Mullis et al. (2006) claim that "PIRLS will provide a wealth of information that can be used
not only to improve the reading curriculum and instruction for younger students, but also help in
interpreting the results for 15-year-olds in PISA" (p. 102). However, there is no evidence that students´
achievement in PIRLS is related to literacy instruction (Shiel & Eivers, 2009). In addition, although
the relationship between students´ reading scores and some background variables at the student,
household, school and class within school levels have been investigated, more research is needed to
identify the effects of the factors associated with reading achievement. Thus, and in order to contribute
to evidence-based policy implications, we ran a secondary analysis of the PIRLS 2006 dataset for 20
EU countries to measure the effects of specific variables identified in previous investigations using
PIRLS data as well as variables identified in psycholinguistic research as predictors of reading
attainment. Specifically, this study addresses the following questions: 1) what is the relationship
between students´ scores on the purposes and processes of reading and related curriculum and
instructional coverage in the Program for International Reading Literacy Study (PIRLS) 2006
participating countries? And 2) which variables explain reading achievement in the PIRLS 2006
study? First, we ran correlations that indicate that the relationships between achievement and
curriculum and instructional emphasis are very weak. The results show that the curriculum information
on the PIRLS Encyclopaedia regarding curriculum emphasis as well as that reported by national
representatives contributes very little to explain reading achievement. Second, we used multilevel
analysis including three levels pertaining to: i) Student background characteristics, ii) Class
characteristics and iii) School characteristics. Findings indicate that our model, controlled for country
effects, explains 43% of the variance in students’ achievement and that the variables with the highest
impact on students´ overall reading score relate to home resources and practices, students´ pre -
reading knowledge and their attitudes and to school compositional effects. Country-level analysis
confirms that the variables identified for the model with the 20 countries as having a strong influence
on students’ reading achievement are also statistically significant in all countries, except one. These
findings have important policy implications as they show which factors can be addressed by policy
measures to improve students´ performance. For example, measures related to curriculum and
instruction and to social equity can be implemented by national governments to reduce educational
inequality. In sum, this report offers a detailed account of the reading research related to the
assessment of reading literacy, explains the methodological procedures used in the analysis to answer
the research questions and, after the presentation of the results, it discusses policy implications.
DE SOUSA LOBO BORGES DE ARAUJO Luisa;
DINIS MOTA DA COSTA Patricia;
2012-05-24
Publications Office of the European Union
JRC66894
978-92-79-21381-6 (print),
978-92-79-21382-3,
1018-5593 (print),
1831-9424 (online),
EUR 24949 EN,
OP LB-NA-24949-EN-C (print),
OP LB-NA-24949-EN-N (online),
https://publications.jrc.ec.europa.eu/repository/handle/JRC66894,
10.2788/75147 (print),
10.2788/77242 (online),
