Accuracy and Prediction in Throwing: A Taxonomic Analysis of Children's Performance

Abstract

The throwing and prediction performances of first, third and fifth grade boys and girls were analyzed within the framework of a four-part taxonomy originally conceived by Fitts (1965). Throwing performance was assessed under task conditions which varied the motion states of the thrower's body and the target (stationary or moving) by use of a dual pendulum apparatus. Accuracy scores were highest in a condition where both body and target were stationary, and lowest where both body and target were moving. Task conditions requiring motion of only target or of body were of intermediate difficulty, and scores for these tasks were not significantly different from each other. There was evidence of learning across trial blocks for all tasks, but no indication that rates of acquisition differed for the task types. Likewise, significant main effects were observed for age levels but no age X task type interactions were disclosed. Boys were more accurate than girls across task conditions, most noticeably on the two most difficult tasks. Comparison of subjects' ability to predict, from a stationary body position, the coincidence of the moving target with a standard reference point, and their ability to predict the coincidence of their moving body with the same reference point revealed lower error scores on the former prediction task.