Speed-accuracy tradeoffs in a targeted Box and Block Test

Abstract The Box and Block Test (BBT) is a clinical assessment of gross manual dexterity, requiring participants to move small blocks between two boxes. It is scored using the number of blocks transferred in 60 s; a high score requires rapid arm movements and accurate grasping, with no constraints on block placement accuracy. A low score cannot distinguish inaccurate movements from slow movements. To address this, we explored two variations of the standard BBT using 16 blocks moved in a predefined sequence. The insert BBT requires inserting blocks into recesses that constrain accuracy. The targeted BBT requires placing each block onto a target, allowing measurement of movement timing and accuracy. Across participants, movement times for the insert and targeted BBT were strongly correlated, but only weakly correlated with the standard BBT scores. Two speed-accuracy tradeoffs were evident: within participants, slower placements were generally more accurate; and across participants, those who moved more slowly were generally more accurate. Surprisingly, locations associated with faster movements were also associated with more accurate block placements, highlighting the spatial affordances of the block locations. Our tasks capture performance features absent in the standard BBT and emphasize the value of designing dexterity tasks that evaluate both speed and accuracy.