on a 5-point rating scale (1 = notat all; 5 = to a great extent). When there were two kindergartenteachers in a class, both of them evaluated half of the childrenin the group. In small classrooms only one kindergarten teacherevaluated all the children in the class. A summary score was calcu-lated on the basis of the five teacher-rated items. The Cronbach ˛for task-avoidant behavior was 0.91. The task-focused versus task-avoidant scale of the BSR has been shown to correlate moderatelywith children’s self-reported task-focused behavior (0.30) (Nurmi& Aunola, 2000; Onatsu & Nurmi, 1997) and also with observers’rating of it (0.42) (Nurmi & Aunola, 2000). It has also shown goodtest-retest reliability (Aunola et al., 2003). Further, task-avoidantbehavior has been found to predict negatively the developmentof math (Aunola et al., 2003; Onatsu-Arvilommi et al., 2002) andreading skills (Aunola et al., 2002; Onatsu-Arvilommi et al., 2002)

7.3.1.2. Pre-math skills. Children’s pre-math skills at Time 1 wereassessed by using the number sequences test (for similar tasks,see Aunola et al., 2004; Koponen, Aunola, Ahonen, & Nurmi, 2007;Räsänen, Salminen, Wilson, Aunio, & Dehaene, 2009). Knowledgeof number sequences before school age has been shown to pre-dict later success in arithmetic skills (Kinnunen, Lehtinen, & Vauras,1994; Räsänen et al., 2009). Knowledge of number sequences wasassessed by means of four subtasks in which children were askedto count aloud forward and backward (counting forward from 1 to31, counting backward from 12 to 7, counting backward from 23 to18, counting forward from 6 to 13). In each of these four subtasks,children got one point when they made only a small error in count-ing aloud (e.g., one number missing in a sequence), and two pointswhen they counted aloud exactly correctly (the maximum score 8points). The split-half reliability for pre-math skills was 0.63 andtest–retest reliability 0.68.

7.3.1.3. Math skills. Children’s math skills at Time 3 were assessedby using two tests (i.e., math screen number concept test andarithmetic test). Jordan, Kaplan, Nabors Oláh, and Locuniak (2006)suggest that a composite of early numerical skills should com-prise the following types of tasks: magnitude comparison, counting,identification of numbers, and simple arithmetic Children’s numberconcept was assessed using a part of the national school readi-ness test, namely, the math screen number concept test (Elomäki,Huolila, Poskiparta, & Saranpää, 1999). This test consisted of ninesubtasks: children were asked to (1) draw as many balls as wasshown in the picture, (2) draw one ball more than in the picture,(3) draw one ball less than in the picture, (4) draw as many ballsas the number symbol tells, (5 and 6) circle the number which tellshow many balls there are in the box, (7) mark off the first ball, (8)mark off the fourth ball, and (9) mark off the seventh ball. Chil-dren got one point when answering correctly (the maximum score9 points).

Children’s basic arithmetical skills were assessed by means ofthe 3-min Basic Arithmetic Test (Aunola & Räsänen, 2007; see alsoRäsänen et al., 2009) consisting of simple addition and subtrac-tion tasks. The child was allowed to answer either by writing thenumber or by saying the answer aloud. If a child was unable tosolve any calculation tasks with written numbers, an additionaltask with three items was given (that is, a child was asked to countobjects). The score was the total number of correct answers (themaximum score 28 points). Test–retest reliability for this task hasbeen 0.86 (Räsänen et al., 2009). The scores for number concept testand arithmetic skills tests were firstly standardized. Then, a com-posite score was created by averaging the scores of these two tests.The Cronbach ˛ for this Math skill score was 0.70.

7.3.2. Teachers’ measures7.3.2.1. Classroom observations. The kindergarten classrooms wereobserved using the Classroom Assessment Scoring System (theCLASS; La Paro et al., 2004; Pianta et al., 2008). The CLASS con-sists of 10 items measuring three components of classroom quality:(1) Emotional Support (4 items: Positive climate, Negative climate,Teacher sensitivity, and Regard for student perspectives), (2) Class-room Organization (3 items: Behavior management, Productivity,and Instructional learning formats), and (3) Instructional Support(3 items: Concept development, Quality of feedback, and Languagemodeling). In this study, we used nine items from the CLASS (theitem Negative Climate was left out) based on their validation inFinnish kindergarten (Pakarinen et al., 2010). Each item was ratedon a 7-point scale: low (1, 2), moderate (3–5), and high (6, 7). Themanual (Pianta et al., 2008) provides detailed information concern-ing indicators of each item and examples of teacher behaviors andclassroom interactions, as well as reliability and validity informa-tion (see also, Pakarinen et al., 2010).

Page 6

Author's personal copy

380 E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386

The kindergarten classrooms were observed by 17 trainedobservers, all of whom were female university students. Theobservers were carefully trained in advance with live observationsand video observations (see Pakarinen et al., 2010). Each kinder-garten class was observed on two different days (both in springterm) by a pair of observers. The observations were conductedduring February and March (Time 2) during typical days in thekindergarten that did not include any special program. Correlationsbetween the CLASS ratings for the two separate observation daysranged from 0.44 (Productivity) to 0.80 (Teacher sensitivity). Forthe further analyses, the ratings from these two observation dayswere collapsed. On average, 10.89 (SD = 3.35) children were presentduring the observation. Of the observed classrooms 36 (73.5%) werein day care centers and 13 (26.5%) in elementary schools.

