Practice makes transfer of motor skills imperfect

Psychological Research (2012) 76:611–625 DOI 10.1007/s00426-011-0355-2 ORIGINAL ARTICLE Practice makes transfer of motor skills imperfect Arnaud Boutin • Arnaud Badets • Robin N. Salesse Udo Fries • Stefan Panzer • Yannick Blandin • Received: 12 April 2011 / Accepted: 2 June 2011 / Published online: 14 June 2011 Ó Springer-Verlag 2011 Abstract We investigated the practice-effects on motor skill transfer and the associated representational memory changes that occur during the within-practice and betweenpractice phases. In two experiments, participants produced extension–flexion movements with their dominant right arm for a limited or prolonged practice session arranged in either a single- or multi-session format. We tested the ability of participants to transfer the original pattern (extrinsic transformation) or the mirrored one (intrinsic transformation) to the non-dominant left arm, 10 min and 24 h after the practice sessions. Results showed that practice induces rapid motor skill improvements that are nontransferable irrespective of the amount of acquisition trials. Furthermore, the extrinsic component of the skill develops early and remains the dominant coding system during practice. Conversely, we found distinct between-practice memory changes: a limited practice induces an off-line development of the extrinsic component, whereas a prolonged practice session subserves the off-line development Part of this work was presented at the conference of the North American Society for the Psychology of Sport and Physical Activity, Tucson (AZ), 2010. A. Boutin
A. Badets
R. N. Salesse
Y. Blandin (&) National Centre of Scientific Research, Centre de Recherches sur la Cognition et l’Apprentissage, CeRCA, CNRS UMR 6234, MSHS. Bât A5, University of Poitiers, 5, rue Théodore Lefebvre, 86000 Poitiers, France e-mail: [email protected] U. Fries Department of Human Movement Science, University of Leipzig, Leipzig, Germany S. Panzer IfADo, Leibniz Research Centre for Working Environment and Human Factors, Dortmund, Germany of the intrinsic component (Experiment 2). We provided further evidence that the long-term representation of the motor skill also depends on the nature of the practice session itself: the parsing of practice into multiple sessions narrows the effector-transfer capacities in comparison to a single session (Experiment 1). These findings yield theoretical and practical implications that are discussed in the context of recent motor skill learning models. Introduction Growing evidence, spanning behavioural and neurophysiological studies, supports the notion that the acquisition of a new motor skill is mediated by distinct successive phases: a fast learning, within-practice performance improvements phase, followed by a slow learning phase consisting of delayed (off-line), time-dependent performance improvements emerging after practice (e.g., Karni, Meyer, Jezzard, Adams, Turner, & Ungerleider, 1995; Korman, Raz, Flash, & Karni, 2003; Robertson, Pascual-Leone, & Press, 2004b; see also Krakauer & Shadmehr 2006; Stickgold & Walker 2007, for reviews). Current hypotheses state that a specific and labile memory representation of the motor skill is formed during practice. This newly memory trace will then undergo further modifications during a period of consolidation, in the absence of additional practice (e.g., Robertson, Pascual-Leone, & Miall, 2004a; Walker & Stickgold, 2004; see Krakauer & Shadmehr, 2006; Robertson, 2009, for a review). Consolidation refers to the off-line process (see McGaugh, 2000, for a review) that transforms new and initially labile memories into more stable representations that become integrated into the network of pre-existing long-term motor memories ([24 h; see Krakauer & Shadmehr, 2006, for a review). 123 612 Numerous theoretical perspectives proposed that the representation of motor skills relies on distinct and independent coordinate or coding systems (e.g., CriscimagnaHemminger, Donchin, Gazzaniga, & Shadmehr, 2003; Hikosaka, Nakahara, Rand, Sakai, Lu, Nakamura, et al., 1999; Lange, Godde, & Braun, 2004). Each coding system contributes to movement production and could produce specific learning and transfer capacities. While the visually acquired information about movement and target positions are initially encoded in an eye-centred, extrinsic worldbased reference frame, the muscular activation patterns are then encoded in an intrinsic, body-centred reference frame (e.g., Colby & Goldberg, 1999; Soechting & Flanders, 1989). The intrinsic code is represented in terms of an internal model of joint representations (CriscimagnaHemminger et al., 2003), musculoskeletal forces and dynamics (Krakauer, Ghilardi, & Ghez, 1999), which takes into the account of orientation of body segments relative to each other (Lange et al., 2004; Soechting & Flanders, 1989). The intrinsic coding system is supposed to result in an effector-dependent representation of the motor skill (i.e., without effector transfer capacity) to the extent that biomechanical, neurological, and dynamic properties of the effectors used on a transfer test are dissimilar to those used during practice. Conversely, the extrinsic code reflects Cartesian coordinates of the task space with respect to the visual display. This code is represented in an effectorindependent representation (i.e., effector transfer capacity) when the extrinsic coordinates are reinstated during an effector transfer test even though intrinsic characteristics of the required movement are altered (Hikosaka et al., 1999). Many studies have investigated the capacity of individuals to transfer skills between hands/arms using intermanual transfer paradigms (e.g., Cohen, Pascual-Leone, Press, & Robertson, 2005; Kovacs, Boyle, Gruetzmatcher, & Shea, 2010; Kovacs, Han, & Shea, 2009; Panzer, Krueger, Muehlbauer, Kovacs, & Shea, 2009). However, the vast majority of studies have almost exclusively focused on the long-term effects of practice on skill transfer (e.g., Boutin, Fries, Panzer, Shea, & Blandin, 2010; Kovacs, Muehlbauer, & Shea, 2009), on the practice-related memory changes (e.g., Karni, Meyer, Rey-Hipolito, Jezzard, Adams, Turner, et al., 1998; Korman et al., 2003) and on the degree of overnight/overday improvements (e.g., Cohen et al., 2005; Robertson, Press, & PascualLeone, 2005; Witt, Margraf, Bieber, Born, & Deuschl, 2010). But the assumption that the coding systems may be enhanced in distinct learning phases (within- and betweenpractice) has not yet been directly addressed. A protocol is lacking that allows direct assessment of both within- and between-practice session improvements, and their relative contribution to the development of a specific coding system. The main purposes of this study were, therefore, to 123 Psychological Research (2012) 76:611–625 assess the evolution of the coding systems through practice (Experiment 1) and to further investigate the rapid and delayed practice-dependent changes (Experiment 2) that contribute to the formation of long-term motor skill representations. The questions addressed in the current study required the use of a particular experimental design where participants had to perform rehearsed retention and transfer tests (i.e., testing and re-testing sessions). However, numerous studies (see Roediger & Butler, 2010; Roediger & Karpicke, 2006b, for reviews) in the verbal skill literature have shown that if the period of time devoted to learning includes at least one test, performance on a final test is improved. This finding is known as the testing effect (see Roediger & Karpicke, 2006a, for a review). Recent research provided further evidence that testing rehearsal can be used to further promote transfer on the final test (e.g., Butler, 2010; McDaniel, Roediger, & McDermott, 2007, for a review). In their review, Roediger and Butler (2010) argued that testing rehearsal promotes the acquisition and long-term retention of the acquired knowledge that can be flexibly retrieved and transferred to different contexts, thus suggesting that the mnemonics benefits of testing extend beyond the retention of a specific response. The retrieval of information from memory during testing may be the central mechanism that produces learning benefits during a final test (Butler, 2010). One possible explanation is that combining practice with testing session provides greater encoding variability than single practice (McDaniel & Masson, 1985). Thereby, testing rehearsal can be a potent memory enhancer, not just a neutral event as usually wrongly considered (e.g., Carpenter, 2009; Roediger & Butler, 2010; Rohrer, Taylor, & Sholar, 2010, for similar position). It should be noted that testing-effect studies were traditionally