RECOVERY + RAW REPORTING

Resume exactly.
Report honestly.

An interrupted rehearsal stays recoverable on the same device for 7 days. Every restore must match the frozen form, delivery plan, and four native section manifests.
280stable item positions
4native session states
0prompt text in exports
RAWresults only

EXACT INTERRUPTION RECOVERY

Nothing important is reconstructed by guesswork.

The saved envelope retains the outer route and each section workspace independently. A restore is rejected if the form instance, version, timestamp, section manifest, or paired section clock differs—or if any saved native state would need repair.
  1. 01Active segment and position
  2. 02Four native response maps
  3. 03Marks and eliminated choices
  4. 04Section and item time
  5. 05Permanent section locks
  6. 06Break state and remaining time

DELAYED RAW EVIDENCE

Results open only after the administration closes.

Reporting counts stable answer identities. It does not convert performance into an official score, percentile, probability, or readiness prediction.
SECTION

Raw correct / answered / total for each of the four sections

SCIENCE

Biology, General Chemistry, Organic Chemistry, 23 domains, and objectives

READING

Three passage groups and six instructional objectives

PAT + QR

Six PAT families plus Quantitative group, topic, and objective evidence

PRIVACY-MINIMAL EXPORT

Evidence without surveillance.

Included: schema and content versions, stable form identity, completion time, and raw aggregate counts.

Excluded: prompts, option and rationale text, names, emails, device or network identifiers, scaled scores, percentiles, and readiness claims.

RECOVERY FOUNDATION COMPLETE

The exact run is recoverable.
The public beta is open.

The complete full-DAT beta uses this seven-day device-local recovery and raw-reporting boundary. Independent validation and official score conversion remain deliberately separate.