Observation began in the morning when instructional activitystarted (at 9 am) and lasted approximately 3 h (until naptime infull-time programs and until the time children left in half-timeprograms). The two observers assigned their CLASS codings inde-pendently of each other. Inter-rater reliability between the pairsof observers was determined by calculating intra-class correla-tions. To calculate the intra-class correlations, we used a two-waymixed effect model (measure fixed, observers random), the abso-lute agreement definition, and the average measure intra-classcorrelation (which assumes no interaction effect). McGraw andWong (1996) recommend this approach when ratings are on con-tinuous scales and the observers vary from classroom to classroom.The inter-rater reliabilities between the pairs of observers var-ied mainly between 0.80 and 0.94. For the final analyses, a meanscore for each item was calculated from the ratings of the twoobservers. Summary scores were calculated taking the mean of theitems belonging to each specific factor, namely, Emotional support,Classroom organization, and Instructional support. The Cronbach ˛sfor the scales were as follows: Emotional support 0.93; Classroomorganization 0.88; Instructional support 0.90.

7.3.2.2. Professional experience. Kindergarten teachers were askedto indicate their professional experience separately in day care set-ting and in elementary school setting on a 6-point scale (0 = not atall, 1 = less than a year, 2 = 1–5 years, 3 = 6–10 years, 4 = 11–15 years,5 = more than 15 years). A new variable describing kindergartenteachers’ total amount of professional experience was formulatedby adding experience in kindergarten to experience in elementaryschool setting.

7.4. Parents’ measures

7.4.1. Mothers’ level of educationMothers were asked about their level of vocational education

Out of the 1268 mothers, 1028 returned the questionnaire. Fivemothers out of those who returned the questionnaire did not reporttheir vocational education. Altogether, we did not have data oneducation of 245 mothers. Despite of the amount of missing data,the results of MCAR tests showed that the missingness was ran-dom both at the individual-level: �2(76) = 94.05, p = 0.08 and at theclass-level: �2(24) = 31.45, p = 0.14.

Five and half percent of mothers had no vocational education,25.2% had a vocational school degree, 18.7% had a vocational collegedegree, 9.8% had a polytechnic degree or Bachelor’s degree, and21.4% had a university degree.

8. Results

8.1. Analysis strategy

The present study aimed to examine the extent to which kinder-garten classrooms differ with respect to children’s task-avoidant

behavior, mathematical skills, and the extent to which observedclassroom quality would predict these classroom differences. Mul-tilevel modeling (Duncan et al., 1997; Heck & Thomas, 2009) isan excellent tool for answering these research questions. First, itenables the variance in the observed variables to be divided intotwo components: (1) variation due to differences between differ-ent kindergarten classrooms (between-kindergarten class variation)and (2) variation due to individual differences after taking intoaccount kindergarten class membership (within-kindergarten classvariation). Second, multilevel modeling enables one to enter vari-ous predictors both at the kindergarten class-level (between-level)and at the level of the individual children (within-level).

The analyses were carried out as follows. First, intraclass cor-relation coefficients were calculated in order to determine thebetween-classroom variation in task-avoidant behavior and math-ematics. Only those individual-level variables in which classroomdifferences were statistically significant were included at theclassroom-(between) level in the further multilevel analyses. Next,the correlations between the observed variables were calculatedat the kindergarten class-level (between-level) and at the level ofindividual children (within-level).

As a next step, multilevel models were carried out for task-avoidant behavior and for math skills (see theoretical model inFig. 1). Only statistically significant paths were included in the finalmodels so that only significant predictors would contribute to theproportion of the explained variance. The predictor variables wereallowed to correlate.

All the analyses were performed using the Mplus statisticalpackage (Version 5; Muthén & Muthén, 1998–2009). The parame-ters of the models were estimated using full-information maximumlikelihood estimation with non-normality robust standard errors(MLR estimator; Muthén & Muthén, 1998–2009). The goodness-of-fit of the estimated models was evaluated by four indicators:�2-test, Comparative Fit Index (CFI), Tucker Lewis Index (TLI), RootMean Square Error of Approximation (RMSEA), and StandardizedRoot Mean Square Residual (SRMR).

8.2. Intraclass correlations

First, to examine kindergarten classroom differences in task-avoidant behavior and mathematics, intraclass correlations and thevariance estimates at the between- and within-levels were cal-culated by using the kindergarten class number as a clusteringvariable. The results showed that between-class variation at Time3 both in task-avoidant behavior and math skills was statisticallysignificant: 10% (p < 0.001) of the total variation in task-avoidantbehavior, and 5% (p < 0.05) of the total variation in math skillswas due to kindergarten class membership. The results for thecontrol variables showed further that there was significant kinder-garten class-level variation in pre-math skills (5% of the totalvariance, p < 0.05), and mother’s education (10% of the total vari-ance, p < 0.001). However, the intra-class correlation for gender (1%of the total variance, p = 0.68) and for child’s age in months (1% ofthe total variance, p = 0.72) was non-significant. Consequently, infurther analyses gender and child’s age were treated as individual-level (within-level) variables. Task-avoidant behavior, math skills,and other control variables, in turn, were analyzed at both levels.

8.3. The roles of classroom quality and task-avoidant behavior inmath skills

The within-level (below the diagonal) and between-level (abovethe diagonal) correlations, means, variances, and minimums andmaximums for the observed variables are presented in Table 1.Since the CLASS domains correlated highly with each other andconfirmatory factor analyses indicated three separate factors

Page 7

Author's personal copy

E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386 381

BetweenBetween Task avoidance (T2)Classroomquality (T2)

WithinWithin

Pre-math skills(T1)

Gender

Pre-mathskills (T1)

Math skills (T3)

Mother’seducation (T2)

Mothers’education (T2)

Child’s age(T1)

Professional experience (T2)

Gender

Child’s age(T1)

Task avoidance (T2) Math skills (T3)

Fig. 1. Theoretical multilevel model for kindergarten children’s task-avoidant behavior and math performance.

(Pakarinen et al., 2010), we ran separate multilevel models foreach of the domains (i.e., Emotional support, Classroom organiza-tion, and Instructional support) to determine whether the observedclassroom quality indicators would predict kindergarten classroomdifferences in children’s task-avoidant behavior and math skills,after controlling for professional experience, mother’s education,child’s gender, child’s age, and pre-math skills (see Fig. 1 for a the-oretical model). We also examined whether classroom differencesin task-avoidant behavior would predict classroom differences inlater math skills. We tested also a model for one global factor, thatis, classroom quality consisting of Emotional support, Classroomorganization, and Instructional support, but its associations withtask avoidance and math skills were insignificant. Since profes-sional experience was not associated significantly with any othervariables in the models, it was not included in the final models.

