Research

Current Interests

Directed Forgetting

Often we think of forgetting as an unfortunate side effect of an imperfect memory system. It is frustrating to forget the name of a new colleague, where the car is parked, or the reason you walked into the kitchen. Occasionally there are times we want to forget, though. For example, you might want to forget your best friend's old phone number so it doesn't interfere with the new one. Or, you may want to forget that embarrassing slip-up you made during a recent presentation. How is forgetting achieved in these sorts of situations? Can we ever really forget?

Studies of intentional forgetting suggest that we can forget (for review see MacLeod, 1998). In these experiments words (or occasionally pictures) are presented for encoding and participants are told to remember half of those words, while forgetting the other half. At testing, the participants are asked to recall (or recognize) ALL the words they can remember, even those that they were told to forget. Memory for "remember" items is consistently better than memory for "forget" items, suggesting that forgetting does occur. This effect has been observed across a variety of encoding situations, testing conditions (including implicit and explicit tests), and is not due to demand characteristics (MacLeod, 1999).

The evidence suggests we can forget, but the question of how we forget is still unanswered. Currently there are several prominent theories of directed forgetting including elaborative rehearsal, attentional inhibition, and tagging and selective search. Each theory has received evidential support, as well as some criticisms. Furthermore, it appears that forgetting may be achieved through different means in different situations(Basden, Basden & Gargano, 1993).

As part of my dissertation research, I am examining the mechanisms that underlie directed forgetting by directly comparing several theories, introducing methodological variations and extending the study to novel stimuli. The data thus far are encouraging, and I plan to continue this line of research into the future.

Saccade Induced Retrieval Enhancement (SIRE)

It has recently been discovered that moving your eyes prior to a memory test enhances retrieval (Parker & Dagnall, 2007). Furthermore, this enhancement especially benefits people that are strongly right handed (Lyle, Logan, & Roediger, 2008). There are several theories as to why controlled eye-movements might facilitate memory retrieval, and I have recently embarked on a collaboration with Anthony Barnhart and Keith Lyle to test one of those theories. We have developed a series of experiments (some that span my other interests), and data collection is starting soon!

Previous Projects

Chair Buzz

Can a sense of familiarity be induced through a subliminal physiological stimulus?

In a classic study by Schacter and Singer (1962), participants were given "suproxin" - a drug designed to enhance vision. Actually participants received epinephrine. Some participants were told the drug had some side effects (including increased heart rate), while others were not. After receiving the drug, participants went into a waiting room where they were met by a confederate who acted either happy or angry. The confederate's emotional state influenced the state of the participants, but if the participants knew of the side effects they attributed their change in mood to the drug. Those that were not told of side effects experienced the change as an emotion, and rated themselves happy or angry afterwards. Thus, when the source of the physiological symptoms was unknown, participants attributed it to an emotional experience. We wondered, could a physiological stimulus be used to influence memory?

To answer this question, we had to do some building! Dr. Goldinger (with some help) took an ordinary chair and attached speakers on the bottom. The speakers emitted a low-amplitude 60- Hz triangle wave. (We were able to determine a frequency for the sound wave that would not be noticeable to most participants through pilot testing.) The wireless speakers were controlled by the computer, so that the onset of the stimulus and sound wave could be controlled exactly.

In the experiment we presented participants with lists of words, pictures and faces (blocked) for encoding. Some of the stimuli were easy to remember (i.e. pictures of famous people), and some were more difficult (pictures of medical students). At testing, participants were shown old and new stimuli. They were asked to decide if the object had been shown before, and to rate their confidence in their decision. On half of the testing trials, the speaker would emit a subliminal buzz during the stimulus presentation. On the other half of trials, no buzz was presented. The question was: will the presence of the buzz influence recognition judgments?

