Dual-process theory and signal-detection theory of recognition memory

Wixted, J. T. (2007). Dual-process theory and signal-detection theory of recognition memory. Psychological Review, 114(1), 152-176.

Abstract:
Two influential models of recognition memory, the unequal-variance signal-detection model and a dual-process threshold/detection model, accurately describe the receiver operating characteristic, but only the latter model can provide estimates of recollection and familiarity. Such estimates often accord with those provided by the remember-know procedure, and both methods are now widely used in the neuroscience literature to identify the brain correlates of recollection and familiarity. However, in recent years, a substantial literature has accumulated directly contrasting the signal-detection model against the threshold/detection model, and that literature is almost unanimous in its endorsement of signal-detection theory. A dual-process version of signal-detection theory implies that individual recognition decisions are not process pure, and it suggests new ways to investigate the brain correlates of recognition memory.

Note:
Use the ROC to distinguish different theory on recognition memory.

This paper reviewed two different views on recognition memory.

The first one is dual process theory of recognition memory.
Dual process means familiarity and recollection process in recognition memory.
“Recollection is a relatively slow process that consists of retrieving specific details associated with the prior presentation of an item, whereas familiarity is a relatively fast process that allows one to appreciate the fact that the item was previously encountered even though no contextual detail can be retrieved.”

It is corresponding to unequal variance signal detection model (the author suggested better).
The UVSD has curvilinear ROC and linear zROC (the slope of z ROC means the std (lure) / std (target), when the slope is 1, it shows equal variance among tow distributions. Usually, the slope is 0.8, it means the std of target is 1.25 times std of lure, showing unequal variance.)

The author‘s new model concludes that both familiarity and recollection occurs in an one-dimensional signal detection decision process (dual process UVSD).

The second is the signal detection theory.
à The recognition decision is based on the memory strength in relation to the decision criterion.

1. two criterion model
“if the familiarity of a test item falls above a high criterion value or below a low criterion value, then a fast, familiarity-based decision is made (old or new, respectively). If the value instead falls between the two criteria, then a search process is initiated, which, if successful, leads to a slower, recollection-based old decision. Thus, in this model, the subject is thought to resort to recollection as a backup process whenever familiarity fails to provide a clear answer.”

2. high threshold model
“Recollection is a high-threshold process (i.e., recollection either occurs or does not occur), whereas familiarity is a continuous variable that is governed by an equal-variance detection model.”
“If a target item on a recognition test occasions recollection, then a high-confidence old decision is made. No decision criterion or any other consideration based on signal-detection theory is needed to characterize a recollection process like that. If recollection fails, however, then a familiarity-based decision is made.”
“In this model, the participant is thought to resort to familiarity as a backup process whenever recollection fails to occur.”

The second model (high threshold model” is widely used.

It is corresponding to the dual process signal detection model.
The DPSD (high threshold model) has linear ROC and curvilinear zROC.

Data from part recollection is against the DPSD model, because DPSD suggests that recollection is an all-or-none process.