Computational Thinking Scores by Country: ICILS 2023
Average computational thinking (CT) scale scores for countries and education systems with published ICILS 2023 results. The international comparison uses the 2023 assessment cycle and the CT scale established for trend measurement in ICILS 2018.
ICILS 2023 CT Statistics
Published country means and international summary measures from the IEA computational thinking assessment.
Average Computational Thinking Scores
Country means for the 21 non-benchmarking participants included in the ICILS 2023 international CT average.
ICILS 2023 CT Scores Across Countries
Average CT scale score, ordered from highest to lowest reported mean
The ICILS 2023 international CT average is 483. The United States and North Rhine-Westphalia are retained in the full table with their reporting-status notes rather than included in this comparison chart.
Chart analysis
- Chinese Taipei recorded the highest published country mean at 548.
- Korea, Republic of followed at 537, with the Czech Republic at 527.
- Twelve of the 21 countries contributing to the international average had point estimates above 483.
- Italy, at 482, was one scale point below the published ICILS average.
- Austria and Luxembourg reported the same country mean of 476.
- The lowest published country mean among participants included in the international average was 421 in Uruguay.
- The 127-point distance between Chinese Taipei and Uruguay is slightly greater than one international CT standard deviation reported for the 2023 participating-country distribution.
- Small gaps between country point estimates should not be treated automatically as statistically distinguishable differences.
Full ICILS 2023 Computational Thinking Score Table
All national participants and the benchmarking participant shown in IEA Table 5.5. The source-order distinction for special reporting cases is retained through the coverage-status column.
| Country / system | Average CT | Year | Target grade | Average coverage |
|---|---|---|---|---|
| Chinese Taipei | 548 | 2023 | Grade 8 | Included |
| Korea, Republic of | 537 | 2023 | Grade 8 | Included |
| Czech Republic | 527 | 2023 | Grade 8 | Included |
| Belgium (Flemish) | 509 | 2023 | Grade 8 | Included |
| Denmark | 504 | 2023 | Grade 8 | Included |
| Finland | 502 | 2023 | Grade 8 | Included |
| France | 499 | 2023 | Grade 8 | Included |
| Slovak Republic | 498 | 2023 | Grade 8 | Included |
| Latvia | 495 | 2023 | Grade 8 | Included |
| Sweden | 486 | 2023 | Grade 8 | Included |
| Norway | 485 | 2023 | Grade 9 | Included |
| Portugal | 484 | 2023 | Grade 8 | Included |
| Italy | 482 | 2023 | Grade 8 | Included |
| Germany | 479 | 2023 | Grade 8 | Included |
| Austria | 476 | 2023 | Grade 8 | Included |
| Luxembourg | 476 | 2023 | Grade 8 | Included |
| Slovenia | 448 | 2023 | Grade 8 | Included |
| Malta | 438 | 2023 | Grade 8 | Included |
| Croatia | 429 | 2023 | Grade 8 | Included |
| Serbia | 422 | 2023 | Grade 8 | Included |
| Uruguay | 421 | 2023 | Grade 8 | Included |
| United States | 461 | 2023 | Grade 8 | Excluded: sampling participation caution |
| North Rhine-Westphalia (Germany) | 461 | 2023 | Grade 8 | Benchmarking participant |
Table analysis
- The published national point estimates extend from 421 to 548, a range of 127 scale points.
- Chinese Taipei, Korea, Republic of, and the Czech Republic are the three national participants with means above 520.
- Denmark, Finland, France, the Slovak Republic, and Latvia fall between 495 and 504.
- Sweden, Norway, Portugal, Italy, Germany, Austria, and Luxembourg are clustered between 476 and 486.
- Austria and Luxembourg share the same reported mean of 476.
- Malta, Croatia, Serbia, and Uruguay have country means within the Level 1 score interval defined for individual achievement, while country means should not be treated as country-level proficiency classifications.
- The United States score of 461 is reported by IEA with a sampling-participation caution and is not included in the 483 international average.
- North Rhine-Westphalia is reported as a benchmarking participant rather than as a separate country in the international average.
- All displayed values are point estimates; sampling uncertainty accompanies the source estimates even when only rounded means are reproduced here.
Assessment Cycle Comparison: 2018 and 2023
CT was introduced as an international option in ICILS 2018. Seven countries have directly comparable country means across the 2018 and 2023 cycles; North Rhine-Westphalia also has comparable benchmark data.
