TIMSS 2019 Results

**Scale Creation**

The TIMSS international achievement scales were developed on the basis of item response theory and latent regression population models using plausible values, with each student respondent being assigned five plausible values on each of the achievement scales.

Scale scores range from 0 to 1000.

Each international scale was established by TIMSS 1995 to have a midpoint of 500 located at the mean of the combined achievement distribution across all countries and 100 scale score points corresponding to one standard deviation of the combined distribution.

In order to measure trends over time, the results of all subsequent TIMSS assessments (1999, 2003, 2007, 2011, 2015, and 2019) were transformed into this metric.

With the dual administration mode (paperTIMSS and eTIMSS), the TIMSS scaling approach for 2019 involved additional psychometric analyses. Scaling the TIMSS 2019 achievement data proceeded in two phases:

- Calibrating the paperTIMSS and bridge data on the TIMSS 2019 achievement scales
- Linking the eTIMSS data to these scales via the bridge data

This process enabled the eTIMSS and paperTIMSS achievement results to be reported on the same achievement scale in each grade and subject.

**List of Achievement Scales**

- Mathematics (Grade 4 and Grade 8)
- Science (Grade 4 and Grade 8)
- Content domains
- Mathematics
- Grade 4: number, measurement and geometry, data
- Grade 8: number, algebra, geometry, data and probability

- Science
- Grade 4: life science, physical science, earth science
- Grade 8: biology, chemistry, physics, earth science

- Mathematics
- Cognitive domains (Mathematics and Science, [Grades 4 and 8])
- Knowing
- Applying
- Reasoning

Overall achievement results were reported in terms of the percentages of students reaching the following four benchmarks:

- Advanced International Benchmark (625)
- High International Benchmark (550)
- Intermediate International Benchmark (475)
- Low International Benchmark (400)

The descriptions of the competencies of students reaching each of these benchmarks were determined by means of a scale-anchoring process.

**Scale Creation**

Many of the TIMSS 2019 context questionnaire items were developed to be combined into scales measuring a single underlying latent construct.

Scales were calculated using item response theory (IRT), specifically a Rasch partial credit model.

Each context scale allowed students to be classified into regions corresponding to high, middle, and low values on the construct. To facilitate interpretation of the regions, the cutpoints on the scale delimiting the regions were described in terms of combinations of response categories.

For certain scales that maintained many of the same items across TIMSS 2015 and TIMSS 2019, the scales were linked to allow for trend measurement on the background construct.

**List of Background Scales**

**Questionnaire scales**

- Home environment support scales
- Home Resources for Learning (Mathematics and Science, Grade 4)
- Home Educational Resources (Mathematics and Science, Grade 8)
- Home Early Literacy and Numeracy Activities Before Primary School (Mathematics and Science, Grade 4)
- Could Do Literacy and Numeracy Tasks When Beginning Primary School (Mathematics and Science, Grade 4)

- School composition and resources scales (Mathematics and Science, Grades 4 and 8)
- School Composition by Socioeconomic Background of the Student Body
- Schools Where Students Enter the Primary Grades with Literacy and Numeracy Skills
- Instruction Affected by Mathematics Resource Shortages
- Instruction Affected by Science Resource Shortages

- School climate scales (Mathematics and Science, Grades 4 and 8)
- School Emphasis on Academic Success – Principals’ Reports
- Parents’ Perceptions of Their Child’s School
- Student’s Sense of School Belonging

- School discipline and safety scales (Mathematics and Science, Grades 4 and 8)
- School Discipline
- Safe and Orderly School – Teachers’ Reports
- Student Bullying

- Teacher preparation, professional development, and job satisfaction scales (Mathematics/Science, Grade 4 and 8)
- Teachers’ Job Satisfaction

- Challenges to teaching and learning scales (Mathematics and Science, Grade 4 and 8)
- Classroom Teaching Limited by Students not Ready for Instruction

- Students’ attitudes scales
- Students Like Learning Mathematics/Science (Grades 4 and 8)
- Students Confident in Mathematics/Science (Grades 4 and 8)
- Students Value Mathematics/Science (Grade 8)

- Curriculum and instruction scales (Grades 4 and 8)
- Instructional Time Spent on Mathematics/Science
- Percentages of Students Taught the TIMSS Mathematics/Science Topics
- Instructional Clarity in Mathematics/Science Lessons – Students’ Reports
- Disorderly Behavior During Mathematics Lessons
- Teacher’s Emphasis on Science Investigation

**Mathematics and Science**

- Countries (average achievement)
- In mathematics at Grades 4 and 8, East Asian countries were the top performers. Singapore, Chinese Taipei, Korea, Japan, and Hong Kong SAR outperformed the other countries by a substantial margin.
- In science at both grades, Singapore, Chinese Taipei, Korea, and Japan performed well and were joined by the Russian Federation, and Finland.

