Introduction
Recent psychological research recognizes that people are inextricably linked to their social environments and to those around them. For example, people report a stronger preference for spending time with others rather than being alone and do so for a majority of their waking hours (Kahneman, Krueger, Schkade, Schwarz, & Stone, 2004). Moreover, people are more likely to experience a wide variety of health problems when lonely or isolated (see Cacioppo & Patrick, 2008). However, this is a paradox of sorts: Although people cannot seem to live without one another, they also sometimes cheat and manipulate each other, are physically aggressive and verbally offensive, lie, steal, and exhibit a number of other socially deleterious tendencies.
Given the prevalence of conflicted, antisocial, and otherwise unpleasant interactions with other people, researchers have been interested in factors that promote cooperative, prosocial, and satisfying relationships. Our focus in this article is specifically on empathy. In general, empathy seems to enable people to relate to others in a way that promotes cooperation and unity rather than conflict and isolation. Thus, an examination of potential changes in empathy over time affords new insights into how and why people help and relate positively to one another. Temporal changes in empathy might help explain certain interpersonal and societal trends that suggest people today are not as empathic as previous generations.
In the current article, we use cross-temporal meta-an*lytic methods to examine changes over time in American college students' dispositional empathy scores. We do so by using the time-lag method, which separates the effects of birth cohort from age by an*lyzing samples of people of the same age at different points in time. In this study, we compare college students from the late 1970s and early 1980s to college students in the 1990s and 2000s. By studying college students at each of these time periods, we are able to collect data from people who are from the same age group but different birth cohorts. Birth cohorts can be seen as sociocultural milieus (Stewart & Healy, 1989; Twenge, 2000), in that children growing up in the 1970s in the United States were exposed to different sociocultural norms than those growing up in the 2000s, despite being physically located in the same country. The logic underlying this approach is similar to that used in cross-cultural psychology to examine similarities and differences in the self-construals, traits, and behaviors of people across different sociocultural regions of the world (e.g., Choi, Nisbett, & Norenzayan, 1999; Heine & Lehman, 1997; Markus & Kitayama, 1991), except that we instead a**ess differences between birth cohort groups (rather than cultures). Several studies have used this method to find birth cohort differences in traits such as anxiety, self-esteem, narcissism, locus of control, and s**ual behavior (respectively, Twenge, 2000; Twenge & Campbell, 2001; Twenge, Konrath, Foster, Campbell, & Bushman, 2008; Twenge, Zhang, & Im, 2004; Wells & Twenge, 2005). These studies used metaan*lytic methods to compare samples of college students or children who completed the same psychological questionnaires at different points in time. In the method of crosstemporal meta-an*lysis, researchers correlate the mean scores on a measure with the year of data collection, weighting for sample size, to a**ess changes over time on particular measures (e.g., Twenge, 2000).