TIMSS 1999


					CLOSE WINDOW

Sampling

The sampling objective for the TIMSS 1999 Video Study was to videotape a representative group of eighth-grade mathematics lessons in each participating country. In general, the sampling plan for the TIMSS 1999 Video Study followed the standards and procedures agreed to and implemented for the TIMSS 1999 assessments. That is, the school sample was required to be a Probability Proportionate to Size (PPS) sample. A PPS sample assigns probabilities of selection to each school proportional to the number of eligible students in the 8th-grade in schools countrywide. Then, one mathematics and/or one science 8th-grade class per school was sampled, depending on the subject(s) to be studied in each country.

Most of the participating countries drew separate samples for the Video Study and the assessments. For this and other reasons, the TIMSS 1999 assessment data cannot be directly linked to the video database.

Within the specified guidelines, the participating countries each developed their own strategy for obtaining a random sample of eighth-grade lessons to videotape for the TIMSS 1999 Video Study. Under the international sample design, the first sampling stage in each country was the sampling of schools. The school sampling frame in principle included all schools in the country that had eligible students in eighth grade. If originally selected schools declined to participate, in some countries replacement schools were randomly selected using the same procedure. The next stage following school selection was classroom selection. One mathematics and/or one science eighth-grade class per school was sampled.

Only one mathematics and/or science class was randomly selected within each school. No substitutions of teachers or class periods were allowed. The designated class was videotaped once, in its entirety, without regard to the particular topic being taught or type of activity taking place. The only exception was that teachers were not videotaped on days they planned to give a test for the entire class period. Teachers were asked to do nothing special for the videotape session, and to conduct the class as they had planned.

The TIMSS 1999 Video Study final sample included 638 eighth-grade mathematics lessons and 439 eighth-grade science lessons.

Country Number of mathematics lessons Number of science lessons

Australia 87 87

Czech Republic 100 88

Hong Kong SAR 100 NA

Japan 50¹ 95

Netherlands 78 81

Switzerland 140 NA

United States 83 88

¹Japanese mathematics data collected for the TIMSS 1995 Video Study were re-analyzed as part of the TIMSS 1999 Video Study.

TIMSS 1999 Video Study participating countries and their average score on the TIMSS 1995 and TIMSS 1999 mathematics assessments.

Country TIMSS 1995 mathematics score TIMSS 1999 mathematics score

Average Standard error Average Standard error

Australia* 519 3.8 525 4.8

Czech Republic 546 4.5 520 4.2

Hong Kong SAR 569 6.1 582 4.3

Japan 581 1.6 579 1.7

Netherlands* 529 6.1 540 7.1

Switzerland 534 2.7 - -

United States 492 4.7 502 4.0

* Nation did not meet international sampling and/or other guidelines in 1995. See Beaton et al. (1996) for details.
- Data not available.
NOTE: Rescaled TIMSS 1995 mathematics scores are reported here [Gonzales et al. 2000]. Switzerland did not participate in the TIMSS 1999 assessment.
SOURCE: Gonzales, P., Calsyn, C., Jocelyn, L., Mak, K., Kastberg, D., Arafeh, S., Williams, T., and Tsen, W. (2000). Pursuing Excellence: Comparisons of International Eighth-Grade Mathematics and Science Achievement from a U.S. Perspective, 1995 and 1999 (NCES 2001- 028). U.S. Department of Education. Washington, DC: National Center for Education Statistics.