The key consideration when designing practice activity is the retention and transfer of learning from that activity to real-world performance (4). The manner in which practice activity is organized can affect the performance, learning, and transfer of the skills being practiced. The contextual interference (CI) effect predicts that practice scheduled in a random order (high CI) leads to more errors during practice, but superior learning and transfer of skills, compared to practice scheduled in a blocked order (low CI) (31). A key skill possessed by expert performers in many domains is the ability to anticipate upcoming events (38). However, researchers have yet to examine the effect of practice order on the performance, learning, and transfer of anticipatory judgments. In this article, we examine the CI effect during the practice of anticipatory judgments through simulation techniques and its transfer to applied sport performance in a dynamic, temporally constrained tennis task.
The CI effect has been extensively examined in a variety of motor learning tasks (25). In a study by Shea and Morgan (31), participants practiced three patterns of a simple barrier knockdown motor task under a blocked schedule of practice (e.g., AAA–BBB–CCC) or a random schedule of practice (e.g., ACB–BCA–CAB). After acquisition, participants completed retention tests in which half of the trials were administered in a blocked order and the other half of the trials were administered in a random order. Two transfer tests involving different barrier knockdown tasks were included. During early acquisition trials, the blocked practice group had a significantly faster total movement time compared to the random group, indicating superior performance during practice. However, on retention and transfer tests, the random practice group had a significantly faster total movement time compared to the blocked group, indicating superior learning and the CI effect (31). The CI effect has been replicated in many studies examining motor skill tasks across various contexts, as described in several review articles (20,25,28).
Two main theories have been forwarded to explain the CI effect. First, the reconstruction hypothesis proposes that a random practice order leads to interference between the tasks being practiced, causing them to be forgotten in the short term between trials. As a consequence, participants are required to reconstruct an action plan in order to execute each attempt at a new task, promoting more memorable internal representations for the tasks. In contrast, in a blocked practice order, the same task is practiced on consecutive trials and the same action plan is used, removing the need to reconstruct an action plan each time (30). Second, the elaboration hypothesis proposes that a random practice order promotes more memorable representations through greater comparative and contrastive analyses between the tasks, compared to the repetitive nature of blocked practice (32). Although both these theories differ in the mechanisms that underpin the CI effect, they both attribute the robust findings to the greater cognitive effort and increased neural activity occurring during random—as opposed to blocked—practice (9,18,24).
In most domains, performance involves perceptual–cognitive skills, such as anticipatory judgments and decision making, as well as motor skill execution. Perceptual–cognitive skill refers to the ability of performers to search, identify, process, and integrate environmental information with existing knowledge and current motor capabilities to facilitate the selection of appropriate responses (26). Anticipation is the ability to recognize the outcome of the actions of other athletes before those actions are executed (38). Researchers have reported that experts are superior to novices in perceptual–cognitive skills in a range of domains, including law enforcement (37), medicine (6), military (36), and sport (40). Moreover, perceptual and cognitive skills can be trained using simulation methods (8). For example, Smeeton et al. (33) investigated the relative effectiveness of video-based simulation training combined with various instructional techniques for enhancing anticipatory judgments in tennis. Participants viewed videos of tennis shots occluded at ball–racket contact and were required to predict shot direction. The training groups improved their anticipatory judgment performance from pretest to posttest compared with a control group, and these skills transferred to quicker decision times (DT) in a field-based transfer test (1).
