with physical limitations and have smaller pools of financial resources, reliance on families to provide care in later life is prevalent. Understanding the effects of family cohesion among these groups will become increasingly important as more and more families become responsible for care.

Affective Forecasting

There has been a recent proliferation in research examining affective forecasting—the ability to predict future feelings (Gilbert, Driver-Linn, and Wilson, 2002; Kahneman, 1994; Loewenstein and Frederick, 1997; Loewenstein and Schkade, 1999; Mellers and McGraw, 2001; Wilson and Gilbert, 2003). Most recently, researchers have measured both predicted and experienced affective responses in an effort to identify and explain inaccuracies in affective forecasting. Virtually all of the studies examining different aspects of affective forecasting, however, have involved young or, in some cases, middle-aged adult samples. And with the exception of quite a large body of research on affective forecasting involving changes in health conditions, most studies have focused on the types of events that are especially important to young and middle-aged adults, such as romantic break-ups or professional setbacks. Even when it comes to health, however, there has been virtually no systematic research on affective forecasting by older people.

The research on affective forecasting suggests that individuals are generally accurate in making predictions about the valence (positive or negative) of future emotional experiences (Wilson and Gilbert, 2003) and about the specific discrete emotion they will experience (Robinson and Clore, 2001). However, these results must be qualified by other findings that suggest that people have overly simplistic, schema-driven perceptions of the emotional responses to events when thinking about the distant as opposed to near future (Liberman, Sagristano, and Trope, 2002). Furthermore, research suggests that for events that elicit a more complex blend of emotions, individuals may not be as good at predicting the specific mixture of emotions they will experience (Wilson and Gilbert, 2003). Overall, though, people are relatively accurate in identifying specific future emotions.

In contrast, individuals are not accurate at predicting the intensity and duration of their emotional experiences (Baron, 1992; Coughlan and Connoly, 2001; Mellers and McGraw, 2001; Wilson and Gilbert, 2003). Individuals tend to overestimate the enduring effect that future events will have on their emotional well-being. Measurement of this bias has typically used self-report scales and behavioral forecasting measures. This bias is important because evidence suggests that social behaviors and decisions are influenced by affective forecasts.

Recent research has focused on the mechanisms that account for this



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement