What is Item Response Theory (IRT)?

Item Response Theory (IRT) is a statistical framework that models the relationship between a person's latent ability and the probability of answering a question correctly. Unlike classical test theory, IRT accounts for question difficulty, discrimination, and guessing. Eduentry uses a 2-Parameter Logistic (2PL) model — the same model used in PISA, CAT4, and NFER standardised assessments.

How does Eduentry calculate a standardised score?

After each session, Eduentry applies Maximum A Posteriori (MAP) estimation — a Bayesian method — to estimate a child's latent ability (θ) from their pattern of responses. This θ is then linearly transformed to a standardised scale with mean 100 and standard deviation 15, matching the scale used by NFER, GL Assessment, CAT4, and WISC-V.

How does Eduentry select question difficulty?

Questions are chosen adaptively so that each item difficulty matches the child's current estimated ability. This targets the point of maximum information gain — the same principle used in computer-adaptive tests like the Digital SAT and GRE. The result is a shorter, more precise assessment than a fixed-form test.

How accurate is Eduentry's standardised score?

Over 15 adaptive questions per subject, the Standard Error of Measurement (SEM) is approximately ±5–7 standardised score points. This is comparable to a short-form professional assessment. Scores should be interpreted as a meaningful diagnostic estimate rather than a single definitive number, and retesting after a period of practice will track genuine progress.

How does Eduentry compare to CAT4, NFER, or GL Assessment?

Eduentry uses the same IRT methodology and standardised scoring scale as professional tools including CAT4 and GL Assessment. It is a free, accessible diagnostic designed to identify strengths and gaps and track progress. It is not a replacement for invigilated assessments used in formal school selection, but the standardised score is directly interpretable alongside professional results.

Assessment Methodology — How Eduentry Works

Adaptive Testing

Unlike a fixed test where every child answers the same questions, Eduentry adapts in real time. After each answer, the system updates its estimate of the child's ability level and selects the next question to provide maximum measurement precision at that level.

This means a child who answers correctly receives harder questions; a child who struggles receives easier ones. The result: fewer questions are needed to achieve the same measurement accuracy as a traditional fixed-length test.

Key benefit

A 15-question adaptive test can match the accuracy of a 100 question fixed test, because every question is chosen to be maximally informative for that specific child.

Item Response Theory (IRT)

Eduentry uses the 2-Parameter Logistic (2PL) IRT model — the same model used by PISA, large-scale standardised assessments, and professional educational psychologists.

Each question has two parameters:

Difficulty (b) — the ability level at which a child has a 50% probability of answering correctly.
Discrimination (a) — how well the question differentiates between children just below and just above the difficulty threshold.

Ability estimation uses Maximum A Posteriori (MAP) estimationwith a standard normal prior. After each response, the posterior distribution over ability is updated using Bayes' theorem, and the MAP estimate is taken as the current ability estimate.

Scoring Scale

Raw IRT ability estimates (θ) are transformed to a standardised scale with mean 100 and standard deviation 15 — the same scale used by CAT4, the 11+ standardised tests, and professional cognitive assessments such as WISC and WASI.

Score Range	Band	Percentile	SD from Mean
120+	Exceptional	90th+	+1.33σ and above
110–119	Above Average	75th–90th	+0.67σ to +1.33σ
95–109	Average	37th–73rd	−0.33σ to +0.67σ
85–94	Below Average	16th–37th	−1σ to −0.33σ
Below 85	Needs Support	Below 16th	Below −1σ

68% of children score between 85 and 115 (within ±1 standard deviation of the mean).

International Benchmarks

Eduentry maps each score band to four internationally recognised educational frameworks, giving families a global context for their child's performance.

🇬🇧UK National Curriculum

Scores map to: below expected standard / approaching / meeting / above expected / highly able. Competitive grammar school scores are typically 115+.

🇺🇸US Grade Levels

Scores map to: below / approaching / at / above / significantly above grade level, aligned to Common Core and NAEP proficiency descriptors.

🌍PISA (OECD)

Scores map to PISA proficiency Levels 1–6. Average OECD performance corresponds to approximately Level 3, equivalent to a standardised score near 100.

🎓IB Programme

Scores indicate readiness for IB Middle Years Programme, IB Standard Level, or IB Higher Level subjects in the Diploma Programme.

What You Learn from the Results

Unlike a traditional school test that grades a child against a local cohort, a student and their parent who complete the Eduentry assessment will receive a comprehensive results screen and a printable report showing exactly where the child stands globally.

Here is exactly what they will learn from the results:

Standardised Score and Performance Category — a score on the mean-100, SD-15 scale, placed into one of five bands: Needs Support (70–84), Below Average (85–94), Average (95–109), Above Average (110–119), or Exceptional (120+).
Global Percentile Ranking — how the child performs compared to same-age peers worldwide, expressed as a percentile (e.g. top 10%, top 25%).
Detailed Subject Breakdowns — individual scores and performance bands for English, Mathematics, Verbal Reasoning, and Non-Verbal Reasoning, so parents and children can see exactly where strengths and gaps lie.
International Benchmarking — the score mapped to UK National Curriculum levels, US grade-level equivalents, PISA proficiency levels, and IB programme readiness.
AI-Generated Personalised Recommendations— targeted study recommendations based on the child's specific performance profile, highlighting the areas most likely to improve their overall score.

Together, these five outputs give families a clear picture of their child's global standing, specific weaknesses, and actionable steps to improve — all from a single 90-minute assessment.

Subjects Assessed

📚English

Reading comprehension, grammar, vocabulary, spelling, and written language conventions.

🔢Mathematics

Arithmetic, algebra, geometry, statistics, and mathematical problem-solving.

💬Verbal Reasoning

Word analogies, classifications, logical sequences, and language-based problem solving.

🔷Non-Verbal Reasoning

Pattern recognition, spatial reasoning, matrix completion, and visual logic.

Transparency

Eduentry offers a robust and adaptive approach to student assessment, generating results that provide an accurate representation of each learner's academic capabilities.

Traditional assessments are often limited by fixed question sets, where performance can be unduly affected by the difficulty level of the questions rather than the student's actual ability. Eduentry addresses this limitation through an AI-driven adaptive engine that continuously adjusts question difficulty based on each student's responses, ensuring that outcomes reflect true academic potential rather than the structure of the test itself. This produces a consistent, fair, and meaningful measure of each learner's strengths and developmental needs.

While Eduentry's methodology differs from CAT4, GL Assessment, and WISC, its results serve as a reliable and actionable basis for educational planning and decision-making. In cases where formal diagnostic evaluation is required, Eduentry functions as a valuable complement to established psychometric assessment tools.

How Eduentry’s Adaptive Assessment Works