The epistemic structure of instructional tasks and cognitive performance in secondary 3 mathematics and english classes in Singapore: A multi-level SEM approach.

For the past decade or more Singapore has embarked on an ambitious program of pedagogical improvement despite considerable success in international assessments, including TIMMS and PISA, especially in Mathematics and Science. In this paper we will examine one possible source of this successful record -- the nature of the instructional tasks that teachers set for their students in Secondary 3 Mathematics and Science.

The paper will address the following research questions:

1. What are instructional tasks? How might they be conceptually modelled?
2. How might we understand the relationship between instructional tasks, on the one hand, and collateral instructional practices (lesson organization, instructional strategies, classroom talk, classroom management, classroom learning environment), on the other?
3. How do instructional tasks vary by subject?
4. What kinds of knowledge -- factual, procedural, conceptual, epistemic -- do instructional tasks in Secondary 3 Mathematics and English focus on? How are they related to each other? What is their relative importance?
5. What kind of domain-specific knowledge practices are associated with these tasks? How are they related to each other? WHat is their relative importance?
6. How cognitively demanding or complex are these tasks?
7. What is the relationship between epistemic focus, knowledge practice and cognitive complexity?
8. What impact do these aspects of instructional tasks, individually and jointly, have on student cognitive performance in Mathematics and English?
9. What are the implications of these findings for the ability of the Singapore system to improve the quality of teaching and learning in Singaporean classrooms? In particular, what are the implications of these findings for understanding the role of high stakes national assessments of the kind used in Singapore in promoting or inhibiting instructional improvement?

We will address these questions with a data set drawn from a repeated measures survey of Secondary 3 students (n=4000) in Mathematics and English based on a stratified national sample of secondary schools (n=32) in Singapore and all the Secondary 3 classes in Mathematics and English in them (n=120) in 2010. Half the classes in each school focused on Mathematics instruction and learning, the other half on English. Half the students in each Mathematics or English class focused on individual level characteristics; the other half focused on instructional practices at the classroom level. In addition, we also administered a one hour assessment in either Mathematics and English to half the student sample as appropriate.

Statistical analysis includes descriptive statistics, exploratory and confirmatory factor analysis and multi-level SEM modelling.