Cited 0 time in webofscience Cited 0 time in scopus

Role of the executive functions in statistical learning

Role of the executive functions in statistical learning
Park, JungtakJanascek, KarolinaNemeth, DezsoJeon, Hyeon-Ae
DGIST Authors
Jeon, Hyeon-Ae
Issue Date
Neuroscience 2018
Statistical learning (SL) is a fundamental neurocognitive mechanism that enables us to extract complex probabilistic regularities embedded in the environment. However, individual differences in SL have not been comprehensively characterized yet. We hypothesized that SL would be influenced differently by individuals’ executive function (EF) and working memory (WM) capacity. The study included two sessions on two separate days. In Session 1, participants performed various neuropsychological tests known to be involved in EF and WM. In Session 2, participants performed the Alternating Serial Reaction Time (ASRT) task that has previously shown to capture statistical learning of probabilistic associations. Participants were asked to press one out of the four buttons which corresponded to the target presented on the screen. The ASRT task had three different trial types: pattern trials with high probability condition, random trials with high probability condition (random-high), and random trials with low probability condition (random-low). The difference between random-high and random-low trials was defined as ‘SL score.’ As random-high and random-low trials were separated solely by probability, we could examine pure SL effects. The learning task was divided into three periods to investigate the dynamic changes associated with learning. Also, we defined a ratio of bias as the number of incorrect responses when participants pressed the buttons of high-probability target instead of low-probability target. By calculating the bias, we could find whether participants achieved SL or not following the assumption that participants would expect an occurrence of a high-probability target more than a low-probability target over the course of SL. According to our results, participants showed significant SL both in terms of reaction times and accuracy. Regarding the bias, it increased significantly in the second period against the first period, suggesting increased expectation of high-probability targets as learning progressed. Similarly to previous studies, higher SL score seemed to be associated with weaker performance on some WM and EF measures, although these correlations did not reach significance. Interestingly, however, we found that higher SL score was significantly related to better performance (smaller conflict scores) on Stroop task [r = -0.403, p < 0.05], suggesting that inhibitory control may have a different role in SL compared to other aspects of WM and EF. Our findings can contribute to a better understanding of the neurocognitive underpinnings of SL.
Society for Neuroscience
Related Researcher
  • Author Jeon, Hyeon-Ae Laboratory of Cognitive Neuroscience
  • Research Interests fMRI, high-level cognition, brain imaging
There are no files associated with this item.
Department of Brain SciencesLaboratory of Cognitive Neuroscience2. Conference Papers

qrcode mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.