Elsevier

Journal of School Psychology

Volume 42, Issue 5, September–October 2004, Pages 385-401
Journal of School Psychology

Longitudinal effects of preschool behavioral styles on early adolescent school performance

https://doi.org/10.1016/j.jsp.2004.05.002Get rights and content

Abstract

Early behavioral problems and temperament as well as student–teacher relationships each have been shown to be important predictors of school performance. This longitudinal study investigated the relations between children's preschool behaviors and their early adolescent student–teacher relationships and academic achievement. It was hypothesized that preschool behaviors would predict school performance in early adolescents. Results showed that children who were more prone to anxiety as preschoolers tended to have the highest grades as young adolescents. Additionally, children with poor student–teacher relationships, as characterized by dependent and conflictual relationships, had lower grades in school. Thus, in considering the importance of a “child×environment” model for school performance, it appears that it is both the child variable of anxiety and the perhaps environmental variable of student–teacher relationships that are related to school performance.

Introduction

Children's success during their school years is gauged by their competency and behavior in school. Doing well in school is important in building children's self-esteem and feelings of competency (Guerin, Gottfried, Oliver, & Thomas, 1994), and higher academic achievement decreases children's risk of delinquency later in life (Caprara, Barbaranelli, Pastorelli, Bandura, & Zimbardo, 2000). From a developmental perspective, early problem behaviors and temperamental tendencies may represent an early stage in a life-course process (Elder, 1994) that contributes eventually to problems in academic performance and school relationships. Thus, evaluating information about preschool problem behaviors and temperament may provide an important key to understanding problems in the academic arena.

There is an extensive body of research pertaining to early temperament and behavior problems (Guerin, Gottfried, & Thomas, 1997), spanning the preschool period through middle school (e.g., Caspi et al., 1995, Guerin et al., 1994, Rende, 1993, Teglasi & MacMahon, 1990). Some of this research has focused on behaviors specific to the school environment and has used teacher ratings as outcome measures (Alexander, Entwisle, & Kabbani, 2001), and consistently a relation between difficult temperament and behavior problems has emerged. A number of studies on middle-school and high-school students have demonstrated a similar relation between behavior and academic performance (Dishion, 1990, Parker & Asher, 1987, Patterson et al., 1990, Tremblay et al., 1992, Wentzel, 1993). However, research has been more sparse in examining the relations between very young children's early personality and problem behaviors and their subsequent achievement in a school setting (Alexander et al., 2001, Guerin et al., 1994). Children's behaviors in school are a significant determinant of academic achievement (Guerin et al., 1994, McClelland et al., 2000), and therefore it is likely that early temperamental qualities should be a developmental stepping stone to school relationships and academic performance. Prediction of children's school performance from early behavioral factors may help us to assist children in their learning processes in the classroom to augment academic performance. This enhanced “goodness-of-fit” (Alexander et al., 2001) between child personality and the school environment may be essential for many students to make a success of their schooling. Thus, early behavior problems as well as interactions while in school may both be important predictors of school achievement. These behaviors were examined in the present study.

Behavior problems, such as anxiety, aggression, and attention, have been noted as related to school performance in a host of studies (e.g., Gumora & Arsenio, 2002, McDonald, 2001, Normandeau & Guay, 1998, Palisin, 1986, Wentzel, 1993). In addition, certain behavioral styles, or temperament traits, have been identified as correlated with school performance, especially activity, distractibility, persistence, and intensity (Martin, 1989, Nelson et al., 1999). Temperament typically is viewed as reflecting inherent characteristics of the child that are modified by interactions with the environment. This “child×environment” model (Garmezy et al., 1984, Ladd et al., 1999) is useful for understanding that the way a child adjusts to school depends partly on the characteristics he or she brings to the situation and partly on the way in which those characteristics “fit” with the particular environment provided by the teacher and the school. It is important to assess simultaneously child characteristics (the “child” part of the model) and student–teacher relationships (the “school environment” part of the model) to determine their overall impact on school performance scores.

