It Was a Most Unusual Time: How Memory Bias Engenders Nostalgic Preferences

Nostalgic preferences are widespread—people believe past movies, music, television shows, places, and periods of life to have been better than their present counterparts. Three experiments explored the cognitive underpinnings of nostalgic preferences. Participants rated past experiences to have been superior to similar present and recent experiences. These nostalgic preferences appeared to be due to the belief that the atypically positive experiences that participants recalled at the time of judgment were more representative of their past experiences than of their present experiences. Copyright © 2012 John Wiley & Sons, Ltd. key words nostalgia; preferences; availability; judgment and decision making; memory Nostalgic preferences, beliefs that past experiences were better than their present counterparts, are widespread. People believe everything from the general state of their country to the quality of their television programming has declined from its past zenith (CBS News/New York Times, 2006; Princeton Survey Research Associates International, 2005). This more positive evaluation of the past is reflected in a general preference for the music, movies, movie stars, fashion models, and automobiles that were popular during one’s youth (Holbrook & Schindler, 1989; Schindler & Holbrook, 2003). Which past experiences evoke nostalgic preferences varies according to the year in which one was born, not the specific experiences evaluated (Holbrook & Schindler, 1996). Rather than reflect the actual superiority of the past, then, nostalgia and nostalgic preferences have cognitive and motivated origins (e.g., Leboe & Ansons, 2006; Routledge, Arndt, Sedikides, & Wildschut, 2008). In this paper, I suggest that one factor underlying nostalgic preferences may be differences in the perceived representativeness of memories recalled when evaluating past and present experiences. People often base category judgments on a few cognitively accessible exemplars, which tend to be atypical members of the category (Frederickson & Kahneman, 1993; Hastie & Kumar, 1979; Morewedge, Gilbert, & Wilson, 2005; Morewedge & Todorov, ; Risen & Gilovich, 2008; Smith & Zárate, 1992; Tversky & Kahneman, 1973; Wirtz, Kruger, Scollon, & Diener, 2003). When predicting their enjoyment of a football game that they are about to watch, for example, fans tend to recall the best game that they can remember and base their prediction on their enjoyment of that unusually good game. Consequently, category judgments tend to be extreme because people only correct for the atypicality of the instances they remember if their atypicality is obvious or is made obvious (Hamill, Wilson, & Nisbett, 1980; Morewedge et al., 2005). This tendency to base judgments of categories on atypical exemplars may, in part, underlie nostalgic preferences for past experiences. In most domains of memory and attention, bad dominates good (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001; Ochsner, 2000). In the domain of autobiographical memory, however, positive events are better remembered than negative events (Mitchell, Thompson, Peterson, & Cronk, 1997; Walker, Skowronski, & Thompson, 2003; Walker, Vogl, & Thompson, 1997). People exhibit a fading affect bias, for example, such that they more quickly forget negative than positive past experiences (Ritchie et al., 2006). One consequence of biased memory for positive experiences is that if atypically positive exemplars are recalled at the time of judgment, they should seem more representative of past than present categories of which those exemplars are a member. This, in turn, may lead to the belief that categories of past experience were better than their present counterparts. I report three experiments that tested the hypothesis that nostalgic preferences for past experience are, in part, due to (i) a tendency to recall atypically positive memories at the time of judgment, coupled with (ii) a belief that those atypically positive memories are more representative of past than present categories of experience. Experiment 1 tested my hypothesis by examining how better memory for atypically positive experiences influences preferences for television programs of past decades and the present decade. Experiment 2 tested my hypothesis by examining how better memory for atypically positive experiences influences preferences for movies released in the year in which participants graduated from high school, or the most recent full year. Experiment 3 tested whether people see all past experiences as more similar to the atypically positive experiences they recall because people believe there were fewer bad past experiences or because all past experiences are believed to be more similar in quality to those atypically positive experiences. EXPERIMENT 1: PAST AND PRESENT TELEVISION PROGRAMS In the summer of 2005, participants first rated the average quality of television programs of a past decade (i.e., the *Correspondence to: Carey K. Morewedge, Carnegie Mellon University, 5000 Forbes Avenue, Pittsburgh, PA 15213, USA. E-mail: morewedge@ cmu.edu Copyright © 2012 John Wiley & Sons, Ltd. Journal of Behavioral Decision Making, J. Behav. Dec. Making (2012) Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/bdm.1767 1980s or the 1990s) or their present decade (i.e., the 2000s). Next, participants recalled and rated the quality of either any single television program from that decade or the best television program that they could remember from that decade. Finally, all participants rated how representative the program they recalled was of all of television programs of its decade. I expected that participants would exhibit a nostalgic preference for past experiences: Participants would consider television programs of the 1980s and the 1990s to have been better on average than television programs of the 2000s. More important, I predicted that this nostalgic preference would be due to a bias to recall atypically good programs, coupled with the belief that the programs they recalled were more representative of all programs of their past (the 1980s and the 1990s) than of their present (the 2000s).