Familiarity, recollection, and receiver-operating characteristic (ROC) curves in recognition memory

Abstract

The Atkinson-Shiffrin theory describes and explains some of the processes involved in storing and retrieving information in human memory. Here we examine predictions of related models for search and decision processes in recognizing information in long-term memory. In some models, recognition is presumably based on a test item's familiarity judgment, and subsequent decisions follow from the sensitivity and decision parameters of signal detection theory. Other models dispense with the continuous notion of familiarity and base recognition on discrete internal states such as relative certainty that an item has or has not been previously studied, with an intermediate state of uncertainty that produces guesses. Still others are hybrid models with two criteria located along a familiarity continuum defining areas for rapid decisions based on high or low familiarities. For intermediate familiarity values, the decision can be delayed pending the results of search for, and occasional recollection of, relevant episodic information. Here we present the results from a study of human recognition memory for lists of words using both response time and error data to construct receiver-operating characteristic (ROC) curves derived from three standard methods based on the same data set. Models are evaluated against, and parameters estimated from, group as well as individual subjects' behavior. We report substantially different ROC curves when they are based on variations in target-word frequency, confidence judgments, and response latencies. The results indicate that individual versus group data must be used with caution in determining the appropriate theoretical interpretation of recognition memory performance.

High- and low-threshold models of the relationship between response time and confidence.

“…In contrast, the 2LT model was able to account for all major patterns in the data, including the fact that RTs tended to be faster at higher confidence levels for both correct and error responses. Researchers have used RT data to test discrete-state models of recognition-memory in various ways, such as testing for RT equivalencies within an evidence state (e.g., Province & Rouder, 2012) or using RT data to construct ROCs (Juola, Caballero-Sanz, Muñoz-García, Botella, & Suero, 2019; Weidemann & Kahana, 2016). The current study took a different approach by testing whether discrete-state models could simultaneously accommodate confidence ratings and full RT distributions from each level of the confidence scale, and the results show that only the 2LT model succeeded at this challenging task.…”

Neural Mechanisms of Familiarity

“…Global-matching and fluency-attribution models of familiarity suggest that variable levels of memory strength can be experienced for an encountered stimulus, and that familiaritybased recognition decisions are well described by Signal Detection Theory (SDT; see e.g., Gillund & Shiffrin, 1984;Juola et al, 2019;Macmillan & Creelman, 2005;McClelland & Chappell, 1998). Using SDT, familiarity is represented along a continuum of memory strength, with a typical old/new recognition memory task being defined by two normal distributions of memory strengths --one for studied targets (old) and one for unstudied foils (new).…”

“…Although, the two models (SDT and threshold) make slightly different assumptions, they both explain the empirical recognition memory data (Bröder et al, 2013). While the debate remains unresolved, as the current evidence regarding it is mixed (Juola et al, 2019;Malejka & Bröder, 2019), SDT still remains the predominant modelling approach of choice for familiarity.…”

Disentangling Different Aspects of Between-Item Similarity Unveils Evidence Against the Ensemble Model of Lineup Memory

“…However, analyses of data aggregated across participants neglect possible variability in parameters between participants and are thus susceptible to aggregation biases (Morey et al, 2008;Juola et al, 2019;Smith et al, 2017). We therefore additionally conducted a Bayesian hierarchical model analysis (Rouder & Lu, 2005;Rouder et al, 2017), in which the parameter vector for each participant is separately drawn from a population distribution.…”