Think of any movie scene.

Was the scene that came to mind a pleasant one? Something funny or romantic? Or more harrowing?

When I did this exercise, a scene from the horror film, The Ring, came to mind before a scene from the comedy, Ferris Bueller’s Day Off. That is, for me, a negative scene came to mind over a positive scene. If you also conjured up a negative scene, then your experience aligns with a common research finding that very unpleasant experiences are remembered better and with greater detail than very pleasant ones.

Behavioral and neuroimaging studies have shown that we tend to invoke more visual imagery when encoding negative information than when encoding positive information. The extent to which we invoke visual imagery when we try to retrieve details, not just when we are encoding them, was the topic under investigation in a recent article that Chris Brewin and Kirsty Langley published in the Psychonomic Society’s journal Memory and Cognition.

The experiment was in part motivated by the fact that many people with Posttraumatic Stress Disorder (PTSD) have extremely vivid flashbacks of the traumatic event. The observed dichotomy that people with PTSD have heightened perceptual memory but reduced episodic memory may account for these very vivid flashbacks. This, in turn, may mean that imagery is involved during retrieval.

To test whether imagery during retrieval also contributes to the better memory for negative information, Brewin and Langley conducted a memory experiment. In the experiment, participants’ abilities to recall pleasant versus unpleasant scenes in a cued recall task was tested. The cues were either visual or verbal. Participants also completed memory diaries so the imagery that they reported during the cued recall test could be compared to the imagery that may have occurred one week after the test.

Brewin and Langley tested the following main predictions:

1a. Negative images would elicit more vivid imagery than positive ones.

1b. Visual cues would elicit more vivid imagery than verbal ones.

2a. Negative images would be remembered better than positive ones.

2b. Visual cues would aid memory more than verbal ones.

3. Negative images and imagery associated with a visual cue would come to mind more than positive images and imagery associated with a verbal cue during the week after the test.

The figure below shows the procedure. Participants studied a list of images that were very pleasant (positive) or very unpleasant (negative), took part in a brief distractor task, and were tested for the images. Participants were sent away with a diary, in which they were to record whenever memories of the images popped into mind over the next 7 days.

Consider the following example of one trial: the participants viewed a scene of happy women and parrots during the study phase. (A pleasant stimulus). After a brief distractor task, a cue was presented during the cued recall phase. In this example, a picture of a red parrot (a visual cue) or the words “a red parrot” (a verbal cue) was shown. During the test, participants

• described from memory, to the best of their abilities, the original scene that contained the red parrot, and
• said whether the cue brought an image to mind and rated the vividness of that image on a scale from 1 (no imagery) to 4 (“an image as clear as day came to mind”).

If the women and parrots came to mind over the next 7 days, then participants would additionally record that in the diary.

As shown in the figure below on the left, negative images were rated as more vivid than positive ones, but visual cues were not rated as more vivid than verbal ones. As shown in the figure on the right, visual cues generated more correct images. These results are partially in line with the first prediction.

In line with the second prediction, scenes were more accurately recalled if 1) they were negative in valence, 2) they had a visual cue, and 3) imagery occurred.

As shown in the figure below, and in line with the third prediction, the data from the memory diary revealed that more memories of the negative images (across the 7 days) were reported when they were cued with a visual image during the cued recall test phase.

To summarise, participants had better recall for negative scenes, which replicated prior studies. Moreover, imagery contributed to accurate recall―at retrieval―not just at encoding as has been shown previously. Interestingly, imagery contributed to retrieval of negative memories more strongly than positive ones. Brewin and Langley concluded that

“…the findings suggest that the established recall advantage for negative over positive scenes has to do with the fact that they are accompanied in the short term by more vivid imagery, and in the longer term by more persistent imagery.”

Brewin and Langley proposed that patterns of results that arise from fMRI experiments in which participants are scanned during memory recall may reflect imagery, not memory accuracy per se. This idea challenges some past conclusions made about neuroimaging data.

The authors also acknowledge implications for the possible treatment of PTSD. For example, perhaps if steps were taken to reduce imagery of the traumatizing event, it would help reduce vivid and troublesome flashbacks that are common in people with the disorder.

Psychonomics article focused on in this post:

Brewin, C. R., & Langley, K. M. R. (2018). Imagery retrieval may explain why recall of negative scenes contains more accurate detail. Memory & Cognition, DOI: 10.3758/s13421-018-0876-7.