Representativeness in research and practice
While Brunswik's (1956) concept of representative design was originally applied more to experimental designs, Pinder et al. (2011) later adapted it more explicitly for sport practice design. In particular, they drew parallels between issues in both sport research and practice and outlined how they could be solved using more representative designs. For example, as outlined in my previous post, research on visual anticipation processes in sport typically uses paradigms which 1) do not provide natural, 3-dimensional information to participants and 2) require participants to provide verbal or button-pressing responses that are nothing like the complex movements performed in performance settings.
Similarly, given how performance can be seen as an emergent process involving reciprocal and continual interactions between perception and action, practice designs in sport should allow for perception and action to be tightly coupled and refrain from training them independently. The idea here is that, just as researchers would want their experimental tasks to be representative of the real-world behavioural settings (that participants operate in and to which the results of the experiment are intended to apply), coaches should also want their practice tasks to be representative of real-world performance settings (that athletes find themselves in to increase the likelihood of skill transfer).
Representative (Learning) Design
To this end, Pinder et al. (2011) introduce 2 ways in which coaches can increase the representativeness of practice tasks. These include functionality (i.e., the extent to which athletes are exposed to informational sources in training similar to those found in the performance environment) and action fidelity (i.e., the extent to which actions and movements performed in training are similar to those performed in competition). Crudely, functionality and action fidelity simply examine how representative the information presented and actions performed, respectively, in a training task are of a real-world performance setting.
Pinder et al. (2009) provide a nice empirical example for this. In their study, cricket batters faced both a real (person) bowler and a ball projection machine. Comparing the 2 conditions, it can be argued that using the ball projection machine removes any pre-ball release kinematic information from the bowler, information that the batter likely uses to time the loading phase of the swing. In this respect, we can say that a practice task that uses a ball projection machine has limited functionality (i.e., only ball flight information is presented).
Results of the study revealed different temporal structures of the swing between the conditions. Specifically, those facing the machine exhibited earlier back swing initiation, front foot movement, front foot placement, and downswing initiation. This suggests that batters were performing different swing movements between the machine and live bowler conditions, allowing us to argue that the machine task has poor action fidelity. Due to suboptimal levels of functionality and action fidelity, one might then conclude, using this framework, that the machine practice task isn't entirely representative of a real-game situation.
As a side note, functionality and action fidelity aren't the only ways to assess task representativeness! To give an example, Krause et al. (2018) have developed the Representative Practice Assessment Tool (RPAT) for tennis training. Meanwhile, I have also informally developed an IDEAs framework (inspired by Gray (2024)) for representative learning design. Here, the letters correspond to areas in which coaches can seek representativeness, including Information, Decision-Making, Emotions, and Action.
Eye tracking
That the cricket conditions above lead to different movement profiles provides a hint that tasks of different representativeness lead to different perceptual-motor demands from athletes. Might we find similar patterns when looking at other indicators of skilled performance, such as gaze behaviour? To answer this, we turn to van Maarseveen & Oudejans (2018), who got youth basketball players to shoot at baskets while wearing eye-tracking glasses. Crucially, while defenders were present throughout the training task, the defenders only aggressively defended in one condition (i.e., contested condition), and passively remained a distance away from the shooter in the other condition (i.e., uncontested condition).
Linking this to representativeness, the idea here is that because shooters are often exposed to active defenders during a real basketball game, the training task involving contesting defenders would provide a more representative environment for shooters to train in compared to one with more passive defenders. Movement results corroborated those from previous basketball studies -- in the contested condition, participants took faster shots, jumped higher during their jump shots, and shot the ball with higher arcs.
When looking at gaze behaviour, the authors found that those in the contested condition fixated longer on the defender, and also displayed later final fixation onsets. Some jargon here, but a fixation is basically a point of visual focus exhibited by participants. Meanwhile, the final fixation simply refers to the final point of focus before the ball is released. In simpler terms, shooters in the contested condition spent more time looking at the defender, and looked at the final focus point (presumably the basket) at a later time point compared to those in the uncontested condition.
There are more interesting results to be reported, but I'll stop here for now. The main point that I want to get across is that in sports, training tasks of different representativeness can affect not only one's movements, but also their gaze behaviour. Whether and how such changes actually affect performance remains to be seen, as results are quite equivocal and interpretations, in my opinion, have largely been informed by one's theoretical leanings. However, what we can conclude is that there seems to be different perceptual and motor demands placed on athletes based on task representativeness, and this very likely affects the learning and transfer of skill from practice to performance settings.
