Open Data Portal

Citation Information

Type Conference Paper - 4th IEA International Research Conference
Title Changing populations in TIMSS cycles – An alternative approach to reporting trends
Publication (Day/Month/Year) 2010
Page numbers 0-0
City Gothenburg
Country/State Sweden
The IEA Trends in Mathematics and Science Studies (TIMSS) have taken place every four years beginning in 1995. Currently, data is available from 1995, 1999, 2004, and 2007. One of the intentions of TIMSS is to investigate the development over time of mathematics and science achievement in the participating countries. In the international reports, the mean achievement of countries is compared between cycles. The data allows this kind of comparison because it is based on representative samples. A critical point, however, is that these direct comparisons neglect changes in the targeted populations between the cycles. For example, the total and type of enrolment of students in the education systems changed in several countries to a considerable extent from one cycle to the next. One reason for these changes is often the goal to provide “education for all”, in keeping with commitments to the respective demand formulated by UNESCO (2009). Consequently, more children now enter the education systems in these countries, many from population groups that were not enrolled in school in the past. Usually, these students come from disadvantaged backgrounds. For these education systems, no change in observed mean achievement among students between two cycles, or even a small drop, might in fact point to an increase in mean achievement of the total age cohort. This paper reviews trend results from TIMSS, examining changes in the populations of specific participating education systems. Alternative reporting strategies are suggested that consider such changes.

Related studies

Hastedt, Dirk. "Changing populations in TIMSS cycles – An alternative approach to reporting trends." 4th IEA International Research Conference. Gothenburg, Sweden, July, 2010.
Copyright DataFirst, University of Cape Town