8.3.1. Instructional supportThe final model (Fig. 2) for instructional support included statis-

tically significant paths and associations only (�2(11, Nwithin = 1268,Nbetween = 137) = 22.11, p = 0.02; CFI = 0.99, TLI = 0.97; RMSEA = 0.03;SRMRbetween = 0.08, SRMRwithin = 0.02). This model is presented inFig. 2 [kindergarten class-level (between-level) results above thedashed line and individual-level (within-level) results below thedashed line]. The results (Fig. 2) showed that, at the kinder-garten class-level, instructional support predicted negativelytask-avoidant behavior among children: the less instructionalsupport was observed, the higher the children’s teacher-rated task-avoidant behavior shared among the kindergarten class members.Task-avoidant behavior and pre-math skills (T1) further predictedmath skills: the more task-avoidant behavior children were ratedas showing and the lower their pre-math skills were, the lower themath skills (T3) shared among the kindergarten class members.We tested also whether task avoidance mediated the associationbetween instructional support and math skills but these indirecteffects were non-significant (instructional support: estimate ofindirect effect = 0.06; S.E. = 0.04; p = 0.19). The mothers’ education

and pre-math skills typical of the kindergarten class correlated pos-itively with each other: the higher the mothers’ education typicalof the kindergarten class, the higher the pre-math skills typical ofthe kindergarten class. However, mothers’ education and pre-mathskills typical of the kindergarten class had no direct impacts on thechildren’s task-avoidant behavior typical of the kindergarten class.The results at the kindergarten class-level remained the same whencontrolling also for kindergarten class size.

The results (Fig. 2) at the level of individual children(within-level) showed further that, after controlling for class-room differences in task avoidance and math skills, pre-mathskills predicted negatively and gender positively the level of chil-dren’s task-avoidant behavior: the lower their pre-math skills, themore task-avoidant behavior children were rated as showing. Inaddition, children’s age predicted negatively the level of their task-avoidant behavior: younger children were rated as having moretask-avoidant behavior than older ones. Moreover, boys were ratedas showing more task avoidance than girls. In addition, pre-mathskills, children’s age, and children’s task-avoidant behavior furtherpredicted their math skills later on: the better the pre-math skills,the older the child, and the less task-avoidant behavior the childtypically was rated as showing, the better were his or her mathskills later on.

8.3.2. Emotional supportThe final model for emotional support included statistically

significant paths and associations only (�2(13, Nwithin = 1268,Nbetween = 137) = 23.61, p = 0.03; CFI = 0.99, TLI = 0.98; RMSEA = 0.03;SRMRbetween = 0.08, SRMRwithin = 0.02). The results were similar tothose of instructional support, except that emotional support didnot predict task-avoidant behavior typical of the classrooms. How-ever, as reported also earlier, task-avoidant behavior and pre-mathskills (T1) typical of the classroom predicted class members’ latermath skills (Task-avoidant behavior: Standardized Estimate = −0.33,p < 0.05; Pre-math skills: Standardized Estimate = 0.69, p < 0.001):the more task-avoidant behavior children were rated as showing

Page 8

Author's personal copy

382 E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386

Tab

le1

Wit

hin

-lev

el(b

elow

the

dia

gon

al)

and

betw

een

-lev

elco

rrel

atio

ns

(abo

veth

ed

iago

nal

),an

dm

ean

san

dva

rian

ces

ofob

serv

edva

riab

les.

Var

iabl

es1.

2.3.

4.5.

6.7.

8.9.

10.

Mea

nM

inM

axV

arbe

twee

n

Task

avoi

dan

ceT2

11.

000.

11−0

.32

––

−0.2

2−0

.15

−0.3

1c−0

.03

−0.2

52.

321.

005.

000.

11Pr

e-m

ath

skil

lsT1

1−0

.33a

1.00

0.64

a–

–−0

.31d

−0.5

6a−0

.46c

0.06

0.56

a4.

290

8.00

0.42

Mat

hsk

ills

T31

−0.3

6a0.

52a

1.00

––

0.21

−0.1

20.

180.

180.

46b

5.56

013

.00

0.15

Gen

der

20.

24a

0.15

a0.

021.

00–

––

––

–0.

520

1.00

–C

hil

d’s

age

inm

onth

sT1

2−0

.16a

0.23

a0.

22a

0.04

1.00

––

––

–73

.58

68.0

080

.00

–Em

otio

nal

sup

por

tT2

3–

––

––

1.00

0.86

a0.

80a

0.07

−0.1

65.

122.

906.

230.

52C

lass

room

orga

niz

atio

nT2

3–

––

––

–1.

000.

79a

0.29

c−0

.20

5.34

3.19

6.27

0.38

Inst

ruct

ion

alsu

pp

ort

T23

––

––

––

–1.

000.

14−0

.14

3.96

1.64

5.54

0.72

Prof

essi

onal

exp

erie

nce

T23

––

––

––

––

1.00

0.03

4.10

1.00

5.00

1.28

Mot

her

s’ed

uca

tion

1−0

.12b

0.10

b0.

11b

0.02

0.00

––

––

1.00

3.18

1.00

5.00

0.18

Var

wit

hin

0.94

7.77

2.11

0.25

11.5

4–

––

–1.

55

Not

e1:

ap

<0.

001;

bp

<0.

01;

c p<

0.05

;dp

<0.

10.N

ote

2:1V

aria

ble

both

atbe

twee

n-l

evel

and

wit

hin

-lev

el,2

Wit

hin

-lev

elva

riab

le,3

Bet

wee

n-l

evel

vari

able

,–n

otes

tim

ated

.Not

e3:

gen

der

0=

girl

,1=

boy.

and the lower their pre-math skills, the lower the math skills (T3)shared among the kindergarten class members. The results at theindividual-level were similar as in the model of instructional sup-port.

