New direction in decision neuroscience

An article in the new issue of Nature Reviews Neuroscience describes the current endeavour in decision neuroscience to integrate knowledge from ethology and behavioural ecology. This approach already generated some tasks which consider several naturalistic factors and revealed interesting details about human decision making.

Decision neuroscience is an interdisciplinary effort bringing together economics, psychology and neuroscience to study how humans make their choices, and what neural processes underlie these decisions. This is usually done by assessing the behaviour of human subjects in specific, highly constrained tasks. For example, one popular task involves choosing between several “slot machines” on a computer screen, each of which yields a variable amount of points when chosen. The participants have to figure out the value of choosing a given machine to maximize their points. This task is used to examine the switching between the exploitation of a resource and the exploration of other opportunities, a basic situation requiring decision making. The models developed by decision neuroscience provide valuable descriptions on how the brain may compute the choices in these tasks, but these situations are far from real scenarios encountered by humans. Recent research in decision neuroscience, inspired by ethology and behavioral ecology, is starting to consider more and more aspects of real environments, and models developed by these fields are proving highly successful in describing the behaviour of human participants in these novel, more naturalistic tasks.

For example, in natural environments different desired items may require different amounts of energy to collect. The net rate maximization principle asserts that the value of the item and the required effort (the cost) should both be taken into account when making a choice. So, if you have to choose between your favourite food and one that you do not like that much, the fact that you have some of the latter in the fridge, while the former has to be fetched from a not-so-close shop, will probably influence your decision. It has been already showed by decision neuroscience studies that the dorsal anterior cingulate cortex has a role in computing the cost and benefit of different choices, thus the computation suggested by this principle probably occurs in the human brain.

Another realistic consideration is that in natural settings, the value of a resource decreases with exploitation. For example, when all the easily accessible low hanging fruit is taken from a tree, only the higher hanging, hard-to-reach ones remain and thus, the overall value of collecting fruit from that tree decreases. The marginal value theorem (MVT) is able to identify the optimal foraging strategy in environments where the value of the resources decrease predictably, and in a recent study it generated a very good approximation of the human participants’ responses.

Real world foraging may be influenced by other gatherers and this has also been examined in a recent human study with simulated competition. The ideal free distribution (IFD) model takes competition into account, and it was highly successful in describing the behavior of the participants in this study, and also the behavior of mallard ducks in an other experiment. A finding in this ethology experiment sheds light on another important variable: individual differences between the gatherers. Dominant ducks gathered more food than what the IFD model predicted, as they competed more successfully than some of their conspecifics.

Finally, the risk of predation is also an important constraint on decision making. According to the net rate maximization model, when a squirrel acquires a nut it should eat it on the spot, because carrying it away requires extra energy. In reality however, squirrels frequently take the food back to their nests. This makes sense, if one considers that eating it out in the open raises the risk of predation, which can be formalized as a huge cost for the squirrel. In some experiments the human participants had to navigate in a virtual environment where an AI controlled agent chased them and triggered a shock to their hand if it captured them. These experiments revealed the neural structures, which influence decision making under threat.

Thus, ethology and behavioral ecology holds important insights into behavior, which are slowly embraced by decision neuroscience. This mash of approaches proved to be very informative in the past years, and will hopefully bring us closer to understanding behavior in its natural complexity.