Measurement Bias in Representing Accuracy of Movement on Linear-Positioning Tasks

Abstract

A short-term memory paradigm was employed to contrast differences in accuracy of recall of three groups of children who ranged in age from 7 to 16 yr. The experimental task required reproduction of a preselected criterion movement of the arm/hand on a linear positioning device. However, the relatively sizeable age differences between groups were associated with marked disparity in arm lengths. Such disparity is typically not represented by a linear or scalar measure of movement. Since movement of the limbs of the body takes place in an arc of a circle, an angular representation of movement at the primary joint (shoulder) seemed a plausible alternative. This representation was achieved by transposing the raw linear measures into an angle which fully considers differential arm lengths. The results of the linear analyses of absolute error indicated that the 7- to 9-yr.-old group was significantly inferior in accuracy of recall to both the 10- to 12- and 14- to 16-yr.-old groups. Also, the 7- to 9-yr.-old group did not appear to rehearse spontaneously as did the older children. Data transposed into an angular form yielded quite different results from the linear ones reported. In this instance the 7- to 9-yr.-old group remained significantly inferior in accuracy of recall to the 14- to 16-yr.-old group but not the 10- to 12-yr.-olds. Of most interest is that the linear evidence for spontaneity of rehearsal in the two older groups was not sustained by the angular data.