Cómo puntuamos

College Board scores the digital SAT with IRT-based item-level calibration and does not publish raw-to-scaled conversion tables. The section score is an Item Response Theory ability estimate over the specific operational items administered, not a function of the number of correct answers: College Board states that "two students who answer the same number of questions correctly in a test section may earn differing section scores." The IRT item parameters and the theta-to-scale-score transformation are unpublished, so any 400-1600 estimate we produced would be fabricated. We report raw correct counts, per-domain and per-skill accuracy, and timing instead.

ACT converts each section raw score to the 1-36 scale using a FORM-SPECIFIC conversion table built from unpublished equating methodology and item-level IRT parameters. ACT publishes conversion tables only for particular released practice forms, and those tables are valid for those forms alone — they cannot be applied to an original question bank. With no published raw-to-scale method there is no defensible way to estimate a section score, a Composite, a STEM score or an ELA score, so we report raw score, percent correct, per-reporting-category accuracy and pacing instead. ACT percentile ranks and national norms for the enhanced Composite are likewise not published and are suppressed entirely.

LSAC converts the raw score to the 120-180 scale with a form-specific conversion produced by equating after the administration, and does not publish a raw-to-scale table for any live form (for undisclosed forms, never). LSAC’s own example shows the cutoff moving between forms: 56 correct for a 160 on a slightly harder form, 57 on a slightly easier one. With no published raw-to-scale method — and no published count of scored items, so no known maximum raw score — any 120-180 number we produced would be invented.

GMAC publishes no raw-to-scaled conversion for the GMAT. Section scores are described only as depending on "the number of questions answered, whether the answers are correct or incorrect, and the difficulty and other parameters of the questions answered", with no IRT parameters, no ability-estimation method and no conversion table for the 60-90 section scales or the 205-805 Total Score. Any estimate we produced would be invented, so we report raw accuracy, per-domain accuracy and pacing only.

ETS converts a GRE raw score to the 130–170 scale "through a process known as equating", which also absorbs the difficulty difference introduced by section-level adaptation, and publishes no raw-to-scaled conversion table for any form, no equating model, no IRT item parameters and no account of how the delivered second section's difficulty is combined with the raw count. The Analytical Writing 0–6 score additionally depends on e-rater feature weights and a human/e-rater adjudication rule that are likewise unpublished. With none of those inputs available, any 130–170 or 0–6 figure we produced would be a fabrication, so we report raw correct counts and percent correct instead.

Bocconi does not publish, and does not use, any raw-to-scaled conversion for the Online Bocconi Test: the reported result IS the raw penalty-adjusted total out of 50. There is also no published equating between the four permitted attempts, between different test forms, or between the Italian and English versions of the test, and no published formula mapping that total onto the 55% test component of the admission ranking. With no published method there is nothing defensible to estimate, so we report your raw penalty-adjusted total and per-area subscores only.

Bocconi reports the Bocconi Online Test - Law as a penalised raw total out of 50 and publishes no raw-to-scaled conversion at all — there is no scale to convert to. The only equating Bocconi performs is between the Bocconi test, SAT and ACT, via "an internal conversion system based on comparative tables that analyze candidate percentiles" whose tables are not published, and it does not publish how the 55% test weight is normalised before being combined with the 45% GPA weight. With no published method, any estimated scaled score, percentile or SAT/ACT/LSAT equivalent we produced would be fabricated.

There is no scale to estimate: the graduate Online Bocconi test is reported as a penalised total out of 50, and Bocconi publishes no conversion, percentile table or equating between attempts. When an applicant submits more than one accepted test, Bocconi says the system considers the most competitive one, but it does not publish how a Bocconi total is compared with a GMAT or GRE score. With no published method there is nothing defensible to estimate.

Nothing to estimate: PoliTo reports the raw penalty-adjusted score, converted to a scale of 100 by simple proportion. There is no equating to model.

Nothing to estimate: the reported result is the raw penalty-adjusted score.

CISIA reports the absolute score itself; any transformation is done by each university with its own system, which is not national and not modelled here.

CISIA reports the absolute score itself; transformations are each university’s own and are not modelled here.

CISIA reports the absolute score itself; transformations are each university’s own and are not modelled here.