Some researchers have examined the CI effect using perceptual–cognitive tasks, although they did not investigate how the CI effect transfers from training to the real world. Del Rey (10) and Del Rey et al. (11) investigated the CI effect using an anticipatory judgment task involving predicting the arrival of moving lights at a final lamp. A random practice group had significantly lower errors in retention and transfer tests when anticipating the arrival of lights to the final lamp compared to the blocked group, supporting the CI effect. In contrast, Memmert et al. (27) used a more applied badminton simulation task to investigate the CI effect on anticipatory judgments but did not find the classic CI effect. Participants sat in front of a computer screen that showed temporally occluded video footage of a player performing overhead badminton shots to different court locations. Participants were required to predict where the shuttlecock would land on an image of a badminton court, which was also on the computer screen. In six training sessions, one group received the occluded landing locations in a blocked practice order, whereas the other groups received the occluded landing locations in a random practice order. In training, feedback was provided after each trial, whereas learning was assessed in a posttest and a 7-d retention test. There were no differences in the accuracy of anticipatory judgments between random and blocked practice groups across acquisition, posttest, and retention. One possible explanation for this finding concerns the design of the representative task. A simplistic response to a small visual display was used in the task, both of which are thought to limit the expert advantage (39). Researchers have suggested that the coherence of a representative task with its real-world version is vital for appropriate processing to take place, such that decreasing the coherency of the task creates constraints on processing (12). Representative task design involves the use of large screens that allow life-size images to be projected, showing dynamic rather than static images. They allow the performer to complete a response that is the same as, or as similar as possible to, that produced in the actual performance environment (4). The other possible explanation for the lack of group differences is that participants only practiced anticipatory judgments of badminton overhead stroke, but to different landing locations. By definition, CI is the scheduling of practice for a number of different skills, not a single skill (30). Research is needed to examine how practice should be structured during perceptual–cognitive skills training requiring a complex movement response to a number of different skills on a large screen upon which a life-size video is projected.
Retained transfer of learning from practice to real-world performance should be the key consideration when designing practice. To our knowledge, no researchers have examined how practice should be structured during simulation training so that skills transfer more effectively to real-world performance, despite the widespread use of this method (8). For example, Helsdingen et al. (15,16) showed that the CI effect extended to complex police judgment tasks that involved prioritizing the urgency of different case descriptions. The random practice group was significantly more accurate at solving cases compared to the blocked group in a posttest. A transfer test was included in this study, but it was another simulation task in which only structural and surface features differed from training tasks, rather than a transfer to a real-world task. Consequently, there is a need to examine judgments in an applied setting to extend the theory and to verify the translational value of such interventions.
The current study examines the CI effect on simulation training of anticipatory judgments in a temporally constrained task in tennis and the retention and transfer of this ability to applied sport performance in the field. Anticipatory judgments of three different tennis skills were practiced across an acquisition phase in either a random practice order or a blocked practice order, with learning being measured across pretest, retention test, and transfer test. The three skills being anticipated were forehand groundstroke, forehand smash, and forehand volley. It is expected that participants would improve the accuracy of their anticipatory judgments as a result of the training protocol and that this improvement would transfer to the field. In line with the CI effect, it is hypothesized that the blocked practice group will have more accurate anticipatory judgments during the acquisition phase compared to the random practice group. In contrast, the random practice group will have more accurate anticipatory judgments in retention and transfer to a field-based protocol when compared with the blocked practice group, indicating superior learning of the skills.
Based on previous perceptual research (7,33) and taking into account the difficulty of recruiting and keeping participants during an extended period of time, we estimated that nine participants per group were required for this study. Participants were 18 intermediate-level junior tennis players who were divided into either a blocked practice group (n = 9; mean ± SD age, 12.9 ± 1.6 yr) or a random practice group (n = 9; mean ± SD age, 13.2 ± 1.6 yr). Participants in the two groups were matched by ensuring no between-group differences in prior tennis experience, number of hours per week they currently played tennis, laboratory pretest accuracy scores, field pretest accuracy, and field pretest DT (Table 1). Separate independent t-tests on each of these variables showed no between-group differences (all t < 1). Written informed consent was obtained from the participants and their parent or legal guardian before participation, and these documents were stored in the research department of the lead institution. The experiment was conducted in the country of residence and designed in accordance with the 1964 Declaration of Helsinki. Ethical approval was obtained from the lead institution’s research ethics committee.
Test and Training Film Construction
Test and training films were developed for the simulation. Films were made for a pretest, three training sessions, and a 7-d retention test. Video clips of tennis shots were edited using video editing software (Adobe Premier CS5, Adobe Systems, San Jose, CA, USA). Each clip began with a black screen and the trial number. Each film clip consisted of one of three intermediate-level tennis players (mean ± SD age, 19.7 ± 1.2 yr; mean ± SD tennis experience, 8.7 ± 1.2 yr; mean ± SD tennis hours per week, 6.3 ± 1.5 h) on the other side of the net of a standard indoor tennis court. The clips involved the ball arriving at the player from an off-camera feeder player who was one of the two other players and the player moving to the ball, swinging the racket, and hitting the ball back over the net using a predefined shot. The video was filmed from a central position on the baseline of the tennis court at a height of 1.5 m to provide a representative view of the court from the participants’ perspective. Shots were selected for the test film footage when they satisfied three criteria: 1) the ball fed to the player went over the net in a central area so that the player returning it performed similar body movements for each stroke; 2) the returned ball was struck cleanly by the player with the speed of return replicating a game situation; and 3) the returned ball bounced in the intended target location.