8.3.3. Classroom organizationThe final model for classroom organization included statisti-

cally significant paths and associations only (�2(12, Nwithin = 1268,Nbetween = 137) = 19.18, p = 0.08; CFI = 0.99, TLI = 0.98; RMSEA = 0.02;SRMRbetween = 0.07, SRMRwithin = 0.02). The results showed thatclassroom organization did not predict task-avoidance behaviortypical of the classrooms. Other results were similar as in thecases of instructional and emotional support, except that classroomorganization correlated negatively with children’s pre-math skills(Standardized Estimate = −0.40, p < 0.05).

9. Discussion

The present study examined the extent to which classroomquality would contribute to children’s adaptive classroom behav-iors. The results indicated that teachers’ instructional supportpredicted the kinds of achievement behaviors children showed inkindergarten: the more instructional support was observed in theclassroom, the less children were rated as showing task avoidance.This task-avoidant behavior typical of the kindergarten class thenalso predicted children’s math skills in the same classrooms.

The first aim of the present study was to investigate to whatextent children in different kindergarten classrooms might differwith respect to their task-avoidant behavior and math skills. Aswas expected (Hypotheses 1 and 2), the results showed that chil-dren in the same kindergarten class resembled each other in termsof their task avoidance and in their math skills when compared tochildren in other classrooms, that is, ICCs of task avoidance andmath skills were statistically significant. The results showed fur-ther, as was expected (Hypothesis 3), that teachers’ instructionalsupport was associated with children’s task-avoidant behavior: themore instructional support the teacher provided in a classroom,the less task-avoidant behavior children were rated as showing.There are at least two possible explanations for this result. First, inthe classrooms in which teachers help children to develop under-standing and to model their own thinking processes, children arelikely to develop a more task-focused achievement orientation andengage more in learning activities. Second, teachers in these class-rooms may give more positive achievement-related feedback tochildren, which then contributes to children’s self-concept as learn-ers and, subsequently, increases their task-focused behavior. Theseresults are in line with the findings of previous studies amongolder students. Turner et al. (2002, 2003) found that studentsreported a lower incidence of avoidance behavior in classrooms inwhich teachers provided instructional and motivational support forlearning, that is, emphasized learning, effort, and understanding.Teachers providing cognitively challenging tasks and scaffoldinghave also been shown to be more likely to have students showinggreater engagement in learning (Bogner et al., 2002; Dolezal et al.,2003). Further, Ponitz et al. (2009) found that classroom qualitywas associated with student engagement.

The results of the present study add to previous research byshowing that teachers can encourage children’s engagement inlearning and task-focused behavior already in kindergarten settingsbefore formal education commences. By showing that instruc-tional support is connected to children’s task-focused behavior,the present study suggests further that, although Finnish kinder-gartens adopt a play-centered approach and there is no formalteaching of mathematics, kindergarten teachers’ instructional prac-tices are connected with the way children act in learning situations.

Page 9

Author's personal copy

E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386 383

BetweenBetween

WithinWithin

Task avoidance (T2)(R2 = .11)

InstructionalSupport (T2)

Task avoidance (T2)(R2 = .20)

WithinWithin

Pre-mathskills (T1)

Male

.10**

.14***

Pre-mathskills (T1)

-.33*

-.35***

Math skills (T3)(R2 = .61)

Math skills (T3)(R2 = .32)

.30*** -.21***

Mother’seducation (T2) .43***

.69***

-.36*

Mothers’education (T2)

.52**

Child’s age(T1)

-.09**.08***

.23***

Fig. 2. Multilevel model for kindergarten children’s task-avoidant behavior and math performance. The paths and associations between variables are presented as standard-ized estimates. Note 1. *p < 0.05; **p < 0.01; ***p < 0.001.

Our results suggest that the teachers’ role in promoting children’sadaptive behavior in learning-related situations is also importantin play-centered learning activities. Consequently, there is need topay more attention to early educators’ awareness of their influenceon children and the means available to promote children’s focuson task and adaptive behavior. The results of the present studyindicate that teachers can promote children’s preparedness to facedifficult learning situations by giving individualized feedback andoptimally challenging tasks. In addition, they can encourage chil-dren’s efforts and, by doing so, decrease their experience of theanxiety in learning situations that easily leads to task avoidance.

Rimm-Kaufman et al. (2009) indicated that classroom organiza-tion was linked to kindergarteners’ adaptive classroom behavior. Inthe present study, emotional support and classroom organizationcorrelated also negatively with teacher-rated task avoidance. Thecorrelation estimates were similar to those in the Rimm-Kaufmanet al. (2009) and Ponitz et al. (2009) studies, but the modeling failedto find statistically significant associations. Rimm-Kaufman et al.(2009) found also that children who experienced more instruc-tional support from their teachers were rated as lower in cognitiveself-control and lower in positive work habits. On the contrary,in the present study, it was instructional support that was con-nected to children’s adaptive behavior. There are at least twopossible explanations for this. First, Finnish children were one yearolder than children in Rimm-Kaufman et al.’s (2009) study. Olderchildren’s adaptive classroom behaviors may benefit more frominstructional support whereas younger children are more depen-dent on the external environment to support their self-regulatoryskills. A second possible explanation for the difference betweenthe result of the present study and those of Rimm-Kaufman et al. isthat Finnish kindergarten teachers rate children’s behavior differ-ently and have different demands on children than their colleaguesin the US. In the US, teachers who offer more instructional sup-port for children may place more academic demands on children,as suggested by Rimm-Kaufman et al. (2009). In turn, Finnishkindergarten teachers place a strong emphasis on developing thechildren’s positive self-concept and social skills and promotingtheir learning potential (Hytönen, Krokfors, Talts, & Vikat, 2003).