Implications for my BSc thesis
As you might have guessed, I intend to manipulate task representativeness in my study. This is a good time to remind ourselves of the 2 different types of representativeness I have introduced thus far, and how they have informed my project. On the one hand, representative (experimental) design is about creating experimental tasks whose behavioural demands are adequately sampled from real-world settings -- my thoughts on this have pushed me to conduct a fully in situ experiment. On the other hand, representative (learning) design is about creating training tasks whose behavioural demands are adequately sampled from performance and competitive settings -- my thoughts on this have led me to manipulate it as my independent variable.
How will I manipulate representativeness? Following Pinder et al. (2011), I will vary both functionality and action fidelity. Here, I intend to get volleyball players to perform a spiking task as an outside hitter from position 4. Functionality will be operationalised as the presence of blockers -- in some conditions, opposing blockers will be present, whereas in others, they will be absent. Action fidelity will be operationalised as the mode of hand-ball interception -- in some conditions, I will get participants to catch the ball, while in others, they will be asked to fully spike the ball. This yields 4 conditions, namely: catch/no-blockers, catch/blockers, spike/no-blockers, spike/blockers.
The catch/no-blockers condition is of particular interest. Far too often, I have seen coaches use this as one of the first steps to coaching a volleyball spike. My argument here is that while it may seem like a spike, the perceptual-motor demands vary immensely when one is catching a ball vs. hitting it. In other words, I'm saying that this very common drill lacks action fidelity and is an unrepresentative (and hence ineffective at worst, and inefficient at best) way to coach a jumping spike. If I find gaze behavioural differences between this and other conditions, it might add strength to my claim that it places very different perceptual-motor demands on players.
Some challenges
At the moment, this design is all that I'm working with. While I do expect to see differences in terms of gaze behaviour, the specifics of my hypothesis currently elude me. What kind of gaze behaviour measure will I be using (there are many, many to choose from)? In what direction would I expect the changes to be? This will require more reading on my part, and hopefully, the answers to these questions will be made clear by the time I write my preregistration.
Furthermore, the above design represents my current thoughts up till now. Just this week, I realised a major flaw in my conditions -- if some conditions don't have a blocker, how can I put as my hypothesis that I expect to find similar results to van Maarseveen & Oudejans (2018), i.e., that participants would generally focus more on the defenders? If some conditions do have blockers and others don't, it is a foregone conclusion that my blocker conditions would see participants focusing more on them, duh!
With regard to this, I might have to change how I operationalise functionality, perhaps in a way similar to van Maarseveen & Oudejans (2018). For instance, instead of looking at the presence of blockers, I could look at how active the blockers are. Simply put, perhaps blockers passively stand behind the net without jumping or raising their arms in some conditions, while actively jumping and reaching over the net in others. That might work after all!
Concluding remarks
Left alone, my current experiment seems pretty decent, especially for an undergraduate dissertation. But wait, there's more! In the next post, I'll introduce a new element into my research design, this time focusing on the role action capacity and individual differences might play in my study!
References
Brunswik, E. (1956). Perception and the representative design of psychological experiments. In University of California Press eBooks. https://doi.org/10.1525/9780520350519
Krause, L., Farrow, D., Reid, M., Buszard, T., & Pinder, R. (2018). Helping coaches apply the principles of representative learning design: validation of a tennis specific practice assessment tool. Journal of Sports Sciences, 36(11), 1277–1286. https://doi.org/10.1080/02640414.2017.1374684
Pinder, R. A., Davids, K., Renshaw, I., & Araújo, D. (2011). Representative Learning design and functionality of research and practice in sport. Journal of Sport and Exercise Psychology, 33(1), 146–155. https://doi.org/10.1123/jsep.33.1.146
Pinder, R. A., Renshaw, I., & Davids, K. (2009). Information–movement coupling in developing cricketers under changing ecological practice constraints. Human Movement Science, 28(4), 468–479. https://doi.org/10.1016/j.humov.2009.02.003
van Maarseveen, M. J. J., & Oudejans, R. R. D. (2018). Motor and gaze behaviors of youth basketball players taking contested and uncontested jump shots. Frontiers in Psychology, 9. https://doi.org/10.3389/fpsyg.2018.00706
No comments:
Post a Comment