conducted, with very few exceptions, using word lists (e.g., McDaniel & Masson, 1985; Wheeler et al. 2003) or educationally relevant situations (e.g., Roediger & Karpicke, 2006b; Rohrer et al., 2010) as materials. However, to the best of our knowledge, the extent to which testing rehearsal can interact with motor skill learning and transfer remains unknown. Therefore, a condition without testing rehearsal was also included in all experiments. We used a continuous dynamic arm movement task that required participants to make a sequence of extensionflexion movements with their dominant right arm to reproduce a specific pattern of displacement over time. To test whether the contribution of the extrinsic and intrinsic coding systems to skill learning and transfer depends on the amount of practice, performance on retention and transfer tests was systematically evaluated after both limited and prolonged practice sessions. We used a retention test requiring the participants to produce the spatial–temporal Psychological Research (2012) 76:611–625 pattern with the same arm that was used during the practice trials and two effector transfer tests with the unpractised arm (see Kovacs et al., 2009; Panzer et al., 2009 for a similar procedure). In the ‘‘intrinsic transformation’’ test, the original pattern was mirrored so that the sequential movements remain the same when transferred to the non-dominant left arm (i.e., the same pattern of muscle activation and limb joint angles). This test assessed the contribution of the intrinsic component to transfer and learning. In the ‘‘extrinsic transformation’’ test, the original pattern was preserved but performed with the non-dominant left arm (i.e., the same goal movement pattern). This test assessed the contribution of the extrinsic component to transfer and learning because it implicated the same spatial positions but a different pattern of muscle activation compared with that used during practice. It should be noted that ineffective transfer performance for one or both effector transfer tests suggests that the extrinsic and/or intrinsic components of the motor skill were not the dominant coding system(s) for sequence production. Experiment 1 The primary purpose of Experiment 1 was to investigate the evolution of the coding systems with practice. We evaluated retention and effector transfer performances after a limited (18 acquisition trials) and a prolonged (135 additional acquisition trials) practice session. This design involved performing rehearsed retention and transfer tests (repeated testing condition). Therefore, we included a condition that required participants to complete the same amount of practice without retention/transfer tests between-practice sessions (non-repeated testing condition). Our predictions for Experiment 1 were based upon previous theoretical perspectives (e.g., Bapi, Doya, & Harner, 2000; Hikosaka et al., 1999, 2002; Keele, Ivry, Mayr, Hazeltine, & Heuer, 2003; Keele, Jennings, Jones, Caulton, & Cohen, 1995), which suggested that the representation of motor skills evolves from an extrinsic (effector-independent) to an intrinsic (effector-dependent) dominant coding system with practice. Thus, we predicted that a limited practice session would yield better performance on both the retention and extrinsic transformation tests than on the intrinsic transformation test. Conversely, we expected an opposite pattern of results for prolonged practice with better performance on the intrinsic transformation test than on the extrinsic transformation test. Finally, the repeated testing condition should favour incremental gains in performance from testing to retesting and, consequently, could modify the relationship between the amount of practice and the type of coding used to produce the task (e.g., Roediger & Butler, 2010). 613 Method Participants Thirty self-declared right-handed undergraduate students (mean age 19.1 years, SD 0.9 years; 12 women) volunteered to participate in this study. Each participant was requested to read and sign an informed consent form prior to participation in the experiment. None of the participants had previous experience with the experimental task, and they were unaware of the specific purposes of the study. A local ethics committee approved the protocol. Apparatus The apparatus consisted of a horizontal lever supported at one end by a vertical axle that turned in a ball-bearing support in a manner that was almost frictionless. The support was fixed on a table facing the participant, allowing the lever to move in the transversal plane over the table surface. At the other end of the lever, a vertical handle was fixed. The handle’s position could be adjusted so that when the participant grasped the handle, his/her elbow was aligned with the axis of rotation (Fig. 1, top left). A potentiometer was attached to the handle’s axis of rotation to record the position, and its output was sampled at 1000 Hz. The potentiometer data were used to provide lever position information to the participant and stored for later analysis. A wooden cover was placed over the table to prevent participants from seeing the lever and their arm. A video projector was used to display the goal movement pattern, the on-line position of the lever and the knowledge of results (KR) on the wall facing the participant. Participants were seated on a height-adjustable chair, about 2 m from the wall where a 1.64 9 1.23 m image was projected, so that their lower arm was at an approximately 80° angle to the upper arm at the starting position. All aspects of the experiment were programmed with the MatlabÒ R2008b software from the MathWorksÓ (The MathWorks, Inc., Natick, MA), and using the Psychophysics Toolbox extensions (Brainard, 1997; Pelli, 1997; Kleiner, Brainard, & Pelli, 2007). Task, experimental groups and procedures Participants were randomly assigned to the ‘‘repeated testing’’ (RT group; N = 15) or ‘‘non-repeated testing’’ (NRT group; N = 15) practice condition. In the RT practice condition, participants underwent 24-h delayed retention and transfer tests after both limited and prolonged practice sessions. In the NRT practice condition, participants were merely administered 24-h delayed retention and transfer tests following the prolonged practice session. 123 614 Psychological Research (2012) 76:611–625 Groups Acquisition 1 Day 1 Initial testing Day 2 Acquisition 2 Day 2 Subsequent testing Day 2 Repeated testing 2 blocks R - Et - It 15 blocks R - Et - It Non-repeated testing 2 blocks 15 blocks R - Et - It Fig. 1 Schematic illustrating the limb and criterion movement pattern used during acquisition and on the retention and transfer tests (top). Note that the goal movement pattern was projected onto the wall in front of the participant. Participants used their dominant right limb during acquisition and retention. The non-dominant left limb was used for the extrinsic and intrinsic transformation tests. The retention test was conducted first, and the order for the two transfer tests (extrinsic transformation and intrinsic transformation) was counterbalanced Figure 1 (bottom) presents the experimental phases for each practice condition. Participants were individually tested in a silent and dimly lit room. Before beginning the experiment, each participant received written instructions and additional verbal information about the task and procedures. They had to perform a sequence of extension-flexion movements with their dominant right arm to reproduce the spatial and temporal aspects of the goal pattern projected in front of them (Fig. 1, top). The spatial–temporal pattern was created by summing two sine waves with similar periods but different amplitudes. The duration of the goal pattern was 1,500 ms, and the extension–flexion movements required three reversals. The potentiometer output was sampled for 2,000 ms, but only the first 1,500 ms was retained for later analysis. At the beginning of each trial, the goal movement pattern was displayed on the wall, and participants were asked to move the lever to the starting position (1° area at the beginning of the goal movement pattern). One second after achieving the start position, a tone (50 ms in duration, 3 kHz square wave) indicated that they should perform the task. The task required moving the lever with the dominant right arm through a pattern of extension–flexion cycles (three reversals) to produce the required spatial–temporal pattern. As soon as the participant started moving, the goal movement pattern disappeared from the screen, and a cursor representing the position of the lever was displayed. After a 2-s interval following response completion, KR was provided by superimposing the goal movement pattern (white) over the actual pattern (green). In addition, the root-meansquare error (RMSE) of the actual movement from the goal movement was displayed on the screen. RMSE is the deviation of the actual pattern from the goal pattern calculated from the onset of movement until the end of the first 1,500 ms. RMSE is sensitive to both response bias and within-participant variability. When presented, KR was displayed for 5 s. Participants were instructed to reproduce the goal movement pattern as accurately as possible and were requested to reduce the RMSE on subsequent trials. 