Average Achievement (Grade 4)

The 2023 cycle marked the full transition to a computer-based assessment system, incorporating an adaptive booklet rotation to balance the proportion of more and less difficult booklets for countries with very high or very low expected average achievement. The 2023 assessment was carefully designed using a balanced incomplete block design, adhering to industry standards, and analyzed with state-of-the-art psychometric models. Consequently, the TIMSS 2023 science achievement results for all 58 countries and 5 benchmarking participants are reported on the same TIMSS fourth-grade science scale. The TIMSS fourth-grade science achievement scale was established in TIMSS 1995, based on the achievement data across all participating countries, treating each country equally. A centerpoint of 500 was set to correspond to the mean of overall achievement in 1995, with 100 points set to correspond to the standard deviation. Students’ achievement is placed on the scale with each subsequent TIMSS cycle, most recently for TIMSS 2023.

Exhibit 2.1.1 reports the average science achievement estimates at the fourth grade for each of the 58 TIMSS 2023 participating countries and the 5 benchmarking entities. The exhibit also presents the corresponding confidence intervals, and standard deviations for each country, together with the associated standard errors for these estimates (given in parentheses). Additionally, it includes a visualization of the score distribution, illustrating percentiles of the distribution and the confidence interval of the average achievement. The exhibit includes an international average estimate of the mean and a confidence interval based on the 58 countries that participated in TIMSS 2023, with each country contributing equally. The standard error of this estimate reflects the variance of the estimated international average achievement based on these countries.

The exhibit is interactive and can be sorted by different criteria, with average achievement used as the default. Sorted by average achievement, the exhibit shows that Singapore had the highest estimated average achievement, followed by Korea, then by Chinese Taipei and Türkiye who performed similarly on average. Average achievement is high in these countries. For example, the estimated average achievement in Korea is 583 and the 95% confidence interval ranges from 578 to 588. The confidence interval provides a score range around the estimated mean that illustrates the uncertainty in this estimate, which is based on a sample of students taking the assessment in a selection of classrooms within a sample of schools in each country. Intervals constructed in this way around the estimated average have a 95% chance to include the actual average of the country’s achievement.

There is considerable variability in achievement not only across participants, but also within each country, as indicated by the standard deviation, the interquartile range, and the difference between the 10th and 90th percentiles of the achievement distribution shown in the table. Therefore, it would be incorrect to assume that (all) students in any country achieve higher scores than students in another country. While the average achievement of one country may be higher than the average achievement in another country, each country has some lower-achieving students and some higher-achieving students.

While examining the average achievement across countries can provide an initial impression of achievement in participating countries, the data collected by TIMSS 2023 and reported here is more comprehensive. TIMSS also provides information about the range of achievement observed in each country, and in later sections about associations with contextual factors, which provides a deeper understanding of student performance.