| Abstract: | A partial replication of research by Balshan (1962) was undertaken in order to investigate the existence of a general tension factor operating during rest and during auditory stimulation. Thirty female subjects were each exposed to one 20-min session consisting of 15 min of quiet rest followed by 5 min of binaural white noise stimulation. A computer-controlled scanning electromyograph sampled integrated surface EMG activity from a total of eight muscle sites on the head, neck and limbs. Tensional profiles for all eight EMG sites were assembled every 1.84 sec, resulting in an 8×650 data array for each subject. Statistical analyses revealed overall EMG increments at stimulus onset. R-factor analyses on principal components for group data, as in Balshan (1962), revealed group general factors for both prestimulus and stimulus periods. However, P-factoring on principal components of the within-subject data arrays demonstrated that few subjects met minimal loading criteria for a general factor operating in individuals. Such within-subject EMG-site covariations as were observed were reduced considerably when startle responses at stimulus onset were excluded from the principal-component analyses. It is concluded that the group factors obtained in the present experiment and by Balshan (1962) were artifacts of two influences: the erroneous assumption of across-subject equivalence of EMG-site levels, and the summation of uncorrelated EMG activity which biased the group component solutions toward artifactual generality. |