123 Psychological Research (2012) 76:611–625 Participants in the RT and NRT conditions completed the first acquisition phase (Acquisition 1), which was composed of two blocks of nine trials each. During this phase, KR was provided following each trial. Approximately 24 h after the completion of Acquisition 1, participants assigned to the RT practice condition were administered retention and effector transfer tests, while their NRT practice condition counterparts were not. The retention test consisted of one block of nine trials without KR that required the participants to produce the movement pattern with the same limb that was used during the practice trials. With the exception that KR was not provided during retention, all other procedures remained the same as those used for acquisition. In addition, two effector transfer tests without KR were administered in a counterbalanced order after the retention test (Fig. 1, top). The transfer tests were performed with the contra-lateral limb (i.e., nondominant left arm). In the ‘‘intrinsic transformation’’ test, the original pattern becomes mirrored so that the sequential movements remain the same when transferred to the nondominant left arm (i.e., the same pattern of muscle activation and limb joint angles). In the ‘‘extrinsic transformation’’ test, the original pattern was preserved but performed with the non-dominant left arm (i.e., the same goal movement pattern). With the exception that participants in the NRT practice condition did not perform retention and transfer tests on day 2, participants in the RT and NRT conditions continued practice for an additional acquisition phase (Acquisition 2). This acquisition phase consisted of fifteen blocks of nine trials and involved the same experimental procedures as in Acquisition 1. Note that participants in the RT practice condition took a 5-min rest interval following retention and transfer before practicing Acquisition 2. Twenty-four-hour delayed retention and transfer tests (extrinsic transformation and intrinsic transformation) were conducted on day 3 for all participants in the same manner as the initial testing session experienced in the RT condition. 615 pattern was computed at each data point in the time series. Next, differences for each data point in the time series were squared, and means of the squared differences were computed on a trial basis. Finally, the square root of the mean was computed for the final measure of RMSE. Values of RMSE for individual trials were then averaged to yield a global estimate of RMSE for each block (nine trials). Results Acquisition Acquisition data were separately analysed. First, Acquisition 1 data were submitted to a 2 (practice conditions: repeated testing, non-repeated testing) 9 2 (blocks 1–2) analysis of variance (ANOVA) with repeated measures on the last factor. The second analysis was a 2 (practice conditions: repeated testing, non-repeated testing) 9 15 (blocks 3–17) ANOVA with repeated measures on block using data from Acquisition 2. All significant effects were reported at p \ 0.05, unless otherwise stated, and Duncan’s multiple range test was used for post hoc comparisons. Partial eta square (g2p) is the effect size reported for all significant effects (Cohen, 1988). Outliers were removed from the analysis (±2 SD: about 4%). Mean RMSEs during the acquisition, retention and transfer phases is displayed in Fig. 2. Data analysis and measures Data reduction was performed using MatlabÒ (The MathworksÓ, Inc., Natick, MA, USA). The individual trial time series were used to compute lever displacement. Angular displacement time series were filtered with a 2nd order dual-pass Butterworth filter with a cut-off frequency of 10 Hz. Root-mean-square error was computed to estimate performance error in achieving the goal movement pattern. Such a measure is sensitive to both amplitude and timing errors in the produced movement pattern relative to the goal movement pattern. To compute RMSE, the difference between the criterion and the filtered actual movement Fig. 2 Mean RMSEs during acquisition (Acquisition 1 and 2) and on the retention (R), extrinsic transformation (Et) and intrinsic transformation (It) tests for the repeated testing and non-repeated testing groups in Experiment 1. Error bars reflect standard error of the mean 123 616 Acquisition 1 The analysis detected a main effect of block, F(1, 28) = 75.34, g2p = 0.72, indicating higher RMSE on block 1 (M = 100.84) than on block 2 (M = 71.10). The main effect of practice condition, F(1, 28) \ 1, and the practice condition 9 block interaction were not significant, F(1, 28) \ 1. Acquisition 2 The analysis detected a main effect of block, F(14, 392) = 13.43, g2p = .32. Duncan’s multiple range test revealed higher RMSE on block 3 (M = 64.31) compared with blocks 4–7 (from M = 56.04 to M = 51.18, respectively), which were not different from each other but higher than blocks 8–17 (from M = 47.44 to M = 41.43, respectively), which did not differ from each other. The analysis failed to detect neither a main effect of practice condition, F(1, 28) \ 1, nor a practice condition 9 block interaction, F(14, 392) \ 1. Retention and transfer Psychological Research (2012) 76:611–625 repeated testing, non-repeated testing) 9 3 (tests: retention, extrinsic transformation, intrinsic transformation) ANOVA with repeated measures on the last factor. The analysis revealed main effects of practice condition, F(1, 28) = 8.09, g2p = 0.22, test, F(2, 56) = 28.09, g2p = 0.50, and a significant practice condition 9 test interaction, F(2, 56) = 5.03, g2p = 0.15. RMSE on the retention tests were not significantly different between the RT (M = 47.82) and NRT groups (M = 50.59). However, lower RMSE was found for the RT group than for the NRT group on the extrinsic and intrinsic transformation tests. Specifically, for the RT group, post hoc comparisons indicated that RMSE was significantly lower on the retention test (M = 47.82) than on the extrinsic (M = 56.78) and intrinsic transformation tests (M = 59.15), which did not significantly differ from each other. For the NRT group, RMSE was also significantly lower on the retention test (M = 50.59) than on the extrinsic (M = 68.06) and intrinsic transformation tests (M = 79.06). However, in contrast to the RT practice condition, the analysis displayed higher RMSE on the intrinsic transformation test compared with that on the extrinsic transformation test (p \ 0.05) for participants assigned to the NRT group. Repeated testing group: from testing to retesting Discussion To assess the evolution of the long-term representation of motor skills with practice, we compared performances of the RT group on the retention and transfer tests following limited and prolonged practice. Mean RMSEs on the retention and transfer tests were analysed in a 2 (amounts of practice: limited practice, prolonged practice) 9 3 (tests: retention, extrinsic transformation, intrinsic transformation) ANOVA with repeated measures on the two factors. The analysis indicated main effects of amount of practice, F(1, 14) = 31.11, g2p = 0.68, and test, F(2, 28) = 6.17, g2p = 0.30, but failed to detect a significant amount of practice 9 test interaction, F(2, 28) = 1.78. RMSE was higher following the limited (M = 74.59) than the prolonged practice (M = 54.58). For the main effect of test, Duncan’s multiple range test indicated that RMSE was significantly higher on the intrinsic transformation test (M = 71.03) than on the retention (M = 58.26) and extrinsic transformation tests (M = 64.47), which did not differ from each other. Repeated and non-repeated testing groups: subsequent testing Performances of the RT and NRT groups on the retention and transfer tests following the prolonged practice session were compared to show evidence for/against a testing effect. RMSE was analysed in a 2 (practice conditions: 123 Results for the acquisition indicated that the RT and NRT groups performed similarly on the first and second practice sessions (Acquisition 1 and Acquisition 2, respectively), suggesting that participants in both groups improved their performance during practice. Importantly, this finding demonstrates that improvements during Acquisition 2 were not