They may place more social and behavioral demands on childrenas there are no demands on formal teaching and learning standardsin curricula.

We also aimed to determine whether classroom quality indica-tors would predict classroom differences in children performancein mathematics. The results showed (in contrast to what wasexpected in Hypothesis 4a) that teachers’ instructional supporthad no direct or indirect impact on children’s math skills. Thisresult suggests that, as early as in kindergarten, teachers havemore impact on children’s engagement in learning than on theiracademic achievement (see Stipek et al., 1998). The results mighthave been different later on at school when children participate toa greater extent in formal instruction in mathematics. The resultof the present study showing that teachers’ instructional supportdid not impact children’s math skills may, in fact, partly reflect thefact that there is a strong emphasis in Finnish kindergartens on aplay-centered approach and formal instruction in math is scant.However, instructional support was connected to task avoidancewhich was associated with children’s math skills.

Our next aim was to determine whether classroom differencesin children’s task-avoidant behavior would predict classroom dif-ferences in their math skills. As was expected (Hypothesis 4b), theresults showed that the higher the level of children’s task-avoidantbehavior in a particular classroom, the lower were their math skills,even after controlling for entry level math skills. Previous stud-ies (e.g., Aunola et al., 2003; Onatsu-Arvilommi et al., 2002) havefound that task-avoidant behavior has negative consequences forchildren’s later math performance. The present findings contributeto the previous literature by showing that task-avoidant behav-ior which is connected to teachers’ instructional practices, is alsoreflected in children’s math performance.

Results at the within-individual level showed, after controllingfor the between-level results concerning classroom differences,that task-avoidant behavior predicted children’s math skills. Thisfinding is in line with some previous findings (e.g., Aunola et al.,2003; Onatsu-Arvilommi et al., 2002) suggesting that children’smath performance is sensitive to task-avoidant versus task-focused behaviors. Further, less advanced pre-math skills predicted

Page 10

Author's personal copy

384 E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386

task-avoidant behavior suggesting that children’s task-avoidantbehavior develops partly as a result of previous skill development,and related feedback. We also found that boys were rated as show-ing more task avoidance than girls. This finding is consistent withsome previous studies (Midgley & Urdan, 1995; Onatsu-Arvilommi& Nurmi, 2000; Urdan et al., 1998) suggesting that boys may bemore likely to develop maladaptive achievement patterns thangirls. We have to keep in mind that the behavior being valued asdesirable in classrooms is considered to be more typical for girlsthan for boys and gender may influence teacher perceptions of chil-dren (Hamre & Pianta, 2001). The results also showed that the olderthe children at the beginning of their kindergarten year, the lesstask-avoidant behavior they showed and the better were their mathskills later on. This result concerning children’s age suggests thattask-avoidant behavior is partly a question of maturation. Olderchildren are more ready and responsive for instructional activitiesand are more able to concentrate on tasks. Similarly, Jordan et al.(2006) found that older children had better math skills at the endof the kindergarten year.

9.1. Practical implications

The results of the present study have also some practical impli-cations. The results suggest that kindergarten teachers should payparticular attention to the feedback they provide to children andthe quality of scaffolding and language they use in order to supportchildren’s engagement and focus on tasks. By giving individualizedand task-specific feedback instead of just general positive feedback(“right, well done, ok”) teachers would provide a basis for kinder-garteners’ adaptive achievement behaviors. Furthermore, teachersshould aspire to create learning environments where all studentscan feel successful and feel a sense of improvement. This meansthat teachers should be well aware of the skill level of childrenand provide tasks and activities that are appropriate to childrenand promote their learning in their zone of proximal develop-ment (Pressley, Hogan, Wharton-McDonald, Mistretta, & Ettenberg,1996). By providing optimally challenging tasks and activities,teachers may encourage children’s engagement and enthusiasm inlearning.

The development of adaptive achievement strategies during theearly school years is particularly important as they seem to formcumulative cycles with academic skills and school performance(Aunola et al., 2002; Onatsu-Arvilommi & Nurmi, 2000). Conse-quently, the findings of the present study also have consequencesfor teacher education and professional development. For example,teacher education should pay more attention to improving teach-ers’ awareness of characteristics that children bring to kindergartenthat may impact their success. Deeper understanding of the pro-cesses of classroom quality and children’s behavior may provideteachers with better abilities to foster high-quality relationshipswith different kinds of children, thereby promoting their schoolreadiness.

9.2. Limitations

This study has certain limitations that need to be considered.First, although we employed longitudinal data, the procedure wasnot cross-lagged longitudinal, as not all variables were measuredat each time point. Consequently, we cannot be sure whetherinefficient instruction is responsible for increased task-avoidantbehavior in classrooms, or whether children’s achievement-relatedbehavior might have influenced teachers’ instructional practices(see also, Nurmi, 2010). It is possible, for example, that teachersreact to children’s characteristics and adjust their practices accord-ing to children’s characteristics. Consequently, when interpretingthe results, we have to keep in mind that the associations can also

be driven by children. Second, as children’s task avoidance wasbased on teacher-reports, we cannot rule out the possibility thatthe rater-effect might explain some of the associations that werefound. Consequently, there is an evident need for future studiesin which children’s task-avoidant behavior is examined by usingindependent raters, such as observers. Third, the internal consis-tency of the math skills measured was somewhat low. Fourth, inthe present study we were not able to investigate whether gendermoderates the influence of classroom quality on task avoidance.Given that boys typically show more task avoidance than girls do,this is an important question that should be investigated in futurestudies. Finally, the results were found in a particular cultural andeducational setting, that is, in Finnish kindergartens. As there is asubstantial amount of variation in how kindergartens and primaryschools are organized, and the kind of instruction provided, thereis clearly a need to replicate these findings in other cultures andeducational settings. Despite the limitations, the present study pro-vided evidence on the mechanisms by which teacher practices mayaffect children’s task avoidance, which can be tested in the futurewith more nuanced methodologies and measurement approaches.

