Dataset
845 trials
Every recorded attempt, cleaned to 714 usable trials — the messiness is part of the signal, not noise to discard.
London Met Behavioural analysis 2026
Interactive Card Sorting Analysis.
Given KQJ blank cards and an 8×8 board, how do people arrange what matters — and what does the mess reveal about strategy? 845 recorded trials, nine analysis views, one honest read.
Viewer
9 analyses
From clean patterns to extreme cases: move-by-move replay, progression, recovery, speed and repetition views.
Signal
Messiness
Organisation beats raw effort, and blank cards act as scaffolding — recovery attempts are rare but informative.
The data stays private.
Participant trials are covered by the study's ethics and data-handling terms, so the raw dataset isn't shared. What is public: the interactive viewer above, the documented method, and every statistic quoted on this page — I'm happy to talk through how any of them was derived, and how the study could be replicated.