The theory of manifest anxiety as drive (Spielberger, 1966) suggests that high anxiety will affect performance differently for individuals depending on their intelligence. In this view, high levels of anxiety will be harmful to the academic performance of individuals with lower intelligence. Conversely, high levels of anxiety will promote academic achievement in individuals of greater intelligence. Support for this theory was found in a study of 11th and 12th grade Canadian students (McCann & Meen, 1984), which showed that although anxiety did not correlate with academic performance when the sample was taken as a whole, anxiety was correlated with academic performance when the sample was divided based on intelligence. These authors showed that anxiety was correlated with better grades for students who scored higher on intelligence tests, and correlated with lower grades for students who scored lower on intelligence tests, suggesting that anxiety may either be facilitative or debilitative, depending on the individual's intelligence.

The effects of anxiety directly related to test performance on academic performance have received a significant amount of attention (Ashcraft, 2002, Gumora & Arsenio, 2002, McDonald, 2001). For instance, in a study of middle school, middle-to-upper-middle-class students, anxiety was part of a scale assessing negative affect while working on school tasks. This scale was shown to correlate significantly with lower GPA scores (Gumora & Arsenio, 2002). However, less is known about how earlier, general anxiety affects school performance in children. One study of Indian adolescents (Sharma, 1970) found that the relation between anxiety assessed under nonstress conditions and school achievement 4 months later was curvilinear, supporting the inverted-U hypothesis (Yerkes & Dodson, 1908) which would suggest that adolescents who tend to be mildly anxious perform better in school than those who are not anxious and than those who are more severely anxious. Consistent with this, a study of second through fifth graders showed that children rated as highly anxious (2 standard deviations above the average), through both self-report and teacher evaluations, were rated by teachers as having low academic performance in comparison to their nonanxious classmates (Strauss, Frame, & Forehand, 1987). Additionally, a study of nonclinical anxiety in 11-year-old British children found that level of general anxiety was not correlated with performance on the eleven-plus exam, although moderately high anxiety participants were found to have done well on the exam (Eady, 1999). However, a study of kindergarten children showed that teacher-rated anxious-withdrawn behaviors significantly but moderately predicted lower school achievement in first grade (Normandeau & Guay, 1998). Thus, the impact of anxious behaviors on school performance is still not clear. The present study expands the mostly cross-sectional literature by examining parent ratings of preschool anxiety and grade school academic performance approximately 6 to 8 years later.

Aggression also has been examined in relation to school performance. Aggressive children may exhibit behavioral problems in the classroom, making learning more difficult for them. Studies have shown that both prosocial and antisocial behaviors are related to classroom grades even after accounting for a number of other factors, such as sex, family structure, ethnicity, (Wentzel, 1993), and IQ (Normandeau & Guay, 1998, Wentzel, 1993). A study of middle school girls in Jamaica also demonstrated a relation between aggressiveness and poor academic performance (Walker, Grantham-McGregor, Himes, Williams, & Duff, 1998), and teacher-rated aggression in first grade was shown to predict increased drop-out rates (Ensminger & Slusarcick, 1992). It is difficult to determine whether aggression leads to poor academic performance or vice versa when studies examine children who have already begun school (Coie & Dodge, 1998), although Dishion's (1990) analysis suggested that the former is more accurate. The present study assesses aggression rated at age 5 years. If it predicts academic performance in early adolescence, then it is likely that the aggressive behaviors preceded poor academic performance rather than that performance in school caused children to become frustrated and to behave aggressively. This could occur through a variety of channels, including aggressive children being disliked by peers and teachers, becoming disenchanted at school, and therefore performing more poorly academically (Rubin, Bukowski, & Parker, 1998).

