On-Screen Marking (OSM) is digital evaluation of scanned answer scripts. Examiners mark on screen instead of moving physical answer books, with audit trails and institutional control.
Read articleOn-Screen Marking
How digital evaluation of answer scripts works: fictitious-number masking, flap tear-off, overhead or high-speed scanning, private-cloud QA, agency review, packet assignment, OSM marking, and realtime revaluation on the unmarked original.
Foundations
Digital evaluation from fictitious-number masking and flap tear-off, through overhead or high-speed scanning, private-cloud capture, agency page-by-page QA, packet assignment, on-screen marking, and realtime revaluation on the unmarked original.
Read articlePhysical evaluation vs OSM: logistics, audit, scan dependency, and why revaluation is faster when the unmarked original stays in the cloud.
Read articleUniversity OSM: masked copies in a private cloud, agency page review, packet assignment, realtime progress, and revaluation on the unmarked original.
Read articleAI can suggest question-wise marks and reasons on scanned scripts. It should not silently award degrees. Human examiners and institutions remain responsible for final marks.
Read articleSecurity for digital evaluation: fictitious-number masking, flap tear-off, private-cloud copies, role-based packets, unmarked originals for revaluation, and audit — not a slogan.
Read articleHow Avid implements this
First-party detail from the products we built — evaluation screens, workspaces, LMS integrity rules, AI modes. No invented customers.
Production OSM is a script viewer plus a marks panel: page navigation, pen annotations, part-wise marks, compulsory and best-of schemes. How Avid OSM 3.0 implements that for examiners.
Read articleAfter private-cloud ingest, Avid OSM identifies each copy, page count and scan quality. The agency panel reviews every copy; approved packets go to examiners. University view is realtime.
Read articleAvid OSM 3.0 lets a university hide AI, offer on-demand Avid AI Professor suggestions, or run broader AI evaluation. Manual marks take priority. Full AI is a policy choice, not a default.
Read article