Memory and emotion

I am interested in what influences people’s recall of their past emotions. Understanding memory for emotions is important because our conceptions of our lives, especially meaningful (and therefore usually emotional) experiences, are based primarily on what we recall. Additionally, most measures of emotion are retrospective measures, so it is important to know whether such measures accurately reflect momentary experiences. In addition to retrospective measures, I use experience sampling methodology which relies less on recall (Scollon, Kim-Prieto, & Diener, 2003 [pdf]). Respondents in my studies record their emotions on palmtop computers at several random moments each day when signaled. Clear discrepancies have emerged between retrospective reports and momentary or “on-line” reports. Rather than treating inaccuracies in recall as error, however, my research seeks to uncover systematic mechanisms that guide memory for emotions. For example, I found that one’s emotional self-concept predicts memory for emotions, even after controlling for on-line experiences (Scollon, Diener, Oishi, & Biswas-Diener, 2004 [pdf]). In other words, instead of searching their memories and summing up specific instances of emotion to arrive at an estimate of their past emotions, people use heuristic information such as the self-concept to inform their recall. In addition, this effect has been replicated among different cultural groups.

Memory for emotions also determines the choices people make, as my colleagues at Illinois and I have demonstrated. We had participants record their on-line affect several times each day during their Spring Break vacations (Wirtz, Krueger, Scollon, & Diener, 2003 [pdf]). The best predictor of whether students wanted to take a similar vacation in the future was not on-line emotion during the vacation, but rather their memory for emotions. Moreover, expectations about the vacation influenced memories. These findings challenge our conceptions of happiness. Is happiness the sum of pleasant experiences or pleasant memories? Are people aware of the role of memory for emotions in their overall happiness?

Back to top of page

Culture and emotion

By using multiple measures of emotion, my research provides an intricate picture of the emotional lives of individuals in different cultures. In doing so, I hope to uncover sources of cultural differences in subjective well-being (SWB), in addition to illuminating the study of individual differences in emotion. Furthermore, I believe that challenges in cross-cultural measurement can inform the measurement of emotion within cultures.

On-line emotion. A robust and consistent finding in the well-being literature is that individuals from Asian cultures tend to report lower SWB than individuals from North American and Latin societies. However, most cross-cultural comparisons of emotion have been based on recalled reports of emotion that are vulnerable to memory reconstruction. Thus, without on-line measures of emotion, it is unclear whether cultural differences in emotion are due to differences in everyday emotional experience or differences in memory for emotions.

One of my studies which examined on-line and recalled emotions in European Americans, Latino Americans, Asian Americans, Japanese in Japan, and Indians in India, represents one of the few attempts to empirically establish base-rates of specific on-line emotions in different cultures (Scollon et al., 2004). In both on-line and retrospective measures, European and Latino Americans scored higher on reports of pleasant emotion and lower on reports of unpleasant emotion than Asian Americans, Japanese, and Indians. However, on-line methods (especially for negative emotion) showed smaller cultural differences than recall methods, suggesting that cultural norms may have stronger effects on memory than on momentary experience. Groups also differed more in reports of pleasant emotion than in unpleasant emotion, a finding that supports the notion that pleasant emotions may be more influenced by culture and socialization than unpleasant emotions. Particularly striking cultural differences emerged on both on-line and recalled measures of pride. European and Latino Americans reported much more pride than Asian Americans, Japanese, and Indians. This finding resonates with cultural theories that suggest that pride in individualistic societies is highly valued because it emphasizes the uniqueness of the self. Importantly, this research demonstrates that culture influences momentary affective experiences as well as memory for emotions.

Structure of emotion. The intersection of culture and emotion also provides an opportunity to examine important structural issues. One issue that my research addresses is whether the relation between pleasant and unpleasant affect varies by culture and by level of analysis. I have found that, regardless of culture, pleasant and unpleasant feelings are negatively correlated in momentary experience. However, at trait levels (i.e., between-persons), pleasant and unpleasant affect are uncorrelated among European and Latino Americans, and positively correlated among Asian Americans, Japanese, and Indians (Scollon, Diener, Oishi, Biswas-Diener, 2005 [pdf]). In particular, individuals in Asian cultures who experience a great deal of pride also report more guilt, sadness, and irritation, a finding which might explain lower levels of SWB in Asian cultures. This study highlights the importance of studying emotion at both levels. One avenue of future research is to examine why pleasant and unpleasant affect are positively related in Asian cultures.

