Arousal is conceptualized here as a unitary emotional response dimension ranging from sleep to frantic excitement. This dimension is readily assessed with semantic differential measures. Data obtained using this technique have shown chat variations in the arousing quality of situations is indeed a primary and unitary factor. Several nonverbal measures have also been identified which are intercorrelated and essentially define a measure of responsiveness or arousal in social situations. These are vocal activity (including positive as well as negative), facial activity (including positive and negative expressions), speech rate, and speech volume. In considering the problems that are associated with physiological measures of arousal, Berlyne (1967) commented, “All this need not worry us unduly, although it certainly calls for circumspection, as long as we regard arousal as a dimension and not as a phenomenon-not, that is, as a process that goes on in one location in the central nervous system” (p. 12).
Thus, even though there is insufficient knowledge of the relations between the primary and secondary physiological measures of arousal (especially over a variety of situations), Berlyne’s (1960) definition of arousal still has considerable heuristic value. Further, as already noted, there is evidence showing a correlation of verbal reports of arousal or activity with physiological indexes of arousal.

Mehrabian & Russell (1974)