Finally, attention problems have been associated with later poor school achievement in clinical samples (e.g., Barriga et al., 2002, Fergusson & Horwood, 1995) as well as nonclinical samples (e.g., DeJong, 1993, Palisin, 1986). For instance, impaired attention and neurological signs of poor attention have been correlated with poor school achievement in male adolescents (Schonfeld, Shaffer, & Barmack, 1989). More generally, attentional capacity is part of a set of learning behaviors (Yen, Konold, & McDermott, 2004) determining what a student will perceive and learn in the classroom (Renninger, 1998). Children with longer attention spans are likely to be better problem solvers (Palisin, 1986).

In addition to the specific problem behaviors above, temperament may also provide an early means by which to identify children at risk for problems in school. Children possess unique temperaments that allot them a special way of adapting and reacting to situations. Understanding and accepting these differences can assist schoolteachers and other school officials to create innovative ways to help children adapt socially and academically to the classroom environment so that there is a better child–environment fit. In a number of studies that investigated temperament's role in school performance, temperament has been found to be a significant predictor of later academic achievement (Martin et al., 1988, Martin et al., 1994, Nelson et al., 1999). In one study (Nelson et al., 1999), parent ratings of children's 5-year-old temperaments, specifically task persistence, emotional intensity, and activity, were predictive of poor school performance and problem behaviors almost 4 years later. Similarly, Miller (2000) found that preschool intensity and distractibility were related to teacher-assigned grades in fourth and seventh grades. Activity specifically may reflect an impulsivity that prohibits learning and impedes performance (Palisin, 1986). A combination of early temperament ratings and problem behaviors from preschool-aged children should increase our ability to predict potential problems later in school.

The teacher–student relationship should provide an essential aspect of the learning environment for many children and should be an important variable in predicting academic success or failure. Similar measures as those already discussed for school achievement have been found to correlate with student–teacher relationships (Rubin et al., 1998). For instance, in a first grade sample of low-income children (Birch & Ladd, 1998), antisocial children had more dependent and conflictual relationships with their teachers and were less close to them. Additionally, a measure of asocial behavior that included anxiety was shown to predict increased dependent relationships between students and teachers (Birch & Ladd, 1998). A study of adolescents showed that a subset of aggressive children who were rated as rejected-aggressive were perceived by teachers as being more dependent (Wentzel & Asher, 1995). Although it is important to recognize that the relationship between teachers and students is bidirectional (Pianta, Hamre, & Stuhlman, 2003), the present study focuses on what the students bring to the classroom, in terms of early temperament and behaviors, that can affect or predict the relationships with their teachers.

Teacher's attitudes regarding students are influential in their judgments, grading, and behaviors toward students (Birch & Ladd, 1997, Hamre & Pianta, 2001, Martin et al., 1994). Teachers, on average, spend more waking hours with children than do parents, especially in elementary school years (Pianta, Steinberg, & Rollins, 1995). Therefore, teachers' grades rather than standardized test scores should be more related to student–teacher relationships because grades are more subjective and the student–teacher relationship may play a more important role in assigning grades (Alexander, Entwisle, & Dauber, 1993).

Several studies on very young children have demonstrated a link between student–teacher relationships and student scholastic performance. Studies have shown that teachers prefer to work in the classroom with children who are less active, less distractible, and more persistent (Martin, 1989). Teachers may have closer, warmer relationship with their students who are more pleasurable to work with and may offer more assistance when those students are having trouble with a task, may be more encouraging, and may have more patience with these students. These behaviors can increase children's confidence and academic self-concepts, which can, in turn, enhance their academic performance (Hamre & Pianta, 2001). In a study on kindergarten adjustment and achievement, Ladd et al. (1999) found that children's behaviors at the advent of kindergarten were important influences on the types of relationships they formed with their teachers, and these relationships impacted their scholastic achievement. Specifically, teacher–child closeness and teachers' emotional tone when interacting with the children were significantly related to school achievement.