- International benchmarks
- Across countries, only small percentages of students reached the Advanced International Benchmark.
- Top performing countries had highest percentages of students reaching the advanced level.
- Most TIMSS countries are educated high percentages of their students to at least the Low International Benchmark

- Trend
- In both grades, more countries had increases than declines in average mathematics and science achievements in both the long term (since 1995 or 2007, respectively) and the short term (since 2015); the exception was fourth grade science where there were as many declines as improvements in the short term (2015–2019).

- Achievement and gender
- There was gender equity in average mathematics and science achievement in almost half of the countries.
- Boys had higher mathematics achievement in more countries than girls in fourth grade.
- Girls had higher science achievement in 18 countries at the fourth grade and 15 countries at the eighth grade.

**Context for education**

- Parents, home resources and early activities
- Students with many home resources for learning (e.g., books, internet access) and parents with higher levels of education had higher average achievement in mathematics and a science at both grades.
- Students whose parents often engaged them in early literacy and numeracy activities at home had higher levels of mathematics and science achievement on average at the Grade 4 than did students whose parents never or almost never did.
- Grade 4 students showed higher achievement, on average, when they had attended preprimary education or when they entered primary school with already acquired literacy and numeracy skills.

- Attitudes toward learning
- A liking for learning mathematics and science, respectively, was strongly associated with higher average achievement at both grades. The percentage of students who did not like learning mathematics or science, respectively, was higher in eighth grade than fourth grade.
- A feeling of confidence in mathematics or science, respectively, was also strongly associated with higher average achievement at both grades. The percentage of students who were not confident in mathematics or science was higher in eighth grade than fourth grade.
- Valuing mathematics or science, respectively, was associated with higher average achievement. Most eighth-grade students reported valuing mathematics or science, respectively, at least somewhat.

- Teachers and teaching
- Higher student achievement was associated with greater clarity of instruction in both subjects and grades.
- More fourth grade students reported a high clarity of instruction than eighth grade students.
- Teachers reported a sizable gap between their professional development needs and their recent professional development opportunities when asked about integrating technology into instruction and improving students’ critical thinking skills.

- Technology for instruction
- There was wide variation across countries in computer availability. Having computers available for instruction was associated with higher achievement, on average, in both mathematics and science and at both grades.
- Teachers reported doing computer activities more frequently in science lessons than in mathematics lessons.
- Though few students were in classes where computers were used on a frequent basis, there was a substantial range across countries.

- School environment
- Students attending schools with fewer resource shortages had higher average achievement in mathematics and science at both grades.
- Students attending schools with a higher emphasis on academic success had higher average achievement in mathematics and science at both grades.
- Students with a high sense of school belonging had higher average mathematics and science achievement at both grades.
- Students who attended schools with fewer discipline problems and safer and more orderly environments had higher average achievement in mathematics and science at both grades.
- Students who reported experiencing bullying on a weekly basis had considerably lower average achievement than students bullied monthly or never or almost never.

- Methods and procedures (MP): TIMSS 2019 technical report
- MP: TIMSS 2019 TP, ch. 10: reviewing achievement item statistics
- MP: TIMSS 2019 TP, ch. 11: scaling methodology
- MP: TIMSS 2019 TP, ch. 12: implementing the scaling methodology
- MP: TIMSS 2019 TP, ch. 13: examining eTIMSS country differences between eTIMSS …
- MP: TIMSS 2019 TP, ch. 14: estimating standard errors
- MP: TIMSS 2019 TP, ch. 15: scale anchoring to interpret achievement scales
- MP: TIMSS 2019 TP, ch. 16: creating context questionnaire scales