engleski

Control theory model in a two-dimensional pursuit tracking task

Computer models of neural processes represent a hypothetical organization and architecture of real neural processes. A branch of applied control theory – Perceptual Control Theory (PCT) offers such computer models that can be quantitatively compared to behaviors of the subjects. The aim of this study was to apply PCT and to do quantitative analysis of models’ performance in a two-dimensional pursuit tracking task, as well as to show the possibilities of using PCT in experimental psychology studies. The study included 14 subjects. Models’ parameters were adjusted for each subject. Subjects and models went through four situations of pursuit tracking – two random generated, circular and a diagonal path. A quantitative comparison of subject’s behavior and model’s behavior was made. The results demonstrate that the models reproduced subjects’ behavior very accurately – in the first situation the correlation between subjects’ and models’ behavior was r = 0, 99. Same model was used to predict subjects’ behavior in new situations and they did it very accurately. In random moving target situation, the error was 2, 12% of the moving span, in the random fast moving target situation, the error was 2, 64% ; in the circular 2, 4% ; and the diagonal 2, 3%. Models followed the target more accurately than the subjects in random moving target situation: (t=4, 304, df=26, p=0, 00021) and (t=6, 190, df=26, p=0, 000002) respectively. There were no statisticaly significant differences found in the third and fourth situations: (t=1, 597, df=26, p=0, 1224), (t=-0, 936, df=26, p=0, 357). The results show a good plausibility of the model of the nervous system that is responsible for the pursuit tracking task.

perceptual control theory ; pursuit tracking task ; computer model ; coordination

nije evidentirano