9.3. Conclusion

The results of the present study contribute to our understandingof the role that teachers’ instructional practices may play in chil-dren’s achievement-related behavior. In the classrooms in whichteachers show high levels of instructional support, children arerated as evidencing more task-focused behavior, which then alsocontributes to their math performance. The findings emphasizethe importance of teachers’ instructional practices in kindergartencontexts for the promotion of children’s preparedness to face chal-lenging learning situations and to engage in learning.

References

Ackerman, P. T., Anhalt, J. M., & Dykman, R. A. (1986). Arithmetic automatization fail-ure in children with attention and reading disorders: Associations and sequelae.Journal of Learning Disabilities, 19, 222–232.

Anderman, E. M., Eccles, J. S., Yoon, K. S., Roeser, R., Wigfield, A., & Blumenfeld, P.(2001). Learning to value mathematics and reading: Relations to mastery andperformance-oriented instructional practices. Contemporary Educational Psy-chology, 26, 76–95.

Aunola, K., Leskinen, E., Lerkkanen, M.-K., & Nurmi, J.-E. (2004). Developmentaldynamics of math performance from kindergarten to grade 2. Journal of Edu-cational Psychology, 96, 699–713.

Aunola, K., Leskinen, E., & Nurmi, J.-E. (2006). Developmental dynamics betweenmathematical performance, task-motivation, and teachers’ goals during thetransition to primary school. British Journal of Educational Psychology, 76, 21–40.

Aunola, K., Nurmi, J.-E., Lerkkanen, M.-K., & Rasku-Puttonen, H. (2003). The roles ofachievement-related behaviors and parental beliefs in children’s mathematicalperformance. Educational Psychology, 23, 403–421.

Aunola, K., Nurmi, J.-E., Niemi, P., Lerkkanen, M.-K., & Rasku-Puttonen, H. (2002).Developmental dynamics of achievement strategies, reading performance, andparental beliefs. Reading Research Quarterly, 37, 310–327.

Aunola, K., & Räsänen, P. (2007). The 3-minutes Basic Arithmetic Test. Unpublishedtest material. Jyväskylä, Finland.

Bogner, K., Raphael, L., & Pressley, M. (2002). How grade 1 teachers motivate literateactivity by their students. Scientific Studies of Reading, 6, 135–165.

Butkowsky, I. S., & Willows, D. M. (1980). Cognitive-motivational characteristics ofchildren varying in reading ability: Evidence for learned helplessness in poorreaders. Journal of Educational Psychology, 72, 408–482.

Caprara, G. V., Barbanelli, C., Steca, P., & Malone, P. S. (2006). Teachers’ self-efficacybeliefs as determinants of job satisfaction and student’s academic achievement:A study at the school level. Journal of School Psychology, 44, 473–490.

Chard, D., Clarke, B., Baker, S., Otterstedt, J., Braun, D., & Katz, R. (2005). Using mea-sures of number sense to screen for difficulties in mathematics: Preliminaryfindings. Assessment for Effective Intervention, 30, 3–14.

Clarke, B., & Shinn, M. (2004). A preliminary investigation into the identificationand development of early mathematics curriculum-based measurement. SchoolPsychology Review, 33, 234–248.

De Corte, E. (1995). Fostering cognitive growth: A perspective from research onmathematics learning and instruction. Educational Psychologist, 30, 37–46.

Desoete, A., & Grégoire, J. (2006). Numerical competence in young children and inchildren with mathematics learning disabilities. Learning and Individual Differ-ences, 16, 351–367.

Page 11

Author's personal copy

E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386 385

Desoete, A., Roeyers, H., & Buysse, A. (2001). Metacognition and mathematical prob-lem solving in Grade 3. Journal of Learning Disabilities, 34, 435–449.

Diener, C. I., & Dweck, C. S. (1978). An analysis of learned helplessness: Continuouschanges in performance, strategy, and achievement cognitions following failure.Journal of Personality and Social Psychology, 36, 451–462.

Dolezal, S. E., Welsh, L. M., Pressley, M., & Vincent, M. M. (2003). How nine third-grade teachers motivate student academic engagement. The Elementary SchoolJournal, 103, 239–269.

Donlan, C. (1993). Basic numeracy in children with specific language impairment.Child Language Teaching and Therapy, 9, 95–104.

Dowker, A. (1998). Individual differences in normal arithmetical development. In C.Donlan (Ed.), The development of mathematical skills (pp. 275–302). East Sussex,U.K.: Psychology Press.

Downer, J. T., Rimm-Kaufman, S. E., & Pianta, R. C. (2007). How do classroom condi-tions and children’s risk for school problems contribute to children’s behavioralengagement in learning? School Psychology Review, 36, 413–432.

Duncan, T. E., Duncan, S. C., Alpert, A., Hops, H., Stoolmiller, M., & Muthén, B. (1997).Latent variable modeling of longitudinal and multilevel substance use data.Multivariate Behavioral Research, 32, 275–318.

Dweck, C. S. (1986). Motivational processes affecting learning. American Psychologist,41, 1040–1048.

Elomäki, T., Huolila, R., Poskiparta, E., & Saranpää, P. (1999). (A group-based assess-ment of school readiness) Kouluvalmiuden arviointi ryhmässä. Turku, Finland:Publications of the Social Service Centre for the City of Turku.

Emmer, E. T., & Stough, L. (2001). Classroom management: A critical part ofeducational psychology, with implications for teacher education. EducationalPsychologist, 36, 103–112.

Fazio, B. B. (1996). Mathematical abilities of children with specific language impair-ment: A 2-year follow-up. Journal of Speech, Language and Hearing Research, 39,839–849.

Geary, D. C. (1993). Mathematical disabilities: Cognitive, neuropsychological, andgenetic components. Psychological Bulletin, 114, 345–362.

Hamre, B. K., & Pianta, R. C. (2001). Early teacher–child relationships and the trajec-tory of children’s school outcomes through eight grade. Child Development, 72,625–638.

