Working Memory and Emotional Subsystems

Introduction

The way human beings perceive and understand the world involves both bottom-up as well as top-down processing. Our perceptual and cognitive systems not only detect signals and process information pre-attentively, they also store and retrieve it, form connections between both prior and new knowledge, and learn to use that information metacognitively. Thus, cognition involves receiving sensory input and stimuli and also integrating that information with relevant data stored in our long-term memory (Lazarus, 1984). 

However, when it comes to carrying out day-to day tasks, something else plays a critical role in how we make sense of our environment. This element is what scholars have called the working memory. 

Working memory is used to hold information briefly and hence it can be defined as a “temporary storage system under attentional control that underpins our capacity for complex thoughts” (Baddeley, 2007). It can be seen as a mental scratch pad which brings about relevant elements from both sensory as well as long-term memory and, in effect, lies at their juncture. It is “minute in its ability to process new material but massive in its ability to process very extensive and complex, previously learned information” (Sweller, 2003). 

Thus, the three components of the memory system—sensory, working, and long-term memory—help us engage in all kinds of cognitive activities that can range from fairly simple to highly intricate and complex tasks.

This paper is divided into two parts. It primarily focuses on working memory and its emotional subsystems. In the first part of the paper, different models and theories of working memory are explored. Additionally, cognitive load and the limitations of working memory is also discussed in detail. These aspects are connected to how emotions like anxiety and motivation can affect working memory. In the second part of the paper, building on these theoretical aspects, a self-checkout kiosk of Walmart Superstore has been reviewed.

Working Memory: Models and Theories

In the late 1960s, Atkinson and Shiffrin (1968) proposed the multi-store model that talked about sensory register, short-term memory, and long-term memory. Although popular, this model received criticism due its simplistic representation of short-term memory. In response, Baddeley and Hitch (1974) postulated a three-component model for working memory that replaced the notion of unitary, short-term memory (Baddeley, 2012). 

According to this model, working memory consists of: (1) phonological loop, (2) visuo-spatial sketchpad, and (3) central executive. The central executive is the chief attentional control system that manages and allocates data to phonological loop and visuo-spatial sketchpad. While the phonological loop deals with written and spoken input, the visuo-spatial sketchpad stores and processes information in a spatial or visual form. Although the initial model was useful, episodic buffer was added to it (Baddeley, 2012) so as to integrate the components of working memory with long-term memory. According to Baddeley (2010), this component “is capable of holding multi-dimensional episodes or chunks, that may combine visual and auditory information possibly also with smell and taste.”

Cowan (1988) proposed another model of working memory which, in a way, specified the interaction between episodic buffer and central executive. This model posits that the main feature of working memory is an attentional system that has the capacity of four chunks and outside this, the short-term storage depends on long-term memory (Cowan, 2000). While the two models are different, they both lay emphasis on the importance of the interaction of working memory with long-term memory.

Limitations of Working Memory and Cognitive Load

            Unlike long-term memory that can handle and organize large swathes of information, working memory comes with certain limitations. Working memory is limited in capacity and duration and is considered to be highly volatile. Although it employs attentional resources and can handle cognitive load of varying levels, overloading it with information and tasks can affect performance (Engle, 2002). When the task at hand requires more cognitive load than the person can handle, it leads to mistakes and abandonment.

Limited Capacity

            Working memory can carry only limited amount of information as it has finite capacity. George Miller (1956), in his seminal work, talked about the limited capacity of memory and our ability to combine or chunk information to reduce cognitive load. He postulated that we only have the capacity and memory span to retain seven, plus or minus two chunks of information. Recognizing this, Sweller, in the late 1980s, came up with the cognitive load theory that can be used to explain the limited capacity of working memory. According to this theory, a person’s knowledge base is made up of a combination of elements or cognitive structures called schema (Sweller, 2011). Schemas can be acquired over a long period of time, and this is what differentiates an expert from a novice. Cognitive load can be: (1) intrinsic – minimum load that cannot be reduced, (2) germane – load associated with construction and commitment of schema to long-term memory, or (3) extraneous – load that is irrelevant and does not aid in schema formation. Since working memory has limited capacity, it is important to reduce extraneous cognitive load and focus on the task at hand.

In contrast to Miller’s work, recent research (Baddeley, 2010; Cowan, 2010) suggests that memory capacity is limited to only three to five chunks. This capacity varies from person to person depending on their age, emotional state, and intellectual abilities. (Conway, 1996; Luck & Vogel, 2013; Miller & Buschman, 2015). It ebbs with old age, anxiety, learning disabilities, and fatigue.

Limited Duration

            Working memory is limited in its time duration. Information decays over time and it can easily get lost without recall strategies or active rehearsal (Brown, 1958; Peterson & Peterson, 1959). This can happen as quickly as twenty to thirty seconds. Complex tasks, interference, and unnecessary thoughts can also add to the cognitive load, and thus important information in the working memory can easily get lost. 

