On the Fundamental Processes of Recognition Memory
| dc.contributor.advisor | Shiffrin, Richard M. | |
| dc.contributor.author | Mohamed, Zainab Rajab | |
| dc.date.accessioned | 2026-10-29T18:32:29Z | |
| dc.date.available | 2026-10-29T18:32:29Z | |
| dc.date.issued | 2026-10 | |
| dc.description | Thesis (Ph.D.) - Indiana University, Psychological and Brain Sciences, Cognitive Science, 2026 | |
| dc.description.abstract | This dissertation integrates two complementing projects: both aim at testing the ability of REM (Shiffrin & Steyvers, 1997), a process model for recognition memory, to generalize to different recognition memory paradigms. REM was developed to predict response patterns of single item recognition for words. In this dissertation, four experiments were conducted that varied stimulus types (words and pictures), list length (12 and 24), judgment types (oldness and newness), and test tasks, which included single item recognition, two- and three- alternative forced choice and old/new classification of multiple items. Part I tested the ability of the classic REM model to generalize and predict choice probabilities of all the conditions from the four studies. The model did a great job predicting the behavioral patterns in the data with minimal modifications to its original parameters and assumptions. We argue that the most critical assumption in REM is the basis for its recognition decision: REM has the form of a signal detection model and shares with attention likelihood theory (Glanzer & Adams, 1985) an assumption that likelihood ratios rather than gaussian strengths are the basis of the decision process. Part 2 extended REM to jointly predict response patterns and response times for the explored tasks building on the dynamic REM framework (Cox & Shiffrin, 2017), which assumes that features are sampled over time and contribute to the accumulated evidence gradually till the evidence reaches a decision threshold. It successfully predicted the accuracy and median response times in the data while holding many parameters constant across conditions. The success of REM and dynamic REM to account for the large number of explored conditions suggests that the processes they assume are sufficient to produce a good approximation to the way humans recognize, and could be of fundamental importance. These findings encourage further evaluations of the model and testing its generalizability across a broader range of memory paradigms. | |
| dc.identifier.uri | https://hdl.handle.net/2022/38624 | |
| dc.language.iso | en_US | |
| dc.publisher | [Bloomington, Ind.] : Indiana University | |
| dc.rights | This work is protected by copyright unless stated otherwise. | |
| dc.rights.uri | ||
| dc.subject | Cognitive psychology | |
| dc.subject | computational modeling | |
| dc.subject | Recognition memory | |
| dc.subject | cross-paradigm generalization | |
| dc.title | On the Fundamental Processes of Recognition Memory | |
| dc.type | Doctoral Dissertation |
Files
Original bundle
1 - 1 of 1
Loading...
- Name:
- Thesis-ZainabMoahmed-261006.pdf
- Size:
- 7.04 MB
- Format:
- Adobe Portable Document Format
Collections
Can’t use the file because of accessibility barriers? Contact us