Comparable Country Scores in 2018 and 2023
Average CT scale scores for the seven countries with comparable data in both cycles
The source table reports official cycle differences using unrounded estimates. Displayed rounded means can therefore produce arithmetic differences that differ slightly from the reported change column.
Trend analysis
- Denmark moved from 527 in 2018 to 504 in 2023; IEA reports a change of -23 points and flags the cycle difference in its statistical comparison.
- Luxembourg moved from 460 to 476, with an IEA-reported change of +16 points that is also flagged in the source table.
- Finland moved from 508 to 502 and Germany from 486 to 479; neither cycle difference is flagged by IEA in Table 5.7.
- France changed from 501 to 499, while Portugal changed from 482 to 484.
- Korea, Republic of, is shown at 536 in 2018 and 537 in 2023; the official change column reports 0 after calculation from the underlying unrounded estimates.
- The available CT history consists of assessment cycles rather than an annual time series.
| Country / system | 2018 | 2023 | IEA change | IEA statistical flag |
|---|---|---|---|---|
| Denmark | 527 | 504 | -23 | Flagged |
| Finland | 508 | 502 | -7 | Not flagged |
| France | 501 | 499 | -2 | Not flagged |
| Germany | 486 | 479 | -7 | Not flagged |
| Korea, Republic of | 536 | 537 | 0 | Not flagged |
| Luxembourg | 460 | 476 | +16 | Flagged |
| Portugal | 482 | 484 | +2 | Not flagged |
| North Rhine-Westphalia (Germany) | 485 | 461 | -25 | Flagged; benchmark |
Assessment analysis
- Comparable national trend coverage is limited to seven countries because CT was an optional ICILS component and 2018 was the first international CT cycle.
- Luxembourg has the largest positive reported country change in the comparable set at +16 points.
- Denmark has the largest negative reported country change at -23 points.
- The North Rhine-Westphalia benchmark shows a reported change of -25 points.
- IEA flags Denmark, Luxembourg, and North Rhine-Westphalia in the source comparison table; the remaining displayed cycle differences are not flagged.
- Trend comparisons use CT scale scores. The proficiency-level descriptions introduced for ICILS 2023 should not be substituted for the preliminary 2018 level descriptions when comparing proficiency percentages over time.
ICILS Computational Thinking Proficiency Scale
The 2023 CT achievement scale uses four proficiency levels plus a category below Level 1.
| CT level | Lower boundary | Score interval | IEA descriptor |
|---|---|---|---|
| Below Level 1 | Below 330 | 330 or below | Basic interaction |
| Level 1 | Above 330 | >330 to 440 | Fundamental sequencing |
| Level 2 | Above 440 | >440 to 550 | Structured problem solving |
| Level 3 | Above 550 | >550 to 660 | Integrated problem solving |
| Level 4 | Above 660 | >660 | Systems thinking |
Table analysis
- The CT proficiency boundaries are 330, 440, 550, and 660 scale-score points.
- The published 2023 country means range from 421 to 548, so the national means themselves lie within the Level 1 or Level 2 score intervals.
- A country mean located inside a proficiency interval does not mean every student in that country performs at that level.
- Across the ICILS 2023 average distribution, 37% of students were at Level 2, 23% at Level 3, and 6% at Level 4.
- The same international distribution placed 34% of students below Level 2.
Score Interpretation and Coverage
Technical points affecting comparisons of the published country means.
2023 international mean
The CT average of 483 is an equally weighted international comparison based on the non-benchmarking participants that met the study’s sampling-participation requirements.
Scale reference
The CT reporting scale was established in ICILS 2018 with an international reference mean of 500 and standard deviation of 100 for the seven participating countries meeting the required sampling conditions in that cycle.
Within-country variation
IEA reports that variation among students within countries is larger than the spread between national means. Country averages therefore summarize only one part of the achievement distribution.
Participation caution
The United States achieved at least 50% overall sampling participation but did not meet the stated participation guideline. Its reported score is retained with that caution and excluded from the ICILS 2023 CT average.
Source and Methodology
The full country table reproduces the published IEA point estimates rather than creating a composite score or converting CT results into percentages. The main chart excludes the United States and the benchmarking participant because neither contributes to the reported 483 international average. No values are imputed for non-participating or non-reportable systems.
ICILS scores are population estimates with sampling uncertainty. Ordering countries by rounded point estimates is not equivalent to proving that every adjacent country differs statistically. The United States carries a sampling-participation caution, North Rhine-Westphalia is a benchmarking participant, and the published CT results apply only to education systems with internationally reportable ICILS 2023 computational thinking data.