Episodic visual recognition memory and aging: A case of preserved function Robert Sekuler, Michael J. Kahana and Arthur Wingfield Volen National Center for Complex Systems Brandeis University Visual episodic recognition memory was assessed in young (mean age 22.5 years) and older (mean age 74.1 years) participants. To isolate purely visual memory, recognition was tested with sets of briefly-presented compound sinusoidal gratings, which minimized age-related dif- ferences in visual processing, and resisted verbal encoding. Recognition was measured after delays of 1, 2 or 4 seconds. Overall, neither accuracy of recognition nor speed of response differed significantly between groups or with probe delay, which strengthens recent claims that visual memory tends to be spared during the course of normal aging. Studies with rehearsable verbal stimuli such as letters, words or sentences suggest that normal adult aging is as- sociated with diminished short-term memory (e.g., Kausler, 1994; Wingfield & Kahana, 2002; Wingfield & Stine- Morrow, 2000). However, recent research raises questions about whether this association extends to memory for other kinds of test materials. For example, using a match-to- sample procedure, McIntosh et al. (1999) found that younger and older observers had virtually identical short-term mem- ory for a single, briefly-presented, vertical sinusoidal grating. Functional neuroimaging studies of older and younger partic- ipants’ brains led McIntosh et al. to speculate that a compen- satory reorganization of neural circuits helped to preserve vi- sual memory in their older participants. Subsequently, with a slightly modified test procedure, Bennett, Sekuler, McIn- tosh, and Della-Maggiore (2001) confirmed the absence of age-related differences in short-term memory for sinusoidal gratings. Such stimuli can be valuable probes of memory be- cause their representations in early vision are known, and be- cause they resist naming or other symbolic encoding, which promotes reliance on a form of short-term memory that is purely visual (Phillips & Christie, 1977; Phillips, 1974; Ka- hana & Sekuler, 2002). Given the importance of this finding one should ask whether the simplicity of the task used by McIntosh et al. and by Bennett et al. may have masked age-related changes in short-term visual memory. For example, in those stud- The authors acknowledge support from the Keck Foundation, and National Institutes of Health grants MH55687 and AG15852. Invaluable assistance was provided by Julie Golomb and Chris McLaughlin. We thank Allison B. Sekuler for excellent com- ments on an earlier version of this manuscript. Correspon- dence concerning this article may be addressed to any of the authors, Volen National Center for Complex Systems, MS 013, Brandeis University, Waltham, MA 02254-9110. e-mail may be sent to sekuler@brandeis.edu, kahana@brandeis.edu or wing- field@brandeis.edu ies observers had to encode and remember only one stimu- lus per trial. The temporal durability of memory for a sin- gle stimulus is no guarantee that the memory would not be vulnerable to disruption by other, succeeding stimuli (Hole, 1996). Moreover, observers in the McIntosh et al. and Ben- nett et al. studies had to encode just a single attribute of each trial’s single stimulus, i.e., the stimulus’ vertical spatial frequency. We conjectured that age-related differences in vi- sual memory might be better revealed by a more challenging task. So we adapted Sternberg’s visual recognition paradigm to study episodic recognition memory for two separate, two- dimensional sinsuoidal gratings (Kahana & Sekuler, 2002; Zhou, Kahana, & Sekuler, 2003). Our stimuli and procedures were patterned after ones that Kahana and Sekuler used with young adults. We took two steps in order to increase task difficulty over that in McIntosh et al. (1999) and in Bennett et al. (2001). First, we presented not one, but two, briefly-presented Study stimuli per trial. These study items, which varied from trial to trial, were followed by a Probe (p), which the participant judged as having been or not been among the just-seen study stimuli. Also, this task forced participants to encode not just one, but two separate aspects of each stimulus. Each stimulus was generated by summing one vertically-oriented grating, and one horizontally-oriented grating. On each trial, the stimuli differed from one another either in vertical spa- tial frequency (with horizontal spatial frequency held con- stant), or vice versa. (Examples of both types of trials are shown in Figure 1.) The random variation in which orien- tation differentiated the stimuli forced participants to encode the spatial frequency information contained in both orienta- tions. For young adults, at least, this requirement was not expected to have much impact; with young participants, Vin- cent and Regan (1995) showed that orthogonal orientations in a compound grating could be processed independently of one another. Whether the same would hold for older partici- pants is an open question. With young participants, delaying the presentation of a p stimulus for several seconds (Magnussen, Greenlee, As-