Indigenous emotions. Translation adds an extra challenge to the study of emotion across cultures. For instance, finding the same structure of emotion in different cultures is not sufficient evidence for measurement invariance if the structures are based on English or translated English emotions. Many cultures have emotions that do not have English equivalents and cannot be directly translated. By neglecting these culturally “indigenous emotions,” we might inflate the chances of detecting universal structure. Although indigenous emotions have received treatment in ethnographic studies, my research is the first to use experience sampling to track indigenous emotions (Scollon et al., 2004). I cluster analyzed emotions in Japan and India and found that indigenous emotions did not form distinct clusters from the pleasant and unpleasant dimensions that are often found in studies using translated English emotions. This is important because it shows that pleasantness and unpleasantness capture emotional experience in different cultures, and that indigenous emotions are fairly well-represented by English emotion words.

Back to top of page

Personality development: Increasing happiness and emotional stability over time

One obstacle to happiness is high negative affect or neuroticism. Thus, one way to become happier would be to reduce one’s level of neuroticism. Unfortunately, it is a commonly held belief (one that has also been promoted by trait theorists for decades) that traits such as neuroticism are “fixed like plaster.” Similarly, high stability coefficients of well-being measures have been taken as evidence that a person’s happiness is merely a reflection of personality. According to this bleak view then, both happiness and emotional stability levels are immutable. However, such conclusions are based on a limited approach to change. Until recently, most studies of change have focused on either stability or mean-levels of traits, both of which mask important processes that may be occurring at the individual level. Few studies have measured change at the within-person or individual level. In a reanalysis of data from a longitudinal study (see Headey & Wearing, 1989) that tracked the SWB of over 1,000 Australians over 9 years, I use latent growth modeling to estimate intraindividual differences in extraversion, neuroticism, and life satisfaction over time. I have found that personality does indeed change at the individual level. More importantly, I have identified systematic ways in which change occurs. First, growth in marital satisfaction and work satisfaction accompany increases in life satisfaction and decreases in neuroticism over time. In addition, increased work satisfaction is associated with increases in extraversion over time. I believe this is an exciting and important new line of research because of its potential to inform interventions aimed at improving well-being. Moreover, changes that occur at the individual level are, arguably, the changes that matter most to people’s lives. Click here for the full manuscript.

Back to top of page

Folk theories of the Good Life

Presumably people seek to maximize those aspects that define the good life; therefore, identifying cultural conceptions of the good life is fundamental to a science of well-being. My research examines folk theories of what makes life worth living. Rather than treating folk theories as error-ridden and uninformative, my research views them as a rich source of meaning, representing the intersection of shared beliefs and individual histories.

In a series of studies conducted with Dr. Laura King, we asked participants to make ratings about the desirability and moral goodness of a life as a function of its happiness, meaning, and wealth (King & Napa, 1998 [pdf]). People rated happy and meaningful lives as most desirable and morally good, and they rated happy people as more likely to go to heaven. Whereas researchers previously believed that people value money over happiness, my research shows that people want happiness and meaning more than wealth. In another set of studies, I included effort or challenge as another feature of the good life (Scollon & King, 2004 [pdf]). I found that folk theories equated the good life with the easy life when effort was conceptualized as number hours of work. Effort was rated an important part of a meaningful and happy life only when effort was framed as active engagement that was not energy depleting nor time-consuming. These folk concepts of effort suggest the possibility that people value hard work in the context of the good life, but only if active engagement does not come at the cost of other life interests.

These studies were the first empirical attempts at mapping folk theories of the good life onto the SWB literature, and I look forward to expanding this program of research to investigations of other cultures’ folk theories. In fact, a cross-cultural study of folk concepts of the good life could shed light on cross-cultural differences in SWB. In addition, findings with regard to effort suggest intriguing possibilities about the role of leisure in the good life.

Back to top of page