Data Methodology, Gradometer

The value of this site is entirely dependent on whether its data is right. This page documents how we source it, how we grade our own confidence, and how we keep it from going stale.

Source hierarchy

We prefer sources in this order, and we record which tier each record came from:

Official primary documents: The university's own student handbook, academic regulations, examination by-laws or published grading policy. Marked high confidence.
Official secondary pages. A department or campus page restating the policy. Usually high or medium.
Regulatory frameworks. HEC, UGC India, UGC Bangladesh. Useful as a fallback, but national frameworks routinely differ from institutional practice. Marked medium.
Reputable secondary sources. Used only where nothing official is accessible, and always marked low confidence with a visible warning.

Confidence levels shown on every page

Each grading record carries a badge. Verified against official source means we traced it to the institution's own document. Verified, secondary source means the data is consistent across credible sources but we have not confirmed it against a primary document. Indicative. Needs verification means treat it as a starting point only.

Review intervals

Different data decays at different speeds. Each record has a review interval, and records past due are surfaced in an internal verification queue.

Data type	Review interval
University name, city, type	365 days
Founding year	730 days
Grading scale and grade points	180 days
Credit and degree requirements	180 days
Conversion formulas	180 days
Tuition and fees	90 days
Admission requirements	30–90 days
Scholarship records	30 days
Scholarship deadlines near a due date	7 days

Handling conflicting sources

Conflicts are common. COMSATS' own regulations print A− as 3.66 while many calculator sites use 3.67. NUST is associated with two different percentage formulas that differ by 12.5 points. Our rule: publish the institution's own figure, note the discrepancy openly, and tell the student to trust their transcript above any website, including this one.

Relative grading is recorded as such

Where a university grades relatively and publishes no fixed percentage bands, we set has_percentage_bands: false and the site refuses to display a marks column. It is better to show nothing than to invent a table that does not exist. This is the single most common error we found in competing calculators.

Scholarship lifecycle

Scholarship cycles are versioned rather than overwritten. A closed 2026/27 call stays on the page next to an expected 2027/28 one, because historical opening and closing dates are the best available predictor of when the next call will appear. Statuses are: open, upcoming (announced), expected (recurring but unannounced), closed, and archived.

Corrections

If something here is wrong, we want to know. Every correction is checked against an official source before it goes live, we do not change data on assertion alone. Report an update.