How Variability Increases Learning

Variability and specificity differ over time frames

To get good at something you have to practice that specific skill. That makes sense - but how much can variability contribute to learning and how different can the stimulus be. There are no clear answers to some of these questions but what is clear is that variability is, long-term, likely much more effective strategy for learning anything.

There are a couple of theories of learning – one is that of specificity. This means you get good at precisely what you practice, learn, or train.

This is also neatly explained by neural pathways and neuroplasticity. As we train or practice a specific skill, or specific knowledge, this activates very precise networks in the brain. Because of this precision this entrains these networks and leads to plasticity.

As you may, or may not know, the brain is plastic i.e. it grows and shrinks. There are a number of ways this happens:

1. Activation and reactivation of synapses. Your synapses are the connections between brain cells and what enables you brain to pass on the signal to the next neuron or a further neurons, or pass it along a different network. What most people don’t know, however, is that many of these synapses may not be active. Your brain basically builds these synapses and if you don’t use them enough, deactivates them. So activity activates more synapses – this is a rapid process.

2. Building of new synapses. I outline some of the processes in this article here. Basically, high activation will stimulate the growth of new synapses increasing the strength of a given pathway or building new more refined pathways.

3. Automating networks. The networks that are used for complex tasks are then activated simultaneously – this means a network for a function can almost be seen as one operational unit rather than a collection of various neurons.

4. Precise coordination. Research into precise motor skills shows that the difference even between some elite athletes is in how precisely motor units coordinate and operate together. For example, a piece of research last year showed that the differences in skating speed of elite cross-country skiers was in how well they transferred power – which is a function of precise minute coordination in, and between, the muscles.

So, all of this, in short, means that practising skills improves these skills, irrespective of what skills. This can be a cognitive skill such as doing Sudoku, speaking a language, or various reasoning exercises. It can also be a motor skill, such as playing tennis, running (as I do), boxing, or juggling. It can also be perceptional skills such as identifying emotions, identifying breeds of dogs, or recognising faces.

However, this specificity is also problematic. For, example there has been plenty of research into cognitive games. You know the type of thing - they claim it will keep your brain young and boost your IQ. Or so many companies claimed until one of the most well-known companies, Lumosity, was fined for over exaggerating the claims. And what were they over exaggerating? They were over exaggerating the benefit in everyday life i.e. the transfer of these specific skills to general living.

Many of these games test and practice various cognitive abilities such as short-term memory. Short-term memory has been shown to contribute to fluid intelligence - but also it is of interest to those who are aging because short-term memory is one of the first things to decline as you age (tell me about it).

However, a lot of the initial research showed that participants who played these games improved on the games getting ever better scores but there was little transfer of this to real-world tasks - such as remembering what to get at the shops or remembering the name of the person you were just introduced to. However, the general consensus now is that they are generally beneficial and do have some transfer but don’t expect miracles.

Variability

This then brings us to the alternative concept, called many things, and I will call variability. Variability refers to the concept of how much variation you should have in the stimulus, the stimulus which we hope will trigger adaptation, i.e. learning.

In come Limor Raviv and colleagues at the Max Planck Institute for Psycholinguistics who have recently conducted a meta-review and reviewed 150 studies across different domains. First off, it is clear that the general concept is widely used but that it is called different things in different places. And also these diverse fields and domains have not necessarily communicated with each other. So, the concept of variability in different names is researched in areas such as: