BML Loop 1 — Measure
Hypothesis: Catalogers will increase their understanding and productivity using a Graph RDF UI rather than a form-based Linked Data editor.
Metric
Median time to fully catalog the same representative resource (fixed fixture record) using each of: Graph Editor, Sinopia, Marva. Same catalogers, order counterbalanced across sessions to cancel out learning effects.
Why
Direct productivity signal — this is the core of "productivity" in the hypothesis, not a vanity metric.
Method
Depends on #2 (task-timing instrumentation) for the Graph Editor side; timing for Sinopia/Marva sessions via moderated stopwatch/screen recording since we don't control their instrumentation.
Owner
@jermnelson
Andon
See #8 for the abandonment/failure threshold that should pause this pilot.
BML Loop 1 — Measure
Hypothesis: Catalogers will increase their understanding and productivity using a Graph RDF UI rather than a form-based Linked Data editor.
Metric
Median time to fully catalog the same representative resource (fixed fixture record) using each of: Graph Editor, Sinopia, Marva. Same catalogers, order counterbalanced across sessions to cancel out learning effects.
Why
Direct productivity signal — this is the core of "productivity" in the hypothesis, not a vanity metric.
Method
Depends on #2 (task-timing instrumentation) for the Graph Editor side; timing for Sinopia/Marva sessions via moderated stopwatch/screen recording since we don't control their instrumentation.
Owner
@jermnelson
Andon
See #8 for the abandonment/failure threshold that should pause this pilot.