Motivation
I am currently working on a project trying to enhance extinction learning via additional top-down control processes, i.e. suppression. This is a relevant article investigating whether episodic retrieval suppression of a memory prevents that memory from having an implicit effect on future cognition.
Introduction
Psychologists have long theorized that unwanted and suppressed memories may still affect cognition, and that volitional suppression is an ineffective way to deal with traumatic experience. Some recent research has suggested that this is not the case, and that suppressing retrieval of a memory may actually completely remove the memory’s effect on cognition. However, the current authors point out that the methods used to reach this claim are not fool-proof, and the memory tests used could have been contaminated by explicit retrieval. The authors present a novel behavioral paradigm designed to directly measure the implict effect of suppressed memories, without explicit contamination.
Method
The authors combine a think/no-think (TNT) paradigm, with a remote associates test (RAT) in order to measure suppression induced forgetting.
TNT overview
- Subjects first learn cue-target pairs of words (learning phase)
- Cues are presented, and subjects are instructed to either think on the target, or suppress (think/no-think phase)
- results in 2 types of target words:
- pairs that were only studied, no TNT (Baseline)
- pairs in which the target was suppressed (No-think)
RAT overview
- RAT consists of presenting three words (i.e. OPERA, HAND, DISH) with the goal of coming up with a solution word that fits all 3 cues (i.e. SOAP)
- Critically, the solutions words were either novel words, or words that were encountered during the TNT (Baseline or No-think)
The experiment was presented to subjects deliberatly as two distinct sub-experiments in order to reduce the chances of subs noticing TNT targets being used as RAT solutions. The dependent measure is then the percentage of RAT problems solved with each type of word (Novel, Baseline, No-think). If retrieval suppression truly removes implicit effects of a memory, then participants should be less able to access No-think RAT solution words compared to Baseline words that were studied, and the Novel words serve as a control.
Results - Experiment 1
They found that participants found significantly fewer RAT solutions from the No-think category compared to Baseline, and significantly more from Baseline as compared to No-think and Novel. This demonstrates:
- The enhancement TNT studying has on implicit priming (Baseline > Novel)
- The success of TNT suppression on removing implicit priming (Baseline > No-think)
- Suppression competely removes studying effect (No-think = Novel)
Experiment 2 motivation
Briefly, authors wanted to be sure that there was no way participants were using explicit retrieval strategies to recall No-think words. Experiment 2 consisted of two groups, an implicit group identical to Exp. 1, and an explicit group that was told that some RAT solutions were studied during the TNT, and to try and remember those words as solutions.
Results - Experiment 2
As expected, the implicit group replicated results from Experiment 1. Providing subjects with explicit instructions completely abolishes the suppression effect, there was no difference between Baseline and No-think words in this group.
There were further analyses in support of the masking hypothesis proposed by the authors, but they don’t really add to the impact of the presented results.
Discussion
This paper presents solid evidence that retrieval suppression removes a memory’s ability to implicitly influence cognition. Authors suggest that this is achieved through direct inhibition of both hippocampal and neocortical representations of a memory. They contrast suppression to avoidance, which may in fact leave implicit associations intact, since people learn to avoide illiciting the representation at all.
Impression
This is a great, elegant article that hopefully will lead to some advancements/tweaks in my own project