University of Groningen The long and the short of priming in visual search Kruijne, Wouter; Meeter, Martijn Published in: Attention, Perception & Psychophysics DOI: 10.3758/s13414-015-0860-2 IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below. Document Version Publisher's PDF, also known as Version of record Publication date: 2015 Link to publication in University of Groningen/UMCG research database Citation for published version (APA): Kruijne, W., & Meeter, M. (2015). The long and the short of priming in visual search. Attention, Perception & Psychophysics, 77(5), 1558-1573. https://doi.org/10.3758/s13414-015-0860-2 Copyright Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons). The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license. More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne- amendment. Take-down policy If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim. Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum. Download date: 29-09-2021 Atten Percept Psychophys (2015) 77:1558–1573 DOI 10.3758/s13414-015-0860-2 The long and the short of priming in visual search Wouter Kruijne · Martijn Meeter Published online: 2 April 2015 © The Author(s) 2015. This article is published with open access at Springerlink.com Abstract Memory affects visual search, as is particularly Introduction evident from findings that when target features are repeated from one trial to the next, selection is faster. Two views A powerful factor determining where we look and what have emerged on the nature of the memory representations we attend is where we have looked and what we have and mechanisms that cause these intertrial priming effects: attended before. The effects of our previous overt and covert independent feature weighting versus episodic retrieval of shifts of attention on our current ones are often investi- previous trials. Previous research has attempted to disen- gated by comparing visual search in which targets must tangle these views focusing on short term effects. Here, be found with either the same features as on previous tri- we illustrate that the episodic retrieval models make the als, or with different features. When compared to feature unique prediction of long-term priming: biasing one target switch trials, feature repetitions have been found to shorten type will result in priming of this target type for a much response times (RT) and decrease the amount of errors longer time, well after the bias has disappeared. We demon- (Maljkovic and Nakayama, 1994). In addition, repetitions strate that such long-term priming is indeed found for the cause shorter saccade latencies (Becker, 2008; McPeek et visual feature of color, but only in conjunction search and al., 1999) and bias target selection (Brascamp et al., 2011a; not in singleton search. Two follow-up experiments showed Meeter and Van der Stigchel, 2013). Interestingly, such rep- that it was the kind of search (conjunction versus singleton) etition effects have been found to affect vision largely out and not the difficulty, that determined whether long-term of the observers control (Maljkovic and Nakayama, 1994; priming occurred. Long term priming persisted unaltered Huang et al., 2004; Hillstrom, 2000). Collectively, these for at least 200 trials, and could not be explained as the effects are called intertrial priming. result of explicit strategy. We propose that episodic mem- A wealth of priming research over the past decades ory may affect search more consistently than previously (Kristjansson´ and Campana, 2010) has as of yet failed to thought, and that the mechanisms for intertrial priming yield consensus on its underlying mechanisms. The primary may be qualitatively different for singleton and conjunction dichotomy (Thomson and Milliken, 2013) appears to be search. between the feature-weighting account of intertrial priming and the episodic retrieval account. The feature-weighting Keywords Visual search · Priming · Long-term memory · account entails that the processing of a trial enhances the Implicit memory activation of those visual features that identify the tar- get, and in addition may suppresses distractor features. Electronic supplementary material The online version of this This activation effectively changes how these features are article (doi:10.3758/s13414-015-0860-2) contains supplementary ‘weighted’ on the next trial, which yields repetition benefits material, which is available to authorized users. on subsequent trials that will decay over time. The feature- weighting view is intuitive, and the idea that trials can W. Kruijne () · M. Meeter Vrije Universiteit Amsterdam, Amsterdam, The Netherlands produce ‘lingering’ activity that affects subsequent perfor- e-mail: [email protected] mance is supported by several neurophysiological findings Atten Percept Psychophys (2015) 77:1558–1573 1559 (Kristjansson´ and Campana, 2010; Yeung et al., 2006;de that episodic retrieval of past trials affected ‘late’ stages, Lange et al., 2013). Similarly, the idea that such weighting yielding these interactions. is subject to decay is in line with the observation that facili- These effects are thus explained within both hybrid and tation effects have been found to rapidly disappear over the episodic retrieval accounts by reference to retrieval of past course of some 5–8 trials (Maljkovic and Nakayama, 1994; trial episodes. However, debate has focused very little on Hillstrom, 2000), and that long intertrial intervals can how these results relate to the mechanisms of episodic attenuate or abolish priming effects (Maljkovic and retrieval. For example, an unexplored issue is whether Nakayama, 2000; Thomson and Milliken, 2012). Note that episodic retrieval – which is generally probed at large time different properties of a search trial might independently scales – can be reconciled with the observed time course of contribute to priming effects: most notably, repetitions of priming. To explore this issue, we have simulated priming the response, position, and target-defining feature on a experiments with a mathematical model of episodic mem- search trial might independently produce repetition bene- ory (SAM, Search of Associative Memory, Raaijmakers fits or switch costs (Meeter and Olivers, 2006;Lamyet and Shiffrin, 1981; Mensink and Raaijmakers, 1988; al., 2010; Tollner et al., 2008; Gokce et al., 2014). A math- Raaijmakers, 2003), which implements some very general ematical implementation of the feature weighting account associative memory principles: has been put forward by Maljkovic and Martini (2005), and (a) memory for an item is acquired by forming traces in updated in Martini (2010). which the item is associated with the temporal context In contrast to the independent feature weighting view, active during acquisition. the episodic retrieval account assumes that every trial is (b) The activation of such traces by the context active stored as a bound episodic memory, and that automatic at retrieval determines the probability that an item is retrieval of these memories affects performance on the cur- retrieved; and rent trial. Retrieval of matching trial traces facilitates the (c) context randomly changes over time, gradually render- current trial, whereas nonmatching traces do not. Evidence ing items less accessible. supporting episodic retrieval over the feature weighting account came from the finding that repetitions of target We subjected SAM to sequences of items that reflected features are not independent, but interact with intertrial rep- search displays with alternating target colors. The details etitions or switches of response-features or task-irrelevant and results of these simulations are reported as Supplemen- features, causing super- and under-additive priming effects tary material (available online); to summarize, SAM could respectively (Hillstrom, 2000; Huang et al., 2004). Evidence to a surprising extent capture results from multiple experi- for this account is not limited to such interaction effects: ments exploring the time-course of priming (Maljkovic and recently, Thomson and Milliken (2013) reported that when Nakayama, 1994, Experiment 5; Brascamp et al., 2011b, trials were occasionally paired with a different task, these Experiment 1). Its fits were comparable to those of a ‘rare’ trials would prime the next rare trial, 16 trials into the descriptive feature-weighting model with only short-term future. Not only does this ‘context dependence’ in priming facilitation (Martini, 2010). contrast with independent feature weighting, but the pro- Further simulations with SAM revealed a prediction longed time course of this effect also highlights the link with unique to the episodic retrieval account: as learning occurs associative retrieval from memory. on each presentation, more associated traces are formed for Of note, these two