Cognitive Load: Looking Into a Crucial Element of UX
If you work in the user experience field, you’ve probably come across the term “cognitive load” at some stage, either in exchanges with colleagues, at conferences, or in one of the countless articles online. To understand the term and its meaning correctly, however, it’s worth taking a closer look at the underlying theory and its background in applied psychology.
There are countless situations in daily life when we feel mentally overwhelmed — be it at university, where you are trying to pay attention to the lecturer and want to take notes at the same time, or on a website that seems so overloaded that your first instinct is to just click out of it again. The list could go on and on, but we understand the problem — too many stimuli vying for the limited resources of our working memory at the same time.
We can identify two critical factors that underlie our understanding of cognitive stress: our knowledge of both thought and attention processes.
Although the first concepts relating to our thinking developed in ancient Greece, it was George A. Miller who laid the foundation for modern cognitive psychology. Since 1956, his research on short-term memory has led to fundamental empirical findings on how our thought processes function. He discovered that on average, only 7 ± 2 units of information could be stored — which even training can’t improve. It was the starting point for further research in the field of cognitive psychology and, in the meantime, has become the most quoted psychological work.
Baddeley & Hitch, for example, published their Working Memory Model in 1974, after it was recognized that there is a difference between merely retaining information in short-term memory and manipulating it. It was also known that varying sensory impressions are processed differently and independently of each other. The models were refined to explain the observations from experiments further. Since the end of the 1990s, attention processes have been increasingly taken into account, and working memory has been understood as a cognitive process that keeps information in a particularly accessible state.
As a result, it quickly became apparent that information can be processed better if its presentation takes into account the limited capacity of our working memory. Sweller’s “Cognitive Load Theory” is based on empirical studies and has this aim precisely; to optimally prepare learning material or information. It divides the total cognitive load into three categories: the “intrinsic cognitive load,” the “extraneous cognitive load,” and the “germane cognitive load.”
The “intrinsic cognitive load” relies on learning content and complexity. The more complex and challenging the learning content, the more cognitive resources have to be used. The actual material to be conveyed and the load itself cannot be changed; individual learner knowledge levels alone can reduce this part of the cognitive load. Teaching addition is more straightforward than teaching the Gaussian integral error function.
The “germane cognitive load” describes the necessary and desirable part of the cognitive load that helps us build new schemas or link information with what is already known. It is so to speak a vital resource for learning and understanding.
The “extraneous cognitive load” is determined by the way information is presented. Any form of presentation that is not conducive or relevant to the process of understanding increases the load and distracts from the actual task.
Since our working memory capacity is restricted, we have to reduce noise to help detect the signal. We achieve this by keeping the extraneous load as low as possible to free up more capacity for the germane load. Furthermore, this load can be measured psychophysiologically in the laboratory by pupil reaction.
Whether we design learning materials, websites, or apps, the composition of the total cognitive load determines how easily our users find their way around, perform tasks, or decide on a product. Therefore, it makes sense to follow a few general principles which we would like to outline briefly here:
This refers to the intrinsic cognitive load. Indeed, we can’t change this load in general, but we can split it up. If you’re building a house and therefore need to get 5 tons of bricks from A to B, you usually will do this bit by bit. The total amount of bricks you transported will not change, and all of them are needed, but you split them up into batches to reduce workload at a given time.
Noise is everything that distracts the user and takes their attention away from the essential information we want to transport. Therefore avoid visual clutter in any form, may it be inappropriate images or design elements, redundant links, and needless text or meaningless typography choices. Valuable things should be as salient as possible.
You should take what your users know and what they expect into consideration. If your presentation violates their concepts of reality, you force them to reevaluate their models — a very uncomfortable task as it uses up a lot of energy.
Researchers have found a set of common effects in learning situations. We should consider them regularly as they can help us identify the possible problems a user may be confronted with.