testing-dependent. Indeed, participants in the NRT practice condition did not express worse performance during Acquisition 2 compared with their RT practice condition counterparts. Our findings revealed that testing rehearsal did not interact with motor skill acquisition. Similar conclusions can be drawn regarding motor skill retention. As revealed by performances during retention at subsequent testing, RMSE on retention tests were identical for the RT and NRT groups. However, different patterns of results were observed at subsequent transfer tests for the RT and NRT groups. This indicates that the initial testing session, performed in the RT practice condition, has influenced effector transfer capacities. This is in line with previous research in the testing-effect literature (e.g., Butler, 2010; Rohrer et al., 2010; see Roediger & Butler, 2010; Roediger & Karpicke, 2006a, for reviews), and extend the findings from the verbal to the motor domain. Specifically, transfer data for the NRT group indicated that practice produced an effector-dependent representation (lower RMSE on the retention Psychological Research (2012) 76:611–625 test compared with the transfer tests) that remained extrinsic (lower RMSE on the extrinsic than on the intrinsic transformation test). Conversely, testing rehearsal resulted in a multiple coding representation (similar performances for the RT group on the extrinsic and intrinsic transformation tests). Interestingly, this finding suggests that testing rehearsal promotes the development of a multiple coding representation that can be flexibly used from dominant to non-dominant limb. Because testing rehearsal interacts with motor skill transfer, it is therefore not surprising that we did not find any amount of practice 9 test interaction when analysing data of the RT group. Indeed, one could assume that retrieval practice at initial testing might have enhanced motor skill transfer at subsequent testing, thus precluding the possibility of additional analysis regarding the evolution of the motor skill representation from testing to retesting. Therefore, to assess the motor skill representation after limited practice, we further conducted a 3 (tests: retention, extrinsic transformation and intrinsic transformation) ANOVA with repeated measures on test using the RT group’s data at initial testing. Analysis revealed a main effect of test, F(2, 28) = 4.47, g2p = 0.24, with lower RMSE on the retention (M = 68.71) and extrinsic transformation tests (M = 72.16) than on the intrinsic transformation test (M = 82.91). These results agree with the theoretical perspective that an effector-independent representation of an extrinsically transformed skill is developed after few practice trials (Hikosaka et al., 1999, 2002; Keele et al., 2003, 1995). To summarise, a key finding of Experiment 1 is that the long-term representation of the motor skill after multisession practice was qualitatively (i.e., reliance on a particular coding system) different when participants were confronted with testing rehearsal or with a single delayed testing session. Neither the delay between testing sessions nor the multisession format per se could explain such differences because the experimental design was exactly the same for the RT and NRT practice conditions. Two issues could be considered when interpreting these results: the opportunity for testing rehearsal and/or the process of motor memory consolidation that took place during the 24-h interval following the first and second acquisition phase. Therefore, Experiment 2 sought to further clarify the relative contribution of testing rehearsal and/or the process of memory consolidation to motor skills coding. An additional purpose of Experiment 2 was to investigate the rapid and delayed practice-related memory changes that contribute to the formation of long-term motor skill representations, using single practice sessions composed of either 18 acquisition trials (limited practice) or 153 acquisition trials (18 ? 135; prolonged practice). 617 Experiment 2 Experiment 2 was designed to extend the findings of Experiment 1 from multisession to single session and to evaluate the off-line practice-related memory changes following a single practice session. The results of Experiment 1 revealed that testing rehearsal might be an effective memory enhancer when the testing sessions are administered between-practice sessions. Experiment 2 aimed at determining whether off-line testing-dependent memory enhancements of the motor skill could be observed from testing to retesting in the absence of further task engagement. Participants from two groups were required to reproduce the spatial–temporal pattern used in Experiment 1 for either 18 or 153 acquisition trials in a single practice session before being evaluated on both 10-min and 24-h delayed retention and transfer tests. As in Experiment 1, two other groups practised the same goal pattern for either 18 or 153 acquisition trials but with the exception that no 10-min retention/transfer tests were conducted, thus avoiding a potential testing effect from 10-min testing to 24-h delayed retesting. We expected rapid representational memory changes with practice, as evaluated by 10-min retention and transfer tests, with the dominance of the extrinsic coding system after limited practice and conversely of the intrinsic one after prolonged practice. Additionally, we predicted that the rapid and practice-dependent memory changes would extend beyond the practice session, i.e., from 10-min to 24-h testing. More specifically, a limited practice session would favour the off-line development of the extrinsic component of the motor skill while in contrast, a prolonged practice session would emphasize the off-line development of the intrinsic coding system. Method Participants Sixty self-declared right-handed undergraduate students (mean age 21.1 years, SD 1.5 years; 21 women) volunteered to participate in this study. Each participant was requested to read and sign an informed consent form prior to participation in the experiment. None of the participants had previous experience with the experimental task, and they were unaware of the specific purposes of the study. A local ethics committee approved the protocol. Apparatus The apparatus was identical to that used in Experiment 1. 123 618 Psychological Research (2012) 76:611–625 Day 1 Immediate testing Day 1 24-h testing Day 2 Partial eta square (g2p) is the effect size reported for all significant effects (Cohen, 1988). Outliers were removed from the analysis (±2 SD: about 4%). Mean RMSEs during acquisition for the limited and prolonged practice groups are displayed in Fig. 3. Repeated testing 2 Blocks R–Et–It R–Et–It Limited practice Non-repeated testing 2 Blocks Table 1 Experimental phases for the repeated testing and nonrepeated testing groups in Experiment 2 Groups Acquisition Limited practice R–Et–It Prolonged practice Repeated testing 17 Blocks Non-repeated testing 17 Blocks R–Et–It R–Et–It R–Et–It R retention, Et extrinsic transformation, It intrinsic transformation (order counterbalanced) The analysis revealed a main effect of block, F(1, 28) = 59.89, g2p = 0.68, indicating that RMSE was higher on block 1 (M = 102.01) than on block 2 (M = 68.59). The main effect of practice condition, F(1, 28) \ 1, and the practice condition 9 block interaction, F(1, 28) \ 1, were not significant. Prolonged practice Task, experimental groups and procedures Participants were randomly assigned to one of four groups (N = 15) defined by the practice condition (repeated testing or non-repeated testing) and the amount of practice (limited or prolonged). While assigned to the ‘‘repeated testing’’ (RT) practice condition, participants underwent both immediate and 24-h delayed retention and transfer tests following either a limited or a prolonged practice session. Conversely, while assigned to the ‘‘non-repeated testing’’ (NRT) practice condition, participants were merely administered 24-h delayed retention and transfer tests following either a limited or a prolonged practice session (see Table 1). Task and procedures for the acquisition, retention and transfer tests were conducted in the same manner as described in Experiment 1, except as noted. The analysis detected a main effect of block, F(16, 448) = 36.29, g2p = 0.56. Duncan’s multiple range test revealed that RMSE was higher on block 1 (M = 98.54) than on block 2 (M = 69.59). Further, analysis found higher RMSE on block 2 than on blocks 3–4 (from M = 63.18 to M = 57.63, respectively), which was not different from each other but higher than on blocks 5–13 (from M = 53.90 to M = 47.12, respectively), which did not differ from each other. From block 14, RMSE remained stable until the end of acquisition (from M = 45.02 to M = 43.91, for blocks 14–17 respectively). The analysis failed to detect neither a main effect of practice condition, F(1, 28) = 2.37, nor a practice condition 9 block interaction, F(16, 448) \ 1. Retention and transfer Data analysis and measures Data reduction was performed in the same manner as in Experiment 1. Results Acquisition Acquisition data were separately analysed. First, the limited practice data were submitted to a 2 (practice conditions: repeated testing, non-repeated testing) 9 2 (blocks 1–2) ANOVA with repeated measures on the last factor. The second analysis was a 2 (practice conditions: repeated testing, non-repeated testing) 9 17 (blocks 1–17) ANOVA with repeated measures on block using data from the prolonged practice groups. All significant effects were reported at p \ 0.05, unless otherwise stated, and Duncan’s multiple range test was used for post hoc comparisons. 