Every number on Clinical Asset Appraisal is built from real observed listings. These are the lay definitions for the terms we use to describe that data. For the full methodology, see Methodology.
The middle price. Half the listings cost more, half cost less. We lead with median because used clinical pricing is full of outliers and the mean (average) gets pulled around by a single $80,000 dealer ask or a $10 surplus lot.
Comparable listings we've observed. "19 comps" means our median is built from 19 real listings we've seen. More comps = more reliable signal.
A bucket of listings of the same type. A complete-system peer median is computed only from complete-system listings — never mixed with parts.
How much data we have. STRONG = 50+ listings across 3+ sources. GOOD = 20+ across 2+. THIN = 5+. VERY THIN = fewer than 5. THIN means "directional, verify yourself".
How sure our classifier is. We only publish numbers when classification confidence is ≥ 0.65, and only show a listing in the deal feed when device-match confidence is ≥ 0.75 and classification confidence is ≥ 0.65.
Quartile range. p25 is the 25th-percentile price (cheap end). p75 is the 75th-percentile (high end). Together they show the typical price band.
A fully assembled, ready-to-deploy unit. The whole machine, not a part.
A significant subsystem — a probe, a detector, a vaporizer. Not the whole machine, but more than a small part.
A component, cable, battery, board, or accessory. Sold individually.
Non-functional or salvage. The seller says it doesn't work or is sold as-is.