Differential effects of aging on processes underlying task switching Robert West a, * , Stephanie Travers b a Department of Psychology, Iowa State University, W112 Lagomarcino Hall, Ames, IA 50010, USA b Department of Psychology, Luther College, Decorah, IA, USA article info Article history: Accepted 2 March 2008 Available online 9 April 2008 Keywords: Aging Task switching Event-related brain potentials Cognitive control Partial Least Squares Analysis abstract In this study, we used event-related brain potentials (ERPs) to examine the effects of aging on processes underlying task switching. The response time data revealed an age-related increase in mixing costs before controlling for general slowing and no effect of aging on switching costs. In the cue-locked epoch, the ERP data revealed little effect of age on the parietal P3 related to cue encoding, an age-related decrease in parietal activity related to cue retrieval, and an age-related increase in the amplitude of the parietal and frontal activity related to task set configuration and rule mapping. In the target-locked epoch, there was differential neural recruitment in younger and older adults in response to task mixing. These data are consistent with the idea that older adults may not fully implement task set before onset of the target stimulus. Ó 2008 Elsevier Inc. All rights reserved. 1. Introduction The idea that later adulthood is associated with a decline in the efficiency of executive or cognitive control processes has a long history in the cognitive aging and neuropsychological literatures (Albert & Kaplin, 1980; Braver et al., 2001; Hasher & Zacks, 1979; Moscovitch & Winocur, 1992; West, 1996). Support for this posi- tion has been garnered from behavioral studies demonstrating significant age-related declines in performance on neuropsycho- logical tests that are considered general indices of executive functions (e.g., the Wisconsin Card Sorting Test (Rhodes, 2004) and the self-ordered pointing test (Daigneault & Braun, 1993)) and from experimental tasks that are designed to target-specific control processes such as goal maintenance (Braver et al., 2001) and focus switching (Verhaeghen & Basak, 2005; Verhaeghen & Hoyer, 2007). This behavioral work is complimented by studies incorporating electrophysiological and functional neuroimaging techniques that reveal effects of aging on the neural correlates of specific control processes underlying action monitoring (Falken- stein, Hoorman, & Hohnbein, 2001), task preparation (DiGirolamo et al., 2001; Kray, Eppinger, & Mecklinger, 2005), and interference control (Jonides et al., 2000). The current study used event-related brain potentials (ERPs) to extend the findings of recent work examining the effects of aging on processes underlying task switching (Kray et al., 2005; West, 2004; West & Moore, 2005). Specifically, we considered the effects of aging on the neural corre- lates of processes associated with task mixing, cue retrieval, task set configuration, and rule mapping in an explicit cue task switch- ing paradigm. The task switching paradigm has been used increasingly often to examine the effects of aging on the ability to manage the sequential processing demands of alternating between two or more tasks, and the ability to implement an appropriate task set in anticipation of an upcoming event (Kramer, Hahn, & Gopher, 1999; Mayr, 2001; Verhaeghen, Cerella, Bopp, & Basak, 2005). The first of these abilities is typically operationalized as task mixing costs that represent an increase in response time or decrease in response accuracy when performance of two tasks is mixed within a block of trials relative to when a single task is performed in a block of trials. The second ability is typically operationalized as task switching costs that represent an increase in response time or decrease in response accuracy that occurs when one alternates between tasks in a mixed block relative to when the same task is performed on consecutive trials in a mixed block. Behavioral and electrophysiological evidence reveals that switching costs may arise from the recruitment of diverse processes associated with cue retrieval (Logan & Bundesen, 2003), rule mapping (Logan & Bundesen, 2004; Mayr & Kliegel, 2003; Travers & West, in press), and task set configuration (Monsell & Mizon, 2006; Nicholson, Karayanidis, Bumak, Poboka, & Michie, 2006; Travers & West, in press). The extant literature often reveals differential effects of aging on behavioral indices of mixing and switching costs. The magni- tude of mixing costs is typically larger for older adults than for younger adults (Mayr, 2001; Verhaeghen et al., 2005). The age- related increase in mixing costs is generally larger than expected by general slowing (Verhaeghen et al., 2005), remains significant 0278-2626/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bandc.2008.03.001 * Corresponding author. Fax: +1 515 294 6424. E-mail address: rwest@iastate.edu (R. West). Brain and Cognition 68 (2008) 67–80 Contents lists available at ScienceDirect Brain and Cognition journal homepage: www.elsevier.com/locate/b&c