123 Mean RMSEs on the retention and transfer tests were analysed in several ways. To assess the coding systems’ time course during practice and memory consolidation, we separately analysed performances of the RT groups (a) at 10-min testing and performances of the RT and NRT groups at 24-h testing after (b) limited and (c) prolonged practice (Fig. 3). Additionally, to investigate the contribution of the consolidation process to the quantitative and qualitative memory changes, we compared performance (d) at 10-min testing to performance at 24-h retesting using data from the RT groups (Fig. 4). Repeated groups: 10-min testing Mean RMSEs on the retention and transfer tests were analysed in a 2 (amounts of practice: limited practice, prolonged practice) 9 3 (tests: retention, extrinsic transformation, intrinsic transformation) ANOVA with repeated measures on the last factor. The analysis indicated a main Psychological Research (2012) 76:611–625 619 Fig. 3 Mean RMSEs during acquisition and on the retention (R), extrinsic transformation (Et) and intrinsic transformation (It) tests for the repeated testing and non-repeated testing groups performing a limited (blocks 1–2; left panel) and a prolonged practice session (blocks 1–17; right panel) in Experiment 2. Error bars reflect standard error of the mean effect of test, F(2, 56) = 30.05, g2p = 0.52. Post hoc comparisons revealed lower RMSE on the retention test (M = 52.92) compared with that on the extrinsic transformation (M = 69.87), which is also significantly lower than on the intrinsic transformation test (M = 78.49). The analysis failed to detect neither a main effect of amount of practice, F(1, 28) \ 1, nor an amount of practice 9 test interaction, F(2, 56) = 2.10. Limited practice: repeated and non-repeated testing groups (24-h testing) Mean RMSEs on the retention and transfer tests were analysed in a 2 (practice conditions: repeated testing, nonrepeated testing) 9 3 (tests: retention, extrinsic transformation, intrinsic transformation) ANOVA with repeated measures on the last factor. The analysis indicated a main effect of test, F(2, 56) = 11.27, g2p = 0.28. Duncan’s multiple range test revealed that RMSE was higher on the intrinsic transformation test (M = 77.71) than on the retention (M = 63.66) and extrinsic transformation tests (M = 66.70), which did not differ from each other. The analysis failed to detect neither a main effect of practice condition, F(1, 28) = 2.64, nor a practice condition 9 test interaction, F(2, 56) \ 1. Prolonged practice: repeated and non-repeated testing groups (24-h testing) Fig. 4 Off-line improvements are defined as the average RMSE difference (10-min testing minus 24-h testing) on the retention (R), extrinsic transformation (Et) and intrinsic transformation (It) tests following a limited (left panel) and a prolonged practice session (right panel) for the repeated testing groups. Error bars reflect standard error of the mean Mean RMSEs on the retention and transfer tests were analysed in a 2 (practice conditions: repeated testing, nonrepeated testing) 9 3 (tests: retention, extrinsic transformation, intrinsic transformation) ANOVA with repeated measures on the last factor. The analysis indicated a main effect of test, F(2, 56) = 8.47, g2p = 0.23, with lower RMSE on the retention test (M = 53.40) than on the extrinsic transformation (M = 64.26) and intrinsic transformation tests (M = 65.39), which did not differ from each other. The main effect of practice condition, F(1, 123 620 28) \ 1, and the practice condition 9 test interaction, F(2, 56) \ 1, was not significant. Off-line improvements (from testing to retesting): repeated testing groups We compared performance on tests conducted immediately and tests conducted following a 24-h delay among participants who had received either a limited or prolonged practice session. The difference between these measures (testing minus retesting) highlighted the off-line development of the motor skill; a positive value reflects improved performance. We measured off-line improvements for the RT groups using two amounts of practice (limited practice, prolonged practice) and three retention/transfer tests (retention, extrinsic transformation, intrinsic transformation), resulting in a 2 9 3 factorial design with repeated measures on the last factor. Results are depicted in Fig. 4. The analysis indicated a main effect of test, F(2, 56) = 6.70, g2p = 0.19, and a significant amount of practice 9 test interaction, F(2, 56) = 3.10, g2p = 0.10. Duncan’s multiple range test revealed higher off-line improvements after limited practice on the extrinsic transformation test (M = 10.58) compared with the retention (M = -0.09) and intrinsic transformation tests (M = 4.71), which did not differ from each other. In contrast, following prolonged practice, post hoc comparisons revealed higher off-line improvements on the intrinsic transformation test (M = 17.63) than on the retention (M = -1.67) and extrinsic transformation tests (M = 7.67), which did not differ from each other. Discussion Participants demonstrated a significant decrease in RMSE across acquisition blocks, suggesting that RT and NRT groups similarly improved their performances during both limited and prolonged practice sessions. In addition, analysis of the retention/transfer data at 24-h testing failed to detect a significant interaction between the practice conditions and the tests for both limited and prolonged practice. This suggests that testing rehearsal did not interact with the off-line development of the coding systems. Our findings argue for the evidence that testing rehearsal per se does not induce representational memory changes of the motor skill if no additional practice trials are performed between testing sessions. To our knowledge, this is the first demonstration of the interactive influences between multisession practice and testing rehearsal on the evolution of motor skill memories. However, although the off-line learning phase was not influenced by testing rehearsal following a single practice 123 Psychological Research (2012) 76:611–625 session, we found evidence that the long-term representation of motor skills depends on the amount of prior practice. Following a limited practice session, we showed a significant advantage at 24-h testing of effector transfer for an extrinsic transformation of the goal movement pattern (i.e., the same spatial–temporal pattern) compared with an intrinsic transformation (i.e., the same pattern of muscle activation and limb joint angles). In contrast, no transfer capacities were observed for an extrinsic or intrinsic transformation of the original pattern following a prolonged practice session (similar performances on the extrinsic and intrinsic transformation tests). Interestingly, these findings revealed that the long-term representation of a motor skill is based on an extrinsic coding system and at an effector-independent level after a limited single practice session, whereas a prolonged single practice session resulted in a multiple coding and effector-dependent representation. Distinct consolidating mechanisms appear therefore to be dependent on prior practice experience (see also Trempe & Proteau, 2010). More specifically, we found robust delayed gains expressed at retesting on the extrinsic transformation test after a limited single practice session and on the intrinsic transformation test after a prolonged single practice session. The present data are consistent with our initial predictions, which were based on the rationale that distinct coding systems develop offline and at different practice scales. We showed between-practice memory enhancements involving an off-line developing effector-independent component that is represented in extrinsic coordinates early in practice, and an off-line developing effector-dependent