Learning Painting Styles: Spacing is Advantageous when it Promotes Discriminative Contrast SEAN H. K. KANG * and HAROLD PASHLER University of California, San Diego, La Jolla, CA, USA Summary: Repetitions that are distributed over time benefit long‐term retention more than when massed. Recent research has suggested that the advantage of spacing may extend to induction learning–learners were better able to identify the artists of previously unseen paintings when, during training, artists’ paintings were spaced (paintings by different artists were interleaved) rather than massed (a given artist’s paintings were blocked and presented consecutively). Increasing temporal spacing between paintings while maintaining a presentation sequence that was blocked by artist produced test performance no better than massed presentation (both worse than interleaved presentation) (Experiment 1). Displaying paintings by different artists simultaneously produced test performance as good as interleaved presentation and better than massed presentation (Experiment 2). Our findings argue that spacing benefits perceptual induction learning not because of increased temporal spacing per se but rather because interleaving paintings by different artists enhances discriminative contrast between the artists’ styles. Copyright © 2011 John Wiley & Sons, Ltd. Keywords: spacing effect; distributed practice; perceptual category learning; inductive learning Performance on a wide variety of tasks is improved when repetition of study or practice is distributed over time rather than massed, even when total study or acquisition time is held constant. Referred to as the spacing effect, it is considered one of the most robust and replicable phenomena in behavioral science (Dempster, 1996; Cepeda, Pashler, Vul, Wixted, & Rohrer, 2006), and has been demonstrated in domains as diverse as memory for verbal material—e.g. nonsense syllables, words and sentences (e.g. Underwood, 1970), memory for pictures (e.g. Hintzman & Rogers, 1973), arithmetic skill acquisition (e.g. Rickard, Lau, & Pashler, 2008), and motor or procedural learning (e.g. Baddeley & Longman, 1978). Although the spacing effect appears to be robust and important for educational practice, one might argue that in many (if not most) real‐world situations, the importance of learning and retaining specific instances or episodes from the past is limited because the probability of encountering the exact same event or circumstance again in the future is low. More important, one might say, is to identify abstract patterns and principles from past examples and to transfer this knowledge to new examples. This line of reasoning would argue for stressing inductive learning (i.e. learning to generalize from relevant prior encounters). Until recently, however, there was no evidence to suggest that spacing could benefit inductive learning. Indeed, some have suggested that, whereas spacing enhances memory, it might be detrimental to induction (Rothkopf, as cited in Kornell & Bjork, 2008)—since spacing exemplars from a given category or concept could hinder the noticing of common features that define that category or concept (e.g. Underwood, 1952; Kurtz & Hovland, 1956). Kornell and Bjork (2008) investigated this directly using a task of learning the painting style of individual painters. Their results showed that, contrary to the suggestions just described, spacing substantially enhanced performance in this inductive learning task. In their study, they presented subjects with paintings by 12 artists (displayed one at a time), with the instruction to learn the style of each artist. In the massed condition, the paintings were blocked by artist, such that a number of paintings by a given artist would appear consecutively. In the spaced condition, paintings by different artists were interleaved, such that no two paintings by a given artist appeared consecutively. Subjects were subsequently pre- sented with paintings by the same 12 artists, which they had not previously been shown, and had to pick which of the 12 artists was responsible for each painting. Across two experiments, test performance was markedly better in the spaced as compared with the massed condition (see Kornell, Castel, Eich, & Bjork, 2010, for a replication of this result in older adults). One potentially important question that these striking findings of Kornell and colleagues leave unanswered is whether the advantage of spacing was due to increased temporal spacing between paintings by the same artist or due to the interleaving of paintings by different artists (both variables were manipulated at the same time). Also, by using paintings from 12 artists, it is possible that the spacing benefit they observed was at least partly due to the effect of spacing on memory (i.e. learning which style is associated with which of the 12 artists) and less so to any effect on induction per se. The aims of the present study were the following: (1) to determine whether temporal spacing or interleaving (or perhaps both) underlies the spacing effect seen in the learning of artists’ painting styles and (2) to examine whether the spacing advantage observed by Kornell and colleagues persists when the memory load is substantially reduced. Across two experiments, we compared the impact of a number of study conditions on learning of the painting styles of three artists. In all conditions, the set of paintings shown, and the total time spent viewing *Correspondence to: Sean H. K. Kang, Department of Psychology, University of California, San Diego, 9500 Gilman Drive, #0109, La Jolla, CA 92093 – 0109, USA. E‐mail: seankang@ucsd.edu Copyright © 2011 John Wiley & Sons, Ltd. Applied Cognitive Psychology, Appl. Cognit. Psychol. 26: 97 – 103 (2012) Published online 2 May 2011 in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/acp.1801