Riding a Bike

This is a response and discussion prompt for Cognition and Learning 6411.

Article Reference

Haith, A. M., & Krakauer, J. W. (2018). The multiple effects of practice: Skill, habit and reduced cognitive load. Current Opinion in Behavioral Sciences, 20, 196-201. doi:10.1016/j.cobeha.2018.01.015

Statement about Article Selection:

Automaticity is found in motor actions (Haith & Krakauer, 2018) mental computational tasks (Willingham, 2009) and social cognition (Hayward, Homer & Sprung, 2018). Riding a bike is a great example of automaticity since it requires practice and skill to become a habit. I didn’t learn to ride a bike until late into my teen years, so there are many biking behaviours that don’t come automatically to me, unlike my husband who’s been riding since he was five. Cognitive load, or the amount of working memory required to complete the actions required to ride a bike, is reduced when automaticity is achieved (Kalyuga & Singh, 2016; Sweller, 2012). Haith & Krakauer (2018) explore the idea of computational caching as “a means of storing direct associations between inputs and outputs in a way that is amenable to instant lookup” (p. 197), thus reducing cognitive load. As you watch this Backward Brain Bike video keep in mind how you understand practice, skill, habit, and cognitive load as a primary means to achieve automaticity in motor functions, but also consider how automaticity can interfere with our ability to identify or break automatic biases and brain barriers.

Connection to Chapter 5: Is Drilling Worth It? 

Willingham (2009) describes practice as a means to gain minimum levels of competence, reinforce basic skills in order to gain more advanced skills, to safeguard against forgetting, and a way to ensure transference. Contingent to practice is a hierarchy of skills, associations that determine actions, and functional relationships between concepts and actions (Haith & Krakauer, 2018; Willingham, 20009). While the focus for Willingham (2009) is on mental processes, Haith & Krakauer (2018) spotlight motor processes, yet both examine practice as the means to reduce the requirements in working memory and to automatically bring together ideas in order to transform them into something new. Chunking (Willingham 2009) and caching (Haith & Krakauer, 2018) are the means to breaking the constraints of human cognitive architecture (Sweller, 2012) imposed by limitations in working memory (Kalyuga & Singh, 2016).

Article Summary

Practice is an essential component in the fluid, rapid, and proficient deployment of behaviours that are the hallmarks of automaticity (Haith & Krakauer, 2018; Keatley, Chan, Caudwell, Chatzisarantis, & Hagger, 2015; Moors & De Houwer, 2006; Willingham, 2009). Automaticity is a change in behavior resulting from practice, and effected by skills, habits, and cognitive load (Haith and Krakauer, 2018). As demonstrated in the backward bike example, actions requiring motor control are decisional problems which demand computational intensity (Haith & Krakauer, 2018). Caching is defined as a memory reserve of frequently applied computations and associations, readily available for rapid retrieval (Haith & Krakauer, 2018). Further to this, retrieval is a form of cacheable computation, since a cache is considered a single memory event, and “one step retrieval should be considered a hallmark of automatic behavior” (Haith & Krakauer, 2018, p. 197). Haith and Krakauer (2018) suggest that the concept of caching unifies theories of practice and automaticity by accounting for improved response speed, tendencies toward habituation, and decreases in cognitive load.

Haith and Krakauer (2018) apply an analogy of driving a car and braking at a red light as a means to describe how cached computations are hierarchically organized and dependent on goal specific associations. Haith and Krakauer (2018) posit that a cascade of caches occurs, rather than a single behavioural event, when making decisions about braking for a red light. The same can be seen in the bike riding analogy. This sequence of automatic actions better describes how skills, habits, and automaticity impact the resulting behaviours, while considering the varying factors (speed, traffic, location, timing) in this stimulus-response scenario (Haith & Krakauer, 2018). According to Haith and Krakauer (2018), these caches of memory are dependent on how they were practiced, how they become associated, and the nature of the overall automaticity of the acquired task. This can be seen in the backward bike example, where cascading memory caches for riding a bike become disassociated, the practice of riding become disconnected from the skills and habits, where cognitive load increases, and the previously acquired automaticity of biking becomes unreliable.

