How should I evaluate Meta Neural Band controls?
Try the actions you will repeat: select, scroll, go back and dismiss. Count unwanted actions and recovery as well as successful inputs. A control is useful when a small mistake stays small.
Include ordinary hand positions
Repeat the same short task seated, standing and with the other hand occupied. This is a proposed trial, not a claim about measured band performance. The aim is to discover whether the controls remain understandable outside a carefully prepared demonstration.
Choose the wrong item on purpose and return to the previous screen. If you lose your place, record where the confusion began. Compare that sequence with the frame controls. The better option for your routine is the one that makes the whole interaction easier, including the part after an error.