Players executed three offensive tennis shots: 1) forehand groundstroke; 2) forehand smash; and 3) forehand volley. These three shot types were selected as researchers have demonstrated that when a player executes them in an attacking manner it promotes the greatest need for anticipatory judgments by their opponent (34). The shots were played to one of four locations on the opponent’s side of the court: 1) left front; 2) left back; 3) right front; and 4) right back. The three skills (groundstroke, volley, and smash) have distinct invariant characteristics in that certain elements of the skill are relatively fixed, such as movement patterns. However, variations are possible within the skills, such as the speed and height of ball flight, which are described as parameters of the invariant skill (30). Within an applied sport setting, it is difficult to control every parameter within an invariant skill. For example, Hall et al. (13) demonstrated the CI effect on a baseball-hitting task using three invariant skills or pitches (fastballs, curveballs, and change-ups) received in either a blocked practice order or a random practice order. Within the three different types of pitches, the parameters varied somewhat, such as the speed and height of each pitch. In our study, the blocked and random schedules of practice were created using the three relatively different invariant skills, as per the majority of other research in this area (13,14,23), and not the different parameters within a skill (27).
For the test and training clips, the video occluded at three points, which were selected based on previous research examining anticipatory judgments in sport (17,34). The occlusion points were 80 ms before ball–racket contact, at ball–racket contact (0 ms), and 80 ms after ball–racket contact. At the occlusion point, the screen went black, and the phrase “Respond” appeared in large font, which allowed 3 s for the participant to respond before the next trial number appeared. Each trial lasted approximately 9 s. Across the pretest (n = 108 trials), three training sessions (n = 72 trials per session), and 7-d retention test (n = 108 trials), shot type, shot landing location, and occlusion condition were balanced so that there was an equal number of each condition. To ensure that the structure of practice in the pretest and retention test did not favor either of the groups, we used both blocked and random practice structures (54 trials in each), which were counterbalanced across participants. For the blocked practice group, the three skills were completed so that, in each practice session, all trials of one shot were completed before moving on to all trials for the next shot, with the order of the three shots being counterbalanced across participants. The server, end location of the shot, and occlusion point used varied across these blocks. For the random practice group, the quasi-random order meant that one of the three tennis shots was not played more than twice in a row. In the 7-d retention test, 50% of the clips were repeated from the pretest and 50% were new clips, and these were balanced equally across the random and blocked conditions. These new clips were used to ensure that participants were not completely familiar with the clips after completing the pretest. The old clips were so the level of difficulty of the clips was kept constant between tests. The old and new clips were balanced equally across both blocked and random conditions.
Apparatus and Procedure
The experiment consisted of a pretest in both the laboratory and the field, three laboratory-based training sessions separated by 7 d, and a 7-d retention test in both the laboratory and the field. All sessions were completed alongside the regular tennis training sessions of the participants. It was arranged with the coach that no other anticipation training would occur during the study period.
Laboratory pretest and retention test
Figure 1A presents an overhead illustration of the experimental setup. Participants stood 4 m from the center of a large portable projection screen (2.74 × 3.66 m, Cinefold Projection Sheet; Draper Inc., Spiceland, IN, USA) on which the test films were backprojected (CP-X345; Hitachi, Yokohama, Japan). The size of the image was representative of the proportions normally experienced in game situations when participants are positioned on the baseline of the court. Participants held a tennis racket and were required to respond to the onscreen shot by simulating a return shot and physically moving to one of four markers that were on the floor 1 m from them in four directions corresponding to the four locations where the ball could bounce. Hand notation was used to record the movement response from each trial. The laboratory pretest and the 7-d retention test took approximately 15 min each to complete.
