An evaluation of the minimal constraining information during observation for movement reproduction

An important question in the field of imitation is what information is used for movement
reproduction. While it is argued that relative motion information is perceived and minimised,
direct evidence is lacking. In this experiment relative motion was manipulated to convey a
novel kicking action. Twenty-four adults were assigned to one of the three impoverished relative
motion display groups showing only the TOE, FOOT or LEG. After practise with partial
information, participants watched a full-body display and in a final condition performed the
action with an additional context constraint (i.e., a ball). Movement kinematics were collected
and difference scores between participants and the model were analysed. The groups did not
differ in terms of knee–ankle coordination. For hip–knee, the TOE group performed more like
the model than the FOOT and LEG group. When transferred to the full-body display, there
were no significant improvements. End-point trajectory information can provide sufficient
information to reproduce key characteristics of the movement form.