Well, the quick and dirty answer is: Yes! Now for the more nuanced answer: The buzz did not always affect judgments. The buzz increased the likelihood of saying "old" both correctly and incorrectly, but only for the hard stimuli. Thus, when the stimuli were more difficult to remember, people interpreted the subliminal cue as a general feeling of familiarity. The presence of the buzz also affected confidence ratings. When people made the wrong decision (false alarm) they were relatively confident. Conversely, when they made the correct decision (hits), participants were less confident. It turns out that internal beliefs about one's own memory caused the buzz to be interpreted in different ways. When participants saw a new (unfamiliar) word and received the buzz they attributed the physiological sensation to familiarity. When participants saw an old (familiar) word and received the buzz they interpreted the same sensation as doubt. Thus, confidence ratings were high in one situation, and lower in the other.

To answer the original question - it appears that memory can be influenced by a physiological stimulus, through attributions we make about what memory "feels" like!

(PDF) This paper was discussed on the psychology blog Cognitive Daily

Taboo

How do individual differences in working memory capacity affect performance in a naturalistic, group activity?

Individual differences in working memory capacity (WMC) are correlated with performance in other cognitive tasks including reading comprehension, vocabulary learning, and following directions. It is also correlated with the ability to suppress natural responses. This led us to think about real-world situations where it is important to suppress information, or a behavior, or even a thought. For example, imagine you are invited to a surprise party for a friend. The day before the event you run into your friend, and immediately think about the party! Of course, you must be careful not to say anything that will tip him off. Even something as small as "see you tomorrow" could ruin the surprise! Your job is to avoid the natural tendency to talk about the party, even though it is all you can think about. We thought that these types of natural interactions, where important information needs to be suppressed, should be sensitive to differences in WMC.

To test this theory, we chose to use the game Taboo®. In Taboo you must try to get your teammates to guess a target word by giving them clues. Unfortunately for the clue-giver (and the guessers) each target word has a list of 5 "taboo" words that cannot be used in the clue-giving process. This makes the game difficult (and fun), because the taboo words are selected to be the best clues! Furthermore, the off-limit words are right under the target word, staring the clue-giver in the face. Thus, the clue-giver must pay attention to the Taboo words, but suppress the urge to use them as clues.

Consistent with our predictions, we found that performance on this task was related to WMC. We found that high-span participants made fewer Taboo errors, and repeated clues less often. These findings were expected, and consistent with current theories of WMC (Unsworth & Engle, 2007). We also found that were better guesses, and better clue-givers, something that is not predicted by most theories of WMC. Overall, high-spans were better players! From a scientific perspective, these results provide evidence that WMC can impact performance even in a naturalistic, group task. From a game-playing perspective, these results suggest you should try to find some high-span friends for your team!

(An interesting side-note: we found that high-spans were more likely to have previously played Taboo in their free time. Out of the 24 participants that had played before our experiment, 23 were high-spans. In our final analysis, we excluded all the players that had previous experience.)

(PDF)

Direct Response Switching - My working in WMC led me to a general interest in cognitive control and multi-tasking. Beginning with my Masters Thesis, I used the Directed Response Switching task (DRS), developed by myself Stephen Goldinger, and Tamiko Azuma, to probe into these issues. In this task, participants are shown words individually on the screen and have to respond to the words as quickly as possible. The color of the word provides the cue for the correct response. For example, the response for blue and green words might be "name the word aloud", while the response for purple or brown words might be say "bam". Across trials, participants must maintain both responses, but quickly select the correct response each time. In the target "bam" trials, participants have to suppress the prepotent naming response, and execute the secondary response. Reaction times are used to estimate the cost of switching, and provide insight into the underlying control processing.

We have manipulated this basic design to answer several questions about cognitive control. We have varied the nature of the secondary response, making it fairly easy in some situations (press a button) and difficult in other conditions (think of a new word that begins with a specific letter). We have also manipulated the nature of the cue from perceptual (color) to conceptual (group membership). We have examined performance across WMC span groups, and altered the design to answer questions about proactive and reactive control mechanisms (Braver, Gray & Burgess, 2007). This task is simple and flexible, making it easily adapted to new situations. Work using this paradigm has been presented as a poster at Pscyhonomics, and as a talk at Western Psychological Association. Manuscripts are currently in progress.

(POSTER Psychonomics 2006 - Individual Differences in WMC)

(POSTER Psychonomics 2009 - Conceptual/Perceputal Cue)