Field pretest and retention test
Figure 1B presents an illustration of the experimental setup for the field test. Participants were required to respond to shots played by an opposing intermediate-level tennis player (mean age, 22 yr; mean tennis experience, 7 yr; mean tennis hours per week, 4 h) on a standard indoor tennis court. The player was not part of the laboratory test or training film. The shots performed by the player were the same three as used in the laboratory test films. A second skilled tennis player who projected the ball to the player was positioned slightly off court to the right of the participant. Upon receiving the feed ball, the player on court was required to execute each shot to one of the four locations used in the film on the participant’s side of the court. The lead experimenter briefed the feeder and player on which shots to be performed across the tests so that each skill was counterbalanced for all participants. There were 36 trials for each participant in the field-based protocol, which were divided into two sets of 18 trials. In one set, participants received the shots in a blocked order, where all trials on one skill were completed before starting all trials on the next skill. In the other set, they received the shots in a random order, in which no shot type was repeated more than twice in a row. The order of presentation of the two sets was counterbalanced across participants. Any shots that did not reach the intended target or failed to go over the net were discarded, and the trial was repeated at the end of the session, where they were placed in their respective practice orders.
Participant responses were filmed using a video camera (XM-2; Canon, Tokyo, Japan) with wide-angled lens at a sampling frequency of 50 Hz. The camera was located behind and to the left of the participant. It recorded the moment of ball–racket contact and the movements of the participant. The field pretest and the 7-d retention test took approximately 15 min each.
The training phase consisted of three laboratory sessions that occurred once each week during a 3-wk period between the pretest and the 7-d retention test. Participants watched two presentations of the same shot during each trial. First, the video footage occluded at one of the three time points, and they were required to respond to the anticipated location of the ball bounce, as in the pretest and the retention test. Second, the same video clip was shown in full, enabling the participants to view the ball flight and shot outcome in terms of where on the court the ball bounced. No verbal instructions were given regarding the information on screen or participant movements and responses. Each training session consisted of 72 trials and took approximately 15 min to complete. The 72 trials consisted of 24 trials of each shot, with each shot equally divided into the three occlusion points and four locations.
Dependent measures and statistical analysis
For the laboratory tests and training, response accuracy (RA) was the primary dependent variable. Responses were deemed as being accurate when the movement response of the participant was to the same location as the bounce of the ball on their side of the court. Data from the laboratory and the field were analyzed separately. In the field, both RA and DT were recorded. DT was defined as the time period from ball–racket contact by the opponent to the initiation of movement by the participant (in milliseconds). The movement initiation of the participant was used as response. Movement initiation was defined as “the first frame where there was an observable and significant lateral motion to the right or left of the racket, the hips, the shoulder or the feet, which was made in order to move to the future location of the next strike” (34, p. 822). Movement initiation in tennis usually occurs during or just after a player executes a split-step/landing sequence (35). Responses initiated before ball contact received negative values. Footage of the field tests was analyzed through Adobe Premier CS5 software. A participant from each group dropped out from the field transfer test due to an injury and a time scheduling issue, so they were excluded from the field test data set. Measures of interobserver and intraobserver reliability were obtained for DT by using intraclass correlation techniques (3) on the data from two participants (144 trials), one from each group. The obtained correlation coefficients for the interobserver (0.938) and intraobserver (0.876) measures demonstrated the reliability of data analysis.
RA across training sessions was analyzed using a 2 group (random, blocked) × 3 training (training 1, training 2, training 3) × 3 occlusion (80 ms before, ball–racket contact, 80 ms after) mixed-design ANOVA, with repeated measures on the last two factors. To examine learning in the laboratory, we analyzed RA using a 2 group (random, blocked) × 2 test (pretest, retention) × 3 occlusion (80 ms before, ball–racket contact, 80 ms after) mixed-design ANOVA, with repeated measures on the last two factors. Bonferroni post hoc procedure was used for any significant within-participant main effects. Tukey HSD post hoc procedure was used for any significant interactions. Performance on the field-based protocol was analyzed using a factorial multivariate ANOVA (MANOVA) in which group (blocked, random) was a between-participant variable, test (pretest, retention) was the within-participant variable, and RA and DT were dependent measures. Planned comparisons were carried out to compare the performance of both groups on each dependent measure, respectively. The α level for significance was set at P < 0.05 for all tests, and partial η2 (ηp2) was used as a measure of effect size.