In a kindergarten sample of low-income children, a dependent student–teacher relationship was found to correlate with poorer school performance (Birch & Ladd, 1997). Similarly, in a high-risk sample of preschoolers who were tested through second grade, Burchinal, Peisner-Feinberg, Pianta, and Howes (2002) showed that children who had developed closer relationships with their teachers scored higher on language and reading achievement tests over time. In a separate, nonhigh-risk sample, Hamre and Pianta (2001) demonstrated a negative relation between kindergarten student–teacher conflict and a composite of math and language achievement at grades 1 through 9. Early student–teacher closeness was not related to later achievement, suggesting that close relationships may be more important in a high-risk sample.

The present study examined the long-term effects of parent-rated child preschool temperament and problem behaviors on academic performance in early adolescence. If we can predict how children's early temperament and problem behaviors affect how well they do in school later in life, this information can be used to identify children who may need special help in learning how to interact with teachers and peers. Additionally, if the relational child–teacher unit predicts child outcome beyond that predicted by the child characteristics, then the importance of a focus on this relational unit in teaching paradigms is supported (Pianta et al., 2003). By considering students' individual differences, intervention and prevention programs can be more carefully tailored and their effectiveness should greatly improve.

This study had two sets of hypotheses, one regarding prediction of student–teacher relationships and one regarding prediction of school achievement. The first hypothesis was that children with 5-year-old ratings of high activity, intensity, and distractibility and low persistence would have poor student–teacher relationships in grade school. In addition, preschool attention and aggression problem behaviors at age 5 were hypothesized to predict more negative student–teacher relationships.

The second set of hypotheses was geared toward the prediction of academic performance from temperament variables, early problem behaviors, and student–teacher relationships. High preschool ratings of activity, intensity, and distractibility and low ratings of persistence were expected to negatively predict late elementary school academic achievement. It was also hypothesized that aggression and attention problems would predict lower school achievement, whereas higher anxiety would predict greater school achievement because the present sample is nonclinical and therefore anxiety levels were expected to be low to moderate, showing a linear relation with school achievement, and not high enough to cause decreased achievement, as predicted by Yerkes and Dodson's (1908) inverted-U hypothesis. Finally, children with closer student–teacher relationships were predicted to have higher academic achievement, and negative student–teacher relationships were expected to be related to poorer school performance.

Section snippets

Participants

Subjects for this study previously participated in a Play Study at age 5 years (DiLalla, 1998). Participants for this follow-up study were recruited from 146 children (78 boys and 68 girls) originally tested 6 to 8 years earlier at age 5. All children were from southern Illinois and surrounding regions. Twenty-seven (18%) addresses for participants could not be found. Fifty-one parents (43% of those with addresses) agreed to participate in the study, and of those, 45 teachers returned completed

Results

First, the effects of grade in school were assessed to determine whether children from younger grades differed from the older children. There was a significant effect for STRS Closeness [F(3,43)=9.15, p<.001], with children in grades 5 and 6 scoring higher than children in grades 7 and 8. However, when regression analyses (described below) were repeated for only children in grades 5 and 6, the results were the same. Analyses could not be repeated for the older children alone because there were

Discussion

Results of this study demonstrate that early general anxiety is related to later school performance, as are poor concurrent student–teacher relationships. This study contributes an unusual perspective by demonstrating that anxiety from age 5 is predictive of school performance as many as 6 to 8 years later, suggesting that this early behavior either is highly stable or that it has set in motion consequences in school that are maintained over this long period of time. This supports the original

Acknowledgments

This project was made possible by a senior honors thesis grant to the second author from the Southern Illinois University Office of Research and Development Administration, as well as grants to the first author from the Central Research Committee of the Southern Illinois University School of Medicine. The authors would like to thank all families and teachers who donated their time to participate in this project.

References (62)

  • C.-J. Yen et al.

    Does learning behavior augment cognitive ability as an indicator of academic achievement?

    Journal of School Psychology

    (2004)
  • T.M. Achenbach

    Manual for the child behavior checklist/4–18 and 1991 profile

    (1991)
  • T.M. Achenbach

    Manual for the teacher's report form and 1991 profile

    (1991)
  • K.L. Alexander et al.

