Many forms of mental illness can affect our moods. But that isn’t all they do: they can also damage our willpower. Problems such as depression, post-traumatic stress disorder, attention-deficit hyperactivity disorder and some forms of brain injury are known to have an effect on motivation and the ability to summon mental effort.
Researchers often investigate these conditions using tests that tweak the level of mental effort involved. They might present their subjects with a choice between two tasks, one of which is mentally strenuous but highly rewarding while the other is less rewarding but easier.
Tests of this sort allow us to investigate potential treatments for these motivation-impairing disorders.
But if you really want to get into the basic neural mechanisms, at some point you need to get a brain out of its skull and under the microscope. As most people tend to object to the idea of having their brain extracted, this sort of research is normally conducted with non-human animals.
But this presents a difficulty: how do you test motivation or mental effort in a rat?
Historically, researchers have dodged this issue by substituting physical effort instead. They may give a rat the choice of two different paths through a maze, only one of which requires clambering over obstacles. But is this physical-effort-based task really an appropriate match for the mental-effort-based tasks used with humans?
Paul Cocker of the University of British Columbia didn’t think so. In recently published research, Cocker and his colleagues present a new approach that they believe provides a better equivalent to the tests used with humans.
Rats were trained on an original “cognitive effort task”. First, the rat needed to choose between a hard or easy task, by pressing one of two levers. If they chose the hard task, a light inside one of many nosepoke ports would flash on very briefly (just one-fifth of a second). If they chose the easy task, the signal light would remain illuminated for much longer (a full second).
In each case, if the rat stuck his nose into the appropriate nosepoke port within five seconds of the light coming on, it would be rewarded with a sugar pellet.
But while the easy task was rewarded with one sugar pellet, the hard task earned two. The idea is that the rat would need to concentrate more intently to follow the hard signal than would be the case with the easy signal.
There are a few potential criticisms of this task. If the rats were less accurate at choosing the appropriate nosepoke port in the hard task, they may end up choosing the easier task because it is a more reliable way of getting the tasty treats, rather than because it requires less cognitive effort.
Twice the payout isn’t very appealing if it’s delivered less than half of the time. But Cocker and colleagues did a thorough job of foreseeing and countering these objections, with a variety of clever controls and comparisons.
One of these bits of cleverness was the use of a set of “yoked” (joined together) control animals. These rats were given a “super-easy” task: their signal lights stayed on for however long it took for the animal to stick his nose into one of the ports, regardless of whether they were going for the high- or low-reward version of the test.
Instead of being rewarded with sugar pellets every time they stuck their noses into the appropriate spot, these rats were only rewarded some of the time.
This is where the “yoke” comes in: each of the yoked control animals was matched with one of the rats getting the normal version of the test, and their reward probabilities were determined by the success rate of their partner from the main group.
So, if the main group rat only chose the correct nosepoke port 80% of the time, his yoked control partner would only have his correct responses rewarded 80% of the time.
In this way, the experimenters teased apart the two meanings of “difficulty”: the “how much effort does this require?” meaning and the “how likely am I to succeed?” meaning.
In the normal group, the rat’s responses could have been driven by either factor. In the yoked control group, the “effort” component was removed, leaving just the “chance of success” part.
Examining the behaviour of the yoked control animals allowed the researchers to argue that the main group really was being driven by “effort” rather than “chance of success”.
So how does this get put to use? While the major focus of Cocker and colleagues’ paper is on explaining and justifying their new testing method, they also demonstrate some potential applications.
One thing they did was to dose their critters with various levels of alcohol, caffeine or amphetamine in order to see what effect this might have on their preferences for the high-effort/high-reward task versus the low-effort/low-reward task.
Without the drugs, rats tended to prefer the high-effort/high-reward option, and each rat tended to be consistent in its preference. Rats that strongly preferred the hard task on one day also tended to prefer the hard task on the next day.
With the drugs, if you looked at all of the rats together, things stayed pretty much the same. But if you split the rats into two groups based on how strong their preference for the hard task was, some interesting things appeared.
Under the influence of amphetamine, the rats with a strong preference for the hard task (“workers”) started edging more towards the easy task. But the rats that had previously not had a strong preference for the hard task (“slackers”) began behaving more like the “workers”.
With caffeine, there was a similar increase in the “slacker-ness” of the “workers”, but no increase in the “worker-ness” of the “slackers”.
This tells us a few things: these drugs can have different effects on different individuals, and the effects they have can be shaped by the underlying traits of the individuals involved.
This may sound obvious to anyone who’s seen the exact same quantity of beer transform one person into the life of the party and another into an obnoxious thug, but it’s still a fairly new thing to demonstrate in a rat.
As mentioned above, having ratty models of these features of the mind allows us to experiment in ways that would not be acceptable in humans: molecular-scale analysis of the mechanisms driving these behaviours, and early testing of cutting-edge treatments.
These hard working rats may be the first step on a long road towards a better understanding of a range of psychological problems.