Highly Volatile

            Another limitation of working memory is that it is highly volatile. Information in working memory can evolve as well as get corrupted easily. It can easily be affected by prior experiences, emotions, attentional state, and external events (Engle, 2002). While experts seem to have better working memory than novices, their abilities can be explained by the fact that they can quickly access information reserved in their long-term memory. Older people find it particularly difficult to retain thoughts in the working memory because their ability to suppress extraneous cognitive load diminishes with age. This, in turn, places extra cognitive load on their working memory and makes it volatile. It is therefore important to be mindful of limitations of working memory when designing products. 

Emotional Subsystems

            According to Dolan (2002), “the cognitive domain where the influence of emotion is best understood is memory.” Emotions have the ability to affect working memory both positively as well as negatively (Engel, 2002). While positive emotions like motivation direct executive control, negative emotions like anxiety can increase cognitive load (Pessoa, 2009; Derakshan & Eysenck, 2010). Although we experience various emotions like trust, challenge, pleasure, stress, and fear, the dominant emotions in human beings that affect working memory are motivation and anxiety. Both these emotions are forms of arousal, which is required to take action (Staal, 2004). While motivation is a positive emotional state, anxiety is usually considered a negative emotion.

Motivation

            Motivation can either be personal or imposed. People are usually motivated by reward, gratification, efficiency, and pleasure. Motivation improves cognitive performance and efficiency. It influences different working memory processes like selective attention, active maintenance, and encoding (Krawczyk & D’esposito, 2013). 

Anxiety

Anxiety in small doses can help us stay engaged and focused, but too much anxiety can negatively impact working memory. It has been posited (Moran, 2016) that anxiety restricts the capacity of working memory. A lot of factors can induce anxiety. Some of them include fatigue, frustration, sensitive topics, anger, threat, loss, or even the environment. According to Miller and Bichsel (2004), there are two types of anxiety: (1) trait, and (2) state. People with trait anxiety tend to feel anxious in every situation (Sorg & Whitney, 1992), while people with state anxiety feel it only in particularly stressful or difficult situations. Although both affect task performance, people with high trait anxiety suffer from working memory deficit and perform poorly in stressful situations.

Design Review: Walmart Self-Checkout Kiosk

            Walmart has a number of self-checkout kiosks in its stores across the US. It allows customers to avoid long queues and quickly purchase items without having to interact with or wait for a cashier. For the purpose of this paper, a self-checkout kiosk at a Walmart Superstore has been reviewed.

Scenario

An old woman is shopping at Walmart and has some items in her shopping cart. It’s the holiday season and regular cashiers are busy attending several customers. The queue is long and therefore the woman decides to use a self-checkout kiosk. Although she has shopped at Walmart a number of times, she can be considered a novice since has never used such a self-checkout kiosk before. She is motivated because she feels she would be able to quickly scan her items, buy them in a matter of minutes, and leave for home to spend time with her family and friends. 

The kiosk is conveniently located close to the exit and the bright self-checkout lights indicate its exact location. Quick checkout and easy access to the exit are some of the rewards a shopper can associate with self-checkout kiosks. These would motive the old shopper to use the service and, in turn, improve her cognitive performance and efficiency.

The home screen of the kiosk has low information density and would ideally not put cognitive load on a shopper’s working memory. As recent research has suggested, memory capacity is limited to three to five chunks. This screen can be pre-attentively categorized and chunked as four categories: (1) the receipt, (2) accessibility and sound buttons, (3) visual instructions on how to scan an item, and (4) six buttons. Although the six buttons might not put cognitive load on expert shoppers, the language and “Request Help” button should be chunked together with the accessibility and sound buttons. This would ensure that aged shoppers or those with disabilities can get help and change their preferences without inconvenience. Since all the other buttons are associated with shopping and checking out, chunking them together with preferences like language and assistance does not make sense. Moreover, “Quick Lookup” and “Lookup by Name” are similar categories and can easily be chunked together as a single category. This will not only help reduce the number of options, but it will also prevent shoppers from feeling anxious when selecting an option.

Some items like fruits and vegetables cannot be scanned directly by the machine. A shopper would either have to look up the item or enter the item number to proceed. This could particularly increase the cognitive load of an older shopper and make her anxious. For example, when buying a packet of grapes, if the shopper wants to enter the item number, she would have to search for it on the packet. The item number is not clearly visible and there is a lot of extraneous information on the bag. This can not only increase a shopper’s cognitive load, but it can also make her anxious and prone to errors. If there is a queue waiting behind her, she would become all the more anxious and quickly but erroneously enter the item number. The recommendation would be to add a sticker on the packet that has good contrast and clearly states the item number. Another recommendation would be to pack the grapes in a box and add a sticker with barcode on it for easy scanning.

