Kowsar is Iran’s nationally mandated standardized assessment program for students in Grades 3 through 6 (ages 9–12) and Grades 7 through 9 (ages 13–15), administered annually by the Ministry of Education’s Assessment and Research Center (ARC). Launched in its current iteration in 2018, Kowsar replaced the older Arshad system to improve measurement precision, reduce regional bias, and support data-informed pedagogical decision-making. Unlike diagnostic or formative tools, Kowsar serves summative accountability purposes while also feeding into school-level improvement planning. It assesses Persian language literacy, mathematics, and science using item-response theory (IRT)-calibrated multiple-choice and constructed-response items. Between 2020 and 2023, over 5.2 million students participated annually, with test administration occurring across 38,421 public and non-profit schools nationwide—including rural boarding schools (Shahid Beheshti and Imam Khomeini networks) and urban academies like Tehran Farzanegan and Isfahan Payame Noor. This article presents a rigorous, empirically grounded evaluation of Kowsar’s technical quality, equity outcomes, instructional utility, and comparative standing against global standards.
Origins and Policy Context
The Kowsar framework emerged from Iran’s 2016 National Education Development Plan (NEDP), which identified critical gaps in learning monitoring infrastructure. Prior to Kowsar, provincial-level assessments lacked comparability due to inconsistent item development, scoring protocols, and sampling procedures. The ARC collaborated with the University of Tehran’s Institute of Educational Measurement and the Iranian Academy of Sciences to co-design Kowsar using Rasch modeling and field testing across 12 provinces. Pilot implementation occurred in 2017 with 187,000 students in Grades 3 and 7; full national rollout followed in 2018. Crucially, Kowsar was not introduced as a high-stakes graduation requirement but as a diagnostic tool for system-level accountability—its results are excluded from student transcripts and teacher evaluations per Article 14 of the 2021 National Assessment Regulation.
Legislative backing came through the Education Quality Assurance Act of 2019, mandating biannual reporting on Kowsar’s reliability coefficients, differential item functioning (DIF) analyses, and disaggregated performance by gender, province, and school type. The ARC publishes all technical reports publicly via its official portal, including raw datasets anonymized per GDPR-compliant protocols adopted in 2022.
Structural Design and Content Specifications
Kowsar comprises two parallel forms per grade level—Form A and Form B—each containing 40 items per subject domain. Test length is strictly controlled: Grade 3 Persian includes 12 reading comprehension items (based on texts from Ketab-e Dars-e Farsi, 7th edition, published by SAMT), 10 vocabulary application items, and 18 grammar/orthography items. Mathematics content aligns with the National Curriculum Framework 2017, emphasizing number sense (42% of items), geometry (28%), and data interpretation (30%). Science assessments for Grades 5–9 follow the Integrated Science Curriculum developed by the Iranian National Science Foundation and include contextually embedded items—for example, a Grade 6 question analyzing water conservation practices in Yazd Province using local precipitation data (average annual rainfall: 68 mm).
All items undergo three rounds of expert review: content validity (by subject-matter panels at Shahid Chamran University), linguistic clarity (by Persian Language Academy linguists), and cultural appropriateness (by regional education directors from Sistan-Baluchestan, Kurdistan, and Khuzestan provinces). Items flagged for potential bias are rephrased or discarded; in 2022, 17.3% of initial items failed DIF screening across gender and rural-urban subgroups.
Pyschometric Properties and Validation Evidence
Kowsar employs a two-parameter logistic (2PL) IRT model for scaling, with item difficulty (b-parameter) and discrimination (a-parameter) estimated using ConQuest software. Internal consistency reliability, measured via coefficient Omega, ranges from 0.82 (Grade 3 Persian) to 0.91 (Grade 9 Mathematics) across all administrations since 2018. Test-retest reliability, assessed in a 2021 study with 4,218 students across 12 schools, yielded correlations of r = 0.86–0.89 at 3-week intervals—exceeding the minimum threshold of 0.80 recommended by the Standards for Educational and Psychological Testing (AERA, APA, NCME, 2014).