Hamre, B. K., & Pianta, R. C. (2005). Can instructional and emotional support in thefirst grade classroom make a difference for children at risk of school failure?Child Development, 76, 949–967.

Hecht, S. A., Torgesen, J. K., Wagner, R. K., & Rashotte, C. A. (2001). The relationbetween phonological processing abilities and emerging individual differencesin mathematical computation skills: A longitudinal study from second to fifthgrades. Journal of Experimental Child Psychology, 79, 192–227.

Heck, R. H., & Thomas, S. L. (2009). An introduction to multilevel modeling techniques.New York, NY: Routledge.

Howes, C., Burchinal, M., Pianta, R. C., Bryant, D., Early, D., Clifford, R., et al. (2008).Ready to learn? Children’s pre-academic achievement in pre-Kindergarten pro-grams. Early Childhood Research Quarterly, 23, 27–50.

Hytönen, J., Krokfors, L., Talts, L., & Vikat, M. (2003). What educational objectives areconsidered important by preschool teachers in Helsinki and Tallinn? Trames.Journal of Humanities and Social Sciences, 7, 257–268.

Jones, E. E., & Berglas, S. (1978). Control of attributions about the self through self-handicapping: The appeal of alcohol and the rate of underachievement. Person-ality and Social Psychology Bulletin, 4, 200–206.

Jordan, N. C., Hanich, L. B., & Kaplan, D. (2003). A longitudinal study of mathematicalcompetencies in children with specific mathematics difficulties versus chil-dren with comorbid mathematics and reading difficulties. Child Development,74, 834–850.

Jordan, N. C., Kaplan, D., Nabors Oláh, L., & Locuniak, M. N. (2006). Number sensegrowth in kindergarten: A longitudinal investigation of children at risk for math-ematics difficulties. Child Development, 77, 153–175.

Kinnunen, R., Lehtinen, E., & Vauras, M. (1994). Matemaattisen taidon arviointi. InM. Vauras, E. Poskiparta, & P. Niemi (Eds.), Kognitiivisten taitojen arviointi koulu-tulokkailla ja 1. luokan oppilailla (Assessment of cognitive skills of first graders)(pp. 55–76). Finland: Center for Learning Research, University of Turku.

Koponen, T., Aunola, K., Ahonen, T., & Nurmi, J.-E. (2007). Cognitive predictors ofsingle-digit and procedural calculation skills and their covariation with readingskill. Journal of Experimental Child Psychology, 97, 220–241.

La Paro, K. M., Pianta, R. C., & Stuhlman, M. (2004). The Classroom Assessment ScoringSystem: Findings from the prekindergarten year. The Elementary School Journal,104, 409–426.

Lau, S., & Nie, Y. (2008). Interplay between personal goals and classroom goal struc-tures in predicting student outcomes: A multilevel analysis of person-contextinteractions. Journal of Educational Psychology, 100, 15–29.

Lerkkanen, M. -K., Niemi, P., Poikkeus, A. -M., Poskiparta, M., Siekkinen, M., & Nurmi,J. -E., (2006). The First Steps Study. Unpublished data. Finland: University ofJyväskylä.

Little, R. J. A. (1988). A test of missing completely at random for multivariatedata with missing values. Journal of the American Statistical Association, 83,1198–1202.

Mashburn, A. J., Pianta, R. C., Hamre, B. K., Downer, J. T., Barbarin, O. A., Bryant, D.,et al. (2008). Measures of classroom quality in prekindergarten and children’sdevelopment of academic, language, and social skills. Child Development, 79,732–749.

McGraw, K. O., & Wong, S. P. (1996). Forming inferences about some intraclasscorrelation coefficients. Psychological Methods, 1, 30–46.

McWilliam, R. A., Scarborough, A. A., & Kim, H. (2003). Adult interactions and childengagement. Early Education and Development, 14, 7–27.

Midgley, C., & Urdan, T. (1995). Predictors of middle school students’ use of self-handicapping strategies. Journal of Early Adolescence, 15, 389–411.

Muthén, L., & Muthén, B. O. (1998–2009). Mplus users’ guide and Mplus version 5.0.Retrieved from http://www.statmodel.com/index2.html.

National Board of Education. (2000). Esiopetuksen opetussuunnitelman perusteet2000 [Core curriculum for preschool education in Finland]. Helsinki, Finland:National Board of Education.

Nicholls, J. G., Cheung, P. C., Lauer, J., & Patashnick, M. (1989). Individual differencesin academic motivation: Perceived ability, goals, beliefs, and values. Learningand Individual Differences, 1, 63–84.

Nolen, S. B., & Haladyna, T. M. (1990). Personal and environmental influences onstudents’ beliefs about effective study strategies. Contemporary Educational Psy-chology, 15, 116–130.

Nurmi, J. E. (2010). Do students have an impact on teachers’ instruction and classroompractice? Manuscript submitted for publication.

Nurmi, J.-E., & Aunola, K. (2000). The Task-avoidance scale: Reliability and validityinformation. Finland.: University of Jyväskylä.

Nurmi, J.-E., Onatsu, T., & Haavisto, T. (1995). Underachievers’ cognitive andbehavioral strategies: Self-handicapping at school. Contemporary EducationalPsychology, 20, 188–200.

Onatsu, T., & Nurmi, J.-E. (1995). The Behavior Strategy Rating Scale. Helsinki: Univer-sity of Helsinki, Finland.

Onatsu, T., & Nurmi, J.-E. (1997). Task-avoidance scale: Reliability and validity infor-mation. Helsinki, Finland: University of Helsinki.

Onatsu-Arvilommi, T., & Nurmi, J.-E. (2000). The role of task-avoidant and task-focussed behaviors in the development of reading and mathematical skillsduring the first school year: A cross-lagged longitudinal study. Journal of Educa-tional Psychology, 92, 478–491.

