İşleyen bellek içeriğinin bilişsel kontrol

Working memory (WM) is a dynamic system that allows us to actively maintain recently encountered information so that it is available for ongoing cognitive operations. Information held in an accessible state in WM is used to guide many complex decisions and behavior. However, WM has capacity limitations. Efficient use of this capacity-limited system is the key to successful execution of complex tasks and requires control over the contents maintained in WM. There are several processes that control WM such as updating active content according to task goals, focusing attention on the relevant content, removing irrelevant information, and controlled retrieval of relevant information. In this dissertation, I focus on one of these processes: controlled retrieval of relevant information. The set of studies conducted in this dissertation use convergent methodologies to test possible factors that influence successful controlled retrieval. Chapters I and II introduce two behavioral studies employing the speed-accuracy-trade-off (SAT) procedure to independently measure the speed and accuracy of working memory retrieval. Chapter III introduces a study that employs functional magnetic resonance imaging (fMRI) to further investigate the findings from Chapter II and complement the behavioral measures with neural measures. Chapter I examines how individual differences in working memory capacity (WMC) is an important indicator of the efficiency of controlled retrieval from working memory. Individuals who fell in the upper and lower quartile on a WMC measurement task were to retrieve temporal order information of the recently presented items, which required a controlled serial search through the contents of WM. The prediction was that those who rank higher on WMC measures would perform well at this search, which would result in both faster and more accurate retrieval when compared to those with lower WMC. Findings from this chapter revealed that WMC was indeed successful at predicting retrieval efficiency such that individuals who fell in upper and lower quartile in the WMC measurement task differed in both speed and accuracy while retrieving temporal order information from WM. In Chapters II and III, I considered another possible factor that might affect retrieval efficiency: affective content of to be remembered material. I used two different tasks in which controlled retrieval demands were increased by manipulating the presence of proactive interference (PI). Participants needed to retrieve additional diagnostic information about the test item either at the item level or context level to correctly recognize it in the presence of interference. I evaluated the differences between neutral and emotional materials, foreseeing possible differences in the process in which PI builds up and the process in which it is resolved via controlled retrieval. There was less PI build up for emotional materials in both tasks, but controlled retrieval was less efficient if resolving PI required context retrieval; on the other hand, if participants could recover relevant information by retrieving details about the item, there was no such differences in controlled retrieval across emotional and neutral information. Together, these Chapters demonstrate that the efficiency of controlled retrieval depends on (1) an individual's ability to maintain relevant materials in an accessible state while concurrently executing a task (i.e., greater WMC) and (2) whether the encountered information is emotional or neutral. Studies presented here add up to a growing literature of studies investigating how we can utilize use of working memory which determines our intellectual abilities.