Detection of the number of changes in a display in working memory.

In the past few years, we have seen an explosion of research on visual working memory, following an article by Luck and Vogel (1997) that introduced a procedure in which, on each trial, a briefly-studied array of objects is followed by a probe display testing memory of one or more features of at least one object from the studied array (see also Phillips, 1974; Sperling, 1960). The recent research has addressed a variety of interlocking issues, including the basis of individual and group differences in capacity (e.g., Cowan, Morey, AuBuchon, Zwilling, & Gilchrist, 2010; Gold et al., 2006; Vogel, McCollough, & Machizawa, 2005), the role of feature binding in working memory (e.g., Allen, Hitch, Mate, & Baddeley, 2012; Cowan, Blume, & Saults, 2013; Logie, Brockmole, & Jaswal, 2011; Oberauer & Eichenberger, 2013), the sharing of capacity between different modalities and codes (e.g., Fougnie & Marois, 2011; Morey & Mall, 2012; Saults & Cowan, 2007; Stevanovski & Jolicoeur, 2007; Vergauwe, Barrouillet, & Camos, 2010), the proper expression of capacity limits in terms of discrete items versus a continuous resource (e.g., Anderson, Vogel, & Awh, 2011; Bae & Flombaum, 2013; Bays & Husain, 2008; Cowan, 2001; Donkin, Nosofsky, Gold, & Shiffrin, 2013; Rouder et al., 2008; Thiele, Pratte, & Rouder, 2011; van den Berg, Shin, Chou, George, & Ma, 2012; Zhang & Luck, 2008), and processes that are used to manage the information in one’s working memory (Barrouillet, Portrat, & Camos, 2011; Camos, Mora, & Oberauer, 2011). Most tests of working memory for arrays have involved memory probes with at most one change in an item compared to the studied array (but see Gibson, Wasserman, & Luck, 2011; Wilken & Ma, 2004, Experiments 4–6). This limitation in method, however, has been for convenience rather than as a reflection of what is supposedly most interesting or important in the world. Many real-world comparisons of two displays do involve multiple differences between them; this is the case, for example, when one compares two paintings by the same artist to discern their similarities and differences, compares two mobile phones to determine which one has better features, or compares before-and-after pictures. As an initial foray into the topic of multi-change detection, we examine relatively small displays with 5, 7, or 9 objects; a real-world analogue might be the comparison of two recipes to estimate how many of their ingredients they have in common. After describing our task and theoretical analysis of it, but before reporting data, we discuss additional background concerning two topics: 1) literature on change detection for scenes, and 2) literature related to auxiliary tasks that we used to assess some individual differences in memory and metamemory (in particular, awareness of one’s working memory) in the present study.