Date of Award

Summer 8-15-2018

Author's School

Graduate School of Arts and Sciences

Author's Department


Degree Name

Doctor of Philosophy (PhD)

Degree Type



The parietal memory network (PMN) is a functional brain network that has been recently described through the convergence of task-based fMRI and resting-state functional MRI studies (Gilmore et al., 2015). The networkճ characteristic encoding/retrieval flip (deactivation at encoding and activation at later retrieval, discussed by Gilmore and colleagues) and its manifestation of a negative subsequent memory effect (greater deactivation at encoding for items that will later be recognizedѩ.e., subsequent hitsѴhan for those that will notѳubsequent misses) (Cabeza et al., 2004; Daselaar, Prince, & Cabeza, 2004; De Chastelaine & Rugg, 2014; Elman, Rosner, Cohn-Sheehy, Cerreta, & Shimamura, 2013; Kim, 2011; Otten & Rugg, 2001) and other memory-related contrasts suggest possible ties to effective encoding of memory. This study sought to use task-based functional MRI to further investigate this tie between the PMN and encoding through the use of the levels of processing paradigm (Craik & Lockhart, 1972; Craik & Tulving, 1975). Specifically, I hypothesized that the PMNѷhich demonstrates the negative subsequent memory effectѷould deactivate more for conditions that lead to more effective encoding such as a deeper level of processing.

Partial support for greater deactivation in the PMN for deeper processing was observed in two of the three regions that form the network (specifically within precuneus and mid cingulate): Precuneus (and to a lesser extent) mid cingulate deactivated for deeper, semantic processing than for the more shallow orthographic processing. However, the two regions did not show differential activity between semantic and phonological processing (presumably shallower than semantic processing), despite behavioral differences; The third region of the network, PIPL, did not show a consistent levels of processing effect in the univariate analyses, nor did it show the negative subsequent memory effect despite all other PMN regions showing the effect. Linear-mixed effect modeling of PMN regions showed that trial-by-trial variation in levels of BOLD activity in precuneus and mid cingulate predicted subsequent memory, above and beyond the level of processing manipulation as well as contribution from left inferior frontal gyrus (IFG), a region consistently identified in subsequent memory studies (Kim, 2011). Attempts to use multivariate pattern analysis to classify subsequent memory using only individual PMN regions led to above chance classification (hit or miss) for all PMN regions. The ability to predict subsequent memory using only activity from PMN regions (and beyond the contribution of left IFG) supports the role of the PMN in encoding, showing that a sufficient level of deactivation in PMN regions is associated with successful encoding regardless of the level of processing. Overall, the study supported the conclusion that two members of the PMNѴhe precuneus and mid cingulateѣontribute to effective encoding of memory.


English (en)

Chair and Committee

Kathleen B. McDermott

Committee Members

Todd S. Braver, Ian G. Dobbins, David A. Balota, Nico U. Dosenbach,


Permanent URL: https://doi.org/10.7936/e75a-4w81