Cognitive load is impacted by the nature of the cached computations and actions. Reduction in cognitive load occurs when critical decisions with high computational requirements become habitual e.g. turning bike handles left or right to balance. Further to this, skills that play an “auxiliary role that is nevertheless cognitively demanding and important for overall task accomplishment” (Haith & Krakauer, 2018, p. 200), such as pedaling a bike, can also reduce the cognitive load required to ride a bike. Yet, these skilled and habitual behaviours that reduce cognitive loads, also result in inflexibility, outdated cached information, leading to the persistence of old patterns of behaviour (Haith & Krakauer, 2018) that interfere with the acquisition of new practices and automatic behaviours, as modelled in the backward bike scenario. Perhaps that explains why my bike riding habits and skills have not gained the level of automaticity exhibited by my husband. My behavioural computations, as demonstrated by those attempting to ride a backward bike, require high levels of cognition whenever I’m riding on two wheels! Practice has not yet improved my cascading cached computations.

Article Critique

Automaticity encompasses more than the rapid deployment of actions, through the development of skills, habits, and practice (Hayward et al., 2018; Keatley et al., 2015; Moors, 2016; Moors & De Houwer, 2006; Servant, Cassey, Woodman, & Logan, 2017; Willingham, 2009).  Haith and Krakauer (2018) explore caching as a mechanism of automaticity but fail to provide a working definition of cognitive load (Sweller, 2012); ignore factors of automaticity (Bargh, 1994); disregard the underlying components of automaticity (Moors, 2016; Moors & De Houwer, 2006); ignore skill transference that results from practice (Willingham, 2009), and fail to connect the causal and mechanistic explanations of automaticity (Moors, 2016).

Without a working definition of cognitive load, Haith and Krakauer (2018) don’t fully reveal how this essential element of their model improves automaticity and supports their conception of cached memory. If a clear conception of cognitive load was provided, the interplay between working memory and long term memory would support an understanding of cognitive architectures involved in caching computational actions into memory (Merrenboer & Sweller, 2005).

Haith and Krakauer (2018) provide no mention in their article to instrumental research by Bargh (1994) who describes four factors of automaticity – awareness, intention, efficiency, and control. Each of these elements would explicitly have an impact on practice, as part of an autonomous sequence of actions (Moors & De Houwer, 2006). As seen in the backward bike example, Bargh’s (1994) factors have direct impact on the cached computations that provide automaticity in riding a bike. Bargh’s (1994) factors should be considered in the creation of cognitive caching, how it is sustained through ongoing practice, or how it is retained over time.

While a deeper discussion of Moors and De Houwer’s (2006) analysis of the concepts of automaticity is beyond the scope of this critique, their pivotal examination of goal defined and stimulus driven features of automaticity have corresponding impacts on the development, sustainability, and retention of computational memory caches as described by Haith & Krakauer (2018). Since Moors and De Houwer (2006) posit that automaticity “can be diagnosed by the presence of features” (p. 320), it would be logical that Haith & Krakauer (2018) would explicitly explore which features are present in the practices that lead to the development of computational caching. Cache development and retrieval (Haith & Krakauer, 2018) can be seen as intentional and goal directed (Moors & De Hauwer, 2006), as suggested by the example of driving a car and stopping at a red light.

Willingham (2009) indicates that practice helps transference since surface structures of problems wills be seen as similar, underlying structures of similar problems will be recognized, and recognizing functional relationships leads to obvious connections within deeper structures. Haith & Krakauer (2018) don’t identify transference as a result of practice. Their focus on behavioural, motor actions may have prevented them from examining the role practice has in transference of skills and habits of other, associated or similar contexts. For example, skills in automatically stopping a bike at a red light could lead to quicker cache development when transferring this to driving a car. Further examination of this computational memory retrieval (Haith & Krakauer, 2018) should include transferability of skills and habits when practicing and retrieving computational action sequences. The implications for caching when it comes to practicing mental computational sequences, such as those exemplified in chess mastery (Willingham, 2009) should also be explored.

