UC Berkeley

My thesis research has focused broadly on how the environment shapes the structure and function of prefrontal cortex, for better or worse. I am interested in understanding in how experience-dependent plasticity can be harnessed to boost, or in some cases, remediate prefrontal function. I focused on training reasoning, the ability to solve novel problems, for two reasons: 1) reasoning is highly predictive of academic outcomes, and 2) reasoning was originally conceptualized as a fixed trait, and many people, both in science and in the general public, still believe that reasoning is set in stone. I conducted two studies, one in children and one in adults, to test the predictions that reasoning ability is malleable, and that repeated practice with reasoning problems strengthens connectivity within the frontoparietal network that supports high-level cognition.

In my first study, children from low socioeconomic status backgrounds participated in 8 weeks of training with computerized and non-computerized games that targeted either reasoning ability or cognitive speed. Training led to dissociable behavioral outcomes: children in the reasoning training group improved substantially on a matrix reasoning test, while children in the speed training group improved specifically on a measure of cognitive speed. Reasoning gains corresponded to roughly 10 points in performance IQ. These results indicated that reasoning is modifiable by training in this population.

In my second study, I capitalized on the fact that adults are taught reasoning skills by courses aimed at improving performance on the Law School Admissions Test (LSAT). I recruited young adults who were enrolled in an LSAT preparation course, and age- and IQ-matched controls intending to take the LSAT in the future. Magnetic resonance imaging (MRI) measures were collected for all subjects during two scanning sessions separated by 90 days. I first tested whether reasoning training altered functional connectivity at rest. I found strengthened fronto-parietal and parietal-striatal functional connectivity, particularly between hemispheres. I next investigated whether structural connectivity, as measured by diffusion in white matter, had been altered by training. I found training-related radial diffusivity (RD) decreases in white matter connecting frontal cortices, and mean diffusivity (MD) decreases closer to cortex in white matter in left frontal lobe and right parietal lobe. In conclusion, my thesis research has demonstrated that training can improve reasoning ability and, in adults, alter functional and structural connectivity. These results provide evidence for neural plasticity in large-scale networks supporting high-level cognition.