Normal aging impairs the representation and integration (binding) of spatial and temporal context in episodic memory. We directly compare age differences in episodic memory in relation to processing spatial and temporal context. As part of the COGITO study, 101 younger and 103 older participants trained an object-location serial recall task for 100 sessions. Training exacerbated the recall deficit of older relative to younger adults. Younger adults improved in recall performance on both spatial and temporal dimensions. In contrast, older adults improved on the spatial dimension only. Individual differences in pretest performance and change were positively correlated across dimensions among younger adults but negatively related among older adults. We conclude that older adults are impaired at simultaneously processing spatial and temporal context and preferentially process spatial at the expense of temporal context.
Keywords: aging, training, context binding, latent change score model, representational precisionEpisodic memories include information about the content of an event, such as the name of a person, and context, such as where and when this person has been met (Johnson, Hashtroudi, & Lindsay, 1993;Light, 1991;Tulving, 1983). Cross-sectional (Cansino, 2009;Nilsson et al., 1997;Salthouse, 1998) and longitudinal (de Frias, Lövdén, Lindenberger, &Nilsson, 2007; Rönn-lund, Nyberg, Bäckman, & Nilsson, 2005) research has shown that episodic memory performance declines with adult age. In general, older individuals have particular difficulties in recollecting contextual information, whereas memory for content is less affected (Spencer & Raz, 1994. However, several findings challenge such a general conclusion and ask for detailed analysis of the characteristics of context, including how it relates to content (Spencer & Raz, 1995). For example, Chalfonte and Johnson (1996) segmented complex memories into memory for features (e.g., item identity, spatial, and color context) and memory for associations (i.e., binding) between features. When assessing these single features in separation, the authors found age-related differences in memory for location but not in memory for identity or color. Thus, age-related differences in source memory depend on the characteristics of the feature dimension. In addition, Chalfonte and Johnson (1996) also observed that feature binding is particularly impaired in older adults (see also Bayen, Phelps, & Spaniol, 2000;Mitchell, Johnson, Raye, Mather, & D'Esposito, 2000;Naveh-Benjamin, 2000;Naveh-Benjamin et al., 2009;Shing et al., 2010;Shing, Werkle-Bergner, Li, & Lindenberger, 2008).Spatial and temporal coordinates are the most prominent aspects of context (e.g., Brown & McCormack, 2006;Gallistel, 1990;Kahneman, Treisman, & Gibbs, 1992;Tulving, 1983;Xu & Chun, 2009). In a seminal article, Hasher and Zacks (1979) argued that age-related differences in memory for both spatial and temporalorder information were minor because such information is processed automatically. In fact, several att...