[ad_1]
Rewards are frequently used to promote learning, but rewards can actually hide real knowledge, says a new Johns Hopkins University study with rodents and ferrets.
The results, published on May 14 in Nature Communications, distinguish between knowledge and performance, and explain how the environment can affect both.
"Most learning research focuses on how humans and other animals learn" content "or knowledge, and here we suggest that there are two parallel learning processes: one for the content and the other for the context or environment.If we can separate the operation of these two pathways, maybe we can find ways to improve performance, "says Kishore Kuchibhotla , Assistant Professor in the Department of Psychological and Brain Sciences at Johns Hopkins University, and lead author of the study.
Researchers know that the presence of reinforcement, or reward, can change the behavior of animals, but it is unclear exactly how rewards affect performance-related learning.
Kuchibhotla explains that an example of the difference between learning and performance is the difference between a student student and knowing the answers at home, and a student demonstrating that knowledge during a test at the school.
"What we know at one time may be different from what we show, and what interests us is to be able to access that knowledge in the right environment," he says.
To study what animals know in the hope of better understanding the learning, Kuchibhotla and the research team trained mice, rats and ferrets to a series of tasks and measured accuracy with which they performed these tasks with or without reward.
For the first experiment, the team trained mice to lick water with the help of a lick tube after hearing a tone and not licking after hearing a another tone, without reward. It takes two weeks for the mice to learn this in the presence of the water reward. At an early time of learning, around days 3 to 5, the mice performed the task at random (about 50%) when the licking tube / reward was present. When the team completely removed the licking tube during these early days, the mice performed the task with an accuracy of over 90%. The mice therefore seemed to understand the task several days before expressing their knowledge in the presence of a reward.
To confirm this conclusion with other tasks and animals, the team also asked mice to press a lever to get water when they heard a certain tone; induced the rats to look for food in a cup if they heard a tone, but not if a light appeared before the tone; the rats pressed a lever for sweet water when a light was presented before a tone; Rats pushed the lever to get sweet water when they heard a certain tone and urged ferrets to differentiate two different sounds to get water. In all the experiments, the animals had better results when the rewards were not available.
"The rewards, it seems, help to improve learning gradually, but can mask the knowledge that animals have actually gained, especially at the beginning of learning," Kuchibhotla said. In addition, the finding that the performance of all animals improved without reward, suggests that variability in learning rates may be due to differences in animal sensitivity to reward context rather than to differences in intelligence. .
The researchers suggest that the dissociation between learning and performance might one day help us isolate the root causes of poor performance. Although the study only concerns rodents and ferrets, Mr. Kuchibhotla believes that it might be possible to someday help animals, as well as humans, to gain better access to the content when they need it, if the proper mechanisms in the brain can be identified and manipulated.
For humans, this could help people with Alzheimer's disease to maintain lucidity for longer periods and to improve the test environment of schoolchildren.
Funding for this study was provided by the National Institute of Deafness and Other Communication Disorders (DC009635, DC012557, DC05014), a career scholarship Hirsch / Weill-Caulier, a scholarship of the faculty of the Howard Hughes Medical Institute, Emergences of City of Paris, National Agency. of research (ANR-17-ERC2-0005, ANR-16-CE37-0016), program "Investments of the future" (ANR-10-LABX-0087 IEC, ANR-11-IDEX-0001-02) , PSL Research University and the National Institutes of Health's Neuroscience Computing Training Program (R90DA043849).
[ad_2]
Source link