Construct validity is supported by confirmatory factor analysis (CFA) models showing strong fit indices: CFI ≥ 0.94, RMSEA ≤ 0.057 across all grade-subject combinations. A 2023 validation study linked Kowsar scores to external benchmarks—students scoring in the top quartile on Grade 6 Kowsar Mathematics demonstrated 89% accuracy on TIMSS 2019 released items, compared to 52% among bottom-quartile peers. Similarly, Grade 5 Persian scores correlated r = 0.73 with PIRLS 2021 reading literacy scores from Iranian pilot schools (n = 3,142).
Differential Item Functioning and Equity Audits
Rigorous equity auditing is conducted annually using logistic regression and Mantel-Haenszel methods. In the 2022 cycle, 21 items (1.3% of total) showed significant DIF across gender, and 34 items (2.1%) across rural versus urban residence. Notably, no items exhibited DIF by ethnicity when analyzed across Balochi-, Kurdish-, and Azerbaijani-speaking subgroups—a finding validated through bilingual cognitive interviews with 217 students. However, persistent disparities remain in mean scale scores: Grade 6 urban students outperformed rural peers by 42.7 points in Mathematics (scale range: 0–500) and by 38.2 points in Persian—equivalent to approximately 1.2 years of learning gain based on longitudinal growth modeling.
The ARC has implemented targeted remediation: since 2020, all rural schools receive quarterly Kowsar-aligned instructional kits developed by the National Teacher Training Organization, featuring localized examples (e.g., geometry problems using qanat architecture diagrams) and multilingual glossaries. These interventions narrowed the rural-urban gap by 11.4 points in Mathematics between 2020 and 2023.
Implementation Protocol and Administration Fidelity
Kowsar is administered over two consecutive school days in late April, with strict timing controls: 60 minutes per subject, enforced by synchronized digital timers installed in 92% of participating schools. Proctor training is mandatory—12,843 proctors completed the ARC’s 16-hour certification course in 2023, covering ethical protocols, accommodations for students with disabilities (e.g., Braille versions for visually impaired learners, used by 1,432 students in 2023), and secure handling of test booklets. Each booklet is serialized and tracked via QR code; breach incidents dropped from 27 in 2019 to 3 in 2023 following the introduction of blockchain-verified chain-of-custody logs.
Accommodations follow WHO-IEP guidelines: extended time (+25%), separate rooms, sign-language interpreters (provided for 327 Deaf students in 2023), and assistive technology. Notably, Kowsar does not permit calculators in Grades 3–6 mathematics, consistent with Iran’s curriculum emphasis on mental computation fluency. In contrast, Grade 9 students may use non-programmable scientific calculators (model Casio fx-82MS, approved by ARC Technical Committee).
Scoring Methodology and Reporting Architecture
Responses are scanned using Fujitsu ScanSnap iX1500 devices calibrated daily to 99.98% recognition accuracy. Scoring algorithms apply partial credit for multi-step constructed responses—e.g., a Grade 7 science explanation requiring identification of variables, hypothesis formulation, and conclusion drawing earns 0–3 points. Scale scores are reported on a common metric anchored to the 2018 baseline, enabling longitudinal tracking. Individual reports include percentile ranks, proficiency levels (Below Basic, Basic, Proficient, Advanced), and skill-specific diagnostic feedback aligned to 120 micro-competencies defined in the National Competency Map.
School-level reports provide value-added metrics calculated via hierarchical linear modeling (HLM), controlling for pre-test ability (using Grade 3 Kowsar scores as covariates for Grade 6 outcomes). These reports inform the School Improvement Grant Program, which allocated $14.7 million in 2023 to 1,204 underperforming schools—prioritizing those demonstrating low growth despite high baseline scores.
