Sleep-Associated and Retrieval-Associated Memory Consolidation Inaugural-Dissertation zur Erlangung der Doktorwu¨rde der Philosophischen Fakulta¨t II (Psychologie, Pa¨dagogik und Sportwissenschaft) der Universita¨t Regensburg vorgelegt von Magdalena Abel aus Regensburg Regensburg 2011 Erstgutachter: Prof. Dr. Karl-Heinz T. B¨auml Zweitgutachter: Prof. Dr. Klaus W. Lange Acknowledgement This work would never have been finished without the support that I received from many different people over the years. First and foremost, I would like to thank Prof. Karl-Heinz Ba¨uml, the initiator and supervisor of this thesis. He provided me not only with the infrastructurenecessaryforaprojectlikethisone,butalsogavemeprofessional and scientific input as well as motivating words, whenever I needed them. Second, I want to thank everybody at the B¨auml lab for sharing their thoughts and helpful comments, but even more so for sharing coffee breaks and good times. My (current and former) colleagues Alp Aslan, Bernhard Pasto¨tter, Ina Dobler, Oliver Kliegl, Andreas Schlichting, Christoph Holterman, Margit Frimberger, Anuscheh Samenieh, Tobias Staudigl, Maria Wimber,andSimonHanslmayrclearlymadeuniversityamoreenjoyableplace. Moreover,withoutthelab’sresearchassistants(e.g.,DoroBodmann,Elisabeth Rose, Karin Aumu¨ller, Leon Kroczek, Lisa Kugler, Marie Fellner, Michael Wirth, Sabine Krug, and Sandra Kaltner) the data collections this thesis is built on would never have been possible. Last but not least, props go out to my friends and relatives, who kept me company throughout the years. I would like to dedicate this work to my ‘extended family’, and especially to my parents. Your love and support never cease to amaze me, I owe you everything. 2 Preface When somebody is facing a tough decision, a suggestion frequently made is to advise the person to sleep on the respective problem. As this phrase is part of our conventional language use, most people seem to agree or to have made the experience themselves that things can fall into place during the night, and that one may rise with a new perspective the next day. Several anecdotes provided by scientists and artists support the idea that sleep can add something to our cognition; the chemist Kekule, for instance, reported that the potential ring-shaped form of the benzene molecule occurred to him in a dream in which he saw a snake bite into its own tail (for a list of similar anecdotes, see Maquet & Ruby, 2004). However, the proposal that the contents of our minds are further processed during periods of sleep does not rely solely on introspective feelingsandreports.Asamatteroffact,aconsiderableamountofexperimental research has been devoted to disentangling the effects that sleep may exert on cognition. A strong focus has thereby been put on the effect of sleep on memory consolidation. In general, the process of memory consolidation is assumed to stabilize memory contents and, thus, to be beneficial for later remembering. Consolidated memories are further assumed to show less time-dependent forgetting and to be less susceptible to the detrimental effects of retroactive interference (e.g., Wixted, 2004). During the last two decades, corresponding evidence has been brought forward that links sleep, and more specifically, certain sleep stages and sleep parameters to the effective consolidation of memories (for a review, see Diekelmann & Born, 2010). While we sleep, 3 Preface 4 memory contents are assumed to be reactivated, and thereby to be stabilized and to be made more resistant to forgetting. Such and similar processes have been suggested to underly the familiar phenomenon of insight after sleep (e.g., Cai, Mednick, Harrison, Kanady, & Mednick, 2009; Wagner, Gais, Haider, Verleger, & Born, 2004). The first part of this thesis will be dedicated to investigating sleep-associated memory consolidation in more detail. Indeed, previous researchindicatesthatsleepmaynotbenefitallmemoriesequally.Forinstance, ithasbeenfoundthatsleep-relatedbenefitsaremorepronouncedformemories that are harder to access (Drosopoulos, Schulze, Fischer, & Born, 2007). If so, sleep may not only counteract time-dependent forgetting by stabilizing memory contents, but may also counteract experimentally induced forms of forgetting that are assumed to rely on a reduced accessibility of memories. By investigating interference effects and list-method directed forgetting, the prior work is to be extended to different conditions and a different paradigm. Irrespective of the vast amount of research on sleep-associated memory consolidation, there are other factors as well that impact upon the contents of our minds. While sleep is mostly regarded as a pleasant state connected with rest and inactivity, active retrieval from memory constitutes quite the opposite: It takes place during wakefulness and requires deliberate efforts. Yet, research on the so-called testing effect has previously shown that such retrieval from memory is also beneficial for long-term retention. In particular, less time-dependent forgetting has been observed after retrieval in comparison to restudy or distractor conditions (for a review, see Roediger & Butler, 2011). As reduced time-dependent forgetting is seen as one criterion for memory consolidation (e.g., Wixted, 2004), one could speculate that retrieval from memory might be connected to the consolidation of memory contents, too. Nevertheless, another branch of research shows that retrieval can also entail detrimental effects for related, but nonretrieved material. The term retrieval-induced forgetting circumscribes the finding that retrieval of a subset of material prompts forgetting of