component that is represented in intrinsic coordinates late in practice. General discussion The main questions addressed by the present study were (1) whether motor skill learning is mediated by different coding systems, and (2) how those coding systems evolve during the rapid and slow learning phases depending on the organisation of practice and testing sessions. This study sought to determine the underlying processes that mediate the formation of motor skill memories. As expected, in two experiments we showed that practice triggers within-practice and between-practice session reorganisations of the motor skill representation, yielding distinct transfer capacities. Our findings revealed that practice induces rapid and non-transferable (i.e., effector-dependent representation) motor skill improvements irrespective of the amount of acquisition trials. We found that the extrinsic component of the skill develops early and remains the dominant coding Psychological Research (2012) 76:611–625 system during practice. Conversely, results revealed latent and practice-dependent reorganisations of the motor skill during the off-line period: a limited practice induced an off-line development of the extrinsic component, whereas a prolonged practice session improved an off-line development of the intrinsic component. Additionally, we provided evidence that the long-term representation of the newly acquired skill is also dependent on the nature of the practice session itself. Specifically, the parsing of practice into multiple sessions provides participants with a stronger intrinsic skill representation, which is less susceptible to transfer in comparison to the outcome of a single practice session. Importantly, however, we found interactive influences between multisession practice and testing rehearsal. Results suggest that the alternation between-practice and testing sessions provides greater transfer capacities than does the mere repetition of practice sessions themselves. Each of these issues will be discussed in the following sections. Practice and its effects on motor skill transfer To investigate the rapid (10-min testing) and latent (24-h testing) reorganisations of the motor skill representation and the practice-effects on motor skill transfer, we evaluated the ability of participants to transfer the original spatial–temporal pattern (extrinsic transformation) or the mirrored one (intrinsic transformation) to the unpractised non-dominant left arm following limited and prolonged practice. It is important to note that the present findings are considered for dominant arm practice but may not generalize to practice with the non-dominant arm. Indeed, we used unidirectional transfer tests from the dominant practised arm to the non-dominant unpractised arm. Consequently, the present finding can only be interpreted for a motor improvement of the dominant arm. In the same vein, recent studies (e.g., Kovacs et al., 2009; Sainburg, 2005; Stöckel, Weigelt, & Krug, 2011; Wang & Sainburg, 2006) provided evidence for inter-manual transfer asymmetries depending on the limb used during practice, i.e., from the dominant limb to the non-dominant limb or vice versa. Thus, on the one hand, our results indicated an early setting-up of an effector-dependent representation, as revealed by the 10-min retention/transfer tests (Experiment 2). Specifically, the better performances observed on the retention test compared with the extrinsic and intrinsic transformation tests provide evidence for within-practice improvements during the acquisition of skilled performance, which is a fast and non-transferable (unpractised arm) process. This rapid motor skill’s reorganisation arises irrespective of the amount of practice. On the other hand, 621 unlike the rapid learning phase, we showed latent and practice-dependent reorganisations of the motor skill. After limited practice, the slow learning phase enables the latent reorganisation of the skill memory towards an effector-independent representation of the extrinsically transformed skill. Conversely, following prolonged practice, the slow learning phase maintained the transfer capacities to an effector-dependent level based on a multiple coding representation of the motor task (similar performances on the extrinsic and intrinsic transformation tests). Lange et al. (2004) reported findings about the neural processes underlying the acquisition and effector transfer of motor skills. For instance, when participants performed the original motor task with their unpractised left arm (here referred to as the extrinsic transformation test), they found increased neural activity in restricted cortical regions and decreased activation of cortical networks. They concluded that processes involved in the modification of the intrinsic motor skill component included fronto-central areas as well as the supplementary motor area. Conversely, when participants performed the mirrored motor task with their left arm (here referred to as the intrinsic transformation test), they indicated that the processes underlying the modification of the extrinsic motor skill component are controlled by intensified parieto-frontal information flow. The production of the mirrored motor task was associated with suppressed cortical activity, which presumably reflects the inhibited recall of the learned extrinsic motor skill component. This interpretation gains support from our results but only during the rapid learning phase. Indeed, we showed that the extrinsic component of the motor skill develops early and remains the dominant coding system during practice. Because the acquired extrinsic component is not the appropriate coding system to perform the intrinsic transformation test, the worse performances observed on the intrinsic transformation test might reflect the inhibition of the learned extrinsic component. In contrast, we observed distinct and practice-dependent representations of the motor skill in long-term memory, suggesting that practice initiates latent and distinct processes during consolidation. The process of memory consolidation during the off-line period is therefore differently affected by the amount of prior practice experience. Distinct consolidating mechanisms The present findings confirmed the relevance of the slow learning phase for the formation of long-term motor skill representations. The effectiveness of the consolidation process for motor skill learning has been extensively 123 622 documented in the literature (e.g., Diekelmann & Born, 2010; McGaugh, 2000, for reviews). It is suggested that consolidation refers to a process that transforms new and labile memories encoded during practice into more stable representations that become integrated into the network of pre-existing long-term memories. In this view, the present study shows for the first time that practice triggers distinct consolidating mechanisms during the off-line period. Indeed, we observed an off-line developing effector-independent component that is represented in extrinsic coordinates following few practice trials, and an off-line developing effector-dependent component that is represented in intrinsic coordinates with extensive practice. Recent research using imaging techniques suggested that motor memory consolidation involves a reorganisation of the motor skill representation residing predominantly in M1 (e.g., Li et al. 2001; Sanes & Donoghue, 2000). The off-line representational memory changes, presumably initiated during practice, could be mediated by functionally distinct subpopulations of neurons within M1 (Steele & Penhune, 2010) or by the recruitment of additional M1 units into a local network that specifically represents the trained motor skill with prolonged practice (Karni et al., 1995). We therefore propose that the motor memory consolidation is not a single process. There might be multiple routes to support off-line learning and engagement of these distinct mechanisms (coding systems) is determined by prior practice experience. It is plausible that both of these mechanisms operate together to mediate off-line memory processing, with one tuned to the more rapidly learned extrinsic component of the task (following limited practice) and the other tuned to the more slowly learned intrinsic component of the task (following prolonged practice). This off-line memory processing could provide the key basis for the formation of motor skill representations in long-term memory. Therefore, the current study provides a critical insight into the off-line memory processing framework. Observing motor skill memory improvements after a single practice session challenges the prevailing notion that the released time is critical for motor memory consolidation. However, it appears