Figure 2 shows RA across the two groups on the three training sessions. A 2 group × 3 training × 3 occlusion ANOVA revealed no main effect for group in RA during acquisition [F(1, 16) = 0.10, P = 0.76, ηp2 = 0.01]. There was a significant improvement in RA across the training phase [F(2, 32) = 10.25, P < 0.01, ηp2 = 0.39]. Post hoc analysis indicated that RA in the third training session (mean ± SD, 54.8 ± 4.8%) was significantly higher than that in the first training session (mean ± SD, 47.1 ± 3.4%; P < 0.01). However, RA in the second training session (mean ± SD, 51.2 ± 7.4%) was not significantly different from either of the other two training sessions (all P > 0.05). Figure 3 shows RA at each occlusion point from the three training sessions. There was an occlusion main effect [F(2, 32) = 153.38, P < 0.01, ηp2 = 0.91]. Post hoc analysis indicated that RA in the trials occluded 80 ms after ball–racket contact (mean ± SD, 66.6 ± 6.6%) was significantly higher than RA in trials occluded at ball–racket contact (mean ± SD, 48.2 ± 5.5%) and 80 ms before ball–racket contact (mean ± SD, 38.7 ± 3.0%; P < 0.01). Furthermore, RA on trials occluded at ball–racket contact was significantly greater than that on trials occluded 80 ms before ball–racket contact (P < 0.01). No interaction effects were observed.
Laboratory pretest and retention test
Figure 2 shows RA across the two groups on the pretest and the 7-d retention test. A 2 group × 2 test × 3 occlusion ANOVA revealed no main effect for group [F(1, 16) = 2.90, P = 0.11, ηp2 = 0.15]. However, RA significantly improved from pretest (mean ± SD, 50.9 ± 6.9%) to retention test (mean ± SD, 67.5 ± 6.9%) [F(1, 16) = 113.74, P < 0.01, ηp2 = 0.88]. There was also a significant group–test interaction [F(1, 16) = 6.03, P = 0.03, ηp2 = 0.27]. Post hoc analysis revealed no significant difference in RA in the pretest between the blocked group (mean ± SD, 50.5 ± 8.6%) and the random group (mean ± SD, 51.2 ± 5.6%). However, in the retention test, the random group (mean ± SD, 71.7 ± 5.3%) had a significantly higher RA compared to the blocked group (mean ± SD, 63.3 ± 6.0%).
Figure 3 shows RA at each occlusion point on the pretest and the 7-d retention test. There was an occlusion main effect [F(2, 32) = 70.63, P < 0.01, ηp2 = 0.82]. Post hoc analysis indicated that RA was significantly higher in the trials occluded 80 ms after ball–racket contact (mean ± SD, 70.7 ± 5.8%) compared to trials occluded at ball–racket contact (mean ± SD, 58.0 ± 8.2%) and 80 ms before ball–racket contact (mean ± SD, 48.8 ± 7.7%; P < 0.01). Moreover, RA on trials occluded at ball–racket contact was significantly greater than that on trials occluded 80 ms before ball–racket contact (P < 0.01). There was a test–occlusion interaction [F(2, 32) = 5.01, P < 0.01, ηp2 = 0.24]. Post hoc analysis revealed that, in the pretest, RA was not different between trials occluded at ball–racket contact (mean ± SD, 47.7 ± 9.9%) compared to trials occluded 80 ms before ball–racket contact (mean ± SD, 42.8 ± 8.7%). However, in the retention test, RA was significantly higher on trials occluded at ball–racket contact (mean ± SD, 68.4 ± 9.0%) compared to trials occluded 80 ms before ball–racket contact (mean ± SD, 54.8 ± 10.8%). The group–test–occlusion interaction was not significant.
Field pretest and transfer test
Figure 4 shows RA and DT across the two groups on the pretest and the 7-d retention test. The results of MANOVA used to analyze performance on the field-based protocol, with RA and DT as dependent measures, are presented in Table 2. Planned comparisons indicated that there was no significant difference in RA or DT between groups in the pretest (all P > 0.05). Furthermore, in the retention test, no significant difference in RA was found between the blocked group (mean ± SD, 88.5 ± 7.0%) and the random group (mean ± SD, 88.0 ± 3.3%) [F(1, 14) = 0.03, P = 0.86, ηp2 = 0.02]. However, there was a significant difference in DT on the retention test [F(1, 14) = 7.19, P = 0.02, ηp2 = 0.34]. The random group (mean ± SD, 98 ± 89 ms) had a significantly faster DT in the retention test compared to the blocked group (mean ± SD, 238 ± 118 ms) and the pretest, suggesting greater transfer of learning.
