Mood and Memory

Do you ever wonder if your mood affects the way you think? I'm not talking about behaving more aggressively when you're angry or more passively when you're sad; I'm talking about the subtler impact on cognitive processing. Some recent research has indicated that we process things differently depending on whether we're in a positive or negative mood. People in good moods tend to make more connections between related items, while people in bad moods generally focus on what's in front of them.

Justin Storbeck and Gerald L. Clore realized that there may be a connection between this research on emotion and other research on false memories (presumably they were in a good mood when they made the connection!). Memory research (as we've discussed here) makes a distinction between item-specific and relational processing in memories, each of which are activated in different circumstances. Isn't it possible, then, that we might remember different things depending on whether we're in a good or bad mood?

One method of studying false memories is to present people with a list of items, all related to a critical word, which is not included in the list. For example, the list might include boat, wind, mast, yaw, boom, and water, but not the critical word sail. Then participants are asked to write down as many of the items as possible which appeared on the list. People are often more likely to list the critical word (the one not on the list) than any of the words they actually saw.

Storbeck and Clore wondered: if we're more likely to see relationships between items when we're in a good mood, then are we more likely to demonstrate false memories, which seem to derive from those same relationsips? They used music to induce positive or negative moods in people, then gave them the false memory test. In a second experiment, they expanded on the basic false memory paradigm by asking participants to not only list words they had seen, but any related words they could think of (they were instructed to place a check mark next to words they hadn't actually seen). Here are the results:

i-1effd420297010470cbd93465e288684-falsemood.gif

The results were as predicted: people in good moods given the basic test falsely recalled the critical words more often than those in bad moods. When they were allowed to write down related words, people in negative moods still showed a lower incidence of false memory. They also were less likely to write down the critical word as a related word—it indeed appears that they made fewer connections between the related words.

Storbeck and Clore also found that there was no difference between the positive and negative mood groups in true recall—they remembered the words actually appearing on the list with equal accuracy. They argue that this result supports the fuzzy-trace theory of memory, which says that we remember things in two ways—verbatim and gist memory. Normally, we form both types of memories for everything, but negative moods appear to inhibit gist memory, so only verbatim memories are formed.

Storbeck, J., & Clore, G.L. (2005). With sadness comes accuracy, with happiness, false memory. Psychological Science 16(10), 785-791.

More like this

I tend to find that I have clearer memories (thinking back on them) of times when I have been in the extremes of emotion.

I also tend to have an easier time recalling information clearly when i am really angry.

Another thing I would wonder about would be the affect of using emotionally charged words.

I'd also wonder about the results which would be gained from presenting words to a participant without the participant expecting a memory test.

Additional, it would be nice to be able to set up an experiment that does not involve words memory.

Rather than merely measuring inclusion errors in recall tasks, wouldn't it be better to test other memory tasks as well while using different performance measures?

How about a recognition task? The test would present word pairs from the presentation. The word pairs would be varied. Some would be actual targets, some would not be actual targets presented during the presentation. Some of the non-targets would be semantically related to the targets presented during the presentation.

Then, as a part of the analysis, the measure would be d', the overall performance. (rise in hits, sink in misses? rise in hits and misses? etc.)

There are other factors which could be involved which have to be accounted for, such as frequency characteristics of the words chosen for the task, and all manner of things.

This, of course, would not be testing an exact "false memory" claim; but it would be testing a related, more general, claim.

MS Office and open source office productivity apps

In the desktop Linux survey published last week we saw that email and office productivity suites were the applications critical to Linux desktop deployments. Now that I have had a few days to go back and think it through, I...