Skip to content Skip to navigation

Affect Valuation Index


The Affect Valuation Index (AVI) was developed in 2001 by Jeanne Tsai and Brian Knutson to distinguish between “ideal affect” (the affective states that people value or ideally want to feel) and “actual affect” (the affective states that people actually feel). At the time Tsai and Knutson began their work, most researchers had neglected “ideal affect,” and therefore, Tsai and Knutson had to develop a way of measuring it. 


Development. Because Tsai and Knutson were interested in using their measure in different cultural contexts, they based their measure on the affective circumplex (the mapping of different emotional states into the dimensions of arousal and valence; see Feldman Barrett & Russell, 1999 for discussion; Watson & Tellegen, 1985), which has been replicated across cultures and languages (e.g., Russell, Lewicka, & Nitt, 1989; Russell, 1991). The authors based the AVI on existing measures of actual affect that sampled all of the octants of the affective circumplex (e.g., Feldman-Barrett, 1996; Larsen & Diener, 1992). They experimented with various forms of the AVI before settling on the current version. 

Description and Use. The AVI includes a measure of actual affect and a measure of ideal affect. The measure of actual affect asks respondents to use a 5-point rating scale (1=not at all, 5 = an extreme amount) to indicate how much they actually feel a number of states on average (other time frames can be used such as “over the course of a typical week” “right now”). This measure is almost identical to measures developed by other researchers. The measure of ideal affect asks respondents to use the same rating scale to indicate how much they ideally want to feel the same states on average (other time frames can be used such as “over the course of a typical week” “right now”). In a typical study, we counterbalance the order in which the actual and ideal affect measures are administered. 

Analysis. Because of cultural differences in response styles, we ipsatize participants’ scores (standardize within individual). The AVI includes some items that are not equivalent across cultures and therefore serve as “filler” items in our studies (i.e., they are not calculated in the aggregate scores), which allows us to ipsatize scores. We then create aggregate scores for each octant of the affective circumplex (i.e., high arousal states, high arousal positive states, positive states, low arousal positive states, low arousal states, low arousal negative states, negative states, high arousal negative states) for actual affect and for ideal affect. As a result, we end up with sixteen scores (eight for actual affect, eight for ideal affect). In our studies to date, we have primarily focused on actual and ideal high arousal positive states (actual HAP, ideal HAP) and actual and ideal low arousal positive states (actual LAP, ideal LAP). We recommend using ipsatized scores for cultural group comparisons, but using raw scores when conducting correlational analyses (because ipsatizing can significantly reduce variance). 

Psychometric Properties.  Each actual and ideal affect aggregate shows high internal consistency and moderate-to-high test-retest reliability (one-month and two month). The measure of ideal affect also shows good discriminant validity. Specifically, ideal affect is either weakly or not significantly correlated with prevention focus, promotion focus, sensation seeking, arousal seeking tendency. Consistent with previous research, actual affect is correlated with these individual difference measures. Ideal affect also shows good convergent validity: self-reports of ideal affect are consistent with behavioral measures of ideal affect (e.g., consumer product choices that reflect different affective states). “Ideal affect” is moderately correlated with “should affect” (how people think they should feel), suggesting that they are related, but different constructs. The initial psychometric properties are reported in Tsai, Knutson, & Fung (2006). 

Cultures. The AVI has been translated into Chinese, Japanese, Korean, French, and German. Although it has mainly been used with college students, it has also been used with community samples of diverse ages (18-80). 


Several researchers are currently using the AVI in their work. From our lab, the following papers report findings using the AVI: 

Tsai, J.L. (2007). Ideal affect: Cultural causes and behavioral consequences. Perspectives on Psychological Science, 2, 242-259.   

Tsai, J.L., Miao, F., Seppala, E., Fung, H., & Yeung, D. (2007). Influence and adjustment goals: Sources of cultural differences in ideal affect. Journal of Personality and Social Psychology, 92, 1102-1117.   

Tsai, J.L., Miao, F., & Seppala, E. (2007). Good feelings in Christianity and Buddhism: Religious differences in ideal affect. Personality and Social Psychology Bulletin,33, 409-421.   

Tsai, J.L. Knutson, B., & Fung, H. H. (2006). Cultural variation in affect valuation. Journal of Personality and Social Psychology, 90, 288-307.


We are currently collecting more data to examine the generalizability of the AVI for samples that vary in cultural background, age, and socioeconomic status. We are also developing a measure of avoided affect (i.e., the affective states that people want to avoid). 


Like any instrument, the AVI has strengths and limitations. Researchers should seriously consider both before deciding to use the AVI.   

The AVI has several strengths, including: (1) good psychometric properties for the samples studied, (2) easy administration, and (3) easy scoring. In addition, it has been used successfully in various studies and yielded important findings. 

The AVI has several limitations, including: (1) it cannot be used with young children; therefore, we have developed other methods to measure ideal affect in children between the ages of 2-5, (2) at times, participants may be confused about whether they are completing the measure of actual or ideal affect; therefore, it is important to make sure that participants realize that the questionnaires are asking different questions, (3) research assistants may make errors when entering the data, so data should be checked several times, (4) the measure is relatively long and therefore, more complex states like shame, compassion cannot be easily added, and (5) in some cases, researchers may want to describe the meaning of some of the states before administering the instrument. 


You are welcome to use the AVI in your research. We ask that in exchange for using the AVI, you let us know the findings of your studies, and send us any publications of your work with the AVI. To download a copy of the AVI, click here.

To obtain one of the translated versions of the AVI, please contact Lab. 



Feldman Barrett, L. (1996). Hedonic tone, perceived arousal, and item desirability: Three components of self-reported mood. Cognition and Emotion, 19, 47-68. 

Feldman Barrett, L., & Russell, J. A. (1999). The structure of current affect: Controversies and emerging consensus. Psychological Science, 8, 10-14. 

Larsen, R. J. & Diener, E. (1992). Promises and problems with the circumplex model of emotion. In M. S. Clark (Ed.), Review of personality and social psychology: Emotion (pp. 25-59). Newbury Park: Sage. 

Russell, J. A. (1991). Culture and the categorization of emotions. Psychological Bulletin, 110, 426-450. 

Russell, J.A. (2003). Core affect and the psychological construction of emotion. Psychological Review, 110, 145-172. 

Russell, J.A., Lewicka, M., & Nitt, T. (1989). A cross-cultural study of a circumplex model of affect. Journal of Personality and Social Psychology, 57, 848-856. 

Tsai, J.L. Knutson, B., & Fung, H. H. (2006). Cultural variation in affect valuation. Journal of Personality and Social Psychology, 90, 288-307. 

Watson, D., & Tellegen, A. (1985). Toward a consensual structure of mood. Psychological Bulletin, 98, 219-235.