We investigated the CI effect on the acquisition of anticipatory judgments in a dynamic, temporally constrained environment. Furthermore, we reported a novel attempt to examine whether the structure of practice during simulation training affects the transfer of these skills to an applied sport setting. Specifically, we investigated the effects of a random schedule and a blocked schedule of practice on the acquisition, retention, and transfer of anticipatory judgments in tennis acquired through simulation training. The main hypothesis of the CI effect was that the random group would demonstrate significantly greater improvements in judgments from the laboratory pretest to the retention test compared to the blocked group. Our data provide evidence for the CI effect, as the random group demonstrated significantly more accurate judgments in the 7-d laboratory-based retention test compared to the blocked group and the pretest. These data support previous research investigating the CI effect on motor and perceptual–cognitive skills (10,11,15,16,31). However, these findings contradict those of Memmert et al. (27), who did not provide any support for the CI effect in this domain. Therefore, the current findings provide the first indication that the structure of practice affects the acquisition of anticipatory judgments during simulation training.
It was hypothesized that, in a transfer test to an applied sport situation, the random group would demonstrate more accurate and faster anticipatory judgments compared to the blocked group. In the 7-d transfer test, although RA did not differ between groups, participants in the random group had significantly faster DT compared to the blocked group and the pretest, indicating superior learning. The lack of between-group difference in RA in the transfer tests may be attributable to scores being relatively high across all field tests. In the field-based protocol, participants had access to all of the ball flight information, providing an advantage over the laboratory, where vision of ball flight was not available due to the occlusion paradigm. However, findings for DT indicate that the training intervention led to earlier cue use and that this transferred onto the field, as both groups reduced their DT significantly. Furthermore, the data suggest that the random group was better able to learn early cue use than the blocked group, as they made significantly faster anticipatory judgments in the field-based protocol. Data provide novel evidence that the CI effect transfers from simulation training to an applied sport setting. Our findings support previous literature (13) and extend current understanding by showing that principles from the motor skills literature on the CI effect apply to simulation training to improve anticipatory judgments (8).
In line with the CI effect, we hypothesized that, during the acquisition phase, the blocked group would have more accurate anticipatory judgments compared to the random group. However, contrary to this hypothesis, RA was not significantly different between the two groups during the three acquisition sessions. These data contradict the majority of previous researchers who have investigated the CI effect (25,28). However, some researchers reported a lack of difference between blocked and random practice groups during acquisition but still found the hypothesized differences in the retention and transfer phases (16), somewhat contradicting the “typical” CI effect (20). A possible explanation for this finding is that the three invariant tennis skills contained variable parameters, such as shot location. Other researchers in both applied (13) and laboratory-based settings (23) have examined blocked and random schedules of practice that contain variable parameters, as opposed to constant parameters. Similar to our study, they have shown a lack of differences between practice groups across acquisition, whereas the random practice group has been superior to the blocked practice group in retention and transfer.
The two main theories forwarded to explain the CI effect are the forgetting/reconstruction hypothesis (21,22) and the elaboration/distinctiveness hypothesis (31). They both predict greater cognitive effort during random practice compared with blocked practice, either before or after skill execution, respectively. Another theory is that random practice conditions lead to greater cognitive effort and increased neural activity across practice simply because the task changes often compared to blocked practice (9,18,24). Anticipatory judgments are a perceptual–cognitive process that sometimes may not involve constructing an action plan and executing a motor skill, so the advantage of random practice in these cases may be explained by the elaboration hypothesis. Alternatively, anticipation in sport may be related to the embodiment of actions, which suggests that motor regions of the brain activate via the mirror neuron system when observing a movement (29). Furthermore, the mere knowledge of an upcoming movement is suggested to excite the motor system through a resonant mechanism, enabling people to anticipate, rather than react to, others’ actions (19). Therefore, for experienced tennis players who have the necessary motor skills to perform the observed strokes, anticipation may not just be a perceptual–cognitive process (5). The anticipated opponent action might resonate within the individual’s own motor system, activating an action plan for completing that skill (2). If so, then random practice, compared to blocked practice, would be hypothesized to lead to greater cognitive effort by reconstructing actions plans (21,22). Further research is required to reveal the underlying cognitive mechanisms that lead to the CI effect on anticipatory judgments, perhaps by comparing skilled participants (who have fine-tuned motor–resonance systems) and novice participants (who have not).