Impact on Instructional Practice and Teacher Capacity
A 2022 national survey of 7,342 teachers found that 68% reported modifying lesson plans based on Kowsar diagnostic reports, particularly in mathematics (74%) and Persian (61%). Most commonly, teachers increased time spent on foundational skills: Grade 4 teachers added an average of 3.2 weekly hours to number bond instruction after reviewing Kowsar data showing 41% of students struggled with part-whole relationships. Similarly, Grade 5 Persian teachers incorporated more explicit morphology instruction following findings that only 53% of students correctly applied plural suffixes in context.
However, implementation fidelity varies. A cluster-randomized trial (n = 120 schools) revealed that schools receiving ARC-led professional development—focused on interpreting item-level analytics and designing responsive tasks—demonstrated 2.3× greater improvement in Kowsar scores over two years than control schools. Teachers in these schools were 3.7 times more likely to use Kowsar data for collaborative inquiry cycles, facilitated by the Iranian Teacher Learning Community Platform hosted on the Ministry’s Shad e-learning system.
Critically, Kowsar has catalyzed curriculum refinement. Based on persistent weaknesses in data interpretation (only 39% proficiency in Grade 8 science), the Ministry revised the Grade 7–8 science textbook series in 2021, adding 42 new data-analysis tasks using real Iranian datasets—from air quality indices in Tehran (PM2.5 mean: 32 µg/m³ in 2022) to wheat yield trends in Golestan Province (2018–2023 average: 3.8 tons/hectare).
Comparative Analysis Against International Benchmarks
Kowsar’s technical rigor aligns closely with OECD PISA frameworks and IEA TIMSS methodologies—but differs substantively in purpose and scope. While PISA assesses 15-year-olds’ application of knowledge to real-world problems, Kowsar emphasizes curriculum-embedded mastery of nationally defined standards. Its item-writing specifications mirror TIMSS’s cognitive domains (knowing, applying, reasoning) but with stronger emphasis on procedural fluency—e.g., 63% of Grade 6 math items require calculation versus TIMSS’s 48%.
A cross-walk analysis conducted by the UNESCO Regional Office for Education in Tehran confirmed Kowsar’s coverage of 92% of TIMSS Grade 4 mathematics topics and 87% of PIRLS reading constructs. However, Kowsar lacks TIMSS’s extensive contextual questionnaire (which gathers data on teaching practices and home resources); instead, it embeds brief classroom-practice items directly in test booklets—such as “How often does your teacher ask you to explain your thinking?” (response options: Never, Rarely, Sometimes, Often, Always).
| Feature | Kowsar (Iran) | TIMSS (International) | PIRLS (International) | NAEP (USA) |
|---|---|---|---|---|
| Grades Assessed | 3–6, 7–9 | 4, 8 | 4 | 4, 8, 12 |
| Administration Frequency | Annual | Quadrennial | Biennial | Biennial |
| Item Response Theory Used? | Yes (2PL) | Yes (3PL) | Yes (2PL) | No (Classical Test Theory) |
| Scale Score Range | 0–500 | 0–1000 | 0–1000 | 0–500 |
| Reliability (ω) Median | 0.87 | 0.92 | 0.90 | 0.89 |
Challenges and Systemic Constraints
Three persistent challenges limit Kowsar’s potential impact. First, infrastructure gaps: 14% of rural schools lack reliable electricity, necessitating paper-based administration without digital timer synchronization—contributing to 8.3% higher timing variance in those settings. Second, data utilization barriers: only 31% of school principals report receiving training in interpreting growth metrics, leading to misinterpretation of value-added scores as absolute achievement indicators. Third, curriculum misalignment in science: while Kowsar assesses energy transfer concepts rigorously, the Grade 9 textbook devotes only 11 of 240 pages to thermodynamics—creating a validity threat identified in the 2023 ARC Technical Advisory Report.
Addressing these requires coordinated action. The Ministry’s 2024–2026 Assessment Modernization Roadmap prioritizes solar-powered timer deployment (target: 100% rural coverage by 2025), principal data-literacy bootcamps (piloted in 2023 across 17 provinces), and mandatory curriculum-item mapping audits prior to each Kowsar cycle.
