Journal of Personality and Social Psychology 1999, Vol. 77, No. 3, 600-619 Copyright 1999 by the American Psychological Association, Inc. 0022-3514/99/S3.00 Structure of Self-Reported Current Affect: Integration and Beyond Michelle S. M. Yik and James A. Russell University of British Columbia Lisa Feldman Barrett Boston College Current affect has been described with various dimensions and structures, including J. A. Russell's (1980) circumplex, D. Watson and A. Tellegen's (1985) positive and negative affect, R. E. Thayer's (1989) tense and energetic arousal, and R. J. Larsen and E. Diener's (1992) 8 combinations of pleasantness and activation. These 4 structures each presuppose bipolar dimensions and have been thought of as interchangeable or 45° rotations of one another, but past data were inconsistent. Huge but not perfect overlap among these four structures was found here in 2 studies of self-reported current affect (Ate = 198 and 217) that controlled for random and systematic errors of measurement. The 4 structures were integrated into a common space defined by 2 bipolar dimensions. Current affect has been described with various dimensions and structures, including Russell's (1980) circumplex, Watson and Tellegen's (1985) positive and negative affect, Thayer's (1989) tense and energetic arousal, and Larsen and Diener's (1992) eight combinations of pleasantness and activation. In this article, we argue that it is possible to integrate these various concepts in a simple scheme, thereby taking a large step towards unifying dimensional approaches to affect as a person experiences and reports it. There have been three major approaches to the broad, general dimensions of reported affect. One approach has focused on acti- vation (also known as arousal, energy, tension, and activity). Sleep, drowsiness, alertness, activation, hyperactivation, and, fi- nally, frenzied excitement describe increasing degrees of activa- tion. Activation was prominent in psychological writing through- out most of this century (Berlyne, 1960; Cannon, 1927; Duffy, 1957; Hebb, 1955; Lindsley, 1951; Mandler, 1984; Schachter & Singer, 1962; Zillmann, 1983). Thayer (1989, 1996) has spear- headed a research program on the causes and consequences of everyday affect, conceptualized in terms of types and degrees of activation, demonstrating that self-reported activation is related to everything from diet to personality to neurochemistry. Michelle S. M. Yik and James A. Russell, Department of Psychology, University of British Columbia, Vancouver, British Columbia, Canada; Lisa Feldman Barrett, Department of Psychology, Boston College. Preparation of this article was facilitated by a grant from the Social Sciences and Humanities Research Council of Canada and by a University Graduate Fellowship to Michelle S. M. Yik. We thank Amy Ip and Paula Castellon for their generous help in preparing this article and Steven So for his help in preparing the figures. We profusely thank Michael Browne, Jeremy Miles, and James Steiger for their help on statistical matters. Correspondence concerning this article should be addressed to Michelle S. M. Yik, who is now at the School of Business, Pokfulam Road, University of Hong Kong, Hong Kong, or to James A. Russell, Department of Psychology, University of British Columbia, 2136 West Mall, Vancou- ver, British Columbia, Canada V6T 1Z4. Electronic mail may be sent to myik@business.hku.hk or to jrussell@psych.ubc.ca. In a second and older approach, the emphasis has been on valence. When Wundt, Stumpf, and Titchener introspected, they concluded that pleasure-displeasure was an elementary dimension of conscious feeling (see Reisenzein, 1992; Reisenzein & Schon- pflug, 1992). Pleasure is currently making a comeback as a key concept in diverse areas (Abramson & Pinkerton, 1995; Cabanac, 1995; Piasecki & Kenford, 1997). Valence sometimes is the only general factor found in self-reports of affect (Feldman, 1995; McConville & Cooper, 1992; Wessman & Ricks, 1966; Williams, 1989, 1990). Indeed, in some articles, the words mood, affect, and valence are used interchangeably. A third approach has been to include both valence and activation as separate and equally emphasized dimensions within one de- scriptive structure (Bradley, 1994; Lang, 1978, 1994; Lang, Brad- ley, & Cuthbert, 1992; Osgood, 1966; Osgood, May, & Miron, 1975; Russell, 1978, Wundt, 1924). In this article, we pursue this third approach. Of course, other broad dimensions have been proposed, and we do not want to be misunderstood as claiming that valence and activation are the only dimensions of affect. What we do believe is that these two are essential and that it is essential to study them simultaneously. Doing so provides a conceptual clarity otherwise missing. Psychometric research that has emphasized activation over valence or vice versa has resulted in a puzzling state of affairs. Here are two examples. First, Watson and Tellegen's (1985) psychometric research on positive and negative affect led not to a single dimension, but to a two-dimensional structure (as shown in Figure 1). They interpreted both dimensions in terms of valence—high versus low positive affect and high versus low negative affect—although activation was implicit in their framework. Second, Thayer's (1986) psycho- metric work on activation led not to a single activation- deactivation dimension, but to a two-dimensional structure (also shown in Figure 1). Thus, the study of valence alone may seem to yield two valence dimensions, just as the study of activation alone may seem to yield two activation dimensions. A structure that unites valence and activation resolves such puzzles. According to the hypothesis to be examined in this article, when all dimensions are measured simultaneously, the result will not be four dimen- sions, but two. 600