Recollection can be weak and familiarity can be strong

Abstract

The remember-know procedure is widely used to investigate recollection and familiarity in recognition memory, but almost all of the results obtained with that procedure can be readily accommodated by a unidimensional model based on signal-detection theory. The unidimensional model holds that remember judgments reflect strong memories (associated with high confidence, high accuracy, and fast reaction times), whereas know judgments reflect weaker memories (associated with lower confidence, lower accuracy, and slower reaction times). Although this is invariably true on average, a new 2-dimensional account (the continuous dual-process model) suggests that remember judgments made with low confidence should be associated with lower old-new accuracy but higher source accuracy than know judgments made with high confidence. We tested this prediction--and found evidence to support it--using a modified remember-know procedure in which participants were first asked to indicate a degree of recollection-based or familiarity-based confidence for each word presented on a recognition test and were then asked to recollect the color (red or blue) and screen location (top or bottom) associated with the word at study. For familiarity-based decisions, old-new accuracy increased with old-new confidence, but source accuracy did not (suggesting that stronger old-new memory was supported by higher degrees of familiarity). For recollection-based decisions, both old-new accuracy and source accuracy increased with old-new confidence (suggesting that stronger old-new memory was supported by higher degrees of recollection). These findings suggest that recollection and familiarity are continuous processes and that participants can indicate which process mainly contributed to their recognition decisions.

Figure 1

The signal-detection interpretation of Remember/Know/Guess judgments. Recognition decisions are based on a continuous memory strength scale, and a Remember judgment is made when the memory strength of a target or a lure exceeds a high criterion. A Know judgment is made when memory strength only exceeds a lower criterion, and a Guess judgment is made when memory strength only exceeds a still lower criterion.

Figure 2

The Continuous Dual Process model of Remember/Know/Guess judgments. To make an Old/New decision, continuous recollection and familiarity signals are aggregated to form a continuous memory strength variable. The old/new confidence rating that is made depends on the highest criterion that is exceeded by the memory strength of the test item (the criteria for a 6-point confidence scale are shown). Thus, for an old/new decision, the CDP model is equivalent to the standard unequal-variance signal-detection model. To make a Remember/Know/Guess judgment, the participant first queries memory for evidence of recollection and makes a Remember judgment if the recollection signal exceeds the Remember decision criterion. If not, the participant next queries memory for evidence of familiarity and makes a Know judgment if the recollection signal exceeds the Know decision criterion (else, a Guess judgment is made).

Figure 3

20-point rating scale used in Experiments 1 and 3. A rating of 1 corresponds to the highest level of confidence that the word is new (not on the study list). A rating of 20 corresponds to the highest level of confidence that the presented word was on the study list (old), and is further broken down by a simultaneous Remember (R) versus Familiar (F) judgment. Ratings of 16–19 correspond to varying levels of confidence and also require an R/F judgment. With regard to the CDP model shown in Figure 2, a rating of 20R would be made if aggregated memory strength exceeded the highest confidence criterion on the memory strength axis and if the recollection signal exceeded the Remember criterion on the recollection axis.

Figure 4

Proportion correct for old/new and source accuracy in Experiment 1 for R_16-19 and F₂₀ responses (A), R_16-19 and R₂₀ responses (B), and F_16-19 and F₂₀ responses (C).

Figure 5

State-trace plot for Experiment 1 showing the mean old/new and source accuracies for F_16-19, F₂₀, R_16-19, and R₂₀.

Figure 6

Rating scale used in Experiment 2. A rating of 6 corresponds to the highest level of confidence that the word was on the study list (old) and is further broken down by a Remember/Familiar judgment. Ratings of 4 and 5 reflect lower levels of confidence and also require an R/F judgment. A rating of 1 reflects the highest level of confidence that the word was not previously studied (new).

Figure 7

Proportion correct for old/new and source accuracy in Experiment 2 for R₅ and F₆ responses (A), R₅ and R₆ responses (B), and F₅ and F₆ responses (C).

Figure 8

State-trace plot for Experiment 2 showing the mean old/new and source accuracies for F₅, F₆, R₅, and R₆.

Figure 9

Proportion correct for old/new and source accuracy in Experiment 3 for R_16-19 and F₂₀ responses (A), and state-trace plot showing the mean old/new and source accuracies for F_16-19, F₂₀, R_16-19, and R₂₀.

Figure 10

Two-dimensional decision space for recognition and source memory (with likelihood ratio decision bounds) proposed by Hautus et al. (2008). The numbers represent source confidence ratings using a 6-point scale (6 = “sure source A”; 1 = “sure source B”).