    First-grade classroom behavior: Its short- and long-term consequences for school performance

    Child Development

    (1993)
  • K.L. Alexander et al.

    The dropout process in life course perspective: Early risk factors at home and school

    Teachers College Record

    (October, 2001)
  • M.H. Ashcraft

    Math anxiety: Personal, educational, and cognitive consequences

    Current Directions in Psychological Science

    (2002)
  • R.M. Baron et al.

    The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations

    Journal of Personality and Social Psychology

    (1986)
  • A.Q. Barriga et al.

    Relationships between problem behaviors and academic achievement in adolescents: The unique role of attention problems

    Journal of Emotional and Behavioral Disorders

    (2002)
  • S.H. Birch et al.

    Children's interpersonal behaviors and the teacher–child relationship

    Developmental Psychology

    (1998)
  • C. Bonjean et al.

    Sociological measurement: An inventory of scales and indices

    (1967)
  • G.V. Caprara et al.

    Prosocial foundations of children's academic achievement

    Psychological Science

    (2000)
  • A. Caspi et al.

    Temperamental origins of child and adolescent behavior problems: From age three to age fifteen

    Child Development

    (1995)
  • J.D. Coie et al.

    Aggression and antisocial behavior

  • P.F. deJong

    The relationship between students' behaviour at home and attention and achievement in elementary school

    British Journal of Educational Psychology

    (1993)
  • L.F. DiLalla

    Daycare, child, and family influences on preschoolers' social behaviors in a peer play setting

    Child Study Journal

    (1998)
  • T.J. Dishion

    The family ecology of boys' peer relations in middle childhood

    Child Development

    (1990)
  • S. Eady

    An investigation of possible correlation of general anxiety with performance in eleven-plus scores in year 6 primary school pupils

    Educational Psychology

    (1999)
  • G.H. Elder

    Time, human agency, and social change: Perspectives on the life course

    Social Psychology Quarterly

    (1994)
  • M.E. Ensminger et al.

    Paths to high school graduation or dropout: A longitudinal study of a first-grade cohort

    Sociology of Education

    (April, 1992)
  • D.M. Fergusson et al.

    Early disruptive behavior, IQ, and later school achievement and delinquent behavior

    Journal of Abnormal Child Psychology

    (1995)
  • N. Garmezy et al.

    The study of stress and competence in children: A building block for developmental psychopathology

    Child Development

    (1984)
  • Cited by (77)

    • Trait anxiety and adolescent's academic achievement: The role of executive function

      2021, Learning and Individual Differences
      Citation Excerpt :

      Alternatively, it could be that trait anxiety increased students' motivation to perform well on tasks, resulting in improved performance on academic measures. This possibility is supported by DiLalla et al. (2004) who reported that high anxiety in preschool was a significant predictor of academic success for adolescents. It could also be that anxiety may impair cognitive processes only for children with severe levels of anxiety.

    • The roles of anxious and prosocial behavior in early academic performance: A population-based study examining unique and moderated effects

      2018, Learning and Individual Differences
      Citation Excerpt :

      Among kindergarten students specifically, researchers have shown that anxious behavior is positively associated with loneliness, peer difficulties, hyperactivity, inattention, and aggressiveness, and negatively associated with perceived cognitive competence and kindergarten achievement (Coplan, Closson, & Arbeau, 2007; Duchesne et al., 2010; Heberle, Krill, Briggs-Gowan, & Carter, 2015; Normandeau & Guay, 1998; Pagani, Fitzpatrick, Archambault, & Janosz, 2010). Despite the largely negative implications of anxious behavior, some researchers have shown that in preschool and kindergarten it can have a null (e.g., DiPerna et al., 2007; Haapasalo et al., 2000) or positive (e.g., DiLalla, Marcus, & Wright-Phillips, 2004) association with academic achievement and attendance. These mixed findings highlight the need for additional research to help shed more light on this topic.

    • Invisible Disabilities and Inequality

      2023, Social Psychology Quarterly
    View all citing articles on Scopus
    View full text