Information-integration category-learning task

The evidence is now overwhelming that humans have multiple learning systems that are functionally and anatomically distinct, that evolved at different times for different purposes, that learn in qualitatively different ways, and that are ideally suited to learn different kinds of things.

GRT is a multidimensional extension of signal detection theory, useful for studying whether multiple perceptual elements interact as we make decisions about them.

Repeatedly practicing a skill eventually causes it to be executed automatically. Although initial learning depends on elaborate neural networks, quicker and more direct neural pathways take over as automaticity develops.

Many events that cause sudden mood improvements, such as receiving an unexpected gift, also cause dopamine neurons in the brain to fire, which elevates cortical dopamine levels. Moderate increases in cortical dopamine improve performance in a variety of different cognitive tasks, and it is this effect on brain dopamine levels that seems to account for many of the cognitive benefits associated with positive mood.

Perhaps the most common cognitive laboratory task uses a two-alternative forced-choice design, in which case the accuracy data is just a record of whether the participant responded correctly or incorrectly on each trial. In such experiments, the participant’s response time provides an invaluable extra constraint that can be used to make inferences about the underlying cognitive processes recruited by the task.

Computational neuroscience is an old field, but it focuses almost exclusively on building models of single neurons. Therefore, new methods were required to model activity in networks complex enough to mediate behavior. This emerging new field is called computational cognitive neuroscience (CCN).

"Statistical Analysis of fMRI Data, Second Edition" written by Dr. Ashby - Order on Amazon