SCORING METHODOLOGY
Useful estimates, with the uncertainty left visible.
Compass reports a practice estimate, not an official score, certificate, or immigration decision. Every result names the model and exam-format version used.
Section estimates
Objective sections are calculated from reviewed original items and calibrated against the public scale structure of the selected exam. Writing responses are saved for timed practice but are not automatically marked or included in a score.
Confidence ranges
A short diagnostic has less evidence than a full mock, so its range is wider. More answered items, consistent performance, and representative difficulty can narrow the range without implying certainty.
No invented overall
When an exam provider does not publish a defensible overall conversion—or when an integrated test cannot be reconstructed without speaking—Compass leaves overall blank. Speaking is shown as Not assessed throughout this no-oral product.
Format governance
Each exam pack records its official sources, last verification date, format version, and scoring-model version. Material changes trigger re-review before affected results are compared.