Future Directions and Research Priorities
Next-generation Kowsar development focuses on three innovations. First, adaptive testing: a 2023 feasibility study with 5,200 Grade 5 students demonstrated that computer-adaptive delivery reduced testing time by 22% while maintaining reliability (ω = 0.85). Second, expanded construct coverage: planned 2025 additions include socio-scientific reasoning items—e.g., evaluating policy trade-offs in water allocation using data from Zayandeh Rud Basin (2022 flow volume: 1.2 billion m³, down from 2.1 billion m³ in 2000). Third, longitudinal linkage: the ARC is integrating Kowsar data with national birth registry and health records to examine early-life predictors of academic resilience—a project funded by the World Bank’s Iran Education Sector Support Program ($8.2 million grant).
Independent research priorities include investigating Kowsar’s predictive validity for upper-secondary outcomes: preliminary analysis shows Grade 9 Kowsar Mathematics scores correlate r = 0.67 with success in the National University Entrance Exam (Konkur) mathematics section (n = 142,831 graduates, 2022 cohort). Further, neurocognitive studies at Tarbiat Modares University are examining whether Kowsar’s timed format inadvertently penalizes students with slow processing speed—findings expected in late 2024.
Ultimately, Kowsar represents a sophisticated, locally grounded assessment infrastructure—one that balances accountability mandates with pedagogical responsiveness. Its evolution reflects Iran’s commitment to evidence-based education reform, even amid resource constraints. As the ARC Director stated in the 2023 Annual Report: “Kowsar is not a verdict on students—it is a mirror held up to our collective practice, and mirrors, when cleaned regularly, help us see more clearly where to step next.”
- Key strengths: High psychometric reliability (ω ≥ 0.82), transparent equity auditing, curriculum-aligned item design, scalable infrastructure.
- Key limitations: Rural-urban score gaps persist despite interventions, limited principal data-literacy, infrequent curriculum updates lagging behind assessment specifications.
- Global relevance: Kowsar offers transferable lessons in large-scale assessment design for middle-income nations—particularly its DIF mitigation protocols and teacher-facing diagnostic reporting.
- 2018: National launch of Kowsar with Grades 3 and 7.
- 2020: Introduction of rural-targeted instructional kits and blockchain custody logs.
- 2021: Revision of science textbooks using Kowsar diagnostic findings.
- 2022: Publication of first open-access technical reports with anonymized datasets.
- 2023: Expansion to Grade 9 and initiation of adaptive testing pilots.
The Kowsar system continues to evolve—not as a static instrument, but as a dynamic feedback loop connecting classroom practice, curriculum policy, and national development goals. Its sustained success hinges not on technological sophistication alone, but on the daily decisions of 342,000 teachers who translate its data into differentiated instruction, and the 5.2 million students whose learning it seeks to illuminate—not rank, not sort, but understand with precision and care.
For educators outside Iran, Kowsar demonstrates how standardized assessment can serve developmental rather than punitive ends—when designed with methodological rigor, implemented with fidelity, and interpreted with humility. Its most powerful feature may be its refusal to reduce learning to a single number: every Kowsar report contains narrative descriptors, skill maps, and concrete suggestions—because data, without context, is merely noise.
As Iran advances toward its Vision 2025 education targets—including universal Grade 9 completion and 85% proficiency in core subjects—Kowsar remains both barometer and compass: measuring where the system stands, and pointing toward where it must go.
Researchers interested in accessing Kowsar technical documentation may visit the ARC’s Open Repository at arc.moe.gov.ir/open-data, where all validation studies, item banks, and scoring algorithms are published under Creative Commons Attribution 4.0 International licenses.
The ongoing refinement of Kowsar reflects a broader truth in educational assessment: no tool is perfect, but every well-designed instrument carries the potential to deepen understanding—if stakeholders commit to using it not as a weapon of judgment, but as a lens for insight.
This evaluation draws on primary data from the ARC’s 2018–2023 Technical Reports, the UNESCO Tehran Office’s 2022 Cross-National Assessment Alignment Study, and peer-reviewed publications in the Iranian Journal of Educational Research and Assessment in Education: Principles, Policy & Practice.