Onatsu-Arvilommi, T., Nurmi, J.-E., & Aunola, K. (2002). The development of achieve-ment strategies and academic skills during the first year of primary school.Learning and Instruction, 12, 509–527.

Pajares, F., & Miller, M. D. (1994). Role of self-efficacy and self-concept in mathe-matical problem solving: A path analysis. Journal of Educational Psychology, 86,193–203.

Pakarinen, E., Lerkkanen, M.-K., Poikkeus, A.-M., Kiuru, N., Siekkinen, M., Rasku-Puttonen, H., et al. (2010). A validation of the Classroom Assessment ScoringSystem in Finnish kindergartens. Early Education & Development, 21, 95–124.

Perry, K. E., Donohue, K. M., & Weinstein, R. S. (2007). Teaching practices and thepromotion of achievement and adjustment in first grade. Journal of School Psy-chology, 45, 269–292.

Pianta, R. C., La Paro, K. M., & Hamre, B. K. (2008). The Classroom Assessment ScoringSystem. Manual, Pre-K. Brookes: Baltimore, MD.

Pianta, R. C., La Paro, K. M., Payne, C., Cox, M. J., & Bradley, R. (2002). The relation ofkindergarten classroom environment to teacher, family and school characteris-tics and child outcomes. Elementary School Journal, 102, 225–238.

Pintrich, P. R., & De Groot, E. A. M. (1990). Motivational and self-regulated learningcomponents of classroom academic performance. Journal of Educational Psychol-ogy, 82, 33–40.

Ponitz, C., Rimm-Kaufman, S. E., Grimm, K. J., & Curby, T. W. (2009). Kindergartenclassroom quality, behavioral engagement, and reading achievement. SchoolPsychology Review, 38, 102–120.

Pressley, M., Hogan, K., Wharton-McDonald, R., Mistretta, J., & Ettenberg, S.(1996). The challenges of instructional scaffolding: The challenges of instruc-tion that supports student thinking. Learning Disabilities Research & Practice, 11,38–146.

Räsänen, P., Salminen, J., Wilson, A. J., Aunio, P., & Dehaene, S. (2009).Computer-assisted intervention for children with low numeracy skills. CognitiveDevelopment, 24, 450–472.

Rautanen, R. (2007). (School attendance) Lasten koulunkäynti. In Suomalainen lapsi2007. Helsinki: Statistics Finland., pp. 187–205

Resnick, L. B. (1989). Developing mathematical knowledge. American Psychologist,44, 162–169.

Rimm-Kaufman, S. E., Curby, T. W., Grimm, K. J., Nathanson, L., & Brock, L. L. (2009).The contribution of children’s self-regulation and classroom quality to children’sadaptive behaviors in the kindergarten classroom. Developmental Psychology, 45,958–972.

Rimm-Kaufman, S. E., La Paro, K. M., Downer, J. T., & Pianta, R. C. (2005). The contri-bution of classroom setting and quality of instruction to children’s behavior inkindergarten classrooms. The Elementary School Journal, 105, 377–394.

Ryan, R. M., & Grolnick, W. S. (1986). Origins and paws in the classroom: Self-reportand projective assessments of individual differences in children’s perceptions.Journal of Personality and Social Psychology, 50, 550–558.

Salonen, P., Lepola, J., & Niemi, P. (1998). The development of first graders’ read-ing skill as a function of pre-school motivational orientation and phonemicawareness. European Journal of Psychology of Education, 2, 155–174.

Stephenson, K., Parrila, R., Georgiou, G., & Kirby, J. (2008). Effects of home literacy,parents beliefs and children’s task-focused behavior on emergent literacy andword reading skills. Scientific Studies of Reading, 12, 24–50.

Stipek, D. J., Feiler, R., Byler, P., Ryan, R., Milburn, S., & Miller Salmon, J. (1998).Good beginnings: What difference does the program make in preparing youngchildren for school? Journal of Applied Developmental Psychology, 19, 41–66.

Turner, J. C., Meyer, D. K., Midgley, C., & Patrick, H. (2003). Teacher discourseand sixth graders’ reported affect and achievement behaviors in two high-

Page 12

Author's personal copy

386 E. Pakarinen et al. / Early Childhood Research Quarterly 26 (2011) 376–386

mastery/high-performance mathematics classrooms. The Elementary SchoolJournal, 103, 357–382.

Turner, J. C., Midgley, C., Meyer, D. K., Gheen, M., Anderman, E. M., Yongjin, K., et al.(2002). The classroom environment and students’ reports of avoidance strate-gies in mathematics: A multimethod study. Journal of Educational Psychology, 94,88–106.

Urdan, T., Midgley, C., & Anderman, E. M. (1998). The role of classroom goal structurein students’ use of self-handicapping strategies. American Educational ResearchJournal, 35, 101–122.

Valentine, J. C., DuBois, D. L., & Cooper, H. (2004). The relation between self-beliefsand academic achievement: A meta-analytic review. Educational Psychologist,39, 111–133.

Wharton-McDonald, R., Pressley, M., & Mistretta-Hampston, J. (1998). Literacyinstruction in nine first-grade classrooms: Teacher characteristics and studentachievement. The Elementary School Journal, 99, 101–128.

Wigfield, A., Eccles, J. S., Schiefele, U., Roeser, R. W., & Davis-Kean, P. (2006). Devel-opment of achievement motivation. In N. Eisenberg, W. Damon, & R. M. Lerner(Eds.), Handbook of child psychology: Vol. 3, Social, emotional, and personalitydevelopment (6th ed., pp. 933–1002). Hoboken, NJ: John Wiley & Sons.

Wigfield, A., & Meece, J. L. (1988). Math anxiety in elementary and secondary schoolstudents. Journal of Educational Psychology, 80, 210–216.

Yates, G. C., & Yates, S. M. (1990). Teacher effectiveness research: Towards describinguser-friendly classroom instruction. Educational Psychology, 10, 225–238.