In summary, we report novel data to suggest that the robust CI effect on perceptual–motor skills extends to the development of perceptual–cognitive skills through simulation training and to the transfer of these skills to an applied sport setting. A random practice schedule during simulation training resulted in improved anticipatory judgment accuracy on a 7-d retention test in the laboratory compared to a blocked schedule of practice. Furthermore, the positive changes in anticipatory judgment transferred to a field-based condition where the random schedule of practice resulted in faster DT in the field during the 7-d transfer test compared to a blocked schedule of practice. Overall, the findings support previous researchers (31) by demonstrating that a random schedule of practice, compared to a blocked schedule of practice, leads to superior learning and extends this principle to the training of anticipatory judgments through a video simulation technique.
The authors declare no additional funding sources.
The authors declare no conflicts of interest.
The results of this study do not constitute an endorsement by the American College of Sports Medicine.
1. Abernethy B, Schorer J, Jackson RC, Hagemann N. Perceptual training methods compared: the relative efficacy of different approaches to enhancing sport-specific anticipation. J Exp Psychol Appl
. 2012; 18 (2): 143–53.
2. Aglioti SM, Cesari P, Romani M, Urgesi C. Action anticipation and motor resonance in elite basketball players. Nat Neurosci
. 2008; 11 (9): 1109–16.
3. Atkinson G, Nevill AM. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Med
. 1998; 26 (4): 217–38.
4. Broadbent DP, Causer J, Williams AM, Ford PR. Perceptual–cognitive skill training and its transfer to expert performance in the field: future research directions. Eur J Sport Sci
. 2015; 15 (4): 322–31.
5. Bruce L, Farrow D, Raynor A, Mann D. But I can’t pass that far! The influence of motor skill on decision making. Psychol Sport Exerc
. 2012; 13 (2): 152–61.
6. Causer J, Barach P, Williams AM. Expertise in medicine: using the expert performance approach to improve simulation training. Med Educ
. 2014; 48 (2): 115–23.
7. Causer J, Holmes PS, Williams AM. Quiet eye training in a visuomotor control task. Med Sci Sports Exerc
. 2011; 43 (6): 1042–9.
8. Causer J, Janelle CM, Vickers JN, Williams AM. Perceptual expertise: what can be trained? In: Hodges NJ, Williams AM, eds. Skill Acquisition in Sport: Research, Theory and Practice
, New York: Routledge; 2012. p. 306–24.
9. Cross ES, Schmitt PJ, Grafton ST. Neural substrates of contextual interference during motor learning support a model of active preparation. J Cogn Neurosci
. 2007; 19 (11): 1854–71.
10. Del Rey P. Training and contextual interference effects on memory and transfer. Res Q Exerc Sport
. 1989; 60 (4): 342–7.
11. Del Rey P, Wughalter E, Du Bois D, Carnes MM. Effects of contextual interference and retention intervals on transfer. Percept Mot Skills
. 1982; 54 (2): 467–76.
12. Glockner A, Betsch T. Decisions beyond boundaries: when more information is processed faster than less. Acta Psychol
. 2012; 139 (3): 532–42.
13. Hall KG, Domingues DA, Cavazos R. Contextual interference effects with skilled baseball players. Percept Mot Skills
. 1994; 78 (3): 835–41.
14. Hall KG, Magill RA. Variability of practice and contextual interference in motor skill learning. J Motor Behav
. 1995; 27 (4): 299–309.
15. Helsdingen A, van Gog T, van Merriënboer J. The effects of practice schedule and critical thinking prompts on learning and transfer of a complex judgment task. J Educ Psychol
. 2011; 103 (2): 383–98.
16. Helsdingen A, van Gog T, van Merriënboer J. The effects of practice schedule on learning a complex judgment task. Learn Instr
. 2011; 21 (1): 126–36.
17. Jackson RC, Warren S, Abernethy B. Anticipation skill and susceptibility to deceptive movement. Acta Psychol
. 2006; 123 (3): 355–71.
18. Kantak SS, Sullivan KJ, Fisher BE, Knowlton BJ, Winstein CJ. Neural substrates of motor memory consolidation depend on practice structure. Nat Neurosci
. 2010; 13 (8): 923–5.