If the shopper ends up entering an incorrect item code or wants to cancel an item that she has added to her shopping list, she would need assistance. Since the purpose of such a kiosk is to minimize the time taken to buy items, this step can be somewhat frustrating for a shopper. Most of the times, it has been noted that store attendants are not readily available, and a shopper cannot move on with checking out other items until they have been attended to. If there is a queue waiting for their turn, it can make the shopper feel conscious and anxious and that can negatively impact their working memory. The recommendation would be to allow shoppers to either cancel an item without assistance or to make sure that someone is available for assistance at all times.

The payment screen has low information density and the instructions are easy to follow. Since it does not expect a user to remember anything, it does not impact the limitations of working memory. For anyone who has previously used a pin-pad to make payments, it would be easy to do so using the self-checkout kiosk. This is because they would have this mental model stored in their long-term memory. Since most users, whether they are old or young, are used to pin-pads, this step would not add to a shopper’s cognitive load. Even if she is a novice user, the clear instructions would ensure that she does not end up making mistakes. 

Conclusion

            The above paper and design review have tried to clearly lay down the characteristics, models, theories, scope, and limitations of working memory. It has also tried to explain the role of emotions like anxiety and motivation and how they affect working memory. For a designer, the challenge is not only to design interfaces that can help users quickly make decisions without much cognitive load, but to also make provisions for general fluctuations in working memory’s capacity in face of anxieties. Though, not easily achieved, this should remain the cherished goal of all design practitioners.

References

Atkinson, R. C., & Shiffrin, R. M. (1968). Human memory: A proposed system and its control processes. In Psychology of learning and motivation (Vol. 2, pp. 89-195). Academic Press.
Baddeley, A. (2007). Working memory, thought, and action (Vol. 45). OUP Oxford.
Baddeley, A. (2010). Working memory. Current biology, 20(4), R136-R140.
Baddeley, A. (2012). Working memory: theories, models, and controversies. Annual review of psychology, 63, 1-29.
Baddeley, A. & Hitch, G. (1974). Working memory. In Psychology of learning and motivation (Vol. 8, pp. 47-89). Academic press.
Brown, J. (1958). Some tests of the decay theory of immediate memory. Quarterly Journal of Experimental Psychology, 10(1), 12-21. 
Conway, A. R. (1996). Individual differences in working memory capacity: More evidence for a general capacity theory. Memory, 4(6), 577-590.
Cowan, N. (1988). Evolving conceptions of memory storage, selective attention, and their mutual constraints within the human information-processing system. Psychological Bulletin, 104(2), 163-191. 
Cowan, N. (2000). Processing limits of selective attention and working memory: Potential implications for interpreting. Interpreting, 5(2), 117-146.
Cowan, N. (2010). The magical mystery four: How is working memory capacity limited, and why? Current directions in psychological science, 19(1), 51-57.
Dolan, R. J. (2002). Emotion, cognition, and behavior. science, 298(5596), 1191-1194.
Engle, R. W. (2002). Working memory capacity as executive attention. Current directions in psychological science, 11(1), 19-23.
Derakshan, N., & Eysenck, M. W. (2010). Introduction to the special issue: Emotional states, attention, and working memory. Cognition and Emotion, 24(2), 189-199.
Krawczyk, D. C., & D'esposito, M. (2013). Modulation of working memory function by motivation through loss‐aversion. Human brain mapping, 34(4), 762-774.
Lazarus, R. S. (1984). On the primacy of cognition.
Luck, S. J., & Vogel, E. K. (1997). The capacity of visual working memory for features and conjunctions. Nature, 390(6657), 279.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological review, 63(2), 81.
Miller, H., & Bichsel, J. (2004). Anxiety, working memory, gender, and math performance. Personality and Individual Differences, 37(3), 591-606.
Miller, E. K., & Buschman, T. J. (2015). Working memory capacity: Limits on the bandwidth of cognition. Daedalus, 144(1), 112-122.
Moran, T. P. (2016). Anxiety and working memory capacity: A meta-analysis and narrative review. Psychological Bulletin, 142(8), 831.
Pessoa, L. (2009). How do emotion and motivation direct executive control?. Trends in cognitive sciences, 13(4), 160-166.
Peterson, L., & Peterson, M. J. (1959). Short-term retention of individual verbal items. Journal of experimental psychology, 58(3), 193.
Sorg, B. A., & Whitney, P. (1992). The effect of trait anxiety and situational stress on working memory capacity. Journal of Research in Personality, 26(3), 235-241.
Sweller, J. (2003). Evolution of human cognitive architecture. Psychology of learning and motivation, 43, 216-266.
Sweller, J. (2011). Cognitive load theory. In Psychology of learning and motivation (Vol. 55, pp. 37-76). Academic Press.