An analysis of TIMSS 2003

Between grades differences addressed as a selection-bias problem

How large-scale studies like TIMSS can be used to investigate effects of instructional quality

What can they learn from Singapore?