5. Predictors of Numeracy Performance in NAPLAN

Predictors of Numeracy Performance in National Testing Programs

Authors and Source

  • Colin Carmichael, Amy MacDonald, and Laura McFarland-Piazza

  • Charles Sturt University, Australia

  • Article based on an exploratory study examining predictors of children's numeracy performance in the Australian National Assessment Program—Literacy and Numeracy (NAPLAN).

Ecological Theoretical Model

  • The study utilizes an ecological theoretical model that examines child, home, and school variables affecting NAPLAN numeracy performance.

Data Overview

  • Sample size: 2450 children from the Longitudinal Study of Australian Children (LSAC).

  • NAPLAN Year 3 numeracy assessment analyzed through bivariate relationships and linear regression models.

  • Results emphasize the importance of a supportive home-school relationship.

Introduction

  • Increased global emphasis on education accountability through national/international testing (Smeed, 2010).

  • Countries using testing include the UK and US (Polesel et al., 2012).

  • Australia introduced NAPLAN in 2008, increasing pressure on schools regarding student performance in literacy and numeracy.

Focus on Numeracy Education

  • The trend of "teaching to the test" potentially undermines understanding of the factors influencing numeracy performance, highlighted by LSAC as a means to explore contextual influences on outcomes.

  • LSAC follows two cohorts: Birth cohort (approx. 5000, 6-12 months old) and Kindergarten cohort (approx. 5000, aged 4.5-5 years).

Bronfenbrenner's Ecological Model of Child Development

  • Five nested systems:

    1. Microsystem: Settings in which a child interacts directly (family, school).

    2. Mesosystem: Interconnections among microsystems (e.g., parent–school interactions).

    3. Exosystem: Settings that indirectly impact the child (e.g., parents' workplaces).

    4. Macrosystem: Cultural patterns and ideologies affecting a child's environment.

    5. Chronosystem: Changes and transitions over time that affect child development.

  • Strong home-school mesosystem indicated as beneficial for children's school success (Garbarino et al., 1992).

Components of the Ecological Model in Study

  1. Child Factors: Characteristics like gender, IQ, etc.

  2. Home-Community Microsystem: Socioeconomic position (SEP), parental education, etc.

  3. School Microsystem: Teacher qualifications, learning environment, etc.

  4. Home-School Mesosystem: Nature of parent-school interactions.

Research Questions

  • How does the theoretical model predict children’s NAPLAN numeracy performance?

    • Home-school mesosystem's strength prediction.

    • Comparisons of home-community vs. school microsystems on influence.

    • Identification of primary predictors among child factors.

  • Sub-questions on collective influences of various factors.

Literature Review

  • Child Factors

    • Gender:

    • Notable gender differences in mathematics achievement (Atweh et al., 2012), with a widening gap favoring males; effect size of 0.1 (Vale et al., 2011).

    • Indigeneity:

    • Indigenous status not necessarily predicting poor performance; issues with standardized testing validity for Indigenous students (Meaney et al., 2012).

    • Culturally and Linguistically Diverse (CALD) Background:

    • Non-English backgrounds show varied performance; e.g., Chinese-Australian students outperforming peers (Zhao and Singh, 2010).

    • Diverse Needs:

    • 6-8% of students face difficulties; differences between learning difficulties and disabilities noted (Diezmann et al., 2012).

  • Home-Community Microsystem Factors:

    • Parental Education:

    • Maternal education strongly linked to competency (Liaw & Brooks-Gunn, 1994; Downer & Pianta, 2006).

    • Socioeconomic Status (SES):

    • Higher SES correlates with better academic outcomes (Sirin, 2005).

    • Quality of Home Environment:

    • Strong predictor for academic achievement (Sammons et al., 2008).

    • Parental Attitudes:

    • Expectations tied to performance (Balli, 1998; Halle et al., 1997).

  • School Microsystem Factors:

    • Teacher Characteristics:

    • Positive relationship with student achievement reported; teacher qualifications significant for mathematics instruction (Wayne & Youngs, 2003).

    • Learning Environment:

    • Quality articulated as crucial for learning (Dorman, 2001).

    • Attendance and Attitude Towards School/Math:

    • Positive correlation found with student achievement (Roby, 2004; Ladd et al., 2000).

  • Home-School Mesosystem Factors:

    • Parental Academic Support:

    • Involvement in homework tied to performance; direct help can negatively affect scores (Cooper et al., 2000; Patall et al., 2008).

    • School Involvement:

    • Active participation leads to positive academic outcomes (Marcon, 1999; Jeynes, 2005).

    • Parent-Teacher Interaction:

    • Significant for children's performance (Topor et al., 2010).

Methodology

  • Participants: 4331 children in Wave 3 of LSAC; focused on the 2450 children taking NAPLAN for the first time.

  • Data Collection: Conducted from April 2008 to April 2009, children aged 8.25 to 10 years.

Measures

  • Numeracy Outcome: Total numeracy score from NAPLAN, mean for Year 3 in Australia (396.9) and for participants (421.6).

  • Child Factors: Gender, IQ via matrix reasoning test, diverse needs.

  • Home-Community Factors: Parental education and SES; quality of home environment.

  • School Factors: Teacher qualifications, learning environment, attendance, attitude.

  • Home-School Factors: Academic support, communication practices.

Analysis Plan

  • Descriptive analyses to assess frequencies/means.

  • Bivariate relationships between predictor and NAPLAN scores using Pearson correlations and t-tests.

  • Multivariate analysis employed using linear regression models.

Results

  • Significant Bivariate Relationships:

    • Child factors (matrix reasoning, gender, etc.) show significant associations with numeracy performance.

    • Home-community factors relate to SES, parental education, hostility; home microsystem more predictive than school microsystem.

Multivariate Analysis Findings

  • The child factors model explained most variance (Adj. R² = 0.315).

  • Home-community microsystem contributed 10.5% variance; school microsystem contributed minimal variance (3.8%).

  • Home-school mesosystem raised predictive power with 11.0% explained variance.

Combined Model

  • Eight significant predictors identified across four systems contributing to NAPLAN numeracy performance:

    • IQ, family SES, provision of specialized services, gender, parental help with homework, teacher’s perception of parental involvement, attitude towards mathematics, parental education.

Discussion

Child Factors

  • Strong correlation between matrix reasoning scores and numeracy performance; IQ as a critical predictor.

  • Gender differences observed with girls performing lower than boys in numeracy (effect size of 0.23).

Home-Community Factors

  • SES and parental education as significant predictors, emphasizing disparity in academic outcomes based on social and economic status.

School Microsystem Factors

  • Limited predictive power of teacher characteristics; importance of positive teacher-child relationships and student attitudes on performance noted.

Home-School Mesosystem Factors

  • Communication quality and parental involvement shown to mediate child outcomes; mixed results regarding direct homework help.

Conclusion

  • Study presents comprehensive insights into the intricacies of factors impacting children's numerical performance.

  • Highlights the significance of familial support and socio-economic influences over school environment in academic achievements.

  • Calls for further investigation into methods of enhancing family-school connections to improve student outcomes, especially in contexts focusing on high-stakes standardized testing.