19. Kilner JM, Vargas C, Duval S, Blakemore SJ, Sirigu A. Motor activation prior to observation of a predicted movement. Nat Neurosci
. 2004; 7 (12): 1299–301.
20. Lee TD. Contextual interference: generalizability and limitations. In: Hodges NJ, Williams AM, eds. Skill Acquisition in Sport: Research, Theory and Practice
, New York: Routledge; 2012. p. 79–93.
21. Lee TD, Magill RA. The locus of contextual interference in motor–skill acquisition. J Exp Psychol Learn Mem Cogn
. 1983; 9 (4): 730–46.
22. Lee TD, Magill RA, Weeks DJ. Influence of practice schedule on testing schema theory predictions in adults. J Mot Behav
. 1985; 17 (3): 283–99.
23. Lee TD, Wulf G, Schmidt RA. Contextual interference in motor learning: dissociated effects due to the nature of task variations. Q J Exp Psychol A
. 1992; 44 (4): 627–44.
24. Lin C-H, Fisher BE, Winstein CJ, Wu AD, Gordon J. Contextual interference effect: elaborative processing or forgetting-reconstruction? A post hoc
analysis of transcranial magnetic stimulation-induced effects on motor learning. J Mot Behav
. 2008; 40 (6): 578–86.
25. Magill RA, Hall KG. A review of the contextual interference effect in motor skill acquisition. Hum Mov Sci
. 1990; 9 (3–5): 241–89.
26. Marteniuk RG. Information Processing in Motor Skills
. New York: Holt, Rinehart, and Winston; 1976. 244 p.
27. Memmert D, Hagemann N, Althoetmar R, Geppert S, Seiler D. Conditions of practice in perceptual skill learning. Res Q Exerc Sport
. 2009; 80 (1): 32–43.
28. Merbah S, Meulemans T. Learning a motor skill: effects of blocked versus random practice a review. Psychol Belg
. 2011; 51 (1): 15–48.
29. Rizzolatti G, Craighero L. The mirror–neuron system. Annu Rev Neurosci
. 2004; 27: 169–92.
30. Schmidt RA, Lee TD. Motor Control and Learning: A Behavioural Emphasis
, 5th ed, Champaign (IL): Human Kinetics; 2011. 488 p.
31. Shea JB, Morgan R. Contextual interference effects on the acquisition, retention, and transfer of a motor skill. J Exp Psychol Hum
. 1979; 5 (2): 179–87.
32. Shea JB, Zimny ST. Knowledge incorporation in motor representation. In: Meijer ON, Roth K, editors. Complex Movement Behaviour: The Motor–Action Controversy
. Amsterdam: North Holland; 1988. p. 289–314.
33. Smeeton NJ, Williams AM, Hodges NJ, Ward P. The relative effectiveness of various instructional approaches in developing anticipation skill. J Exp Psychol Appl
. 2005; 11 (2): 98–110.
34. Triolet C, Benguigui N, Le Runigo C, Williams AM. Quantifying the nature of anticipation in professional tennis. J Sports Sci
. 2013; 31 (8): 820–30.
35. Uzu R, Shinya M, Oda S. A split-step shortens the time to perform a choice reaction step-and-reach movement in a simulated tennis task. J Sports Sci
. 2009; 27 (12): 1233–40.
36. Ward P, Farrow D, Harris KR, Williams AM, Eccles DW, Ericsson KA. Training perceptual–cognitive skills: can sport psychology research inform military decision training? Mil Psychol
. 2008; 20 (1): S71–102.
37. Ward P, Suss J, Eccles DW, Williams AM, Harris KR. Skill-based differences in option generation in a complex task: a verbal protocol analysis. Cogn Process
. 2011; 12 (3): 289–300.
38. Williams AM, Ford PR. ‘Game intelligence’: anticipation and decision making. In: Williams AM, ed. Science and Soccer: Developing Elite Performers
, Oxon, UK: Routledge; 2013. p. 105–21.
39. Williams AM, Grant A. Training perceptual skill in sport. Int J Sports Psychol
. 1999; 30 (2): 194–220.
40. Williams AM, Ward P, Knowles JM, Smeeton NJ. Anticipation skill in a real-world task: measurement, training, and transfer in tennis. J Exp Psychol Appl
. 2002; 8 (4): 259–70.