A game throws light on some people’s inability to connect bad decisions with bad consequences.
Why do people do stupid things? No great mystery.
Why do some people keep doing stupid things they know can harm them? Argh. Welcome to the central enigma that makes public health both a necessary discipline and an incredibly frustrating one.
A couple of pathways are known to direct maladaptive decision-making, say psychologists from UNSW and Western Sydney University, in their new study published in PNAS: “a motivational pathway characterised by value distortions that skew the expected utility functions governing action selection, and a behavioural pathway characterised by dominance of autonomous stimulus-response mechanisms”.
But these don’t explain why people who keep making bad choices fail to appreciate the relationship between them and their outcomes, the authors write.
They’ve previously established that when it comes to punishment learning, people are either sensitive, and learn from negative repercussions, or insensitive, having a wrong understanding of the causes of their punishment.
In this paper they propose a third, cognitive pathway, based on the results of experiments involving a game of intergalactic trade and, yes, finally, space pirates!
A group of 167 young people played a game of choosing between two planets for rewards. At varying intervals, clicking on one of the planets would prompt a pirate attack costing the subject 20% of their rewards. At one point this was explicitly revealed to the subjects, some of whom had already figured out to avoid the pirate-infested planet – this minority of subjects was labelled the “sensitive” phenotype.
The rest clustered into “unaware” and “compulsive” phenotypes; these, respectively, showed avoidance only after the reveal, and did not avoid the danerous planet even after being told.
Time delays between choice and punishment made no difference. However, changing the frequency of punishment did: when 40% of bad-planet clicks led to an attack, far fewer people fell into the “compulsive” cluster than when it was 10%.
Some people, the writers conclude, acquire accurate causal beliefs through experience alone; others learn when told and avoid the punishment-generating option; and others fail to learn and avoid even when the causal connection is made explicit to them. The latter, the “compulsive” phenotype, failed to incorporate true evidence to update their beliefs and were less likely to behave differently even if their beliefs were updated.
Subjects were interviewed on how they appraised the outcomes, to rule out the differences being down to some valuing rewards and punishments differently – all wanted loot, and no one liked getting raided by pirates.
“Moreover, participants engaged in effortful and deliberative cognitive strategies to earn reward and avoid punishment,” the authors write. “They formed declarative, internally coherent, mental models of how their actions caused reward and punishment, rather than acting autonomously or habitually relying on stimulus-response procedural knowledge … Whether other trait constructs, such as cognitive flexibility and intelligence, relate to sensitivity phenotypes will be important to determine.”
Study co-author Professor Gavan McNally, a UNSW behavioural neuroscientist, acknowledged that “real life is a lot more stochastic” than this lab experiment.
But the results still show something cognitive is going on other than pure reward valuation or purely compulsive behaviour beyond subjects’ control, he said.
“What we show is there is a cognitive pathway that emerges not from differences in value or awareness, but from failing to understand or appreciate correctly that their own actions are leading them to harm. Our ‘compulsives’ are indeed learning, it’s just that they learn the wrong thing.”
Finally, the importance of frequency of negative outcomes to learning is significant for real life: “The probability that any individual risky action such as speeding, social deception, or substance use will have detectable negative consequences is low. Our findings show that when the costs of actions are rare, learning via experience or information does not always yield veridical causal knowledge or optimum decision-making, even if those costs are severe.”
Sending penny@medicalrepublic.com.au a story tip is a good life choice.