Moors (2016) explores causal and mechanistic explanations of automaticity that have parallels to Haith & Krakauer’s (2018) conception of caching as a form of automaticity. Two causal factors, hardwiring and practice, have direct impact on automatization (Moors, 2016). Just as Haith & Krakauer (2018) explain, when automaticity is achieved, there is a shift from algorithmic computational control to one of single-step memory retrieval, supported by associations in memory (Moors, 2016, Willingham, 2009). Moors (2016) posits a vertical and hierarchical structure, when the complexity of a process includes a varying number of steps and units, which impacts automaticity. This mirrors Haith & Krakauer’s (2018) indication of a cascading cache structure to complex tasks and Willingham’s (2009) conception of deep functional relationships between concepts.  Further, Moor (2016) admits that more recent research supports the notion that complex procedural tasks can be automated, suggesting that perhaps there is merit in Haith & Krakauer’s (2018) conception of a cascading hierarchy of computational caches.

Discussion Question – It’s your CHOICE, pick #1 or #2 to respond to this week.

1. Describe a ‘backward brain bike’ moment you have experienced, when something you automatically knew how to do eluded your memory OR describe a moment when you realized you were functioning on ‘automatic pilot’ and have no recollection of how you completed the actions required for a specific task. How does Haith and Krakauer’s (2018) notion of caching, where the essential components of practice (skills, habit and cognitive load), factor into the conscious or unconscious automatic response you experienced?
2. Do you think Haith and Krakauer’s (2018) notion of caching motor actions can also apply to mental computational chunking that applies to playing chess for example, as described by Willingham (2009), or in social situations e.g. knowing how to respond to social cues at a wedding? Why or why not?

References 

Bargh, J. A. (1994). The four horsemen of automaticity: Awareness, efficiency, intention, and control in social cognition. In R. S. Jr. Wyer, & T. K. Srull (Eds.), Handbook of social cognition, 2^nd Edition. (pp. 1-40). Hillsdale, NJ: Erlbaum.

Haith, A., & Krakauer, J. (2018). The multiple effects of practice: Skill, habit and reduced cognitive load. Current Opinion in Behavioral Sciences, 20, 196-201.

Hayward, E., Homer, B., & Sprung, M. (2018). Developmental trends in flexibility and automaticity of social cognition. Child Development, 89(3), 914-928.

Keatley, D., Chan, D., Caudwell, K., Chatzisarantis, N., & Hagger, M. (2015). A consideration of what is meant by automaticity and better ways to measure it. Frontiers in Psychology, 5, 1-3.

Merrienboer, J. & Sweller, J. (2005). Cognitive load theory and complex learning: Recent developments and future directions. Educational Psychology Review, 17(2), 147-177.

Moors, A. (2016). Automaticity: Componential, causal, and mechanistic explanations. Annual Review of Psychology, 67, 263-287.

Moors, A. & De Houwer, J. (2006). Automaticity: A theoretical and conceptual analysis. Psychological Bulletin, 132(2), 297-326.

Servant, M., Cassey, P., Woodman, G., & Logan, G. (2018). Neural bases of automaticity. Journal of Experimental Psychology: Learning, Memory, and Cognition, 44(3), 440-464.

Smarter Every Day. (2015, April 24). The backwards brain bike (video). Retrieved from https://youtu.be/MFzDaBzBlL0

Sweller, J. (n.d.). Human cognitive architecture. Retrieved from https://pdfs.semanticscholar.org/f7e2/c6b75d4f3cfffb3c7f2a371c0139c49c1bba.pdf 

Willingham, D. (2009). Why don’t students like school? San Francisco, CA: Jossey Bass.

Image Attribution: Photo by Blubel on Unsplash