Jeroen G. W. Raaijmakers, working at the University of Amsterdam, is best known for his collaboration with Richard Shiffrin on the SAM model of memory retrieval. SAM provided a unified mathematical framework for understanding free recall, cued recall, and recognition, and it remains one of the most influential formal models in memory research. Raaijmakers has also made important contributions to the understanding of retrieval-induced forgetting and inhibitory processes in memory.
The SAM Model
Recovery probability: P(recall i|sampled) = 1 - exp(-S(Q,i))
Total strength: S(Q,i) = Product over cues W_k(Q_k, i)
Associative increments during study:
delta a = context-to-item, delta b = item-to-item
The SAM model's two-stage architecture distinguishes between the probability of sampling an item (relative strength) and the probability of successfully recovering it (absolute strength). During study, associative connections are formed between items and context cues, and between co-occurring items. During retrieval, the current context and any recalled items serve as cues that activate stored traces. This framework explains why recall is harder than recognition (recall requires both sampling and recovery, while recognition requires only a familiarity assessment), why earlier recalled items interfere with later retrieval (output interference), and why partial cues can impair rather than help recall (part-list cuing inhibition).
Raaijmakers and Jakab (2013) provided a SAM-based account of retrieval-induced forgetting -- the finding that practicing retrieval of some items causes forgetting of related unpracticed items. Their model explains this through competition at retrieval: practiced items gain associative strength, increasing their sampling probability and reducing the probability of sampling competitors, without requiring an inhibitory mechanism.
Extensions and Applications
Raaijmakers extended the SAM framework to account for directed forgetting, the spacing effect, and the effects of list composition on recall. His work on the relationship between recall and recognition within a single formal framework demonstrated the power of mathematical modeling to unify seemingly disparate memory phenomena under common theoretical principles.
Legacy and Impact
The SAM model established the sampling-and-recovery framework that has been adopted and adapted by numerous subsequent memory models. Its influence extends to models of recognition memory (REM), episodic memory retrieval (Temporal Context Model), and applied settings including eyewitness memory and educational practice. Raaijmakers' contributions to computational modeling methodology, including his work on model fitting and parameter estimation, have strengthened the quantitative foundations of memory research.