that single and multiple practice sessions make distinct contributions to long-term motor skill memory during the off-line period. Single versus multisession practice In the present study, we demonstrated that the long-term representation of motor skills after multisession practice (Experiment 1) was more resistant to transfer compared with a single practice session (Experiment 2) when no testing rehearsal was allowed between-practice sessions 123 Psychological Research (2012) 76:611–625 (i.e., NRT groups).1 The combined results of Experiments 1 and 2 revealed that a single practice session promotes the development of a multiple coding representation of the motor skill, while the parsing of practice into multiple sessions restricts the development of the intrinsic coding system. This demonstrates that the long-lasting multisession practice improvements cannot be conceptualised as the sum of incremental gains expressed following single practice sessions. That is, we postulate that distinct longlasting practice gains would be observed depending on whether the amount of acquisition trials is performed during a unique single practice session or over multiple practice sessions. Furthermore, we showed in Experiment 1 that the longterm representation of the motor skill underwent qualitative changes when comparing participants who rehearsed retention and transfer tests (RT condition) with those who only performed single delayed retention and transfer tests (NRT condition). In accordance with previous research in the testing-effect literature (e.g., Butler, 2010; Rohrer et al., 2010), our results revealed that testing rehearsal promotes the development of a multiple coding representation of the motor task. This suggests that each of the extrinsic and intrinsic codes potentially contribute to retention and transfer performance (see also Kovacs et al., 2009). Thus, performing several practice sessions could sustain the development of a multiple coding representation based on extrinsic and intrinsic components, but only when participants rehearsed retention and transfer tests between-practice sessions. In this condition, the retention and transfer performances were determined by multiple coding systems that acted together to enhance motor skill performance rather than the pure dominance of a single most salient coding system. Our results in Experiment 1 suggest that the reactivation and updating of existing consolidated motor memories (i.e., during the initial retention and transfer tests, respectively) may be critical for further memory modifications. 1 1. We compared retention and transfer performances of the NRT groups following a prolonged practice session arranged in a single and multisession format (Experiments 2 and 1, respectively). We conducted a 2 (practice formats: multisession, single session) 9 3 (tests: retention, extrinsic transformation, intrinsic transformation) ANOVA with repeated measures on test using data from the NRT groups. The analysis revealed a main effect of test, F(2, 56) = 18.28, g2p = 0.39, and a practice format 9 test interaction, F(2, 56) = 3.46, g2p = 0.11. Duncan’s multiple range test indicated that RMSE was significantly higher on the intrinsic transformation test following multisession practice (M = 79.06) than following a single practice session (M = 66.96). No performance difference was observed between groups on the retention (M = 56.20 and 50.59 following a single and multisession practice, respectively) and extrinsic transformation tests (M = 65.69 and 68.06 following a single and multisession practice, respectively). The main effect of practice format was not significant, F(1, 28) \ 1. Psychological Research (2012) 76:611–625 Likewise, recent research suggests that following their initial acquisition and consolidation, memories can be further modified (reconsolidated) after being reactivated during retrieval or after additional practice. Specifically, the reactivation of an existing motor memory yields additional strengthening and off-line gains (Walker et al. 2003). This process, known as ‘‘reconsolidation’’ (e.g., Dudai & Eisenberg, 2004; Przybyslawski & Sara, 1997; Stickgold & Walker, 2007), has been suggested to enable the updating of the internal neural representation of a previously acquired memory (e.g., Dudai, 2006; Sara, 2000). The phenomenon of memory reconsolidation led Censor et al. (2010) to propose a model for human motor memory modification. This model differentiates between what the authors referred to as the ‘‘executing storage domain’’ (primary motor cortex) and the ‘‘core storage domain’’. The brain areas supposed to be involved as a part of the core storage domain which may include the cerebellum, striatum and/or other motor-related cortical areas (e.g., Doyon, Song, Karni, Lalonde, Adams, & Ungerleider, 2002; Ungerleider, Doyon, & Karni 2002; Shadmehr & Holcomb, 1997) and the hippocampus (e.g., Albouy, Sterpenich, Balteau, Vandewalle, Desseilles, Dang-Vu, et al., 2008). According to this model, when the memory is reactivated, recurrent output from the core storage domain to the executing storage domain (which interacts with the environment) enables further memory modifications. In light of this model, our findings suggest that testing rehearsal, when combined to more than a single practice session, enables the modification of existing consolidated motor memories. This implies that testing rehearsal influences the representation of motor skills through the process of memory reconsolidation. Finally, the present findings demonstrated that the combination of practice and testing sessions provides greater effector transfer capacities than does practice alone–an important finding in light of the potential behavioural and motor-related applications of the testing effect. Indeed, our findings are important not only for theoretical but also for practical reasons related to training protocols designed to improve learning, as in the case where one limb is injured, for instance. From a practical standpoint, this research has the potential to impact the design of training and learning therapy programs, and thus to (re)shape the development of new learning and relearning programs. More specifically, when elaborating training protocols on multiple practice sessions, alternating practice and testing sessions seems to be the best condition for long-lasting retention of skill memories and for enhancing transfer capacities. Obviously, the benefits of testing rehearsal on motor skill transfer have yet to be fully exploited. 623 Summary Our findings revealed that practice does trigger rapid (within-practice) and latent (between-practice) reorganisations of the new motor skill representation, yielding distinct transfer capacities. While the extrinsic component of the motor skill develops early and remains the dominant coding system during practice, we showed that practice leads to the off-line development of distinct coding systems. Only few practice trials induces an off-line development of the extrinsic component of the motor skill, whereas extensive practice subserves the off-line development of the intrinsic component. Another important finding is that practice makes transfer of motor skills imperfect, irrespective of the amount of acquisition trials and the practice session format (single session vs. multisession practice). Our results indicated, however, that the alternation between practice and testing sessions provides greater transfer capacities than does the mere repetition of practice sessions. Acknowledgments This work was supported by a grant from the German Research Foundation (PA 774/8-1) and the Agence Nationale de la Recherche (ANR-08-FASHS-14). References Albouy, G., Sterpenich, V., Balteau, E., Vandewalle, G., Desseilles, M., Dang-Vu, T., et al. (2008). Both the hippocampus and striatum are involved in consolidation of motor sequence memory. Neuron, 58, 261–272. Bapi, R. S., Doya, K., & Harner, A. M. (2000). Evidence for effector independent and dependent representations and their differential time course of acquisition during motor sequence learning. Experimental Brain Research, 132, 149–162. Boutin, A., Fries, U., Panzer, S., Shea, C. H., & Blandin, Y. (2010). Role of action observation and action in sequence learning and coding. Acta Psychologica, 135, 240–251. Brainard, D. H. (1997). The psychophysics toolbox. Spatial Vision, 10, 433–436. Butler, A. C. (2010). Repeated testing produces superior transfer of learning relative to repeated studying. Journal of Experimental Psychology. Learning, Memory, and Cognition, 36, 1118–1133. Carpenter, S. K. (2009). Cue strength as a moderator of the testing effect: The benefits of elaborative retrieval. Journal of Experimental Psychology. Learning, Memory, and Cognition, 35, 1563–1569. Censor, N., Dimyan, M. A., & Cohen, L. G. (2010). Modification of existing human motor memories is enabled by primary cortical processing during memory reactivation. Current Biology, 20, 1545–1649. Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hilsdale, Nj: Erlbaum. Cohen, D. A., Pascual-Leone, A., Press, D. Z., & Robertson, E. M. (2005). Off-line learning of motor skill memory: a double dissociation of goal and movement. Proceedings of the National Academy of Sciences of the United States of America, 102, 18237–18241. 123 624 Colby, C. L., & Goldberg, M. E. (1999). Space and attention in parietal cortex. Annual Review of Neuroscience, 22, 319–349. Criscimagna-Hemminger, S. E., Donchin, O., Gazzaniga, M. S., & Shadmehr, R. (2003). Learned dynamics of reaching movements generalize from dominant to nondominant arm. Journal of Neurophysiology, 89, 168–176. Diekelmann, S., & Born, J. (2010). The memory function of sleep. Nature Reviews Neuroscience, 11, 114–126. Doyon, J., Song, A. W., Karni, A., Lalonde, F., Adams, M. M., & Ungerleider, L. G. (2002). Experience-dependent changes in cerebellar contributions to motor sequence learning. Proceedings of the National Academy of Sciences of the United States of America, 99, 1017–1022. Dudai, Y. (2006). Reconsolidation: the advantage of being refocused. Current Opinion in Neurobiology, 16, 174–178. Dudai, Y., & Eisenberg, M. (2004). Rites of passage of the engram: reconsolidation and the lingering consolidation hypothesis. Neuron, 44, 93–100. Hikosaka, O., Nakahara, H., Rand, M. K., Sakai, K., Lu, X., Nakamura, K., et al. (1999). Parallel neural networks for learning sequential procedures. Trends in Neuroscience, 22, 464–471. Hikosaka, O., Rand, M. K., Nakamura, K., Miyachi, S., Kitaguchi, K., Sakai, K., et al. (2002). Long-term retention of motor skill in macaque monkeys and humans. Experimental Brain Research, 147, 494–504. Karni, A., Meyer, G., Jezzard, P., Adams, M. M., Turner, R., & Ungerleider, L. G. (1995). Functional MRI evidence for adult motor cortex plasticity during motor skill learning. Nature, 377, 155–158. Karni, A., Meyer, G., Rey-Hipolito, C., Jezzard, P., Adams, M. M., Turner, R., et al. (1998). The acquisition of skilled motor performance: fast and slow experience-driven changes in primary motor cortex. Proceedings of the National Academy of Sciences of the United States of America, 95, 861–868. Keele, S. W., Ivry, R., Mayr, U., Hazeltine, E., & Heuer, H. (2003). The cognitive and neural architecture of sequence representation. Psychological Review, 110, 316–339. Keele, S. W., Jennings, P., Jones, S., Caulton, D., & Cohen, A. (1995). On the modularity of sequence representation. Journal of Motor Behavior, 27, 17–30. Kleiner, M., Brainard, D., & Pelli, D. G. (2007). What’s new in Psychtoolbox-3? Perception, 36 (ECVP Abstract Supplement). Korman, M., Raz, N., Flash, T., & Karni, A. (2003). Multiple shifts in the representation of a motor sequence during the acquisition of skilled performance. Proceedings of the National Academy of Sciences of the United States of America, 100, 12492–12497. Kovacs, A. J., Boyle, J., Gruetzmatcher, N., & Shea, C. H. (2010). Coding of on-line and pre-planned movement sequences. Acta Psychologica, 133, 119–126. Kovacs, A. J., Han, D.-W., & Shea, C. H. (2009a). The representation of movement sequences is related to task characteristics. Acta Psychologica, 132, 54–61. Kovacs, A. J., Muehlbauer, T., & Shea, C. H. (2009b). The coding of movement sequences. Journal of Experimental Psychology: Human Perception and Performance, 35, 390–407. Krakauer, J. W., Ghilardi, M. F., & Ghez, C. (1999). Independent learning of internal models for kinematic and dynamic control of reaching. Nature Neuroscience, 2, 1026–1031. Krakauer, J. W., & Shadmehr, R. (2006). Consolidation of motor memory. Trends in Neurosciences, 29, 58–64. Lange, R. K., Godde, B., & Braun, C. (2004). EEG correlates of coordinate processing during intermanual transfer. Experimental Brain Research, 159, 161–171. Li, C. S. R., Padoa-Schioppa, C., & Bizzi, E. (2001). Neuronal correlates of motor performance and motor learning in the 123 Psychological Research (2012) 76:611–625 primary motor cortex of monkeys adapting to an external force field. Neuron, 30, 593–607. McDaniel, M. A., & Masson, M. E. J. (1985). Alterning memory representations through retrieval. Journal of Experimental Psychology. Learning, Memory, and Cognition, 11, 371–385. McDaniel, M. A., Roediger, H. L., I. I. I., & McDermott, K. B. (2007). Generalizing test-enhanced learning from the laboratory to the classroom. Psychonomic Bulletin & Review, 14, 200–206. McGaugh, J. L. (2000). Memory—a century of consolidation. Science, 287, 248–251. Panzer, S., Krueger, M., Muehlbauer, T., Kovacs, A. J., & Shea, C. H. (2009). Inter-manual transfer and practice: coding of simple motor sequences. Acta Psychologica, 131, 99–109. Pelli, D. G. (1997). The VideoToolbox software for visual psychophysics: transforming numbers into movies. Spatial Vision, 10, 437–442. Przybyslawski, J., & Sara, S. J. (1997). Reconsolidation of memory after its reactivation. Behavioural Brain Research, 84, 241–246. Robertson, E. M. (2009). From creation to consolidation: A novel framework for memory processing. PLoS Biology, 7, 11–19. Robertson, E. M., Pascual-Leone, A., & Miall, R. C. (2004a). Current concepts in procedural consolidation. Nature Reviews Neuroscience, 5, 576–582. Robertson, E. M., Pascual-Leone, A., & Press, D. Z. (2004b). Awareness modifies the skill-learning benefits of sleep. Current Biology, 14, 208–212. Robertson, E. M., Press, D. Z., & Pascual-Leone, A. (2005). Off-line learning and the primary motor cortex. The Journal of Neuroscience, 25, 6372–6378. Roediger, H. L., I. I. I., & Butler, A. C. (2010). The critical role of retrieval practice in long-term retention. Trends in Cognitive Sciences, 15, 20–27. Roediger, H. L., I. I. I., & Karpicke, J. D. (2006a). Test-enhanced learning: taking memory tests improves long-term retention. Psychological Science, 17, 249–255. Roediger, H. L., I. I. I., & Karpicke, J. D. (2006b). The power of testing memory: basic research and implications for educational practice. Perspectives on Psychological Science, 3, 181–210. Rohrer, D., Taylor, K., & Sholar, B. (2010). Tests enhance the transfer of learning. Journal of Experimental Psychology. Learning, Memory, and Cognition, 36, 233–239. Sainburg, R. L. (2005). Handedness: differential specializations for control of trajectory and position. Exercise and Sport Sciences Reviews, 33, 206–213. Sanes, J. N., & Donoghue, J. P. (2000). Plasticity and primary motor cortex. Annual Review of Neuroscience, 23, 393–415. Sara, S. J. (2000). Retrieval and reconsolidation: toward a neurobiology of remembering. Learning and Memory, 7, 73–84. Shadmehr, R., & Holcomb, H. H. (1997). Neural correlates of motor memory consolidation. Science, 277, 821–825. Soechting, J. F., & Flanders, M. (1989). Errors in pointing are due to approximations in sensorimotor transformations. Journal of Neurophysiology, 62, 595–608. Steele, C. J., & Penhune, V. B. (2010). Specific increases within global decreases: a functional magnetic resonance imaging investigation of five days of motor sequence learning. The Journal of Neuroscience, 30, 8332–8341. Stickgold, R., & Walker, M. P. (2007). Sleep-dependent memory consolidation and reconsolidation. Sleep Medicine, 8, 331–343. Stöckel, T., Weigelt, M., & Krug (2011). Acquisition of a complex basketball dribbling task in school children as a function of bilateral practice order. Research Quarterly for Exercise and Sport, 82, 188–197. Trempe, M., & Proteau, L. (2010). Distinct consolidation outcomes in a visuomotor adaptation task: off-line leaning and persistent after-effect. Brain and Cognition, 73, 135–145. Psychological Research (2012) 76:611–625 Ungerleider, L. G., Doyon, J., & Karni, A. (2002). Imaging brain plasticity during motor skill learning. Neurobiology of Learning and Memory, 78, 553–564. Walker, M. P., Brakefield, T., Hobson, J. A., & Stickgold, R. (2003). Dissociable stages of human memory consolidation and reconsolidation. Nature, 425, 616–620. Walker, M. P., & Stickgold, R. (2004). Sleep-dependent learning and memory consolidation. Neuron, 44, 121–133. Wang, J., & Sainburg, R. L. (2006). Interlimb transfer of visuomotor rotations depends on handedness. Experimental Brain Research, 175, 223–230. 625 Wheeler, M. A., Ewers, M., & Buonanno, J. F. (2003). Different rates of forgetting following study versus test trials. Memory, 11, 571–580. Witt, K., Margraf, N., Bieber, C., Born, J., & Deuschl, G. (2010). Sleep consolidates the effector-independent representation of a motor skill. Neuroscience, 171, 227–234. 123