Abstract:

In recent decades, more than 65 countries conduct international monitoring of student results - TIMSS and PISA. it is believed that they differ by the format of the tasks: TIMSS tests academic knowledge, while PISA investigates real-life problems. This is especially true for the math test.

In this study, we show by the example of 25 countries participating in both studies that 12 countries have average or below average PISA scores with high TIMSS scores, but countries with low TIMSS and high PISA were not found.

Using the example of one of these 12 countries, Russia, applying the methods of automatic text analysis, we evaluated the texts of the tasks of both tests 2009-2019 from the point of view of life problems.

The results showed that the PISA tasks are the most voluminous, it contains more words per task. At the same time, according to most formulas for text complicity, PISA tasks are the simplest for reading than TIMSS texts, and almost do not contain mathematical terms. Thus, the hypothesis that PISA is more focused on real-life problems has been confirmed, but it is also proved that PISA should be easier for students. However, TIMSS results are usually higher in most countries. To sum up, it could be suggested for education policy to include real-life problems in the curriculum to prepare students for further life.

Keywords:

PISA, TIMSS, real-life problems, math, text analysis.