related, nonretrieved material relative Preface 5 to control material (Anderson, Bjork, & Bjork, 1994). Such forgetting is investigated by applying the retrieval-practice paradigm and assumed to rely oninhibitoryprocessesthatareinitiatedduringretrievalinordertoresolvethe interference caused by automatic activation of related material (e.g., B¨auml, 2008). Concerning the longevity of the effect conflicting results have been reported; while some studies found the effect to be lasting, others found it to be vanishing with time (e.g., Garcia-Bajos, Migueles, & Anderson, 2009; MacLeod & Macrae, 2001). The second part of this thesis will be dedicated to investigating whether retrieval may also be connected to processes of memory consolidation. If so, retrieved material should show less time-dependent forgetting and less susceptibility to retroactive interference than nonretrieved control material (e.g., Wixted, 2004). Applying the retrieval-practice paradigm allows not only to replicate and extend previous research on the testing effect by examining time-dependent forgetting and interference susceptibility of practiced memories, it also allows to investigate how related material is affected by retrieval practice. As active retrieval has not only been reported to strengthen the respective memories, but also to weaken related and interfering memories, possible effects of retrieval-associated memory consolidation could also be of interest for the perspective on retrieval-induced forgetting. Finally, the third part of this thesis will summarize the obtained data on effects of sleep and retrieval. The idea that both sleep and retrieval could be related to memory consolidation may beforehand seem like a rather counterintuitive proposition. Sleep, on the one hand, is usually perceived as a passive state and assumed to recur solely because our bodies need rest. Whatever exactly mediates the mnemonic benefits associated with sleep seems to come without the need to actively work for it. Memory retrieval, on the other hand, requires an awake and active mind as well as cognitive efforts. If sleep and retrieval could be connected because they entail the similar basic effects associated with memory consolidation, this could spark new research and deepen our understanding of memory consolidation in a more general way. Contents Abstract 9 I Sleep-Associated Memory Consolidation 10 1 A Current Perspective 11 1.1 Memory Consolidation and its Relation to Sleep . . . . . . . . . 12 1.2 Forms of Sleep-Dependent Memory Consolidation . . . . . . . . 19 2 Interference 22 2.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.2 Experiment 1a: Sleep and Interference (Strong Encoding) . . . . 30 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 2.3 Experiment 1b: Sleep and Interference (Weak Encoding) . . . . 43 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49 2.4 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 3 Directed Forgetting 53 3.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 3.2 Experiment 2: Sleep and Directed Forgetting . . . . . . . . . . . 56 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61 3.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65 II Retrieval-Associated Memory Consolidation 67 4 A Current Perspective 68 4.1 The Testing Effect: Basic Findings . . . . . . . . . . . . . . . . 68 4.2 How Does Retrieval Practice Benefit Retention? . . . . . . . . . 71 5 Retrieval Practice Effects and Normal Forgetting 74 5.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 5.2 Experiment 3: Retrieval Practice Effects, Sleep and Normal Forgetting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 5.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 6 Retrieval Practice Effects and Interference 91 6.1 Experiment 4a: Interference after Retrieval Practice . . . . . . . 92 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 6.2 Experiment 4b: Interference after Restudy . . . . . . . . . . . . 99 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103 6.3 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 III Summary 112 References 122 Abstract After encoding, memory contents need to undergo a phase of stabilization in order to be remembered in the long-term. Such stabilization is referred to as memory consolidation, and is assumed to be observable in less time-dependent forgetting and less susceptibility to interference. Evidence for a role of sleep in the effective consolidation of memories has previously been provided. In addition, research on the so-called testing effect has also indicated a contribution of retrieval practice to long-term memory, as the active retrieval from memory has been shown to be able to boost retention. This thesis investigated effects of sleep and retrieval on memory consolidation. Effects of sleep-associated memory consolidation were examined by inducing interference in paired-associate learning and by applying list-method directed forgetting: Replicating previous work, sleep was found to be of preferential benefit for memories that are hard to access; thereby, sleep counteracted both incidental and intentional forms of forgetting. Applying the retrieval-practice paradigm, effects of retrieval-associated consolidation were observed: Retrieval was found to stabilize directly retrieved and related memories, thereby making them less susceptible to both time-dependent forgetting and retroactive interference. The data presented here indicate that both sleep and retrieval may be associated with memory consolidation. Possible differences and parallels betweensleep-associatedandretrieval-associatedmemoryconsolidationaswell as implications of such a proposal are discussed. 9