Psychophvsiotogy, 32 (1995), 436-443. Cambridge University Press. Printed in the USA. Copyright f^) 1995 Society for Psyehophysiological Research Psychophysiological responses as indices of affective dimensions CHARLOTTE VANOYEN WITVLIET AND SCOTT R. VRANA Department of Psychological Sciences, Purdue University, West Lafayette, IN, USA Abstract The startle reflex, facial electromyogram (EMG), and autonomic nervous sy.slem responses were exaniitied during imagery varying in affective valence and arousal. Subjects (N = 48) imagined affective situations during lone-cucd 8-s trials. Startle blink magnitudes were larger and latencies faster during negatively valent than during positively valent conditions and during high-arousal than during low-arousal conditions. Greatest heart rate acceleration and fastest and largest skin conductance responses to startle probes occurred during high-arousal imagery. Zygo- matic and orbicularis oculi facial muscle activities were higher during high-arousal imagery, whereas corrugator muscle activity was higher during low-arousal imagery. Zygomatic and corrugator activity also varied with emo- tional valence. The startle and facial EMG responses are most parsimoniously organized by the negative affect (NA) and positive affect (PA) dimensions, respectively. This NA/PA framework integrates previous research, dimensional theories of emotional behavior, and physiological assessment of pathological etnotion. Descriptors: Startle, Reflex modulation. Emotion, Imagery, Skin conductance. Heart rate. Facial EMG Most of the variance in emotional judgments and behavior can be accounted for within a two-dimensional affective space (Mehrabian, 1970; Mehrabian & Russell, 1974; Russell, 1980). The space has often been defined by axes labeled valence (pos- itive-negative) and arousal (calm-arou.sing). These dimensions are thought to represent basic strategic dispositions of organ- isms, broadly affecting the direction and intensity of all affec- tive behavior (Lang, 1994). Other emotion theorists describing the same affective space have preferred to rotate the valence and arousal axes 45", labeling the resulting axes negative affect (NA) and positive affect (PA) (Watson & Clark, 1984; Wat.son & Telle- gen, 1985). In this conceptualization, the PA axis extends from highly arousing positive emotion on one end to low-arousal neg- ative emotion on the other; NA extends from highly arousing negative emotion to low-arousal positive emotion (Larsen & Diener, 1992). Recent work has demonstrated covariation between physi- ological variables and emotional dimensions (Greenwald, Cook, & Lang, 1989; Lang, Greenwald, Bradley, & Hamm, 1993b). One of the most important and consistently reported findings This research was supported by National Institute of Mental Health grant MH4686I. This study was completed by C.V.W. under the super- vision of S.R.V. in partial fulfillment of the requirements for a Master's degree. We thank Janice Kelly and Stephen Tiffany for helpful com- ments on earlier drafts of this manuscript, Robert l.evenson for his con- structive comments while serving as action editor for this paper, and Natalie Milbrandt, Steve Branson, and Jennifer Beckham for assistance in data collection. Address reprint requests to: Charlotte vanOyen Witvliet or .Scott K. Vrana, Department of Psychological Sciences, Purdue University, West Lafayette, IN 47907, USA. has been the relationship between the startle reflex eyeblink and the valence dimension. Eyeblink magnitudes to startle probes are larger in the context of negatively valent ernotion than in the context of neutral and positively valent emotion (Bradley, Cuth- bert, & Lang, 1990, 1991; Cook, Hawk, Davis, & Stevcn.son, 1991; Hawk, Stevenson, & Cook, 1992; Vrana & Lang, 1990; Vrana, Spence, & Lang, 1988). However, the affective tnateri- als u.sed in many studies have confounded valence with arousal. Specifically, the negatively valetit materials used were also highly arousing, not allowing for definitive statetnents about the rela- tive modulatory effects of these two dimensions. Recent stud- ies have attempted to disentangle the effects of valence atid arousal on startle modulation. Both Cook et al. (1991) atid Hawk et al. (1992) studied tnod- ulation of the startle blink with affective materials not con- founded on valence and arousal. Startle blink magnitudes were significantly greater during itnagery of negative versus positive emotions in both studies, with arousal having mixed effects on tnagnitude. In contrast, startle latencies were significantly facil- itated (i.e., of shorter duration) during imagery of high-arousal versus low-arousal emotions but were not affected by affective valence. These researchers concluded that startle blink tnagni- tude indexes affective valence, whereas blink latency indexes affective arousal. Cuthbert, Bradley, and Lang (1990) examined startle blink magnitudes under tnultiple levels of arousal and valence protnpted by slides. Startle magnitudes were largest for high-arousal neg- atively valent affect and smallest for high-arousal positively valent affect slides. No differences were found between blinks during negatively valent and positively valent slides at low levels of arousal. Based on the.se data, Lang, Bradley, Cuthbert, and 436